Re: [BuildStream] 1.4 release timeline, targets and breaking changes

[Angelos Evripiotis <angelos evripiotis gmail com>]

On Thu, Nov 1, 2018 at 4:08 PM Agustín Benito Bethencourt via
BuildStream-list <buildstream-list gnome org> wrote:
> On Thursday, 1 November 2018 16:54:59 CET Angelos Evripiotis via BuildStream-list wrote:
> > Can I assume that because it made it into this 1.4 tracking issue that it’s
> > officially welcome?
> > https://gitlab.com/BuildStream/buildstream/issues/743
> To answer your question, in my opinion #743 is not just welcome, it has to happen.

Ah, I think there's a small misunderstanding - I meant to ask if my
new issue #744 was welcome, rather than Javier's :) Fwiw, I vigorously
agree that we should doc the breaking changes, and that issue is a
great way to make sure that happens.

On Thu, Nov 1, 2018 at 5:22 PM Tristan Van Berkom
<tristan vanberkom codethink co uk> wrote:
> On Thu, 2018-11-01 at 15:54 +0000, Angelos Evripiotis via BuildStream-
> list wrote:
> > Hopefully this guy isn’t too late to the 1.4 party (needs some design
> > debate):
> > https://gitlab.com/BuildStream/buildstream/issues/744
>
> This was not added to the list of targets, but that does not mean it
> cannot get in if the effort is made.
>
> Rather we boiled down a list of things we thing should absolutely get
> into the release, and agreed we should focus on those - if other things
> can land before feature freeze then there is no reason to reject them
> (unless perhaps they are wildly unstable and dangerous changes for the
> end of the cycle, which doesnt appear to be the case for 744).

Great! I'll take that as encouragement that it's worth me putting
effort into 744.

> > Can I assume that because it made it into this 1.4 tracking issue
> > that it’s officially welcome?
> > https://gitlab.com/BuildStream/buildstream/issues/743
>
> This issue is Javier rightfully pointing out that we should document
> these breakages somewhere (it is not a complete list, probably we
> should look at the few minor breakages we introduced in 1.2 as well for
> this).

Ok cool, I'll stick to the ML as the Single Point of Truth for
checking whether any further proposed breaks (yuck!) are welcome in
1.4.