[Notes] [Git][BuildStream/buildstream][jjardon/license] 623 commits: Include initialization time in the total session time of the build log

[Javier Jardón <gitlab mg gitlab com>]

Title: GitLab

Javier Jardón pushed to branch jjardon/license at BuildStream / buildstream

Commits:

• 63d99463
  by Tristan Van Berkom at 2018-04-01T11:50:59Z

  Include initialization time in the total session time of the build log
  
  This patch refactors the frontend App object and touches a few internals.
  
    o The scheduler now takes a start time given to it at instantiation time,
      instead of considering the start time to commence in Scheduler.run()
  
    o The App.initialize() method has been swapped out for an
      App.initialized() context manager.
  
      This context manager now takes care of the main start/fail/success
      messages. For convenience and ensured consistency, this context manager
      is now responsible for:
  
        o Printing the startup heading
        o Printing the end of session summary
        o Exiting with a consistent error status in case of errors
  
    o The Pipeline() object no longer prints the start/fail/success messages,
      as they are now taken care of by the App()
  
    o The cli.py frontend code is much simplified by using the new context
      manager, also enforcing consistency in how we handle and report errors.
  
  This fixes issue #186
  

• aa8410d8
  by Tristan Van Berkom at 2018-04-01T12:59:01Z

  _workspaces.py: Dont unconditionally create workspace local state file at startup.
  
  Fixes issue #257
  

• 23cea695
  by Tristan Van Berkom at 2018-04-01T18:57:01Z

  _frontend/widget.py: Add "Strict Build Plan" setting to session heading
  
  Looks like this quite interesting attribute was not getting logged.
  

• 999d168a
  by Tristan Van Berkom at 2018-04-01T19:02:02Z

  _frontend/widget.py: Change heading separator to use '=' instead of '~'
  
  Seems that using `~` conflicts with some markdown when trying to
  paste build logs in, e.g. gitlab issues.
  

• 00ad7629
  by Tristan Van Berkom at 2018-04-02T07:01:54Z

  _frontend/app.py: Renamed from main.py
  
  Since the CLI and the App object was split up, we should name
  the file after the object which it implements.
  

• 6c9c2d42
  by Tristan Van Berkom at 2018-04-02T07:53:45Z

  _frontend/app.py: Split up initialization into two parts.
  
  Created partial_init() function to do a partial initialization
  which does not require loading the Pipeline, which is more expensive.
  
  This is useful for some bst commands which dont operate on a pipeline.
  

• 8b5742dd
  by Tristan Van Berkom at 2018-04-02T10:53:59Z

  _exceptions.py: Adding new AppError exception and error domain
  

• 19e31adb
  by Tristan Van Berkom at 2018-04-02T10:55:57Z

  Refactoring of highlevel workspace code
  
  Move all workspace related code out of Pipeline() and into the
  frontend App() object.
  
  Some changes in transition here include:
  
    o Workspaces() object methods for looking up and deleting workspaces
      now take an element name instead of an element.
  
    o Share code for partial App() initialization between the
      `workspace close` and `workspace list` commands
  
    o No longer require that an element exist in the project
      in order to close a workspace
  
  This fixes issue #249
  

• 202dd7cd
  by Tristan Van Berkom at 2018-04-02T11:00:06Z

  source.py: Remove unused method Source._del_workspace()
  
  This was never particularly useful, there is no circumstances
  under which a workspace needs to be deleted, and a cache key
  invalidated, in the course of a session.
  
  A workspace is deleted only atomically as a part of `bst workspace close`,
  which does not even load a pipeline anymore, so the pipeline state need not
  be adjusted in this case.
  

• 22c6ef17
  by Tristan Van Berkom at 2018-04-02T11:04:59Z

  tests/frontend/workspace.py: Fix inaccurate comments
  

• af719007
  by Tristan Van Berkom at 2018-04-02T11:11:11Z

  tests/frontend/workspace.py: Test that we can remove a workspace for a nonexisting element
  
  Guard against regressions of issue #249
  

• 18b60dbb
  by Tristan Van Berkom at 2018-04-02T11:17:54Z

  tests/frontend/workspace.py: Reduce unneeded coverage here.
  
  Here we are overly testing the same functionality which does
  not require source specific support, e.g. lets not test resetting
  of workspaces for every kind of repo, since we already test opening
  and closing of workspaces for every repo kind, this is redundant
  and slowing down tests.
  

• af08b16e
  by Tristan Van Berkom at 2018-04-03T13:49:10Z

  HACKING.rst: Documenting naming policy for private symbols
  
  This is a part of issue #285
  

• 70c73b93
  by Tristan Van Berkom at 2018-04-03T13:49:10Z

  _workspaces.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 500f4f33
  by Tristan Van Berkom at 2018-04-03T13:49:10Z

  _context.py: Adhere to policy on private symbols
  
  And adjust all surrounding sources for changed symbols.
  
  This is a part of issue #285
  

• 01db4072
  by Tristan Van Berkom at 2018-04-03T13:49:11Z

  _exceptions.py: Adhere to policy on private symbols
  
  And adjust all surrounding sources for changed symbols.
  
  Also, added new LoadErrorReason.UNSUPPORTED_PLUGIN, required
  for changes in how the project will report format version errors
  for plugins at creation time
  
  This is a part of issue #285
  

• af34d716
  by Tristan Van Berkom at 2018-04-03T13:49:11Z

  _project.py: Adhere to policy on private symbols
  
  And adjust all surrounding sources for changed symbols.
  
  Additional details:
  
    o Added Project.get_shell_config() to report the shell configuration,
      instead of making those members all public
  
    o Moved assertions about project specified format versions required
      of plugins out of Plugin.__init__, and into Project.create_element()
      and Project.create_source(), so we can keep more things private
  
  This is a part of issue #285
  

• 5c33a984
  by Tristan Van Berkom at 2018-04-03T14:48:56Z

  _pipeline.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 6c14a05a
  by Tristan Van Berkom at 2018-04-04T04:57:50Z

  _plugincontext.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• aff532d4
  by Tristan Van Berkom at 2018-04-04T05:11:50Z

  _variables.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 355b2263
  by Tristan Van Berkom at 2018-04-04T06:17:26Z

  _artifactcache modules: Adhere to policy on private symbols
  
  Additionally:
  
    o This shares more code by creating ArtifactCache.get_artifact_fullname(),
      which is used for the extract directory relative path in both backends,
      and for the ostree "ref" in the ostree backend.
  
    o Further, this removes some redundant documentation in derived abstract
      methods, and clarifies in both backends which methods are abstract methods,
      we should only be documenting abstract method semantics in one place, where
      they are defined.
  
  This is a part of issue #285
  

• 71f07656
  by Tristan Van Berkom at 2018-04-04T09:16:57Z

  Fix strict setting to be consistent throughout the pipeline.
  
  I originally changed the configuration of strict mode to be on
  a per project basis in the user configuration, because the user should
  be able to set their preference on a per project basis.
  
  I however made the mistake to make the strict mode be considered
  on a per project basis within a single pipeline, this commit corrects
  the behavior such that when you are building a project with junctions,
  only the toplevel project is considered when deciding strict mode
  for the whole pipeline.
  

• 9012a5ed
  by Tristan Van Berkom at 2018-04-05T06:16:15Z

  element.py: Fix _update_state() to consider whether an element is to be built properly
  
  This was previously using self._buildable() to determine whether
  an element is built locally and not to be downloaded, which misses
  out on elements which have already been built, like open workspaces
  which can only have their cache key calculated after the build.
  
  This fixes issue #316
  

• daffe498
  by Tristan Van Berkom at 2018-04-05T06:16:15Z

  tests/frontend/workspace.py: Exercise the build test in strict and non-strict mode
  
  This test fails without the previous patch fixing issue #316,
  so this should guard against regressions of building and caching
  workspace builds in non-strict mode.
  

• 638344e4
  by Tristan Van Berkom at 2018-04-05T06:16:15Z

  element.py: Updating comment for _assemble_done()
  
  This was misinforming that it is only called in a subprocess,
  it also needs to be called in the main process, also point out
  that updating element state is a side effect of this.
  

• cc43127e
  by Tristan Van Berkom at 2018-04-05T06:16:15Z

  _scheduler/buildqueue.py: Dont call Element._update_state() here
  
  This is implied by Element._assemble_done()
  

• 8c8b1c54
  by Tristan Van Berkom at 2018-04-05T06:35:20Z

  _options package: Adhere to policy on private symbols
  
  This also adds a couple of methods to OptionPool to avoid
  exposing some internals to other parts of the codebase, and
  calls those new methods from _project.py and _frontend/widget.py
  
  This is a part of issue #285
  

• b8c68284
  by Tristan Van Berkom at 2018-04-05T06:35:20Z

  _platform package: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• ab63b536
  by Tristan Van Berkom at 2018-04-05T08:13:35Z

  _frontend/widget.py: Defend against empty string message detail
  
  It appears we have some cases of empty detail strings that are
  not None, this was resulting in an IndexError when trying to strip
  the trailing newline from the message.detail.splitlines() result.
  

• 207b986f
  by Tristan Van Berkom at 2018-04-05T08:32:42Z

  _frontend/widget.py: Print the workspace directory for %{workspace-dirs}
  
  Instead of printing the list of workspace directory for each source.
  
  This fixes some fallout from the changes to make workspaces element-wide
  instead of being on a per-source basis, which was merged as merge request !257
  as a part of issue #209.
  

• a4eb8bd2
  by Tristan Van Berkom at 2018-04-05T08:35:03Z

  element.py: Removing Element._workspace_dirs()
  
  This private method is now unused and pointless since we
  made workspaces an element-wide concept as a part of issue #209.
  

• 34210d1e
  by Tristan Van Berkom at 2018-04-05T09:54:24Z

  _platform/linux.py: Fix fallout from context private symbols refactor
  
  This was still calling an outdated version of Context._message(),
  which is now Context.message(), causing stack traces on platforms
  without support for user namespaces.
  
  This was fallout from commit 500f4f330fde8b6001a2f8d8921bd5b8acb79960
  

• 22fe6b74
  by Tristan Van Berkom at 2018-04-05T13:11:58Z

  doc/source/projectrefs.rst: Ammended documentation for project.refs
  
  This was missing the main `projects` toplevel key
  

• 67338ef7
  by Tristan Van Berkom at 2018-04-05T13:13:49Z

  _frontend/app.py: More consistent error reporting
  
  Use App.print_error() in all error exit circumstances, this
  will automatically consider any detail strings reported in BstErrors.
  

• 1280b449
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  _exceptions.py: Added new LoadErrorReasons for junction load failures
  

• 81ec3635
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  _exceptions.py: Allow detail strings in LoadError exceptions
  

• 857751fb
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  _loader.py: Some fixes in how we load sources for junctions
  
    o Ensure that we call Source._load_ref(), and consider project.refs this way
  
    o Ensure that we report a warning in the case that project.refs is in use
      and the loaded junction source has a redundant ref which is going to
      be ignored
  
    o Report more distinct machine readable errors for failures to load
      junction element sources
  
    o Handle Consistency.INCONSISTENT and Consistency.RESOLVED separately:
  
      - The user should be told something different depending on whether
        they need to fetch or whether they need to track.
  
      - It is never possible to automatically fetch the source in the
        case that the source has no ref to begin with.
  
  This also adjusts the test/loader/junctions.py test to expect the new error
  

• 50af604e
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  tests/frontend: Share the configure_project() function
  
  Make buildtrack.py and track.py share the same configure_project() helper.
  

• abb9ef22
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  tests/frontend/show.py: Test behaviors of showing junctioned elements
  
    o Test error conditions for showing an unfetched junctioned project
  
    o Test error conditions for showing an untracked junctioned project
  
  Both tests check both modes of ref-storage.
  
  This adds a new shared helper function `generate_junction`
  

• 24ab5ce7
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  tests/frontend/fetch.py: Added tests for automatically fetching the junctions
  
    o This tests that a Consistency.RESOLVED junction will automatically be fetched
      when `bst fetch` is run on a pipeline which refers to a junction.
  
    o Further, it tests that a Consistency.INCONSISTENT junction will bail
      out with the expected error message
  
  Again testing with both ref-storage modes
  

• ee73a87b
  by Tristan Van Berkom at 2018-04-05T13:14:32Z

  tests/frontend/track.py: Added junction related tests
  
    o Test that we bail out with the expected errors when the
      junction element in question is Consistency.INCONSISTENT
  
    o Test that tracking the junction itself, causes a subsequent
      show of the pipeline to not bail out anymore (tests that
      tracking works and persists for a junction element).
  
  Again these use both ref-storage modes.
  

• 8704363a
  by Tristan Van Berkom at 2018-04-05T13:27:16Z

  tests/frontend/buildcheckout.py: Added junction related tests
  
    o Test that we get the expected error if the junction element
      the pipeline refers to is inconsistent
  
    o Test that the junction element is automatically fetched as
      a part of the build phase
  
  Both tests using both ref-storage modes
  

• 09e52016
  by Tristan Van Berkom at 2018-04-06T07:08:47Z

  _exceptions.py: Added LoadErrorReason.INVALID_SYMBOL_NAME
  

• 1b4307f1
  by Tristan Van Berkom at 2018-04-06T08:20:51Z

  _yaml.py: Added assert_symbol_name() helper function.
  
  This function asserts that a loaded symbol name is a valid one,
  and raises an appropriate and consistent LoadError if an invalid
  symbol name is encountered.
  

• 60f2a320
  by Tristan Van Berkom at 2018-04-06T08:36:13Z

  _project.py: Assert that the loaded project name is a valid symbol name
  
  Fixes issue #339
  

• 337075bb
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  _options/optionpool.py: Assert valid symbol names for option names.
  
  These must be alphanumeric and may not contain dashes.
  

• 550cb369
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  _options/option.py: Assert valid symbol names for variable exports
  

• 420114a0
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  tests/format/project.py: Converted to use CLI test harness
  
  Moved the old style project format tests into the new CLI based
  directory in tests/format/ and converted to use the CLI fixture
  for these tests.
  

• 50b694a3
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  tests/project/plugins.py: Removed this old style test
  
  This is sufficiently covered by other tests in tests/format/project.py
  

• c2500610
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  tests/format/project.py: Added tests for loading of invalid project names
  

• a2bc2a1c
  by Tristan Van Berkom at 2018-04-06T08:36:42Z

  tests/format/options.py: Added tests for invalid option name symbols
  

• 21050433
  by Tristan Van Berkom at 2018-04-06T08:51:16Z

  tests/format/options.py: Added tests for invalid variable names in options
  

• 06b1920b
  by Tristan Van Berkom at 2018-04-06T08:51:16Z

  doc/source/projectconf.rst: Properly document the project name.
  
  This is a part of issue #339
  

• 23edfe5d
  by Tristan Van Berkom at 2018-04-06T08:51:16Z

  doc/source/projectconf.rst: Document restrictions on option and option variable names.
  

• c55dad4b
  by Tristan Van Berkom at 2018-04-06T11:45:13Z

  data/userconfig.yaml: Change default logging to print full shas at init and bst show
  
  This may be obnoxious for `bst show` purposes, which will now also print
  the full artifact cache keys instead of nicely abbreviated ones, but it is a
  more appropriate default for the heading of a build session.
  
  This fixes issue #343
  

• 74c60178
  by Chandan Singh at 2018-04-07T09:42:25Z

  _frontend/cli.py: Add option to close multiple workspaces
  
  At present, it is only possible to close workspaces for elements one at
  a time. This can become slightly tedious process when you have multiple
  workspaces open and you want to close all of them, maybe because you
  just finished working on a set of related elements.
  
  Instead of accepting a single element, accept a list of elements as
  argument for `bst workspace close`. Additionally, add `-a`/`--all`
  option to close all workspaces.
  
  Fixes #337 - Add option to close all workspaces.
  

• 2ca11fde
  by Tristan Van Berkom at 2018-04-07T10:02:13Z

  _frontend/cli.py: Remove --no-checkout option for `bst workspace reset`
  
  This option in `bst workspace reset` is just pointless.
  

• 3ba544b8
  by Tristan Maat at 2018-04-07T10:05:10Z

  Allow 'None' as a default_value for _yaml.node_get
  

• 60dbf19f
  by Tristan Van Berkom at 2018-04-07T10:30:47Z

  Complete the work started to allow None as default in _yaml.node_get()
  
  This completes the work which was started in commit
  3ba544b80f9f268be8ffe62fc8589b30212ec4a2 which only went half
  way towards updating all the source code to be consistent and use
  the new semantic which allows None values.
  

• 59064042
  by Tristan Van Berkom at 2018-04-08T08:23:50Z

  _project.py: Remove BST_WORKSPACE_FORMAT_VERSION
  
  This is now defined in _workspaces.py
  

• 9be2dc5a
  by Tristan Van Berkom at 2018-04-08T08:25:21Z

  _versions.py: Added new file just to hold basic symbolic versions
  
  This is where we're going to store BST_CORE_ARTIFACT_VERSION
  and BST_FORMAT_VERSION from now on.
  
  This was a little bit weirdly defined before, and now we need to
  at least have BST_FORMAT_VERSION available from the frontend/cli
  without onerous imports (we need to import it there, but dont want
  to slow down load time for bash completion runs).
  

• b7ec278d
  by Tristan Van Berkom at 2018-04-08T10:59:31Z

  _yaml.py: Allow None for provenance argument in _yaml.assert_symbol_name()
  
  Since we now also use this to validate user input on the command line.
  

• 5de99191
  by Tristan Van Berkom at 2018-04-08T12:56:33Z

  _frontend/cli.py, _frontend/app.py: Implemented new `bst init` command.
  
  This comes with an interactive mode unless the project name is specified
  on the command line.
  
  This fixes issue #342
  

• 337b7a31
  by Tristan Van Berkom at 2018-04-08T12:57:01Z

  tests/frontend/init.py: New test to test edge cases for new `bst init` command.
  

• 96af6dd5
  by Tristan Van Berkom at 2018-04-08T12:57:01Z

  More specific exceptions when a project.conf is missing.
  
  Since we want to react and start an interactive session when the
  project.conf is missing, we need a more specific error to catch.
  

• 1f57e598
  by Tristan Van Berkom at 2018-04-08T12:57:01Z

  _frontend/app.py: Automatically launch interactive `bst init` when project.conf is absent
  
  When running a command where a project.conf does not exist, ask the
  user if they would like to interactively create a project in the
  said directory.
  
  This is a part of issue #342
  

• 6d7f86fe
  by Tristan Van Berkom at 2018-04-08T13:06:34Z

  NEWS: Adding news entry for new `bst init` command
  

• e0a8b9e5
  by Tristan Van Berkom at 2018-04-08T13:55:03Z

  _versions.py: Fixed copyright year on newly added file.
  
  Oops.
  

• 8bfc7371
  by Jim MacArthur at 2018-04-09T09:49:27Z

  _ostree.py: Mention the remote URL when we fail to fetch remote refs
  

• 68cee1c7
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  tests/frontend/workspace.py: Fixed test_build to make the right assertion
  
  This was removing a file from the workspace, building, checking out,
  and then asserting that the file is *still gone* in the workspace,
  ignoring the same file in the checkout.
  

