Re: [anjuta-list] New developments



Hi,


Le 28/03/2012 18:43, Johannes Schmid a écrit :
* Check usage of Anjuta in gnome-shell
It's basically covered by the G(tk)Application class pretty well and
there shouldn't be any need to use D-Bus directly. However, the main.c,
anjuta.c and anjuta-app.c code is kind of ugly as it has gone through
various transitions or the time being. Cleaning this up and fix some
problems should really be done this cycle.

It will probably need to use the latest changes in GtkApplication (version 3.4), so it's probably better to wait until all major distributions are released.


* Review all bugs
We try to do that from time to time but it is of course always a partial
review. Maybe we can find some day on a weekend where we do a mass
review.

Yes, I think it would be useful but it's not a major point.


* Allow to load more autotools project
* Add more project backend
* Support more options in autotools project
* Support quoting in Makefile.am and configure.ac
Supported even more automake features and making them available to the
user would be really great. I think we have come a long way with
autotools support but we might be even better.

I will probably try to improve both but do you think we should put more effort on read or write capabilities?
Reading more complex projects is probably more useful for experienced users.
While improving writing is better for newcomers those don't know how to write autotools files. It's a more difficult to do though.


* Add missing source files on commit
This is something the project-manager should tell the version control
plugin when it adds files (or directories) and I think we should do that
automatically (the user will see the summary of the files when he
commits something anyway).

Yes, I think it should done for this cycle. I have forgotten a file 2 times within the last months. I already have some patches ready for the project manager part.


* Take care of indentation in project and file wizard generated files
This is reported quite often so it seems to be something that annoys
people. I was thinking about something like a [+TAB+] tag in the
template files but maybe there are better (and easier to maintain)
options for it. This is quite high on the list of things to fix for me.

We can use scheme function, by example something like [+(indent 2)+] instead of [+TAB+][+TAB+]. It allows to use optimize space, by example use one tab instead of 8 space with an indentation of 4 spaces. I don't know if it's really needed though.


* Remember layout in GDL
That would be great but it seems difficult to do as you need deep
knowledge on how gdl and gtk+ work.

Yes, it's not easy. Is there other things to do on GDL which you think are more important? I haven't really worked on GDL.


* Move the template engine in a separate plugin
Probably just create a good common class in libanjuta to handle these
things.

I think libanjuta grows bigger and bigger. It's not a problem for Anjuta, but I think it's more annoying to use libanjuta as a platform. I would prefer to remove things from libanjuta.


* Allow to rename variables and functions
* Support call graph
In the big picture I see this as things that could be done with clang
libraries at least for C/C++. Having a full parser for the code should
simplify this but I don't know how to handle all the details, yet.

I would like to do this for other languages than C/C++. Then, I wondering how to organize all this too.


* Generate distribution package
We could try to support it for the common projects we generate. At least
rpm supports to depend on pkg-config information which we can provide
but I am not sure about deb packages.
In practice it is also quite hard to generate packages for distros you
are not currently on if you don't use build services like OpenSuSE build
service.

Ok


* Improve documentation
This is one of those never-ending stories but we are getting better with
every release so I hope we can at least cover all the core plugins for
the next release.

Ok.


* Make Anjuta starts faster
I need to experiment with GResources a bit which might allow us to do
more efficient IO by compressing resources and loading them as a big
chunk (at least big chunk per plugin). There might be other things we
might be able to optimize.

Ok.


Regards,

Sébastien


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]