• 66cf08fa
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  element.py: Cache source consistency logic and refactor workspace logic into Element
  
  Workspaces are now element wide, so consistency edge cases must
  be handled at the element level instead of the source level.
  

• 26f7f0b8
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  element.py: Factored out pointless extra Element._workspaced() method
  

• 62346702
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  element.py: Manage scheduled tracking state in Element
  
  This removes the scheduled state of tracking from Sources, as
  this is really an element wide thing.
  
  To be consistent with assembly, now this comes with:
  
    o Element._schedule_tracking()
    o Element._tracking_done()
    o Element.__tracking_scheduled
    o Element.__tracking_done
  
  Updated the TrackQueue() to call Element._tracking_done() similarly
  to how we have BuildQueue() call Element._assemble_done().
  

• 0fbb550e
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  element.py, source.py: Removing knowledge of assemble scheduling state from source
  
  And considering the workspace related edge cases in Element instead of Source.
  

• 67d22d76
  by Tristan Van Berkom at 2018-04-09T10:25:36Z

  element.py, source.py: Element consumes the rest of workspace logic.
  
  And Source no longer has any __workspace handle.
  

• 9e74988e
  by Tristan Van Berkom at 2018-04-09T11:24:34Z

  _workspaces.py: Fix cache key calculation regression
  
  This was regressed in the refactor done in commit f761140f
  

• b8b41724
  by Tristan Van Berkom at 2018-04-09T11:24:34Z

  tests/frontend/workspace.py: Test that we detect modifications made to a workspace
  
  This was previously only working for added or removed files and
  broken for modified files.
  

• 8da99651
  by Tristan Van Berkom at 2018-04-10T08:01:51Z

  element.py: Removed useless __workspace pointer
  
  I added this in the last refactor and forgot to remove it.
  

• be6b353d
  by Tristan Van Berkom at 2018-04-10T12:01:55Z

  _workspaces.py: Remove knowledge of Elements completely
  
  This makes workspaces more cleanly separated from everything else.
  
    o Removed some methods from Workspace()
  
    o Added Element._open_workspace() to initialize workspaces on
      sources
  
    o Cleanup some code in the pipeline
  
    o Have the App() call Element._open_workspace() instead of Workspace.open()
  

• 5a36d04e
  by Tristan Van Berkom at 2018-04-10T12:02:15Z

  _workspaces.py: Add generic serialization/deserialization
  
  Also streamline the instantiation code paths to layer the yaml
  parsing on top of simple dictionary deserialization.
  

• 888f0b9f
  by Tristan Van Berkom at 2018-04-10T12:02:15Z

  _workspaces.py: Added Workspaces.update_workspace()
  
  This takes a serialized workspace dictionary as understood
  by Workspace.from_dict() and created by Workspace.to_dict()
  
  Further, this adds a Workspace.differs() method to compare
  the state of two workspace instances.
  

• 0b5809c4
  by Tristan Van Berkom at 2018-04-10T12:11:54Z

  utils.py: Added _is_main_process()
  
  A helper function to identify if we are running in the main process
  or not.
  

• a9cbb0f7
  by Tristan Van Berkom at 2018-04-10T12:11:54Z

  Serialize workspace modifications in the main process.
  
  This patch makes the child process send any workspace state back
  to the main process and saves it there directly before processing
  the result of a build job, such that workspace modifications in
  child tasks work transparently and no special care needs to be taken
  to save them except when doing so in the main process.
  
  Also this removes the line where we update workspace data at staging
  time.
  
  This fixes issue #352
  

• a867f293
  by Tristan Van Berkom at 2018-04-10T12:11:54Z

  _workspaces.py: Assert that configuration is only saved in the main process.
  

• 37d97414
  by Jim MacArthur at 2018-04-10T12:50:53Z

  _frontend/widget.py: Correct log line if logdir is empty
  

• cb386651
  by Tristan Maat at 2018-04-11T06:49:22Z

  HACKING.rst: Add integration and pytest notes
  

• d46fbbcb
  by Tristan Van Berkom at 2018-04-11T09:42:56Z

  element.py: Preserve workspace state in failed build sandboxes.
  
  As reported in regression #346, since we started mounting the workspaces
  we now fail to debug build failures for workspaces, the shell command
  no longer sees the workspaces files to debug.
  
  As the failed sysroot represents a very particular build failure, it
  is incorrect to attempt to remount a workspace again after a failed
  build, as the workspace files may have changed since the build.
  
  Instead, in the case of a build failure on a workspaced element, we
  copy the files from the workspace in place (while automatically preserving
  the mtimes and everything, as utils.safe_copy() does).
  
  When we implement failed build artifacts, this case will have to be
  handled in a very similar way, too.
  
  This fixes issue #346
  

• a1564c4a
  by Tristan Van Berkom at 2018-04-11T09:42:56Z

  tests/integration/shell.py: Test that workspaced files are visible in a bst shell
  
  Test both in regular `bst shell --build`, and on a failed build sysroot
  with `bst shell --sysroot`, the latter being the semantic used when debugging
  a failed build.
  
  This guards against regressions of issue #346
  

• 5217359f
  by Javier Jardón at 2018-04-11T13:58:10Z

  doc/source/install.rst: Update list of base requirements
  

• 274e1831
  by Javier Jardón at 2018-04-11T13:58:10Z

  source/install.rst: Update ArchLinux installation instructions
  

• 6ad88ecb
  by Tristan Van Berkom at 2018-04-11T14:10:54Z

  element.py: Restructure artifact metadata
  
   o Store the artifact.yaml into 4 separate files now
   o Only load on demand what is needed at a given time
   o Add local caching for what was loaded, ensuring we never redundantly load metadata
  
  This breaks artifact format and will require an artifact revision
  
  This should improve performance for issue #294
  

• c94f1264
  by Tristan Van Berkom at 2018-04-11T14:10:55Z

  element.py, source.py: Cleanup how Source cache keys are calculated.
  
  Recently after a refactor we kept the Source adding workspace keys
  to the source keys because, now clean this up to have the workspace
  key added directly in the Element cache key calculation.
  
  This breaks cache keys.
  

• acfe2494
  by Tristan Van Berkom at 2018-04-11T14:10:55Z

  buildstream/_versions.py: Increment BST_CORE_ARTIFACT_VERSION
  
  Now that we've restructured the artifact metadata into
  separate files, we need to rev the artifact version.
  

• be10c324
  by Tristan Van Berkom at 2018-04-12T12:18:13Z

  _frontend/status.py: Fixed status widgets to show full element names.
  

• 95d1dafc
  by Tristan Van Berkom at 2018-04-12T12:37:24Z

  _frontend/widget.py: Fixed regression in logfile abbreviations.
  
  Added required size_request() method to new MessageOrLogFile() widget.
  
  This came up in merge request !377
  

• fece8cc8
  by Tristan Van Berkom at 2018-04-12T12:39:06Z

  _frontend/cli.py, _pipeline.py: Add options for cross junction tracking.
  
  This patch makes cross junction tracking disabled by default, which
  was the initial intention when landing project.refs but never got around
  to doing this (intended to get addressing of junctioned elements via
  command line sorted first, but didnt happen).
  
  This adds the following options to enable cross-junction tracking:
  
    o bst build -J / --track-cross-junctions
  
    o bst fetch -J / --track-cross-junctions
  
    o bst track -J / --cross-junctions
  
  This also fixes `bst fetch --track` which had a bug, it was avoiding
  to track and fetch elements which are in a `cached` consistency state,
  which is wrong when `--track` is specified.
  
  This also updates some test cases which were broken by
  this change.
  
  This fixes issue #354
  

• 337fc965
  by Tristan Van Berkom at 2018-04-12T12:39:06Z

  NEWS: Added news entry detailing that cross-junction tracking is disabled by default
  

• e3e72616
  by Richard Maw at 2018-04-12T13:25:25Z

  mount: Wrap yields in context managers with try-finally
  
  The terminator context will only clean up on a signal,
  so if another exception causes context manager cleanup
  then unmount won't be called unless it's part of another context manager
  or is wrapped in a try block's except or finally.
  
  Everywhere else's unmounts are handled by delegating to another context manager
  but these were what needed to be fixed.
  
  The change in buildstream/_fuse/mount.py would cause lockups
  since the build worker process still having a subprocess
  blocked its termination from completing
  which in turn caused the pipeline manager process to block indefinitely on it.
  

• 4aa482ae
  by Tristan Van Berkom at 2018-04-12T15:01:36Z

  _pipeline.py: Fix inconsistent element error to print element fullnames.
  

• 39c4cd17
  by Tristan Van Berkom at 2018-04-12T15:01:36Z

  Fix disaster while making cross junction tracking optional.
  
  This disaster was introduced in fece8cc81e8d8412e32c6667682a33e7d2f9dafe
  
  When doing `bst build --track`, we were:
  
    o first scheduling every element to be tracked
  
    o later filtering out the cross junction elements
  
    o finally asserting consistency state, which would trigger
      an error because we previously scheduled for tracking.
  
  Fixed this by moving all code which resolves elements to track
  into Pipeline.initialize(), and removing special element list handling
  from the individual build/fetch/track commands.
  

• 522d6538
  by Tristan Van Berkom at 2018-04-12T15:01:36Z

  _scheduler/trackqueue.py: Mark skipped elements as tracked.
  
  This ensures their state will be updated, it is especially important
  for sourceless elements which are scheduled to be tracked, like stack elements.
  
  This fixes a regression from my previous refactor or Sources
  which added Element._tracking_done()
  

• 7c9713ad
  by Richard Maw at 2018-04-12T18:03:09Z

  frontend: capture errors for later use
  
  This will be used to provide an error summary before the pipeline summary.
  

• 30af44e8
  by Richard Maw at 2018-04-12T18:03:09Z

  frontend: Use textwrap.indent to indent
  
  The intent is more obvious by using an existing module.
  

• 7b0a9656
  by Richard Maw at 2018-04-12T18:03:09Z

  frontend: Summarise build log output in --no-verbose mode
  

• bea028a8
  by Richard Maw at 2018-04-12T18:03:09Z

  frontend: Control indentations of multi-line values
  

• fc6d192d
  by Richard Maw at 2018-04-12T18:03:09Z

  frontend: Redisplay failure messages in result summary
  
  It's more convenient than having to search back through all the output
  to find out what the cause of failure was.
  

• eb522a4f
  by Richard Maw at 2018-04-12T18:03:09Z

  tests: Check that output includes log files of failed builds
  

• d94d47c5
  by Richard Maw at 2018-04-12T18:03:09Z

  NEWS: Mention failed job summary
  

• f1567355
  by Javier Jardón at 2018-04-12T22:09:00Z

  .gitlab-ci.yml: Create template for linux-tests job
  

• 9bc2bb04
  by Javier Jardón at 2018-04-12T22:09:00Z

  .gitlab-ci.yml: Run tests in Fedora:27 as well
  

• 1af6f442
  by Javier Jardón at 2018-04-12T22:09:00Z

  .gitlab-ci.yml: Change jobs from <distro>-tests to tests-<distro>
  
  So they show in order nicely in the gitlab CI UI
  

• 8378f5dc
  by Javier Jardón at 2018-04-12T22:15:33Z

  .gitlab-ci.yml: Fix coverage job
  

• 63b7c821
  by Tristan Van Berkom at 2018-04-13T06:01:37Z

  .gitlab-ci.yml: Show which coverage reports are getting combined in CI
  
  This extra line shows which reports will be combined, which helps us
  to more quickly debug problems in the coverage job of the .gitlab-ci.yml
  

• 3b8dd43e
  by Tristan Van Berkom at 2018-04-13T10:32:53Z

  plugin.py: Documentation improvements.
  
  Adding "Abstract Methods" section indicating what implementors
  have to implement.
  
  Also fixed some inconsistencies, and moved the abstract methods
  to the top of the class.
  

• 1ea74544
  by Tristan Van Berkom at 2018-04-13T10:32:53Z

  element.py: Documentation improvements
  
  Added heading section describing the abstract methods, and
  moving the abstract methods to the top of the class.
  

• 03bead6c
  by Tristan Van Berkom at 2018-04-13T10:32:53Z

  source.py: Enhanced documentation
  
  Added new Abstract Methods section here, and rearranged sources
  such that abstract methods are on top.
  

• 596264d1
  by Tristan Van Berkom at 2018-04-13T10:36:27Z

  buildelement.py: Documentation improvements
  
  Document here in terms of what the BuildElement does to implement
  the abstract Element methods.
  

• 299df233
  by Tristan Maat at 2018-04-13T11:51:17Z

  Add element.prepare method
  
  This is one of the tasks of #209
  

• ccb6acc1
  by Tristan Maat at 2018-04-13T11:51:17Z

  integration/workspace.py: Test that configure commands run only once
  

• d1b14086
  by Tristan Maat at 2018-04-13T11:51:17Z

  tests/frontend/workspace.py: Automatically update version
  

• 5632d291
  by Tristan Maat at 2018-04-13T11:51:17Z

  tests/frontend/workspace.py: Test new workspace format version
  

• 2cec9c27
  by Tristan Maat at 2018-04-13T11:54:10Z

  Update plugin documentation
  

• cb702588
  by Tristan Maat at 2018-04-13T12:02:51Z

  Add NEWS entry
  

• a4927ad3
  by Tristan Van Berkom at 2018-04-14T08:32:09Z

  tests/frontend/show.py: Removing some unused variables
  

• ae8543c3
  by Tristan Van Berkom at 2018-04-14T08:32:09Z

  tests/format/variables.py: Ported from old style tests
  

• 188d819d
  by Tristan Van Berkom at 2018-04-14T08:32:09Z

  buildelement.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• d45e12f7
  by Tristan Van Berkom at 2018-04-14T09:13:27Z

  source.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 3ef81d7f
  by Tristan Van Berkom at 2018-04-14T09:14:12Z

  buildelement.py: Use Element.get_variable() instead of poking at private thing
  
  This is the only place where Element._get_variables() was used, which we'll
  now happily remove - also Element implementations in general should not
  have to use private API.
  

• 7db09e2c
  by Tristan Van Berkom at 2018-04-14T10:19:51Z

  element.py: Adhere to policy on private symbols
  
  This involves a lot of reordering of function in the Element class.
  
  This is a part of issue #285
  

• 476526fc
  by Tristan Van Berkom at 2018-04-14T10:23:00Z

  element.py: Rename _get_full_display_key() -> _get_display_key()
  
  The extra "full" was a bit redundant here, the Element does not
  export any "less than full" display key accessor, so lets use
  a more simple name for this.
  

• 1094d62b
  by Tristan Van Berkom at 2018-04-14T10:45:09Z

  _artifactcache/ostreecache.py: Extracting a missing artifact is a bug
  
  Raise an assertion here, not an ArtifactError.
  

• d603051e
  by Tristan Van Berkom at 2018-04-14T10:45:40Z

  _artifactcache/tarcache.py: Extracting a missing artifact is a bug
  
  Raise an assertion here, not an ArtifactError.
  

• d41940f5
  by Abderrahim Kitouni at 2018-04-14T13:53:36Z

  plugins/elements/cmake.yaml: allow using ninja instead of make (#279)
  
  This uses the build tool mode of cmake to have a single command that can call
  either make or ninja.
  
  I've also modified the tests to take the new commands into account
  

• f6f4738c
  by Tristan Van Berkom at 2018-04-15T10:04:07Z

  _frontend/app.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 06da7a02
  by Tristan Van Berkom at 2018-04-15T11:10:01Z

  _frontend/profile.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• 219c7d29
  by Tristan Van Berkom at 2018-04-15T11:10:01Z

  _frontend/status.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• e769beca
  by Tristan Van Berkom at 2018-04-15T11:10:01Z

  _frontend/widget.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• f2edff31
  by Tristan Van Berkom at 2018-04-15T11:48:44Z

  buildstream/plugin.py, doc/source/projectconf.rst: Changed policy for core plugin format revisioning
  
  The policy for format revisioning of core plugins maintained in BuildStream
  now changes such that we revision all plugins under the same core format version.
  
  This means that format extensions in plugins should be revisioned in
  the global _versions.BST_FORMAT_VERSION along with the core format, this
  simplifies parameterization of project.conf for usage of core plugins.
  
  This did not really require any code support as our plugins have never
  been revisioned before.
  

• 43e09ccf
  by Tristan Van Berkom at 2018-04-15T11:50:01Z

  _versions.py: Bump BST_FORMAT_VERSION for new cmake format extension
  
  With commit d41940f516498f827967b8e5d311ce6accb88f56, the cmake plugin
  now supports a new `generator` option which allows one to specify whether
  to use `Unix Makefiles` (default) or `Ninja` (optional).
  
  This requires a format version bump if people want to be sure that
  the feature is supported.
  

• 075b9819
  by Tristan Van Berkom at 2018-04-15T11:52:56Z

  NEWS: Adding news entry for cmake plugin enhancement.
  

• 938ec1f4
  by Tristan Van Berkom at 2018-04-16T06:15:30Z

  doc/source/format.rst: Some enhancements in how we document dependencies
  
   * Added link target in the project.conf documentation for the element-path
  
   * List elements in examples without the leading `elements/` path, this
     is misleading and predates the existence of the element-path
  
   * Specify "element path relative" instead of "project relative", this
     was also outdated. Use these opportunities to link back to the element
     path project.conf docs section
  
   * Document dependency attributes differently
  
     - Use a list of attributes
  
     - Link back to the first example when speaking of shorthand, instead
       of duplication the example in two places
  
     - Enhanced description of how junction attributes are used
  
     - Note that ordering in dependency declaration lists is not meaningful
  
       This is an incorporation of Javier Jardón's patch for issue #358
  
  This fixes issue #358.
  

• 5cedd894
  by Tristan Van Berkom at 2018-04-16T06:50:56Z

  element.py: Document new Element.prepare() method as available since 1.2
  

• 7c2a43d1
  by Tristan Van Berkom at 2018-04-16T06:53:17Z

  element.py: Document new BST_FORBID_SOURCES and BST_FORBID_RDEPENDS as available since 1.2
  
  Also use any() for the check on the generators, instead of expanding the
  whole list exhaustively and checking if the list is not empty.
  

• ca2331c3
  by Tristan Van Berkom at 2018-04-16T07:45:04Z

  source.py, element.py, _pipeline.py: Streamling preflighting.
  
  Instead of having the pipeline preflight all sources separately
  from elements, have the element preflight it's sources.
  
  This is in order to simplify the shared code path for the pipeline
  and the loader to use for instantiating elements.
  
  Also updated tests to expect the new ElementError and SourceError
  instead of the PipelineError which was raised for preflighting before.
  

• cd90fbde
  by Tristan Van Berkom at 2018-04-16T09:26:05Z

  Clean up element/source instantiation code paths.
  
  This removes the extra `kind` parameter from all of the related
  codepaths, it is redundant since the `kind` attribute is already
  stored on the MetaElement and MetaSource objects.
  

• b4f0a52a
  by Tristan Van Berkom at 2018-04-16T10:59:47Z

  element.py, _pipeline.py: Moved instantiation codepath to Element class methods.
  
  This will allow the instantiation codepath to be shared by the Loader
  which also needs to instantiate elements for junctions.
  

• d02a1b4e
  by Tristan Van Berkom at 2018-04-16T11:34:40Z

  _loader.py: Use shared element instantiation codepaths for junction loading.
  
  The consequently fixes issue #292, and you can now use workspaces
  to work on junctioned projects.
  

• e48f2f82
  by Tristan Van Berkom at 2018-04-16T12:14:37Z

  tests/frontend/buildcheckout.py: Added regression tests for workspaced junctions
  
  This guards against regressions of issue #292
  

• 827a1294
  by Tristan Van Berkom at 2018-04-16T12:52:04Z

  _pipeline.py: Stop printing warning when a build plan does not use some existing workspaces
  
  Fixes issue #360
  

• eb0ad4d4
  by Tristan Van Berkom at 2018-04-16T13:30:54Z

  .gitlab-ci.yml: Perform some python code analysis with radon
  
  This adds a new job in the prepare stage which can be viewed in
  the gitlab CI, and also produces an artifacts with the json metrics.
  

• a33ac81e
  by Tristan Van Berkom at 2018-04-16T13:58:26Z

  .gitlab-ci.yml: Add coverage regex
  

• 86587bf9
  by Javier Jardón at 2018-04-16T13:58:26Z

  README.rst: Add coverage badge
  

• 9cf8d35f
  by Tristan Van Berkom at 2018-04-17T07:02:09Z

  _frontend: Fix error messaging for closing workspaces
  
  This fixes issue #182 again
  

• c9b3aad7
  by Tristan Van Berkom at 2018-04-17T07:24:40Z

  _frontend: Earlier assertion for `bst workspace open` when directory is not empty
  
  This just raises the error to the user that the checkout directory
  where they intend to create a workspace in is not empty, a little
  bit earlier, without bothering to initialize the pipeline first.
  

• 548c03f5
  by devcurmudgeon at 2018-04-17T07:47:16Z

  Update README.rst with more detail
  

• 4fc2da9a
  by Tristan Van Berkom at 2018-04-17T08:31:48Z

  element.py: Fix force opening of workspaces.
  
  This was broken because we let the Source objects initialize workspaces
  directly in the target workspace directory, where files might already
  exist.
  
  Instead, initialize the workspace in a tempdir and hardlink the files
  from there.
  
  This fixes issue #364
  

• 098ee3cf
  by Tristan Van Berkom at 2018-04-17T08:31:48Z

  tests/frontend/workspace.py: Added tests for force opening workspaces
  
  Guard against regressions of issue #364
  

• 226880ee
  by Javier Jardón at 2018-04-17T08:49:59Z

  README.rst: Point to new location of x86image example
  

• 880e1913
  by Paul Sherwood at 2018-04-17T12:36:46Z

  Fix typo, add inspiration projects
  

• 6c23e337
  by Chandan Singh at 2018-04-17T13:01:22Z

  refactor: Do not assume there is only one workspace target in app.py
  
  While working with workspaces in _frontend/app.py, we currently assume
  that there is only workspace in the pippeline. This limits our ability
  to support multiple workspaces in a given command.
  
  Remove this assumption from _frontend/app.py which will allow for the
  possibility of supporting multiple workspaces from _frontend/cli.py in
  future. This commit does not change the behavior of these commands
  though as the target is instead passed as an argument.
  

• b264430d
  by Chandan Singh at 2018-04-17T13:01:22Z

  _frontend/cli.py: Add option to reset multiple workspaces
  
  !357 added support for closing multiple workspaces. Similarly, also
  allow `bst workspace reset` to work on multiple workspaces, with `--all`
  as a helper to reset all open workspaces.
  

• 78a86f60
  by Chandan Singh at 2018-04-17T14:53:17Z

  NEWS: workspace close/reset support multiple elements
  

• fb6687b6
  by Tristan Van Berkom at 2018-04-17T15:55:01Z

  _loader package: Creating a new sub package to hold the loader related things.
  

• 5f3da731
  by Tristan Van Berkom at 2018-04-17T15:55:01Z

  _loader package: Splitting up the loader code into separate pieces
  
  This also refactors the loader code to honor private symbol
  naming policy.
  
  This is a part of issue #285
  

• 1159bbc8
  by Tristan Van Berkom at 2018-04-17T15:55:01Z

  _context.py: Removing unused _platform private variable
  

• ea7d6ed2
  by Tristan Van Berkom at 2018-04-17T15:55:01Z

  _context.py: Make `fetch_subprojects` public.
  
  This is actually used by the Loader and is not private.
  
  This is a part of #285
  

• 8027b30d
  by James Ennis at 2018-04-17T16:52:27Z

  index.rst: remove captions
  

• f0b350f1
  by James Ennis at 2018-04-17T16:52:27Z

  main_install.rst: New file with install, docker and artifacts
  
  We now have a main installing page with three sections:
  1. How to install buildstream onto your machine.
  2. How to run buildstream inside docker image.
  3. How to install an artifact server
  

• 4a49cedb
  by James Ennis at 2018-04-17T16:52:27Z

  main_authoring.rst: move to using buildstream
  

• 5d2f6ec6
  by James Ennis at 2018-04-17T16:52:27Z

  Add a general documentation section
  

• 02119f19
  by James Ennis at 2018-04-17T16:52:27Z

  index.rst: Remove unnecessary toctree blocks
  

• 58f8e6bc
  by James Ennis at 2018-04-17T16:52:27Z

  Consistent titling
  

• 516daa5c
  by James Ennis at 2018-04-17T16:52:27Z

  invoking.rst: renamed to commands.rst (anchors also changed)
  

• 5b5c87bf
  by James Ennis at 2018-04-17T16:52:27Z

  config.rst: Renamed to user_config.rst and anchors changed
  

• e4cea3da
  by James Ennis at 2018-04-17T16:52:27Z

  core_framework.rst: New file containing content that was in main_core.rst
  

• 9a3f008e
  by James Ennis at 2018-04-17T16:52:27Z

  main_core.rst: Change title and moved authoring, sandboxing and cachekeys documentation here
  

• 974bd246
  by Tristan Van Berkom at 2018-04-18T03:44:50Z

  _scheduler: Move job.shutdown() calling responsibility inside of Job.
  
  This was being called by Queue if Job succeeds, or inside Job if
  Job fails and retries, lets just keep this inside of Job.
  

• 835e5475
  by Tristan Van Berkom at 2018-04-18T04:49:40Z

  _scheduler: Slightly changing the Job / Queue API contracts
  
  For the Job:
  
     o The action_cb, complete_cb and max_retries are now passed
       into the constructor, and the Job.spawn() API has no arguments.
  
     o The complete_cb() signature has changed to take a success
       boolean instead of a returncode integer, and also now receives
       the result directly.
  
  For the Queue:
  
     o Adapted to new Job API contract changes
  
     o Changed the Queue.done() method to now take a success boolean
       instead of a returncode integer
  
  Updated Queue implementations to handle the success boolean instead
  of the returncode integer.
  

• 6a4949c2
  by Tristan Van Berkom at 2018-04-18T06:04:32Z

  _scheduler/job.py: Adhere to policy on private symbols
  
  This also removes an illegal peek into element._Element__environment
  and uses the public Element.get_environment() accessor instead, removing
  the need to also sanitize the returned dictionary.
  
  Further, this reduces the amount of data passed between processes
  in asyncio callbacks.
  
  This is a part of issue #285
  

• 494fdd65
  by Tristan Van Berkom at 2018-04-18T06:18:43Z

  _scheduler: Adding Queue.dequeue_ready() to avoid peeking into private queue data
  
  And micro-optimizing the scheduler use a generator _expression_ with any()
  to check the result instead of creating a list.
  

• 3b447405
  by Tristan Van Berkom at 2018-04-18T06:44:48Z

  _scheduler/queue.py: Adhere to policy on private symbols
  
  And change the API contract a bit more, now the Queue object
  takes the scheduler in the constructor and the Queue.attach()
  method is removed.
  
  This is a part of issue #285
  

• 853c54eb
  by Tristan Van Berkom at 2018-04-18T09:38:40Z

  _scheduler/job.py: Don't mess with scheduler loop when forking
  
  Instead of unregistering the mainloop handling of SIGTSTP, block
  it (along with SIGTERM) while forking, clear the signal handlers
  and then unblock them in the child process; let the parent handle
  SIGTSTP after unblocking in the main process.
  

• 268340a5
  by Tristan Van Berkom at 2018-04-18T11:02:25Z

  _scheduler/scheduler.py: Adhere to policy on private symbols
  
  Also added Scheduler.jobs_suspended() context manager to
  use in the App instead of exposing Scheduler.suspend_jobs()
  and Scheduler.resume_jobs() along with signal callback connect
  and disconnect APIs, this keeps things more private in the
  scheduler.
  
  This is a part of issue #285
  

• e18b7e5b
  by Tristan Van Berkom at 2018-04-18T11:02:25Z

  _scheduler/fetchqueue.py: Adhere to policy on private symbols
  
  This is a part of issue #285
  

• ee7e89d2
  by Tristan Van Berkom at 2018-04-18T14:13:21Z

  HACKING.rst: Clarify policy that related issues must be mentioned in commit messages.
  
  This was somehow less than obvious.
  

• fd9ee860
  by James Ennis at 2018-04-18T14:16:02Z

  HACKING.rst: Include Documentation Formatting Policy
  

• 149d0859
  by James Ennis at 2018-04-18T14:16:02Z

  HACKING.rst: move location of useful links
  

• 39a7c1f9
  by Tristan Van Berkom at 2018-04-18T14:16:55Z

  HACKING.rst: Fixing profiling section to follow rst formatting policy
  
  Two empty lines before a section title, no empty line after the section title
  

• ae1d5093
  by Javier Jardón at 2018-04-18T14:19:00Z

  README.rst: Add pipeline status badge
  

• 499c6f9d
  by Richard Maw at 2018-04-18T14:50:04Z

  frontend: Log which elements resolved new keys
  
  It can be difficult to pick out important information in all the messages,
  so it's convenient to have classes of them collected into one place.
  
  One such class of important information is when the cache key of an element changes,
  such as when element's source is tracked or cached from a workspace,
  since it can be important to know what the cache key is for that artifact.
  
  Fixes issue #252
  

• a850b8ed
  by Richard Maw at 2018-04-18T14:50:33Z

  Update NEWS
  

• cd75869f
  by Javier Jardón at 2018-04-18T15:25:10Z

  Add about page
  

• 899c3ce0
  by Javier Jardón at 2018-04-18T15:25:10Z

  doc/source/index.rst: Replace welcome paragraph
  

• 64069c8c
  by Tristan Van Berkom at 2018-04-19T06:11:47Z

  _pipeline.py: Added PipelineSelection
  
  Part of a slow, ongoing refactor to unmangle the pipeline into
  something which just creates pipelines of elements for a centerpiece
  to process.
  
  This also renames Pipeline.deps_elements() -> Pipeline.get_selection()
  

• 67437803
  by Tristan Van Berkom at 2018-04-19T06:11:47Z

  _pipeline.py, _frontend/app.py: Added track_selection initialization argument
  
  This informs the pipeline what PipelineSelection mode to use
  for constructing the list of elements to track.
  

• d73d75c4
  by Tristan Van Berkom at 2018-04-19T06:11:47Z

  _frontend/cli.py: Set the track_selection argument to the deps parameter
  
  Tell the pipeline to only track either 'none' or 'all'
  PipelineSelection modes.
  
  This fixes issue #367
  

• 670487be
  by Tristan Van Berkom at 2018-04-19T06:11:47Z

  tests/frontend/track.py: Added test for `bst track --deps none`
  
  This guards against regressions of issue #367
  

• 6fbc81e7
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_artifactcache/pushreceive.py: Fixing unused variables
  

• 480876af
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_context.py: Fixing unused variables
  

• 6004943e
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_frontend/app.py: Fixing unused variables
  

• dc30af8a
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_frontend/status.py: Fixing unused variables
  

• ec2648cb
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_loader/loader.py: Fixing unused variables
  

• cb3a2149
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/_options/optionpool.py: Fixing unused variables
  

• a8c31ff3
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/element.py: Fixing unused variables
  

• f5b99559
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/plugin.py: Fixing unused variables
  

• 849d66e0
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/plugins/sources/_downloadablefilesource.py: Fixing unused variables
  

• 0a718d86
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/sandbox/_mount.py: Fixing unused variables
  

• 0830323d
  by Tristan Van Berkom at 2018-04-19T06:38:09Z

  buildstream/utils.py: Fixing unused variables
  

• c1c09e36
  by Tristan Van Berkom at 2018-04-19T06:38:20Z

  .pylintrc: Enable linting for unused variables
  

• 78bd7f56
  by devcurmudgeon at 2018-04-19T11:11:09Z

  Add example formats
  

• 61889ed5
  by Tristan Van Berkom at 2018-04-20T05:32:05Z

  plugins: Disable attribute-defined-outside-init pylint error
  
  Plugins set their attributes in configure(), because the
  constructor is not public API.
  

• 671a7412
  by Tristan Van Berkom at 2018-04-20T05:34:04Z

  _context.py: Properly declare and document log_message_format in the constructor
  

• be5b26c3
  by Tristan Van Berkom at 2018-04-20T05:35:43Z

  _pipeline.py: Declare tracking related members in the constructor
  

• 1fcfd000
  by Tristan Van Berkom at 2018-04-20T05:37:39Z

  _artifactcache/pushreceive.py: Fixed ProcessWithPipes to declare attributes in constructor
  

• 2282af27
  by Tristan Van Berkom at 2018-04-20T05:40:46Z

  _scheduler/job.py: Disable attribute-defined-outside-init lint error
  
  This is in a derived Process from the standard lib, we add this
  here because we are unsure about the constructor signature.
  

• 5f378fea
  by Tristan Van Berkom at 2018-04-20T05:41:04Z

  .pylintrc: Enable attribute-defined-outside-init messages
  
  This will give us an error if an instance attribute is ever
  declared outside of the constructor, which usually constitutes
  either a bug or a violation of the coding standard, rendering
  code more difficult to read.
  

• 1b7b51ac
  by milloni at 2018-04-20T10:50:06Z

  setup.py: Assert Bubblewrap version
  
  The required version is 0.1.2 based on the fact that we're using
  --hostname and --remount-ro.
  

• 7854fc50
  by milloni at 2018-04-20T10:52:38Z

  doc/source/install.rst: Mention bwrap version requirements
  

• aab0229b
  by Tristan Maat at 2018-04-20T11:00:46Z

  deb.py: Make pylint stop complaining about arpy
  

• 0213102c
  by Tristan Maat at 2018-04-20T11:03:30Z

  HACKING.rst: Add a note about arpy being required for docs
  

• 1598d35e
  by Tristan Van Berkom at 2018-04-23T06:59:16Z

  doc/source/index.rst: Make main documentation title sober again.
  
  Drop the happy welcome phrase with exclamation point.
  

• f7713400
  by Tristan Van Berkom at 2018-04-23T07:37:02Z

  tests/pipeline/preflight-error: Fixing typo in error string formatting
  

• 3199d357
  by Tristan Van Berkom at 2018-04-23T08:19:58Z

  Overwrite sys.excepthook to handle exceptions in the main application.
  
  This uses the existing messaging system to report errors in the main
  application as a BUG type. It requires the use of a global_app variable
  which isn't ideal; this may be replaced in future. This partially addresses
  issue #197.
  
  Theoretically, an exception could occur before Scheduler.loop is set up,
  hence the check for it when terminating all jobs.
  
  NOTE: This was originally submitted by Jim MacArthur, and manually
        reapplied after some refactoring took place.
  

• 4dd72d6c
  by Javier Jardón at 2018-04-23T11:13:35Z

  Revert "doc/source/install.rst: Mention bwrap version requirements"
  
  Seems this was not propperly rebased; libostree and bubblewrap requirements
  are already there.
  Also the paragraph about docker was removed in
  commit f0b350f158ed14c97613dc22b83ed89759b8d255
  
  This reverts commit 7854fc5037204474d16618c657a1cd1ca12c888e.
  

• c939d5f9
  by Javier Jardón at 2018-04-23T11:13:35Z

  doc/source/install.rst: minimum bubblewrap version is 0.1.2
  

• 2fe1010a
  by Tristan Van Berkom at 2018-04-23T13:00:28Z

  doc/source/install.rst: Removing blank line which appeared below title
  
  Sections have 2 blank lines before them, and no blank line after them.
  

• a18fcf0c
  by Jim MacArthur at 2018-04-23T13:03:27Z

  Add artifact cache receive profiling domain
  
  _artifactcache/pushreceive.py: Wrap OSTreeReceiver.do_run in a
    profiling domain.
  _profile.py: Add 'ARTIFACT_RECEIVE' domain.
  

• 1cd3728c
  by Jim MacArthur at 2018-04-23T13:03:27Z

  HACKING.rst: Add a section on BST_PROFILE and the artifact cache.
  

• 15789951
  by Tristan Van Berkom at 2018-04-23T13:04:17Z

  HACKING.rst: Fixing formatting of titles.
  
  Titles have 2 blank lines before them, and no blank lines after them.
  

• 75ebcf56
  by Tristan Maat at 2018-04-23T13:10:18Z

  bst-here: Add '-p' flag to upgrade buildstream (Issue #291)
  

• 4759e2ad
  by Sam Thursfield at 2018-04-23T13:37:06Z

  plugins/elements/compose.py: Use sets instead of lists where appropriate
  
  There is no significant order for the lists of added, removed and
  modified files, so use an unordered set() to store the data.
  

• d423732f
  by Sam Thursfield at 2018-04-23T13:37:06Z

  plugins/elements/compose.py: Only calculate snapshot when needed
  
  This is an optimisation for the case where the compose element doesn't
  do any splitting, and also brings 2 related code fragments closer to
  each other.
  

• cb10e2bd
  by Sam Thursfield at 2018-04-23T13:37:06Z

  plugins/elements/compose.py: Avoid losing files inside directory symlinks
  
  The logic for determining which files were removed by integration
  commands was broken when dealing with files staged within symlink
  directories.
  
  This rather weird scenario is only possible because of the way
  BuildStream layers artifacts. If artifact 1 contains a symlink from
  `/sbin` to `/usr/sbin`, and artifact 2 is staged on top and contains
  a file `/sbin/init`, then the resulting filesystem contains a file at
  `/usr/sbin/init`.
  
  The manifest used by the compose element is generated from the contents
  of the individual artifacts, so it lists the original paths such as
  `/sbin/init`, but would would not contain `/usr/sbin/init` as nothing
  has processed the symlinks.
  
  The path `/sbin/init` is valid inside the composed tree, but filesystem
  traversals that don't follow symlinks will not report that path in their
  results. The compose plugin would look for `/sbin/init` in the results
  of `utils.list_relative_paths()`, find it missing, and would act as if
  some integration command had removed the file. This meant it would not
  end up in the results.
  
  To fix this, I have inverted the logic that processes the results of the
  integration commands. We now work through every path in the manifest
  and check it against the results of the integration commands, rather
  than the other way around, and if any path from the manifest doesn't
  appear in the snapshot we assume that it has staged in a different
  location due to symlinks.
  
  See: https://gitlab.com/BuildStream/buildstream/issues/270
  

• 3ec1cab1
  by Sam Thursfield at 2018-04-23T13:37:06Z

  Add compose-symlinks testcase
  
  This is a regression test for the issue raised here:
  https://gitlab.com/BuildStream/buildstream/issues/270
  

• caa52411
  by Sam Thursfield at 2018-04-23T13:37:06Z

  tests/testutils/integration.py: Avoid inconsistent symlink handling
  
  The output of walk_dir() seemed to be inconsistent in how it traversed
  symlinks. Presumably this is to do with differences in how the filesystem
  return files. If we do an in-place sort of the list of files and directories
  that we get, os.walk() will honour that order which should make the output
  stable.
  

• 84862708
  by Javier Jardón at 2018-04-24T01:45:08Z

  doc/source/install.rst: Update install instruction for Debian strech
  

• dabaf601
  by Javier Jardón at 2018-04-24T01:45:08Z

  doc/source/install.rst: Update build instructions for Debian Testing/Sid
  

• fbcf4ad4
  by Javier Jardón at 2018-04-24T01:45:08Z

  doc/source/install.rst: Update build instructions for Fedora
  

• 4be5f1f6
  by Javier Jardón at 2018-04-24T01:45:08Z

  doc/source/install.rst: Add fuse to the list of dependencies
  
  See #358
  

• 07ac1bfa
  by Javier Jardón at 2018-04-24T01:45:08Z

  doc/source/install.rst: Add requirements to use the bst cache
  
  See #358
  

• b6a64313
  by Javier Jardón at 2018-04-25T08:22:53Z

  Move curent before_script: commands to the only jobs where they are actually needed
  

• ae927b81
  by Javier Jardón at 2018-04-25T08:29:55Z

  .gitlab-ci.yml: Add job to analyze project code quality with Code Climate CLI
  
  See #431
  

• 60029b31
  by Javier Jardón at 2018-04-25T08:29:55Z

  .gitlab-ci.yml: Move radon job to post stage
  

• 38ac4992
  by Tristan Van Berkom at 2018-04-25T13:29:05Z

  _projectrefs.py: Adhere to policy on private symbols
  
  A late fix for issue #285
  

• 7caecc68
  by Tristan Van Berkom at 2018-04-25T14:26:42Z

  _loader: Give the loaded MetaSource objects the kind of the owning element
  
  This is needed so that Sources can derive whether they belong
  to a junction or not, which is needed for separating where
  junction refs are stored.
  

• 719a3fc3
  by Tristan Van Berkom at 2018-04-25T14:26:42Z

  Load and save junctioned source refs from/to junction.refs
  
    o _projectrefs.py: Additional constructor option to choose the base name
  
    o _project.py: Load two ProjectRefs objects, one for the junctions
  
    o source.py: Load and save junctioned source refs with the appropriate ProjectRefs object
  
    o tests: Updated some tests to expect junctions to be stored in junction.refs
  
  This fixes issue #361
  

• 52da8751
  by Tristan Van Berkom at 2018-04-25T14:32:47Z

  doc: Clarify that junction sources are stored separately from project.refs
  

• 54633e92
  by Javier Jardón at 2018-04-25T16:16:18Z

  .gitlab-ci.yml: Run test in current Debian stable (stretch)
  

• 4be462d3
  by Tristan Van Berkom at 2018-04-26T05:17:44Z

  Use versioneer instead of setuptools_scm
  
  Using setuptools_scm had a couple of bad problems:
  
    o Unexpected versioning semantics, setuptools_scm would
      increment the micro version by itself in the case that
      we derive a version number from something that is not a tag,
      making the assumption that we are "leading up to" the next
      micro version.
  
      People mostly dont expect this.
  
    o When installing in developer mode, i.e. with `pip3 install --user -e .`,
      then we were always picking the generated version at install time
      and never again dynamically resolving it.
  
      Many of our users install this way and update through git, so it's
      important that we report more precise versions all the time.
  
  This commit needs to make a series of changes at the same time:
  
    o Adds versioneer.py to the toplevel, this is used by setup.py
      for various activities.
  
      This is modified only to inform the linter to skip
  
    o Adds buildstream/_version.py, which is generated by versioneer
      and gives us the machinery to automatically derive the correct version
  
      This is modified only to inform the linter to skip
  
    o Adds a .gitattributes file which informs git to substitute
      the buildstream/_version.py file, this is just to ensure that
      the versioning output would work if ever we used `git archive`
      to release a tarball.
  
    o Modifies setup.py and setup.cfg for versioneer
  
    o Modifies utils.py and _frontend/cli.py such as to avoid importing
      the derived version when running bash completion mode, we dont
      derive the version at completion time because this can result
      in running a subprocess (when running in developer install mode)
      and is an undesirable overhead.
  
    o Updates tests/frontend/version.py to expect changed version output
  

• 16e719a9
  by Tristan Van Berkom at 2018-04-26T05:23:45Z

  NEWS: Added news entry for versioning fixes
  

• b5c59295
  by Tristan Van Berkom at 2018-04-26T05:47:17Z

  element.py: Added BST_FORBID_BDEPENDS to compliment BST_FORBID_RDEPENDS
  

• 14aa2590
  by Tristan Van Berkom at 2018-04-26T05:47:17Z

  plugins/elements/junction.py: Forbid dependencies
  
  Move this logic into the junction element itself, instead
  of special case erroring for this in the loader.
  

• edfcd0f6
  by Tristan Van Berkom at 2018-04-26T05:47:17Z

  _loader.py: Removing special case error for junctions depending on elements
  
  Leave this error to be handled by preflight.
  
  Updated test case to expect the new ElementError instead of a LoadError
  

• 0b35585b
  by Tristan Van Berkom at 2018-04-26T05:55:17Z

  _versions.py: Bumping format version for changes project.refs behavior
  
  If we want to depend on being able to revision junction.refs and
  project.refs separately, then we are better off just considering the
  project.refs feature as available since the new version 8.
  
  This will not harm projects which depended on it since version 5 instead.
  

• 78ad3b77
  by Tristan Van Berkom at 2018-04-26T05:57:57Z

  doc/source/projectconf.rst: Documenting ref-storage as available since format version 8
  

• d3a991a5
  by Tristan Van Berkom at 2018-04-26T06:11:43Z

  setup.py: Stop installing test cases.
  
  As recommended by Valentin David in issue #332.
  

• 24028499
  by Tristan Van Berkom at 2018-04-26T06:40:13Z

  man: Updated man pages for 1.1.3 release
  

• 7bb59282
  by Tristan Van Berkom at 2018-04-26T06:42:27Z

  README.rst: Removed 'covers many programming languages' from the About BuildStream
  
  This is rather innacurate, you can build software in any programming languange
  given that you have provided the required resources for this in the runtime
  you are using.
  

• 06ae434a
  by Tristan Van Berkom at 2018-04-26T06:43:52Z

  README.rst: Fixed formatting
  
  No empty lines after a section title, 2 empty lines before a section title.
  

• d1c3d705
  by Jürg Billeter at 2018-04-30T09:03:43Z

  _artifactcache/pushreceive.py: Fix commit object check in send_info()
  
  Fixes #325.
  

• 6f1d1552
  by Jürg Billeter at 2018-04-30T09:03:43Z

  _artifactcache/pushreceive.py: Do not return True for skipped push
  

• cb9f318b
  by Jürg Billeter at 2018-04-30T09:06:10Z

  element.py: Print info message when artifact was actually pushed
  
  This matches the pull code path.
  

• 0916d81e
  by Jürg Billeter at 2018-04-30T09:06:10Z

  tests/testutils/runcli.py: Improve detection of pulled/pushed elements
  

• f00a435c
  by Phillip Smyth at 2018-05-01T11:34:19Z

  _artifactcache/pushreceive.py: Modified incorrect error message
  
  changed encode_message push exception from
  'Command must by GLib.Variant'
  to
  'Command must be PushCommand'
  

• eceb7c98
  by Jürg Billeter at 2018-05-03T14:06:55Z

  element.py: Fix buildable check in non-strict mode
  
  Ensure that the strong cache key of each build dependency is available
  before an element is built. Otherwise the strong cache key of the
  element cannot be calculated and caching the artifact produces an
  AssertionError.
  
  In non-strict mode an element's strong cache key may not be available
  yet even though an artifact is available in the local cache. This can
  happen if the pull job is still pending as the remote cache may have an
  artifact that matches the strict cache key, which is preferred over a
  locally cached artifact with a weak cache key match.
  
  Fixes #383.
  

• 494f18fe
  by James Ennis at 2018-05-04T09:24:41Z

  commands.rst: Split up commands to easily allow for further description
  
  An introduction for this section was also added
  

• 451b05fc
  by Valentin David at 2018-05-06T06:47:17Z

  Fix provenance in error message for undefined variables.
  

• 745d0647
  by Valentin David at 2018-05-06T06:47:17Z

  tests/format/variables.py: Add a test for undefined variable error message
  

• 53bff7d6
  by James Ennis at 2018-05-07T10:25:52Z

  app.py: Ensure an elements sub-dir is created upon project initialisation
  

• 2e252e9d
  by Chandan Singh at 2018-05-07T11:21:32Z

  bst-here: Fix typo in help text
  

• 6ffca067
  by Chandan Singh at 2018-05-07T11:21:32Z

  bst-here: Allow users to specify Docker image
  
  Currently, the name of the Docker image is hardcoded in bst-here script.
  This makes it harder for users to override it, which may be needed for
  various reasons (custom plugins, private Docker registries etc.)
  
  Add `-i` option to allow users to specify the base image. When `-i` is
  not provided, default to `BST_HERE_IMAGE` environment variable if it's
  set and otherwise to the current image -
  `buildstream/buildstream-fedora`.
  
  Also, re-order the command-line options in help text and source code in
  alphabetical order to maintain sanity as the number of options is slowly
  growing.
  
  Part of #378.
  

• 49176039
  by Chandan Singh at 2018-05-07T11:21:32Z

  bst-here: Allow users to specify additional volumes to mount
  
  One may want to mount additional volumes to preserve certain directories
  or to share some data between the host and the container. Allow users to
  do so by providing a `-v` option that passes its arguments to
  corresponding `-v`/`--volume` option for `docker run` command.
  
  Part of #378.
  

• e6049105
  by Javier Jardón at 2018-05-07T11:40:54Z

  .gitlab-ci.yml: Remove redundant declaration of dependencies:
  
  "artifacts from all previous stages are passed by default."
  https://docs.gitlab.com/ee/ci/yaml/#dependencies
  

• 03823d12
  by milloni at 2018-05-07T13:51:16Z

  _site.py: Add check_bwrap_version() function
  
  Lazily parse the version of bwrap the first time the function is called.
  On subsequent calls, used cached version info.
  
  See: #373
  

• f8952d6b
  by milloni at 2018-05-07T13:51:16Z

  Replace bwrap checks with calls to check_bwrap_version
  
  See: #373
  

• bbb894bd
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  tests/format/variables.py: Removed testing of stderr
  
  We have machine readable errors for this purpose, and the
  strings happen to change causing tests to break if we test the specific
  UI (reported error strings are UI).
  

• 2390c814
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _stream.py: New Stream object, main calling interface for BuildStream core
  
  This is the first part of the pipeline refactor, at this stage
  all calling interfaces remain the same, except that invocation
  of the scheduler has been moved from Pipline to Stream.
  

• 11dffaef
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _stream.py: Add StreamError exception
  
  Use Stream error for Stream errors.
  

• 76c260b0
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _stream.py: Absorb workspace functionality from App.
  

• c9f67f1e
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _frontend: Early logging initialization
  
    o This makes logging independent from the Pipeline()
  
    o Removed size_request Widget() method, add context to Widget() initializer
  
    o Make the Status() widget derive anything it needs through the Context()
  

• 541cd760
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _stream.py, _frontend: Call Stream() APIs with CLI arguments directly.
  
  This shifts the whole responsibility of interpreting command line
  targets etc to the Stream() object itself. With this commit, the
  Pipeline() truly becomes slaved to the Stream().
  

• c81fb2da
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  Moving 'fetch_subprojects' configuration to stream <--> loader
  
  This was previously decided in CLI, but knowledge of what to initialize
  has been moved to Stream().
  
  Now there is no more point to store this configuration in the Context,
  we just have the Stream() decide it when asking the Pipeline() to
  invoke the Loader().
  

• a542c818
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _frontend/app.py: Remove partially_initialized()
  
  There is no more need for this distinction.
  

• b8e15706
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _artifactcache: Added ArtifactCache.setup_remotes()
  
  This removes some additional initialization code from Pipeline().
  
  Some symbols have changed here, the initialization is now called
  from Stream(), and a test case was also adjusted for this.
  

• 5bdc0a79
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _stream.py, _pipeline.py: Refactoring of the pipeline itself
  
  Here the pipeline becomes essentially stateless, some dangling
  state remains to be factored out because of frontend accesses
  which will be changed in a later commit.
  
  Essentially, the Pipeline.load() method no longer has any knowledge
  of the specific purposes of the loaded targets, and now takes
  a list of target groups and returns a corresponding list of element
  groups.
  
  The Stream() business logic methods now use other pipeline helper
  methods to create and filter lists from the loaded target elements.
  
  The Stream() also finally absorbs the Scheduler frontend facing
  APIs. However Queues are still exposed on the Stream object for
  logging purposes and through callbacks such that the frontend can
  retry elements.
  

• 1ec80edc
  by Tristan Van Berkom at 2018-05-07T18:59:38Z

  _frontend/cli.py: Fetch all elements when tracking a build plan
  
  It makes not sense to type `bst fetch --track --deps all <targets>`,
  because tracking will inevitably modify the build plan.
  
  Stream initialization will not cope with this either, instead of
  silently doing something which does not make any sense, we add an
  assertion that this should not happen.
  
  Unfortunately since `plan` is the default deps type for `bst fetch`,
  this is likely to happen so it's important to correct.
  
  This patch adds a warning in the case tracking of the build plan
  elements is requested, and converts the request to track all elements
  instead.
  

• 32c30766
  by Tristan Van Berkom at 2018-05-10T06:59:47Z

  _frontend/app.py: Fixed ordering of initialization sequence.
  
  We need the logging handler setup before creating the Stream(),
  as the Stream() will expect logging to be already functional.
  

• 8985da07
  by Valentin David at 2018-05-10T11:33:12Z

  buildstream/plugins/sources/_downloadablefilesource.py: Store etag along with cache.
  
  Fixes #377.
  

• edf3e029
  by Tristan Van Berkom at 2018-05-10T12:20:06Z

  Add soft reset functionality for workspaces
  
  Add `--soft` option to `bst workspace reset` which would allow uses to
  reset workspace-related state without affecting its contents. This will
  be useful in case when an user wants to re-run configure-commands for a
  workspaced element.
  
  Patch originally by Chandan Singh, rebased against recent refactor.
  
  Fixes #375.
  

• b2b365fb
  by Chandan Singh at 2018-05-11T01:01:46Z

  _project.py: Remove unused internal function _extract_plugin_paths()
  
  For reference, this method was added in
  commit c41f1093d83eb32606f81005227542699f756b0a, but since
  commit 4912ed5f8922b47c1a16c3752eb92a3f1dea76b2 it has been unused.
  

• e1420437
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _artifactcache/ostreecache.py: Fix global remotes
  
  Counter-intuitively, Python list += mutates the list. Use explicit
  copy() and extend() instead of += to avoid adding project-specific
  remotes to the global remote list.
  

• 6d71826e
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _pipeline.py: Remove unused ignore_cache parameter from _Planner
  

• 37e5362e
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _pipeline.py: Allow pulling strict artifacts in non-strict mode
  
  Remotely cached artifacts matching the strict cache key take precedence
  over locally cached artifacts matching only the weak cache key. However,
  locally cached artifacts were excluded from the build plan, which means
  that BuildStream never even checked whether the strict artifact is
  available in the remote artifact cache.
  
  This changes planning to keep cached elements in the build plan in
  non-strict mode if a remote artifact cache is used.
  

• df17f602
  by Jürg Billeter at 2018-05-11T05:57:37Z

  tests: Do not rely on 'downloadable' status
  
  This is in preparation for moving away from summary files.
  

• da4af798
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _frontend/cli.py: Remove bst show --downloadable option
  
  This is in preparation for moving away from summary files.
  

• 4e512529
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _stream.py: Remove unused downloadable parameter from load_selection()
  

• db503a42
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _artifactcache: Add bool return value to pull()
  
  Calling pull() for a potentially unavailable artifact is no longer
  considered an error.
  

• e2c150f3
  by Jürg Billeter at 2018-05-11T05:57:37Z

  _artifactcache: Add element paramater to has_fetch_remotes()
  

• 17fd0b24
  by Jürg Billeter at 2018-05-11T06:01:12Z

  element.py: Do not query remote artifact caches ahead of pull/push
  

• 095ce5a5
  by Jürg Billeter at 2018-05-11T06:01:12Z

  _artifactcache/artifactcache.py: Remove unused methods
  

• 31ae2f55
  by Jürg Billeter at 2018-05-11T06:01:12Z

  _artifactcache/ostreecache.py: Do not use summary file
  

• 6ba62c57
  by Jürg Billeter at 2018-05-11T06:01:12Z

  doc/source/artifacts.rst: Drop section about summary file updates
  
  Summary files are no longer used.
  

• b78df1e2
  by Jürg Billeter at 2018-05-11T06:01:12Z

  _ostree.py: Remove unused function list_remote_refs()
  

• 270da060
  by Jürg Billeter at 2018-05-11T06:01:12Z

  Do not pull/fetch/build elements that are not required
  
  For `bst build --deps plan`, do not process elements in pull/fetch/build
  queues until they are requested by a reverse dependency.
  

• 3d9491dd
  by Jürg Billeter at 2018-05-11T08:19:54Z

  _stream.py: Do not force `PipelineSelection.ALL` for build with tracking
  
  With the dynamic plan using `_set_required()` we can avoid
  pulling/fetching/building unneeded elements even if elements are being
  tracked.
  

• 01a02a9a
  by Tristan Van Berkom at 2018-05-11T08:45:09Z

  _workspaces.py: Use file mtime for workspace cache keys instead of checksumming
  
  The rationale for this is basically:
  
    o A workspace is completely local
  
    o A cache key for a workspace is used to determine if the
      workspace has changed or not
  
    o Within reason, one should not expect buildstream to detect
      a change without a file's mtime having been modified
  
  On my machine, it takes about 1 minute to checksum 11GB of WebKit checkout,
  that is annoyingly a long time. With this change, it takes only 5 seconds.
  
  Not perfect, but it's a start.
  
  This is related to issues #294, #295 and #392
  

• b9dea666
  by Chandan Singh at 2018-05-11T11:29:45Z

  bst-init: Ensure --element-path is respected by the command
  
  Previously "elements" was hard-coded as the path for the elements
  directory whereas it was supposed to be configurable via the
  `element_path` option to `init_project()`. This led to incorrect
  behavior when `bst init` was run `--element-path` option.
  
  Also, extend tests to test the creation of elements directory.
  
  Fixes #398.
  

• cb24a1d8
  by Tristan Van Berkom at 2018-05-16T06:18:07Z

  doc/source/projectconf.rst: Added link anchors for project option types
  

• cb076de8
  by Tristan Van Berkom at 2018-05-16T06:18:07Z

  doc/source/projectconf.rst: Added link anchor for source aliases
  

• bb8bd1cf
  by albfan at 2018-05-16T07:51:31Z

  doc: Add first example, building on a flatpak SDK
  
  This adds:
    o A ToC area for adding examples
    o The instructive example page for the first example
    o The example project under doc/examples
    o The corresponding integration test in tests/examples
  

• b7e75da3
  by Tristan Van Berkom at 2018-05-16T07:51:31Z

  tests/examples/flatpak-autotools.py: Workaround setuptools bug
  
  Until the setuptools bug is fixed, symlinks are not included
  in source distributions - this works around the problem until which
  point we can use a setuptools version without the bug.
  

• b98791cb
  by Tristan Van Berkom at 2018-05-16T07:51:31Z

  doc/source/examples.rst: Rearranged ToC for examples
  
  We want one example per file, not a huge file full of different examples.
  

• e6f59b63
  by Tristan Van Berkom at 2018-05-16T07:51:31Z

  doc/examples/flatpak-autotools/README.md: Removing
  
  This is redundant with doc/source/examples_flatpak_base.rst, lets
  please only document things once.
  

• c205991e
  by Tristan Van Berkom at 2018-05-16T07:51:31Z

  doc/source/examples_flatpak_autotools.rst: Enhanced example
  
  This has now changed to:
  
    * Have explanations of the project.conf and each element
      in the example, explaining what they are for
  
    * Have links into the reference for the specific features
      this example uses, such that the reader can get familiar
      with the reference manual from example windows
  

• e023beb5
  by Tristan Van Berkom at 2018-05-17T05:46:46Z

  _frontend: Notify failures when on linux
  
  Opens the door to platform specific frontend features,
  and implements a platform specific "notify" method.
  
  This is based on Valentin David's patch on merge request 447,
  and this fixes issue #385.
  

• 84a57b2f
  by Tristan Van Berkom at 2018-05-17T10:56:04Z

  _frontend/linuxapp.py: Fixed crash in previous commit
  
  Commit e023beb56b180316a58f409d48f158e9dc1ece95 introduced
  a crash with an invalid access to os.environ, this fixes it.
  

• 314c25d8
  by Chandan Singh at 2018-05-17T13:21:09Z

  element.py: Fix consistency of workspaced elements when ref is missing
  
  Fixes #393.
  

• ec16561f
  by Jonathan Maw at 2018-05-17T15:45:50Z

  tests: Add workspace tests for filters
  

• 8316a1fe
  by Jonathan Maw at 2018-05-17T15:45:50Z

  tests: Add track tests for the filter element
  

• 38dc2dda
  by Jonathan Maw at 2018-05-17T15:45:50Z

  element.py: Add _get_source_element() method
  

• 6e2e976c
  by Jonathan Maw at 2018-05-17T15:45:50Z

  filter.py: Redirect _get_source_element() calls to its build depend
  
  We do this because trying to open/close/reset a workspace on a filter
  element doesn't make sense, but is probably meant for the element that
  the filter element filters.
  

• c09be483
  by Jonathan Maw at 2018-05-17T15:45:50Z

  Redirect elements when performing tracking and workspace commands
  
  i.e. when an element is specified, it may be replaced with its source element.
  

• 206a7ee8
  by Jonathan Maw at 2018-05-21T09:59:45Z

  NEWS: Add item for element redirection
  

• 77da5945
  by toscalix at 2018-05-22T14:19:19Z

  bst bug template
  

• 3749e44e
  by toscalix at 2018-05-22T14:21:31Z

  bst merge request template
  

• b69e6482
  by toscalix at 2018-05-22T14:25:22Z

  correction :
  

• 4d92c106
  by toscalix at 2018-05-22T14:27:29Z

  correction :
  

• a06778aa
  by James Ennis at 2018-05-23T14:45:48Z

  junction.py: Include note regarding explicit tracking
  

• 9b261f56
  by toscalix at 2018-05-24T19:07:24Z

  bst_task.md: default template also as option
  

• a4cd9d7c
  by toscalix at 2018-05-24T19:11:27Z

  Merge branch 'master' of https://gitlab.com/BuildStream/buildstream
  
  bst_task.md : default template also as option.
  
  This is the default template. It is neccessary to include it in this directory to have it as option which is required in special cases.
  
  This fix issue https://gitlab.com/BuildStream/nosoftware/alignment/issues/12
  

• c0de75e2
  by albfan at 2018-05-25T14:35:18Z

  .gitignore: Ignore autogenerated docs
  

• 4708ee6b
  by Antoine Wacheux at 2018-05-31T12:57:51Z

  _artifactcache/ostreecache.py: Fix artifact cache initialization result tuple
  
  In case of failure, the tuple contained 4 elements instead of 3, causing
  BuildStream to crash.
  

• 065f5ac7
  by Tristan Maat at 2018-06-05T08:47:33Z

  _context.py: Normalize user-defined paths
  

• 279dcb40
  by Tristan Maat at 2018-06-05T08:47:33Z

  tests/context/context.py: Test setting XDG_CACHE_HOME
  

• fe24fc3b
  by Gökçen Nurlu at 2018-06-05T16:23:22Z

  Add test to verify cleanup behaviour after bwrap
  
  This adds `sandbox-bwrap/test-cleanup.bst` to test the cleanup behaviour in
  `sandbox/_sandboxbwrap.py`: There already exists a non-empty `/tmp` folder and
  BuildStream should not try to remove it after a command is executed with bwrap.
  

• 75fe8037
  by Gökçen Nurlu at 2018-06-05T16:23:22Z

  _sandboxbwrap.py: Fix post-bwrap cleanup behaviour
  
  The cleanup was supposed not to remove folders (`/dev`, `/tmp`, `/proc`) if
  they already existed before bwrap but it did the opposite: it tried to remove
  them if they existed before, and didn't remove them if they were created during
  bwrap. This was caused by a `not` clause, and this removes it.
  
  Fixes #379
  

• 8d88b52a
  by Phillip Smyth at 2018-06-05T18:00:03Z

  utils.py: Correcting a typo in safe_remove's comment
  

• 1d694b28
  by Chandan Singh at 2018-06-05T19:34:21Z

  _artifactcache/pushreceive.py: Add Click type for CLI argument 'repo'
  
  The CLI for `bst-artifact-receive` expects a `repo` argument, which is
  supposed to be a directory, but Click currently expects it to be just
  any string. This results in stack traces like the one below when the
  argument provided is not a directory:
  
          $ ~/.local/bin/bst-artifact-receive --pull-url http://foo foobaz
          Traceback (most recent call last):
            File "/root/.local/bin/bst-artifact-receive", line 8, in <module>
              sys.exit(receive_main())
            ...
            File "/src/buildstream/buildstream/_artifactcache/pushreceive.py", line 581, in __init__
              self.repo.open(None)
          GLib.Error: g-io-error-quark: /src/buildstream/43fref: opendir(/src/buildstream/foobaz): No such file or directory (1)
  
  Add types for this argument such that it throws better error messages
  when it receives bad arguments. With the Click types added, it will
  instead fail with messages like these:
  
          $ ~/.local/bin/bst-artifact-receive --pull-url http://foo foobaz
          Usage: bst-artifact-receive [OPTIONS] REPO
  
          Error: Invalid value for "repo": Directory "foobaz" does not exist.
  
          $ ~/.local/bin/bst-artifact-receive --pull-url http://foo setup.py
          Usage: bst-artifact-receive [OPTIONS] REPO
  
          Error: Invalid value for "repo": Directory "setup.py" is a file.
  
  Fixes #409.
  

• 7f9216b3
  by Tristan Van Berkom at 2018-06-06T17:20:24Z

  _frontend/app.py: Remove unused variable at global scope
  

• 5dc8ab98
  by Ed Baunton at 2018-06-06T17:56:59Z

  Add a kind for Make
  

• b36db258
  by Tristan Van Berkom at 2018-06-06T18:50:54Z

  doc/source/authoring.rst: Adding `make` element to the docs
  

• 5ddec56f
  by Tristan Van Berkom at 2018-06-06T18:54:03Z

  _versions.py: Bump format version to 9
  
  For addition of new `make` plugin
  

• 17fa7ba4
  by Tristan Van Berkom at 2018-06-06T18:54:27Z

  make plugin: Ammended documentation to note it's since version.
  

• 0f2fd8c8
  by Tristan Van Berkom at 2018-06-06T18:55:56Z

  NEWS: Mentioning new `make` build element
  

• 55dc3b27
  by Tristan Van Berkom at 2018-06-06T18:59:00Z

  make build element: Add support for PREFIX
  

• 7df95654
  by Chandan Singh at 2018-06-06T21:27:30Z

  _loader/loader.py: Report element-path when failing to load elements
  
  It can be confusing, especially to new BuildStream users, that the CLI
  expects targets to be specified relative to element-path and not the
  current directory. Previously, the CLI would give a generic message
  stating that the file could not be found but it was not obvious that it
  was looking in the `element-path` directory.
  
  Explicitly print the element-path in the summary. Also, try to check if
  the specified element exists in the elements directory and print a hint
  to use the element-path relative paths if that's the case.
  
  Fixes #396.
  
  This is is also related to #341. This commit aims to tackle that issue
  by trying to educate users about element-path.
  

• a5ff465d
  by Chandan Singh at 2018-06-06T22:10:49Z

  _project.py: Allow running bst commands from subdirectories of project root
  
  When initializing the project, BuildStream will continue searching for
  project.conf in parent directories in case it is not found in the
  current directory.
  
  Fixes #368.
  

• 69e59443
  by Chandan Singh at 2018-06-06T22:10:49Z

  _frontend/cli.py: Try to autocomplete element paths when running from a subdirectory
  
  The previous commit added support for running bst commands form
  subdirectories of the project root. Make autocomplete also work in a
  similar way.
  

• e36001f9
  by Ed Baunton at 2018-06-07T13:43:27Z

  Make `bst help` work
  

• b25c31ea
  by Ed Baunton at 2018-06-07T17:34:27Z

  Implement bst help
  

• 8234e9c7
  by Tristan Van Berkom at 2018-06-07T20:14:11Z

  _frontend/cli.py: Allow specifying commands in `bst help`
  
    o This supports deeply nested commands as well as shallow commands
  
    o Automated support for bash completions included
  

• 9c65b908
  by Tristan Van Berkom at 2018-06-07T20:14:11Z

  tests/completions: Added test case for custom `bst help` completions
  

• 4632cb03
  by Tristan Maat at 2018-06-07T20:35:51Z

  _ostree.py: Reintroduce remove()
  

• 60a29f68
  by Tristan Maat at 2018-06-07T20:35:51Z

  _ostree.py: Introduce _list_all_refs() and list_artifacts()
  
  The unused list_remote_refs() function has also been removed as
  part of this commit.
  

• b7191fda
  by James Ennis at 2018-06-07T20:35:51Z

  pushreceive.py: Ensure there is a repopath
  

• 58adbe0b
  by James Ennis at 2018-06-07T20:41:03Z

  pushreceive.py: Remove LRP artifacts from cache, introduce clean_up_cache()
  
  This fixes #136
  

• 8a0c0a9a
  by James Ennis at 2018-06-07T20:41:33Z

  element_generators.py: Add a utility function to create arbitrary size elements
  
  New file consisting of functions that allow us to generate elements
  on the fly
  __init__.py now includes create_element_size()
  

• ed0df6b5
  by James Ennis at 2018-06-07T20:41:33Z

  push.py: Add tests to test remote cache expiry
  
  The tests include a new environment variable: BST_CACHE_QUOTA.
  This variable is used in receive_main() in pushreceive.py.
  Test names: test_artifact_expires, test_large_artifact_is_not_cached_remotely
  test_recently_pulled_artifact_does_not_expire: marked as xfail until
  we implement LRU expiry in remote share
  

• b8a52cbd
  by James Ennis at 2018-06-07T20:41:33Z

  pushreceive.py: Add comments to code and name variables more sensibly
  

• b823a616
  by James Ennis at 2018-06-07T20:41:33Z

  pushreceive.py: Ensure huge artifacts are not pushed
  

• a231d41b
  by James Ennis at 2018-06-07T20:41:33Z

  pushreceive.py: Abstract the buffer_
  

• de194dad
  by Tristan Van Berkom at 2018-06-07T20:41:33Z

  _artifactcache/pushreceive.py: Cleanup reported error when receiving oversized artifacts
  
  This user facing string was redundantly declared in two places, only
  the message when catching the error was ever printed.
  

• 8f2bf4e6
  by Tristan Van Berkom at 2018-06-08T19:48:11Z

  autotools plugin: Dont regenerate existing configure scripts
  
  This closes #256
  

• af10c1ba
  by Valentin David at 2018-06-08T21:07:22Z

  Interpret names as colon separated junction path in loader.
  
  'a.bst:b.bst' gets interpreted as 'b.bst' from junction 'a.bst'.
  
  Part of #359.
  

• ccec163b
  by Valentin David at 2018-06-08T21:07:22Z

  Reword uses of project in Workspaces.
  
  Make it clear we expect the top-level project here as we use it to
  resolve paths relative to project directory.
  

• 130bfbb8
  by Valentin David at 2018-06-08T21:07:22Z

  Handle cross junction elements in workspaces.
  
  Workspaces are now index by colon separated junction path. This
  now allows to create workspaces for elements in external projects.
  
  Workspaces are owned by context instead of root project. However
  it is initialized once top-level project is registered as we need
  to resolve paths relatively to this top-level project.
  
  Part of #359.
  

• acde3ba8
  by Valentin David at 2018-06-08T21:07:22Z

  Allow tracking dependencies within sub-projects.
  
  --track-cross-junctions now concerns crossing junctions rather than
  forbidding elements in sub-project to be tracked.
  
  Part of #359.
  

• 71b78794
  by Tristan Van Berkom at 2018-06-09T19:48:42Z

  doc/source/index.rst: Use only one level depth ToC on main page
  
  The main page has too much information on it otherwise, we want
  a friendly, not overwhelming first page to our docs.
  

• 055b77e8
  by Tristan Van Berkom at 2018-06-09T21:14:10Z

  doc: Added plugins as ToC elements instead of orphaned links
  
    o Now the page titles are declared in plugins, allowing for
      a more descriptive ToC
  
    o Makefile and plugin.rsttemplate updated to not produce the title,
      to no longer use `:orphan:` for plugin pages, and to ignore any
      private modules in the plugin directories.
  
    o Interestingly, now the docs will fail to build if you add
      a new plugin and forget to add it to the documentation.
  

• 5f794829
  by Tristan Van Berkom at 2018-06-09T21:23:23Z

  doc/source/main_core.rst: Use a hidden toctree to hide away some things
  
  Sphinx generates some library style module index, we just include
  it in a hidden toctree and avoid using it altogether.
  

• 96c219d0
  by Tristan Van Berkom at 2018-06-09T21:36:07Z

  doc/source/core_framework.rst: Use a toctree instead of links
  

• 1e98ecb2
  by Tristan Van Berkom at 2018-06-09T22:00:24Z

  doc/source/index.rst: Add resources directly at the bottom of the main page
  
  This is nice to have on the main page, and is only a few links, dont
  like having a whole toplevel ToC entry for this.
  

• 77dd1091
  by Tristan Van Berkom at 2018-06-09T22:08:49Z

  doc: Split out cache keys and sandboxing writings into a sub section
  

• 53653d29
  by Tristan Van Berkom at 2018-06-09T22:09:50Z

  doc/source/authoring.rst: Removed unneeded link anchor from here
  

• 52412db4
  by Tristan Van Berkom at 2018-06-09T22:18:28Z

  .gitlab-ci.yml: Stop depending on specific version of sphinx
  
  This branch fixes #312 by using a hidden toctree to include
  the buildstream package and reducing the amount of allowed :orphan:
  pages.
  

• 8e6f1a9b
  by Tristan Van Berkom at 2018-06-09T23:44:25Z

  doc: Stop generating the modules index
  
  We still have a few unused artifacts in the docs generation,
  this is just one less.
  

• 99fb2d53
  by Tristan Van Berkom at 2018-06-09T23:45:41Z

  doc/examples/flatpak-autotools: Use consistent project name
  
  This was named autotools-flatpak, changed to flatpak-autotools.
  

• 710e1fd4
  by Tristan Van Berkom at 2018-06-09T23:48:09Z

  doc/source/examples_flatpak_autotools.rst: Added some links to the example project on gitlab.
  

• 4af8112d
  by Tristan Van Berkom at 2018-06-10T00:11:07Z

  doc/source/examples.rst: Use a flat ToC here
  
  Here we're really listing examples, a ToC with depth is
  not great here.
  

• d0dc9e64
  by Tristan Van Berkom at 2018-06-10T00:21:47Z

  doc/source/main_using.rst: Reversed ToC here
  
  Seems that the "commands" is taking a lot of space such that
  we can't see the other sections here easily, that is alright
  if "commands" remains at the end.
  

• ea775fa3
  by Tristan Van Berkom at 2018-06-10T01:12:37Z

  doc: Removed some unused link targets
  

• 9f77cb66
  by Tristan Maat at 2018-06-11T18:16:33Z

  Handle missing tags in git repositories correctly
  
  Fixes issue #380
  

• e112a88f
  by Tristan Van Berkom at 2018-06-12T21:38:25Z

  doc/bst2html.py: Added bst2html.py
  
  This baby runs bst and captures the output with colors enabled
  and then generates some html we can include in documentation.
  
  These can be generated in CI continuously and used in the documentation.
  

• 52f70a0d
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  doc/Makefile: Run bst2html.py to generate .html files from .run files
  
  If you need an example output of bst to put in the documentation,
  just add a .run file to the doc/examples directory and it will result
  in a similarly named .html file in doc/source/examples.
  

• 050cf2eb
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  doc/sessions: Added run files describing scenarios to build
  
  The Makefile uses these to run some scenarios which are
  later included by the documentation directly.
  

• c900d42e
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  doc/source/examples_flatpak_autotools.rst: Using generated examples
  
  Show the commands at work in this example.
  

• 4dfc8607
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  doc/source/examples.rst: Restructuring individual examples into an examples subdirectory
  

• 19dddf82
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  doc/source/sessions: Committing generated session files.
  
  This is only to make it easier for people who just want to
  build docs locally and not regenerate the session files.
  
  The session snapshot html files are always generated in CI every time.
  

• f3b19f2c
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  HACKING.rst: Added guidelines for adding examples
  
  And some instructions about generating `bst` output for documentation
  

• d9da078d
  by Tristan Van Berkom at 2018-06-12T21:38:26Z

  .gitlab-ci.yml: Speed up docs build by setting up BST_SOURCE_CACHE
  
  Use the gitlab cache for caching sources needed for building projects
  in order to collect sample output for documentation pages.
  

• 53bc7272
  by Tristan Van Berkom at 2018-06-13T02:41:15Z

  doc: Don't allow setting a height on the generated session html
  
  This was a bad idea and doesnt play well with mobile UIs, better
  off to just include the whole thing even if it's long, and let
  the backing page handle vertical scrolling.
  

• 9d7296eb
  by Tristan Van Berkom at 2018-06-13T04:24:27Z

  doc: Overhaul of page names and titles
  
   o Giving main pages simple word titles
  
     This makes the main page:
  
     * About
     * Installing
     * Using
     * Reference
     * Contributing
  
   o Now named all rst files with their parent page name as a prefix.
  
   o Also changed some titles to make overall consistent titles.
  

• ba018efe
  by Chandan Singh at 2018-06-13T20:24:15Z

  contrib/bst-here: Fix command/arg in usage instructions
  
  As we can only accept arguments when a command is specified, make it
  clear in the usage instructions.
  

• fa73abe4
  by Chandan Singh at 2018-06-13T20:24:15Z

  contrib/bst-here: Remove redundant create_volume_if_not_exists() method
  
  This method doesn't really do anything as this is already the default
  behavior of `docker volume create` so remove it and always call
  `docker volume create` directly. This command will always print the name
  of the volume on STDOUT which is not very interesting so silence that.
  (If it errors out for some reason, that output will go to STDERR so the
  user should still be able to see that.)
  

• 6da05577
  by Tristan Van Berkom at 2018-06-14T03:12:53Z

  doc: Improve the new console output html generator
  
  Before we were creating one description file for each output,
  making it easier to declare a make rule for it - but the result
  was that we would have to build things more and it takes a
  long time.
  
  Instead, now we have session files which describe a series of commands
  to run in a session, and each command optionally produces an output file.
  

• 5ba5415e
  by Tristan Van Berkom at 2018-06-14T22:22:03Z

  doc/Makefile: Fixed loop over session files
  
  This list needs to not be quoted.
  

• a01a5cc8
  by Tristan Van Berkom at 2018-06-16T17:27:49Z

  doc/bst2html.py: Add support for `fake-output` when running commands.
  
  When specifying a fake-output string, we don't really run the command
  or assume it was a `bst` command, and we pretend that `fake-output`
  was the output of the command.
  
  Specifying an empty string explicitly enables the behavior too
  for faking a command that has no stdout/stderr.
  
  This also adds the "remove-files" hack allowing the session scripts
  to remove files before executing commands (kind of unsure if we're
  gonna keep this...)
  

• b592a80b
  by Tristan Van Berkom at 2018-06-16T19:44:56Z

  doc/bst2html.py, doc/Makefile: Added --force option
  
  If --force is not specified, then we'll skip session files in
  the case that all of the outputs exist.
  
  Now setting BST_FORCE_SESSION_REBUILD when building the docs
  will cause the session files to be rebuilt regardless of whether
  they exist or not.
  
  The .gitlab-ci.yml was also changed to use this and force rebuilds.
  

• 1b88e2c6
  by Tristan Van Berkom at 2018-06-16T20:14:04Z

  doc: Adding part 1 of the getting started tutorial
  
    o doc/Makefile: Added new directory to collect rst files from
  
    o doc/examples/first-project: Added the "first-project" example
      project.
  
    o doc/source/sessions/first-project-*.html: Added the generated
      snippets
  
    o doc/source/using_tutorial.rst: Added the new main tutorial page
  
    o doc/source/tutorial/first-project.rst: Added part 1 of the tutorial here
  
    o tests/examples/first-project.py: Added test for the example project
  
  This is largely based on an example by Javier Jardón, which was
  submitted at https://gitlab.com/BuildStream/buildstream/merge_requests/323
  
  Fixes #103
  

• 6fa3f922
  by Tristan Van Berkom at 2018-06-16T20:14:04Z

  doc/Makefile: Stage stored sessions from session-stored
  
  Separate the revisioned provisional session html files such
  that the git tree does not become dirty as a result of a
  documentation build process - which messes up the docs version
  number and the version number printed in some command line output.
  

• 3f560378
  by Tristan Van Berkom at 2018-06-18T00:58:20Z

  doc/source/tutorial/first-project.rst: Adding links
  
  Linking out to the relevant invoking pages for the command line
  reference, and adding a link anchor here for use by the next chapter.
  

• 73b93002
  by Tristan Van Berkom at 2018-06-18T01:37:22Z

  doc: Adding part 2 of the getting started tutorial
  
    o doc/examples/running-commands: New example project of a `manual` build element
  
    o doc/sessions/running-commands.run: New session file to capture bst output
  
    o doc/source/sessions-stored: Added new generated sessions
  
    o doc/source/tutorial/running-commands.rst: New tutorial entry describing how
                                                commands are run in the sandbox
  
    o tests/examples/running-commands.py: Test case validating the tutorial's assertions
  

• 72fbaa1c
  by Tristan Van Berkom at 2018-06-18T19:01:14Z

  doc/source/tutorial/running-commands.rst: Adding summary
  
  And adding some link anchors needed by the incomming chapter.
  

• b4105e8d
  by Tristan Van Berkom at 2018-06-18T19:17:52Z

  doc: Adding part 3 of the getting started tutorial: autotools element
  
  This part of the tutorial uses a lot of the work from Phil Dawson
  and James Ennis, and uses their example submitted on merge request
  499 as a basis to introduce the user to yaml composition and variable
  resolution.
  
  This is a part of issue #103
  

• 9b81f958
  by Tristan Van Berkom at 2018-06-18T19:18:19Z

  doc/source/using_tutorial.rst: Use a numbered ToC for the tutorial
  

• 48b4cb5c
  by Tristan Van Berkom at 2018-06-19T00:19:09Z

  doc/examples/autotools/project.conf: Correcting project name here
  

• 5b6441fa
  by Tristan Van Berkom at 2018-06-19T02:41:50Z

  doc: Use consistent titles when referring to files in an example project
  
  Use the following form across the board:
  
    ``elements/foo.bst``
    ~~~~~~~~~~~~~~~~~~~~
  
    .. literalinclude:: ../path/to/foo.bst
       :language: yaml
  
  Always use an example project relative path, too.
  

• bb223156
  by Tristan Van Berkom at 2018-06-19T02:47:46Z

  doc/source/examples/flatpak-autotools.rst: Reordering of titles.
  
  Place the titles of literally included `bst` files directly before
  the includes, and moved all related text to start below the included
  file for each section.
  

• 4016bec1
  by Gökçen Nurlu at 2018-06-19T04:41:24Z

  Remove shebangs from python files
  
  Fixes #424
  

• 4f168b9b
  by Daniel Playle at 2018-06-20T22:05:19Z

  Restrict version of pylint
  
  pylint >2 is not compatible with pytest_pylint in its current form. As
  such, allowing any version of pylint for testing results in a failure.
  
  This commit restricts down the allowable versions of pylint to those
  that are both compatible with pytest_pylint, and also offer the feature
  set that we require.
  
  See https://gitlab.com/BuildStream/buildstream/issues/427 for further
  details.
  

• 49fb3636
  by Phil Dawson at 2018-06-21T23:32:46Z

  MANIFEST.in: Fix typo when including BuildStream.doap
  
  Correct includsion of buildstream.doap which does not exist
  to 'include BuildStream.doap'.
  
  This commit resolves issue #430
  

• 411dd100
  by Patrick Martin at 2018-06-22T15:41:34Z

  doc/source/index.rst: fix tiny, but impactful typo
  
  Fixes #435
  

• d9a433a7
  by Javier Jardón at 2018-06-22T17:03:44Z

  BuildStream depends on host tools for Source plugins
  

• b5d91794
  by Javier Jardón at 2018-06-22T17:03:44Z

  Source plugin tar depends on host's lzip
  
  Issue #353
  

• 7d97c6d3
  by Javier Jardón at 2018-06-22T17:03:44Z

  Source plugin bzr depends on host's bzr
  
  Issue #353
  

• 3b1e869b
  by Javier Jardón at 2018-06-22T17:03:44Z

  Source plugin git depends on host's git
  
  Issue #353
  

• be92cc8d
  by Javier Jardón at 2018-06-22T17:03:44Z

  Source plugin patch depends on host's patch
  
  Issue #353
  

• e90098e0
  by Javier Jardón at 2018-06-22T17:03:44Z

  Source plugin deb depends on host's arpy python package
  
  Issue #353
  

• 4f1d5a1c
  by Tristan Van Berkom at 2018-06-22T17:05:39Z

  doc/source/core_plugins.rst, doc/source/install_main.rst: Fixing link anchor name
  
  In core_plugins.rst, we are already using _plugins, _plugins_build_elements,
  so lets call this one _plugins_sources to be consistent.
  

• 10d21ff0
  by Valentin David at 2018-06-23T01:17:40Z

  Fix element check for BST_FORBID_BDEPENDS
  

• 6641bcd6
  by Tristan Van Berkom at 2018-06-25T07:04:58Z

  doc/source/main_using.rst: Updating headline for this page.
  
  The purpose of this page has changed with time, better to clarify
  this in the headline of the page.
  

• 6d4cbe0c
  by Tristan Van Berkom at 2018-06-25T07:04:58Z

  doc/source/format_declaring.rst: Improving linkage to other parts of the manual
  
  Adding a couple of important links to relevant material.
  

• 4b499ba3
  by Tristan Van Berkom at 2018-06-25T07:04:58Z

  doc/source/tutorial/autotools.rst: Consistent wording of "chapter"
  
  We've been calling these tutorials "chapters", let's not start
  also calling them "sections".
  

• 1d84bb80
  by Tristan Van Berkom at 2018-06-25T07:04:58Z

  doc/source/tutorial/running-commands.rst: Added tip about using `--` in `bst shell`
  

• a25d683e
  by Tristan Van Berkom at 2018-06-25T07:04:58Z

  doc: Adding part 4 of the getting started tutorial: integration commands
  
  This new section talks about how integration commands work
  and shows them at work.
  

• 91d87e3c
  by Tristan Van Berkom at 2018-06-26T23:40:28Z

  _scheduler/job.py: Added long comment
  
  This explains the nature of a complicated asyncio callback,
  which I've looked up on multiple occasions and is not available
  in python online docs (only in the source can you follow it).
  

• 463698ec
  by Francisco Redondo Marchena at 2018-06-28T14:18:10Z

  Add error message when running commands on directories
  
  Issue #446
  

• 48848f4e
  by James Ennis at 2018-06-29T08:45:12Z

  loader.py: Check whether the dir specified is also a <dir>.bst file
  
    - Slight modification to the wording of existing error message.
    - Closes #446
  

• aaf52d03
  by Tristan Maat at 2018-06-29T15:56:20Z

  cli.py: Make buildstream check element paths instead of click
  
  This is to avoid inconsistencies when dealing with paths inside an
  elements directory
  

• a85aaee6
  by Tiago Gomes at 2018-07-02T14:45:04Z

  Fix documentation typo
  

• 9067e269
  by Tiago Gomes at 2018-07-02T14:45:04Z

  Provide better error message on missing commands
  
  Before running a command in the sandbox, check its existence and fail
  early if it does not.
  
  This fixes issue #289.
  

• c5fe9465
  by Jürg Billeter at 2018-07-08T12:33:46Z

  _signals.py: Guard against use in non-main thread
  

• 3d1e23d6
  by Jürg Billeter at 2018-07-09T13:44:49Z

  tests: Remove unneeded ArtifactShare.update_summary() method
  
  The OSTree summary file is no longer used.
  

• 5f4b9377
  by Jürg Billeter at 2018-07-09T13:44:49Z

  _artifactcache/artifactcache.py: Update remote init error message
  
  initialize_remotes() no longer fetches remote refs.
  

• 9defbb62
  by Jürg Billeter at 2018-07-09T13:44:49Z

  tests/frontend/push.py: Add missing skip annotation for non-Linux
  

• bc5a40e3
  by Phillip Smyth at 2018-07-11T11:06:40Z

  Move _list_dir_contents to __init__.py
  

• ddeac63a
  by Jonathan Maw at 2018-07-16T12:19:14Z

  tests: Fix filter tests not checking whether files should be missing.
  
  They weren't actually catching it if you checked-out the entire depended
  element, instead of just the specified split domains
  

• 4c6512d6
  by Jonathan Maw at 2018-07-16T16:12:08Z

  Only redirect elements for workspace and track operations
  
  It was redirecting for checkout operations
  

• fa2eaba2
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests: Use context manager for ArtifactShare
  

• ba4581f8
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/testutils/artifactshare.py: Add support for statvfs mocking
  

• 8cd81636
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/frontend/push.py: Use ArtifactShare statvfs mocking
  
  This makes it unnecessary to update the free space in the mock object in
  the middle of tests.
  

• 9b63f6f2
  by Jürg Billeter at 2018-07-17T05:56:40Z

  .gitlab-ci.yml: Update images to include protobuf and grpcio
  

• 1a2ee26e
  by Jürg Billeter at 2018-07-17T05:56:40Z

  setup.py: Add grpcio dependency and support for code generation
  
  This allows code generation with ./setup.py build_grpc
  

• b585388e
  by Jürg Billeter at 2018-07-17T05:56:40Z

  HACKING.rst: Add note about protocol buffers and code generation
  

• 383e0586
  by Jürg Billeter at 2018-07-17T05:56:40Z

  Import Google proto dependencies for the Remote Execution API
  
  This imports protos from https://github.com/googleapis/googleapis/ and
  the modules generated with ./setup.py build_grpc.
  

• 73fe9af1
  by Jürg Billeter at 2018-07-17T05:56:40Z

  Import protos for the Bazel Remote Execution API
  
  This imports protos from https://github.com/bazelbuild/remote-apis/ and
  the modules generated with ./setup.py build_grpc.
  

• df9a4949
  by Jürg Billeter at 2018-07-17T05:56:40Z

  remote_execution.proto: Add proposed symlink support
  
  https://github.com/bazelbuild/remote-apis/pull/18/
  

• f13bef56
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache: Add CAS artifact cache
  
  Content Addressable Storage specified in the Remote Execution API.
  
  Fixes #387.
  

• a3bdfc18
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/cascache.py: Add prune() method
  

• 687b9a8b
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/cascache.py: Add remove() method
  

• bed6c800
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/cascache.py: Add list_artifacts() method
  

• 8a0fe273
  by Jürg Billeter at 2018-07-17T05:56:40Z

  Add proto for BuildStream reference storage service
  

• 72284b59
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/cascache.py: Add remote cache support
  

• ff8703c9
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache: Add CAS artifact server
  

• 971606ae
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/casserver.py: Add cache cleanup based on pushreceive
  

• 4b4c5268
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/casserver.py: Add update_mtime parameter to resolve_ref()
  

• 9b879015
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache/casserver.py: Update artifact ref mtime on access
  
  This makes cache cleanup LRU.
  

• 8de8ef1f
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _platform: Use CAS artifact cache
  

• a161c746
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/testutils/artifactshare.py: Use CAS artifact server
  

• d0346e89
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/cachekey: Update expected cache keys after switch to CAS
  

• 05a1a25d
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests: Test push/pull on all platforms
  

• 6fd3f523
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/frontend/push.py: Remove xfail mark from LRU cache test
  
  The CAS server now supports LRU.
  

• b87c696f
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/frontend/pull.py: Add pull test for missing blobs on the server
  

• 99846137
  by Jürg Billeter at 2018-07-17T05:56:40Z

  tests/integration/workspace.py: Test incremental build on all platforms
  

• e35098d3
  by Jürg Billeter at 2018-07-17T05:56:40Z

  doc/source/artifacts.rst: Update documentation for CAS artifact server
  

• 5a7d49dc
  by Jürg Billeter at 2018-07-17T05:56:40Z

  Remove tar artifact cache
  
  No longer used.
  

• 1f8b4aa2
  by Jürg Billeter at 2018-07-17T05:56:40Z

  Remove OSTree artifact cache
  
  No longer used.
  
  Fixes #134, #138, #148, #217, #268, #276, #443, #460.
  

• d87df668
  by Jürg Billeter at 2018-07-17T05:56:40Z

  NEWS: Add news entry for CAS-based artifact cache
  

• 7acc64e4
  by Jürg Billeter at 2018-07-17T05:56:40Z

  element.py: Do not use ArtifactCache.can_diff()
  
  The only existing artifact cache implementation always returns True.
  

• f2484fe4
  by Jürg Billeter at 2018-07-17T05:56:40Z

  _artifactcache: Remove unused method can_diff()
  

• f770ab1c
  by Jürg Billeter at 2018-07-17T08:40:16Z

  element.py: Include fail-on-overlap setting in cache key
  
  Fixes #473.
  

• 7e9282c9
  by Phillip Smyth at 2018-07-17T12:53:55Z

  buildstream/_frontend/cli.py: Added a `--deps` flag to `bst checkout`
  
  buildstream/_stream.py: Added deps param to _prepare_sandbox function call
  buildstream/element.py: Added deps param and logic to _prepare_sandbox function
  

• 2797d1fd
  by Phillip Smyth at 2018-07-17T12:53:55Z

  tests/frontend/buildcheckout.py: Added tests for "Add a `--deps` flag to `bst checkout`"
  

• 70a83ef8
  by William Salmon at 2018-07-17T14:04:16Z

  Adding a helpful link to the example
  

• e4f8fb07
  by devcurmudgeon at 2018-07-17T15:07:05Z

  Fix 'main install' to be explicit that it is for Linux distros only
  

• 87c103d8
  by Jürg Billeter at 2018-07-17T15:07:56Z

  _frontend/cli.py: Fix help text for `bst checkout --deps`
  

• 4c09b686
  by Jürg Billeter at 2018-07-17T17:31:32Z

  _artifactcache/cascache.py: Fix prune()
  
  _reachable_refs_dir() failed to add directory digests to the set of
  reachable objects.
  

• adedd867
  by Tristan Maat at 2018-07-18T05:11:03Z

  Bump required python version to 3.5
  

• b8543f8a
  by Tristan Maat at 2018-07-18T05:11:03Z

  _exceptions.py: Add `detail` to ArtifactErrors
  

• fc79ad2c
  by Tristan Maat at 2018-07-18T05:11:03Z

  utils.py: Allow `list_relative_paths` to list directories
  

• 4f9da15d
  by Tristan Maat at 2018-07-18T05:11:03Z

  _scheduler/*queue.py: Move queues to a subdirectory
  

• 24925634
  by Tristan Maat at 2018-07-18T05:45:59Z

  Make Jobs abstract and element-independent
  

• 33984448
  by Tristan Maat at 2018-07-18T05:45:59Z

  Add cache_quota to user config
  

• 1ec5c7b1
  by Tristan Maat at 2018-07-18T05:45:59Z

  Make elements keep track of their built artifact size
  

• 7229d2e5
  by Tristan Maat at 2018-07-18T05:45:59Z

  Compute the artifact cache size after each build/pull
  

• dc17de38
  by Tristan Maat at 2018-07-18T06:07:24Z

  Clean the artifact cache when we hit the cache quota
  
  When the cache quota is hit, we will remove any elements not required
  for the current build, until our cache is only filled halfway.
  

• 6ff12e5e
  by Tristan Maat at 2018-07-18T06:07:24Z

  tests/artifactcache/expiry.py: Add expiry tests
  

• 308d59be
  by Tristan Maat at 2018-07-18T06:07:24Z

  NEWS: Add news about cache expiry
  

• a950a985
  by Tristan Van Berkom at 2018-07-18T09:59:03Z

  _context.py: Adding apis for message recording and task logging.
  
  Added the new Context.recorded_messages() context manager, this
  causes messages to be logged in a dedicated log file before being
  propagated to the log handler, and also provides an open log handle
  for the capturing and logging of stdout/stderr from child processes.
  
  This comes with the additional accessors:
  
     Context.get_log_handle()
     Context.get_log_filename()
  

• d835c37f
  by Tristan Van Berkom at 2018-07-18T09:59:03Z

  Refactor: Use new logging mechanism from Context
  
    o element.py / plugin.py: Removed supporting logging code, and derive
                              the log handle from Context.
  
    o _scheduler/scheduler.py, _scheduler/queues/queue.py: Adapt to new Job initializer API for the logfile
  
    o _scheduler/jobs/job.py: Run job activities within the new context manager
                              which turns on logging for a job. Also removed a lot
  			    of custom logging abstract methods which are unneeded.
  
    o _scheduler/jobs: Job implementations need not implement so much custom logging.
  

• 26b9f6fa
  by Tristan Van Berkom at 2018-07-18T11:52:43Z

  _stream.py: Clarifying ambiguous FIXME comment about required artifacts
  

• 3814905a
  by Daniel Playle at 2018-07-18T12:08:24Z

  Reorder app.py imports
  
  When testing locally on some systems, it appears that pylint fails on
  the order of imports for `buildstream/_frontend/app.py`. This commit
  changes the order of these imports.
  
  This commit addresses issue !453.
  

• 48f66a3c
  by Chandan Singh at 2018-07-18T18:03:31Z

  .gitlab-ci.yml: Use testsuite images for running tests
  
  Instead of re-using the user-facing `buildstream/buildstream-fedora`
  image for tests (that comes with BuildStream pre-installed), use
  `buildstream/testsuite-fedora` image that is supposed to be used
  exclusively for these tests.
  
  Similarly, use the dedicated `buildstream/testsuite-debian` image
  instead of `buildstream/buildstream-debian`.
  
  Fixes #458.
  
  For related discussion, see
  https://gitlab.com/BuildStream/buildstream-docker-images/issues/8.
  

• 1016dcb4
  by Josh Smith at 2018-07-19T12:52:21Z

  scheduler.py: Correct some anomalies within the docstrings
  

• 709b9bca
  by Francisco Redondo Marchena at 2018-07-19T14:03:12Z

  source-bundle: Enable --except option
  
  Before this option was listed in help but was ignored when
  creating the source-bundle.
  
  Issue #468
  

• 96d07153
  by Jordan Petridis at 2018-07-19T23:54:39Z

  Docs: Update the required build packages for fedora based systems.
  
  `ruamel.yaml` seems to require `Python.h` header file to build.
  
  `python3-devel` is what provides it for Fedora.
  

• 93a77c1d
  by Javier Jardón at 2018-07-19T23:58:28Z

  Docs: Update the required build packages for the rest of the systems
  
  Also remove the recomendation to install psutil as we need to build
  other python modules anyway (like ruamel)
  
  Completes 96d07153b7817cdaeda57dd163eed52b2b1b31e8
  

• 57e0d1c5
  by William Salmon at 2018-07-20T00:03:06Z

  Added dependency to the Docs
  
  These dependencies are needed to use the plugins that are installed
  as part of this guild.
  
  As the dependencies added are not dependencies of the core package
  they have been added to a separate section.
  
  The lzip package is needed to build gnome.
  https://wiki.gnome.org/Newcomers/BuildSystemComponent
  

• 1b6688eb
  by Javier Jardón at 2018-07-20T00:07:23Z

  install_linux_distro.rst: document plugins installed by default
  

• fc63b972
  by Javier Jardón at 2018-07-20T00:12:50Z

  install_linux_distro.rst: Document plugins deps for the rest of the systems
  

• 790fb40b
  by Tiago Gomes at 2018-07-20T09:07:01Z

  Update checkout man page for new --deps option
  

• 9c1f024b
  by Tiago Gomes at 2018-07-20T09:07:01Z

  Add support for creating a tarball on bst checkout
  
  One of the tests added is configured to be skipped for now, as dumping
  binary data is causing a bad descriptor exception when using the pytest
  capture module.
  
  Closes #263.
  

• 0ddbf0e0
  by Tristan Van Berkom at 2018-07-20T11:52:00Z

  setup.py: Specify minimum required version of pytest-cov plugin
  
  This causes the new artifact tests to pass (unless you happened
  to already have a recent enough version of pytest-cov, in which case
  you didn't notice the breakage).
  

• 6158ee5a
  by Sam Thursfield at 2018-07-20T13:17:14Z

  Fix crash when --debug is passed
  
  I hit the following backtrace running `bst --debug push` and `bst
  --debug build`:
  
      pid:16736 id:000[--:--:--][][] START   Push
      pid:16736 id:000[--:--:--][][] START   Loading pipeline
      pid:16736 id:000[00:00:00][][] SUCCESS Loading pipeline
      pid:16736 id:000[--:--:--][][] START   Resolving pipeline
      pid:16736 id:000[--:--:--][][] BUG     'MesonElement' object has no attribute '_Element__cache_key'
  
          Traceback (most recent call last):
            File "/home/sam/.local/bin/bst", line 8, in <module>
              sys.exit(cli())
            File "/home/sam/.local/lib/python3.6/site-packages/click/core.py", line 722, in __call__
              return self.main(*args, **kwargs)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/cli.py", line 162, in override_main
              standalone_mode=standalone_mode, **extra)
            File "/home/sam/.local/lib/python3.6/site-packages/click/core.py", line 697, in main
              rv = self.invoke(ctx)
            File "/home/sam/.local/lib/python3.6/site-packages/click/core.py", line 1066, in invoke
              return _process_result(sub_ctx.command.invoke(sub_ctx))
            File "/home/sam/.local/lib/python3.6/site-packages/click/core.py", line 895, in invoke
              return ctx.invoke(self.callback, **ctx.params)
            File "/home/sam/.local/lib/python3.6/site-packages/click/core.py", line 535, in invoke
              return callback(*args, **kwargs)
            File "/home/sam/.local/lib/python3.6/site-packages/click/decorators.py", line 27, in new_func
              return f(get_current_context().obj, *args, **kwargs)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/cli.py", line 471, in push
              app.stream.push(elements, selection=deps, remote=remote)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_stream.py", line 341, in push
              fetch_subprojects=True)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_stream.py", line 814, in _load
              fetch_subprojects=fetch_subprojects)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_pipeline.py", line 119, in load
              for meta in meta_elements
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_pipeline.py", line 119, in <listcomp>
              for meta in meta_elements
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/element.py", line 894, in _new_from_meta
              element = project.create_element(artifacts, meta)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_project.py", line 184, in create_element
              element = self._element_factory.create(self._context, self, artifacts, meta)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_elementfactory.py", line 57, in create
              return element_type(context, project, artifacts, meta, default_config)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/element.py", line 196, in __init__
              super().__init__(meta.name, context, project, meta.provenance, "element")
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/plugin.py", line 171, in __init__
              self.debug("Created: {}".format(self))
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/plugin.py", line 407, in debug
              self.__message(MessageType.DEBUG, brief, detail=detail)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/plugin.py", line 652, in __message
              self.__context.message(message)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_context.py", line 338, in message
              self._message_handler(message, context=self)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/app.py", line 665, in _message_handler
              text = self.logger.render(message)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/widget.py", line 585, in render
              return self._render(message)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/widget.py", line 620, in _render
              text += widget.render(message)
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/_frontend/widget.py", line 228, in render
              _, key, missing = plugin._get_display_key()
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/element.py", line 1138, in _get_display_key
              cache_key = self._get_cache_key()
            File "/home/sam/.local/lib/python3.6/site-packages/buildstream/element.py", line 991, in _get_cache_key
              return self.__cache_key
          AttributeError: 'MesonElement' object has no attribute '_Element__cache_key'
  
  The issue here is that when `--debug` is enabled, Plugin.__init__()
  tries to log a message like this:
  
          pid:16929 id:001[--:--:--][????????][ main:tracker.bst                   ] DEBUG   Created: meson element at tracker.bst [line 1 column 0]
  
  The log formatter is trying to get the cache key of the element in
  order to show it in the log, but the attribute hasn't yet been set
  because we've not got to the Element() constructor.
  

• f484ab4f
  by Phillip Smyth at 2018-07-20T14:29:44Z

  Updating .bzr plugin to omit the .bzr dir
  
  Fixes issue #455
  

• b662f175
  by Tristan Van Berkom at 2018-07-20T14:29:44Z

  git source plugin: Omit the .git directory completely.
  
  This will break builds which use `git describe` to determine
  their version number until a more elegant solution is implemented,
  and will unblock work on caching of build trees in the artifacts.
  
  Fixes issue #455
  

• 80effb5c
  by Tristan Van Berkom at 2018-07-22T08:30:40Z

  HACKING.rst: No longer requiring issue numbers in commit messages.
  
  Instead, we are enabling the merge commit feature in gitlab and
  hopefully reducing some of the friction in review this way.
  

• c7124c3d
  by Tristan Van Berkom at 2018-07-22T09:46:05Z

  HACKING.rst: Added convention for naming of branches in merge requests.
  

• 282b1092
  by Tristan Van Berkom at 2018-07-23T08:47:36Z

  plugins/sources/git.py: Use --shared instead of --hardlinks
  
  This improves staging performance by avoiding making any
  copies of the objects in the cloned repository in the staging
  area before removing the .git directory.
  
  Fixes #488
  

• fe0df616
  by Tristan Van Berkom at 2018-07-23T09:20:44Z

  Merge branch 'tristan/git-stage-with-shared-clone' into 'master'
  
  plugins/sources/git.py: Use --shared instead of --hardlinks
  
  Closes #488
  
  See merge request BuildStream/buildstream!557

• 6559db4c
  by Josh Smith at 2018-07-23T09:22:39Z

  element.py: Make Element._get_brief_display_key() available to Internal API
  
  This change will allow for more detailed info messages throughout
  the CASCache.
  

• 6b82e476
  by Josh Smith at 2018-07-23T09:22:39Z

  cascache.py: Add push info detailing which remote is being pushed to
  
  Work towards fixing issue #275
  

• 1d36df25
  by Josh Smith at 2018-07-23T09:22:39Z

  cascache.py: Add pull info detailing which remote is being pulled from
  
  Work towards fixing issue #275
  

• 364706f3
  by Tristan Van Berkom at 2018-07-23T09:57:02Z

  Merge branch 'Qinusty/275' into 'master'
  
  Indicate where artifacts are going to and coming from in the log
  
  Closes #275
  
  See merge request BuildStream/buildstream!553

• e70237a5
  by Phillip Smyth at 2018-07-23T13:26:47Z

  Adding caching build trees
  
  buildstream/element.py: Adding build tree to cache
  buildstream/_versions.py: Bumping BST_CORE_ARTIFACT_VERSION
  

• 335afb5b
  by Tristan Van Berkom at 2018-07-23T14:07:07Z

  Merge branch 'caching_build_trees_limited' into 'master'
  
  Adding caching build trees (limited)
  
  See merge request BuildStream/buildstream!560

• 60290f31
  by Javier Jardón at 2018-07-23T15:29:37Z

  doc/source/install_linux_distro.rst: buildstream doesn't depend on ostree or pygobject anymore
  

• de162e43
  by Javier Jardón at 2018-07-24T11:58:16Z

  Merge branch 'jjardon/ostree_repo' into 'master'
  
  doc/source/install_linux_distro.rst: buildstream doesn't hard depend on ostree or pygobject anymore
  
  Closes #492
  
  See merge request BuildStream/buildstream!558

• 6106d657
  by Josh Smith at 2018-07-25T10:50:14Z

  cascache.py: Remove misleading info message
  
  This addresses issue #501.
  

• 4a637d63
  by Jonathan Maw at 2018-07-25T12:05:19Z

  Merge branch 'Qinusty/501' into 'master'
  
  Remove misleading info message
  
  See merge request BuildStream/buildstream!567

• a0ad985f
  by Phil Dawson at 2018-07-25T13:03:23Z

  tests/frontend/show.py: Add test case for maximum recursion depth being exceeded
  
  Add test to ensure gracefull handling of exception thrown while loading
  a pipeline due the python's max recursion depth being exceeded.
  
  This is part of the work for issue #203
  

• 95b4eae4
  by Phil Dawson at 2018-07-25T13:03:23Z

  app.py: Handle exception thrown when recursion limit is exceeded
  
  Because the RecursionError exception was introduced in Python 3.5, until we
  drop support for for Python 3.4, we must use RuntimeError.
  

• 889bf238
  by Jim MacArthur at 2018-07-25T13:43:21Z

  Merge branch 'phil/203-BuildStream-crashes-when-dependency-tree-too-deep' into 'master'
  
  Phil/203 BuildStream crashes when dependency tree too deep
  
  See merge request BuildStream/buildstream!512

• 54fee6c8
  by Phillip Smyth at 2018-07-25T14:15:36Z

  _stream.py: Added functionality for workspace open -f
  
  tests/frontend/workspace.py: Added tests
  

• f62b6cb7
  by Jonathan Maw at 2018-07-25T15:01:33Z

  Merge branch 'bst_workspace_open_force_does_nothing' into 'master'
  
  _stream.py: Added functionality for workspace open -f
  
  See merge request BuildStream/buildstream!549

• c049d021
  by Ed Baunton at 2018-07-25T19:43:51Z

  Add remote source plugin
  
  Add a plugin that supports downloading files verbatim from a source with
  an optional overridable filename and destination directory. Bumps bst
  format version to 10.
  
  Fixes #163
  

• bd1196ef
  by Ed Baunton at 2018-07-26T13:22:52Z

  Provide explicit reason for SourceError exception on path
  
  When the user provides a path for the filename parameter, provide a reason
  

• 9a46c16f
  by Ed Baunton at 2018-07-26T14:06:52Z

  Merge branch 'edbaunton/remote-source' into 'master'
  
  Add remote source plugin
  
  Closes #163
  
  See merge request BuildStream/buildstream!541

• 7ceacfea
  by William Salmon at 2018-07-26T14:32:49Z

  Documentation typos and fixes
  
  Daniel went through the Doc's etc and made a list of suggestions.
  These are the result of the most easily patched suggestions and
  one or two thing that I noticed as well.
  

• 8c9fd9e4
  by Javier Jardón at 2018-07-26T15:32:16Z

  Merge branch 'willsalmon/documentation_form_notes' into 'master'
  
  Documentation typos and fixes
  
  See merge request BuildStream/buildstream!569

• 9ebd7fe1
  by Josh Smith at 2018-07-27T04:41:26Z

  _exceptions.py: Modify BstError API to allow optional retry
  
  job.py: Changes to the logic surrounding retry attempts and child process return codes
  element.py, source.py: ElementError and SourceError also implement this change.
  
  These exceptions now have an optional parameter of temporary which defaults to false. This will potentially break
  backwards compatibility where exceptions were previously raised and a retry was intended.
  
  To trigger a retry, one must now raise their SourceError or ElementError with temporary=True.
  
  This aims to fix #397.
  

• ca224425
  by Josh Smith at 2018-07-27T04:41:26Z

  _downloadablefilesource.py: Update network errors to be temporary
  
  This follows the change in 67ecd97a05279a3b7570ad59f05bf0a5973ef04c.
  

• 0a7cfcdf
  by Josh Smith at 2018-07-27T04:41:26Z

  tests/sources: deb, tar and zip all test for retry behaviour
  
  This is following 5d6418a06a3a0262df5f6b9e6e1a36578515fd4a where
  DownloadableFileSource was modified to raise 'temporary'
  errors when relating to file downloading.
  

• 9b2bc78f
  by Josh Smith at 2018-07-27T04:41:26Z

  cascache.py: Update cache push/pull errors to be temporary
  
  Further work needs to be done for the current grpc exceptions which are reraised.
  

• 4986527f
  by Josh Smith at 2018-07-27T04:41:26Z

  plugin.py: Extend Plugin.call() API
  
  Plugin.call() now takes fail_temporarily as an optional parameter,
  when supplied it will cause subsequent failures to trigger temporary
  errors as opposed to permanent errors.
  
  This also extends Plugin.check_output() which makes use of Plugin.call()
  

• a9a1048e
  by Josh Smith at 2018-07-27T04:41:26Z

  git.py: Update git plugin to fail temporarily on clone failures
  
  The git plugin will now make use of the fail_temporarily parameter
  to Plugin.call(), allowing failures to trigger a retry.
  

• aa071cb6
  by Tristan Van Berkom at 2018-07-27T06:13:53Z

  Merge branch 'tristan/temporary-failures' into 'master'
  
  Temorary failures
  
  Closes #397
  
  See merge request BuildStream/buildstream!574

• 4d86e7c5
  by Tristan Van Berkom at 2018-07-27T06:43:48Z

  element.py, source.py, plugin.py: Adjusting API version since annotations.
  
  Since we backported the temporary/permanent failures patch for #397
  into the `bst-1.2` branch, we need to adjust the since versions in master
  down to 1.2.
  

• fb222ba1
  by Josh Smith at 2018-07-27T08:50:45Z

  cascache.py: Raise ArtifactError on grpc error
  
  This aims to fix issue #502.
  

• 48916b8a
  by Tristan Maat at 2018-07-27T10:22:41Z

  Merge branch 'Qinusty/502' into 'master'
  
  cascache.py: Raise ArtifactError on grpc error
  
  Closes #502
  
  See merge request BuildStream/buildstream!576

• 23e080b9
  by James Ennis at 2018-07-27T11:13:50Z

  README.md: Minor corrections
  

• 196cfffc
  by James Ennis at 2018-07-27T11:13:50Z

  README.md: Direct users towards the tutorial before existing projects:
  
    This patch also removes displaying the projects that live within
    the buildstream-examples repository, as we wish to deprecate this
    repo and migrate the two examples.
    - #438 and #439
  
    In addition to this, both examples do not 'build' in their current
    state.
  

• 2b93574e
  by James Ennis at 2018-07-27T11:13:50Z

  main_install.rst: Add note to refer Windows and macOS users to docker
  

• 63e2320e
  by Josh Smith at 2018-07-27T11:16:55Z

  .gitignore: Ignore .pytest_cache/
  

• 6ea97b17
  by Javier Jardón at 2018-07-27T11:54:20Z

  Merge branch 'jennis/doc_improvements' into 'master'
  
  Refer readers to our tutorial before referring them to existing bst projects
  
  See merge request BuildStream/buildstream!578

• 8b46e874
  by Jonathan Maw at 2018-07-27T12:24:56Z

  project: Parse and store mirrors
  

• a2e9c62a
  by Jonathan Maw at 2018-07-27T12:24:56Z

  Move _ALIAS_SEPARATOR into utils
  
  The separator is useful in source files other than _project.py
  

• 7c993ac0
  by Jonathan Maw at 2018-07-27T12:24:56Z

  source: Store the url aliases or use an override
  
  This is part of a later plan to implement mirroring without forcing
  everyone to update their sources. We use the expected calls to
  Source.translate_url() when running Source.configure() to extract the
  aliases from the URL. Multiple aliases must be extracted because
  sources exist that may fetch from multiple aliases (for example, git
  submodules)
  
  Later, we want to substitute another URI where the alias normally reads
  from the project - We accomplish this by re-instantiating the Source
  with the alias overrides passed as an argument to the constructor.
  

• 2889003c
  by Jonathan Maw at 2018-07-27T12:24:56Z

  source: When fetching, try to fetch from mirrors first
  
  **KLUDGE WARNING**: This involves making the source store its "meta"
  object so that it's possible to create a copy of the source inside the
  fetch queue, instead of back when the pipeline was being loaded.
  
  This adds the SourceFetcher class, which is intended for sources that
  fetch from multiple URLs (e.g. the git source and its submodules)
  
  Fix when fetching
  

• f81e8e7b
  by Jonathan Maw at 2018-07-27T12:24:56Z

  Set default mirror via command-line or user config
  
  In user config (buildstream.conf), it is set with the "default-mirror"
  field.
  
  On the command-line, it is set with "--default-mirror"
  

• 755ed898
  by Jonathan Maw at 2018-07-27T12:24:56Z

  Add mirrored source tracking
  

• 19c01a56
  by Jonathan Maw at 2018-07-27T12:24:56Z

  bzr.py: Improve mirror support
  
  This fixes:
  * Bzr repositories pulling from the branch they were created with.
  * Bzr's _ensure_mirror() not actually checking that it successfully
    mirrored the ref.
  

• 84872141
  by Jonathan Maw at 2018-07-27T12:24:56Z

  git.py: Improve mirror support
  
  When implementing fetching from mirrors, I encountered some problems
  with the git source:
  
  1. The mirror URL was using translate_url()'s output, so if a different
     alias was used, then fetching from the mirror would go to a different
     directory, and be inaccessible.
  2. After fixing that, fetching was unable to pull from a URL other than
     the one used at repository creation, meaning it wouldn't actually
     pull from the mirror.
  

• bd51a0b2
  by Jonathan Maw at 2018-07-27T12:24:56Z

  testutils: Add a helper to copy a testutils repo
  
  This is helpful if you want to test what happens when you have one repo
  that has diverged from another. By copying the repo you're sure they
  start with shared history.
  
  This is especially useful when mirroring.
  

• 909120ab
  by Jonathan Maw at 2018-07-27T12:24:56Z

  tests: Add mirrored fetching and tracking tests
  

• 1cbc2e17
  by Jonathan Maw at 2018-07-27T12:24:56Z

  doc: Add documentation for source mirroring
  

• 202d9d26
  by Jonathan Maw at 2018-07-27T12:24:56Z

  NEWS: Add item for source mirroring
  

• 2b23898d
  by Jonathan Maw at 2018-07-27T12:24:56Z

  doc: Add tutorials for setting up git and tar mirrors
  

• bd4d0355
  by Jonathan Maw at 2018-07-27T13:08:00Z

  Merge branch '328-support-for-downloading-sources-from-mirrors' into 'master'
  
  Resolve "Support for downloading sources from mirrors"
  
  Closes #328
  
  See merge request BuildStream/buildstream!404

• cb4693b2
  by Josh Smith at 2018-07-27T14:00:26Z

  Merge branch 'Qinusty/pytest_cache_gitignore'
  

• 8a96679a
  by Josh Smith at 2018-07-27T14:01:07Z

  Merge branch 'master' of gitlab.com:BuildStream/buildstream
  

• f5c8ff61
  by Josh Smith at 2018-07-27T14:10:45Z

  _context.py: Cache size is now restricted to available disk space
  
  This address issue #491.
  
  When attempting to run buildstream with a configuration specifying
  a cache quota larger than your available disk space, buildstream
  will alert the user and exit.
  
  Note:
  This takes into consideration your current cache usage and
  therefore restricts the overall size of your artifact cache folder.
  

• 32ddb544
  by Qinusty at 2018-07-27T14:57:30Z

  Merge branch 'Qinusty/491' into 'master'
  
  Cache quota is now restricted to available disk space
  
  See merge request BuildStream/buildstream!563

• cb8c99e5
  by Tristan Maat at 2018-07-29T07:52:06Z

  buildelement.py: Only display one activity message (#507)
  

• 08b7562b
  by Tristan Van Berkom at 2018-07-29T10:23:43Z

  Merge branch '507-some-log-lines-appear-to-be-duplicates' into 'master'
  
  Resolve "Some log lines appear to be duplicates"
  
  Closes #507
  
  See merge request BuildStream/buildstream!573

• 3b7158b3
  by Tristan Van Berkom at 2018-07-29T10:25:02Z

  source.py: Bring down since version for source mirroring APIs
  
  Since we have now backported this to `bst-1.2`, the APIs have
  been introduced in 1.2 and not 1.4
  

• e84f8b24
  by Tristan Van Berkom at 2018-07-29T10:25:02Z

  NEWS: Bring down since version of backported features to 1.1.5
  
  These features will be released in 1.1.5 since they have been
  backported, lets reflect this in the NEWS file in master.
  
  Also this adds the backported `remote` plugin to the NEWS
  

• df18d38c
  by Tristan Van Berkom at 2018-07-29T10:26:08Z

  git.py: Handle concurrent download completions properly
  
  Use os.rename() to rename the cloned temporary repository into
  place in the source cache, and issue a STATUS message when discarding
  a duplicate clone, in the case where the same repository is cloned
  twice in parallel.
  
  The problem with using shutil.move() is that it will create the source
  directory in a subdirectory of the destination when the destination
  exists, so it's behavior depends on whether the destination exists.
  
  This shutil.move() behavior has so far hidden the race condition
  where a duplicate repo is created in a subdirectory, as you need
  to have three concurrent downloads of the same repo in order to
  trigger the error.
  
  This fixes issue #503
  

• 92714b03
  by Tristan Van Berkom at 2018-07-29T11:18:52Z

  Merge branch 'tristan/fix-spurious-errors' into 'master'
  
  git.py: Handle concurrent download completions properly
  
  Closes #503
  
  See merge request BuildStream/buildstream!583

• 8568d272
  by Phil Dawson at 2018-07-30T07:58:07Z

  examples: add project demonstrating a junction element
  
  This is part of the work towards #437.
  

• efa24e7e
  by Phil Dawson at 2018-07-30T07:58:07Z

  junctions.run: Add terminal session output for junctions walkthrough.
  
  This is part of the work towards #437.
  

• bf175bf7
  by Phil Dawson at 2018-07-30T07:58:07Z

  junction-elements.rst: Add junctions walkthrough
  
  This is part of the work towards #437.
  

• e8931048
  by James Ennis at 2018-07-30T07:58:07Z

  using_advanced_features.rst: Added file
  

• b4f25190
  by Phil Dawson at 2018-07-30T07:58:07Z

  tests/examples/junctions.py: Add tests for junctions walkthrough
  
  This is part of the work towards issue #437.
  

• aff76ae9
  by Phil Dawson at 2018-07-30T09:15:49Z

  Merge branch 'phil/437-junction-tutorial' into 'master'
  
  Phil/437 junction tutorial
  
  See merge request BuildStream/buildstream!550

• c313ad17
  by Tom Pollard at 2018-07-30T10:57:49Z

  widget.py: Limit failure summary to currently failing elements
  
  Change widget.py print_summary() to only print the failure
  messages of elements in the Failure Summary that failed on the
  current try.
  

• 4a3def5c
  by Tristan Maat at 2018-07-30T11:47:21Z

  Merge branch 'tpollard/386' into 'master'
  
  widget.py: Limit failure summary to currently failing elements
  
  See merge request BuildStream/buildstream!561

• 34015a26
  by Tristan Maat at 2018-07-30T12:19:45Z

  HACKING.rst: Add note about asking for dev permissions
  

• ae47a72b
  by Tristan Van Berkom at 2018-07-30T13:11:34Z

  Merge branch 'tlater/ask-for-dev-permissions' into 'master'
  
  HACKING.rst: Add note about asking for dev permissions
  
  See merge request BuildStream/buildstream!587

• 695c0cb0
  by Phil Dawson at 2018-07-30T16:07:36Z

  junction-elements.rst: Add missing language specifier to literalinclude
  
  A missing language specifier on a literal include directive results in yaml
  being rendered as python code. This commit adds the :language: tag to that
  directive.
  

• fcb87b0a
  by Phil Dawson at 2018-07-30T16:53:37Z

  Merge branch 'phil/fixup-junctions-tutorial' into 'master'
  
  junction-elements.rst: Add missing language specifier to literalinclude
  
  See merge request BuildStream/buildstream!591

• 3d308894
  by Richard Maw at 2018-07-31T16:31:36Z

  element.py: Cache the result of checking whether an artifact is cached weakly
  
  Normally we'd only need it in the case of scheduling a weakly cached build,
  but to allow caching of failed builds we need to be able to distinguish
  between cached successes and cached failures
  for both strong and weak cache keys.
  
  To allow other cache lookup codepaths to look up via the weak key
  requires changes through the call stack to consult which key to use,
  and cache invalidation of the saved state when it changes.
  

• b3a68e28
  by Richard Maw at 2018-07-31T16:31:36Z

  element.py: Add metadata to distinguish between successful and failed builds
  
  This just puts the metadata in place,
  we're adding code paths to add failed builds later.
  

• f4573df3
  by Richard Maw at 2018-07-31T16:31:36Z

  Convert call-sites of Element._cached() that assume success
  
  When we later add cached failures it needs to not treat them as successes.
  

• 553df108
  by Richard Maw at 2018-07-31T16:31:36Z

  _frontend/widget.py: Render cached failures differently to successes
  

• d14d8ee2
  by Richard Maw at 2018-07-31T16:32:35Z

  Cache failed builds
  
  This creates an artifact when element assembly fails too,
  and if it's the right kind of exception uses the now-included install directory
  similarly to if it had returned successfully.
  
  If there's a failure during install the artifact contains any installed files,
  but may contain nothing at all.
  

• d83122bb
  by Richard Maw at 2018-07-31T16:32:35Z

  _scheduler/queues/buildqueue.py: Skip rebuilding cached failures
  
  This flags up a failure and if run in an interactive prompt
  permits the user to attempt a rebuild.
  

• 78944e9a
  by Richard Maw at 2018-07-31T16:32:35Z

  tests: Add tests for cached behaviours
  

• 3fa79d8d
  by Richard Maw at 2018-07-31T16:32:35Z

  _scheduler/queues: Add failed builds to "done" queue
  
  This allows the scheduler to move jobs from the current queue to the next.
  
  As a result of this change later queues than the build queue
  mustn't skip a cached failure, so the logic is specialised to build queues only.
  

• e8cd43dc
  by Richard Maw at 2018-07-31T16:32:58Z

  NEWS: Describe caching of failures
  
  Closes #76.
  

• 88cd61ea
  by knownexus at 2018-07-31T16:32:58Z

  Added the missing `_get_build_log` function
  

• 4fc1f5d1
  by Phillip Smyth at 2018-07-31T17:17:34Z

  Merge branch 'richardmaw/cache-fail' into 'master'
  
  Store failed builds in the cache
  
  Closes #76
  
  See merge request BuildStream/buildstream!475

• 6cd3df09
  by Phil Dawson at 2018-08-01T08:05:45Z

  doc/sessions: Add terminal sessions for use in workspaces tutorial
  
  This is part of the the work towards issue #437
  

• 8c0522ad
  by Phil Dawson at 2018-08-01T08:05:45Z

  Add project for workspaces tutorial
  
  This is part of the work towards issue #437
  

• dc60b713
  by Phil Dawson at 2018-08-01T08:05:45Z

  tutorials: Add workspaces tutorial
  
  This is part of the work towards issue #437
  

• bf8c3dfb
  by Phil Dawson at 2018-08-01T08:05:45Z

  testutils/patch.py: Add methods for applying and removing patches
  

• b69783d4
  by Phil Dawson at 2018-08-01T08:05:45Z

  tests/examples/developing.py: Add tests for workspaces walkthrough
  

• 480565c2
  by Phil Dawson at 2018-08-01T09:04:26Z

  Merge branch 'phil/437-workspaces-tutorial' into 'master'
  
  Phil/437 workspaces tutorial
  
  Closes #437
  
  See merge request BuildStream/buildstream!519

• 87edf67e
  by Martin Blanchard at 2018-08-01T11:07:05Z

  complete.py: Ensure paths get completed from 'element-path'
  
  Element paths should always be completed from the root element folder
  defined by the element-path key in project.conf. Fix complete_path() to
  always search into its given base_directory argument.
  
  See issue BuildStream/buildstream#448
  

• d10e4668
  by Phillip Smyth at 2018-08-01T12:05:40Z

  Merge branch 'mablanch/448-autocompletion-broken-defaults' into 'master'
  
  Fix autocompletion for elements in sub-folders
  
  See merge request BuildStream/buildstream!592

• 6aedc57b
  by Jim MacArthur at 2018-08-01T12:07:32Z

  Add BST_VIRTUAL_DIRECTORY flag for element plugins
  

• 1d3af84d
  by Jim MacArthur at 2018-08-01T12:07:32Z

  utils.py: add getmtime() and magic_timestamp
  
  magic_timestamp is moved into file scope so other classes
  can use it.
  

• 4397e45a
  by Jim MacArthur at 2018-08-01T12:07:32Z

  Sandbox.py: Rename __root to _root.
  
  This is to allow to allow its use by subclasses.
  
  Since access to get_directories is now blocked for some plugins,
  and the subclasses of Sandbox do not have configuration defined
  by YAML files, they need another way to get at the root directory.
  

• 9bf343cf
  by Jim MacArthur at 2018-08-01T12:07:32Z

  Add the virtual directory class 'Directory' and one implementation.
  
  buildstream/storage/Directory.py: New file.
  buildstream/storage/_filebaseddirectory.py: New file.
  buildstream/_exceptions.py: New VIRTUAL_FS exception source.
  

• f124dfec
  by Jim MacArthur at 2018-08-01T12:07:32Z

  storage: Add __init__.py
  

• 42bdce0b
  by Jim MacArthur at 2018-08-01T12:07:32Z

  sandbox.py: Add function to get virtual directory.
  

• fe1a3428
  by Jim MacArthur at 2018-08-01T12:07:33Z

  _stream.py: Convert to virtual directories.
  
  This removes _add_directory_to_tarfile since it is now implemented in
  _filebaseddirectory.py.
  

• 849724ce
  by Jim MacArthur at 2018-08-01T12:07:33Z

  element.py: Conversion to use virtual directories
  

• 270325b4
  by Jim MacArthur at 2018-08-01T12:07:33Z

  plugins/elements/compose.py: Convert to virtual directories
  

• a8c71ba4
  by Jim MacArthur at 2018-08-01T12:07:33Z

  plugins/elements/import.py: Convert to virtual directories
  

• 4e6a9049
  by Jim MacArthur at 2018-08-01T12:07:33Z

  sandbox: Remove instances of get_directory
  
  sandbox/_mount.py, sandbox/_sandboxbwrap.py:
    Remove instances of get_directory
  

• e69f2b23
  by Jim MacArthur at 2018-08-01T12:07:33Z

  plugins/elements/stack.py: Convert to virtual directories
  

• 6877b6fc
  by Jim MacArthur at 2018-08-01T12:07:33Z

  scriptelement.py: Convert to virtual directories
  

• 8edf1f80
  by Jim MacArthur at 2018-08-01T12:07:33Z

  compose.py: Remove unused getmtime()
  

• 02c6c846
  by Jim MacArthur at 2018-08-01T12:07:33Z

  sandbox.py: Use _root in _has_command
  
  This will need more attention when we bring in another virtual
  directory backend, however, we've said it is acceptable for the
  sandbox itself to access the underlying directory, and this is
  the best fix in the meantime.
  

• eabc3899
  by Jim MacArthur at 2018-08-01T13:16:04Z

  Merge branch 'jmac/virtual_directories' into 'master'
  
  Abstract directory class and filesystem-backed implementation
  
  See merge request BuildStream/buildstream!445

• 60ddb193
  by Josh Smith at 2018-08-01T14:17:17Z

  message.py: MessagesType.SKIPPED added
  

• 0756c615
  by Josh Smith at 2018-08-01T14:17:17Z

  cascache.py: Display SKIPPED message for each remote skipped
  

• 62a250fe
  by Qinusty at 2018-08-01T15:01:25Z

  Merge branch 'Qinusty/481' into 'master'
  
  Add SKIPPED message type for actions being skipped
  
  See merge request BuildStream/buildstream!562

• 00ddcbd7
  by Martin Blanchard at 2018-08-01T15:16:33Z

  element.py: Raise an exception on unbuilt element checkout
  
  Trying to checkout an element that has no cached artifacts should be and
  handled failure. See BuildStream/buildstream#447.
  

• 4c739e2a
  by Martin Blanchard at 2018-08-01T15:16:33Z

  _stream.py: Print details on checkout failure
  

• 4a4fed7a
  by Qinusty at 2018-08-01T16:14:50Z

  Merge branch 'mablanch/447-stack-trace-checkout' into 'master'
  
  Handle checkout failure for unbuilt elements
  
  See merge request BuildStream/buildstream!590

• ca25e0c1
  by Javier Jardón at 2018-08-01T16:40:51Z

  Add license: Creative Commons Attribution 4.0 International License
  

18 changed files:

• .coveragerc
• + .gitattributes
• .gitignore
• .gitlab-ci.yml
• + .gitlab/issue_templates/bst_bug.md
• + .gitlab/issue_templates/bst_task.md
• + .gitlab/merge_request_templates/bst_merge_request.md
• .pylintrc
• HACKING.rst
• MANIFEST.in
• NEWS
• README.rst
• buildstream/__init__.py
• buildstream/_artifactcache/__init__.py
• buildstream/_artifactcache/artifactcache.py
• + buildstream/_artifactcache/cascache.py
• + buildstream/_artifactcache/casserver.py
• − buildstream/_artifactcache/ostreecache.py

The diff was not included because it is too large.

—
View it on GitLab.
You're receiving this email because of your account on gitlab.com. If you'd like to receive fewer emails, you can adjust your notification settings.