Hi Trisatan,

Unfortunately, I must disagree with your disagreement.

We have a bar that is set for projects being TLP at Apache and that bar is 
there for a reason.

As we can’t expect every project to instantly meet these expectations, that’s 
what we have the Apache Incubator for.
This is where projects get used to the voting process, how things are handled 
on the mailing-lists, how they get their releases into shape to meet the 
quality-bar of TLPs and so on.

When Perti was initiated, the promise was, that mature projects will be 
mentored under Pertri and its releases will meet the criteria as soon as they 
graduate. Admittedly I was a bit sceptical with this, and this situation seems 
to confirm this.

You have graduated and the same rules apply for BuildStream as they do for all 
other Apache TLP projects.

Sorry for having to be this clear.

Chris


From: Tristan Van Berkom <[email protected]>
Date: Tuesday, 8. November 2022 at 08:26
To: [email protected] <[email protected]>
Subject: Re: [VOTE] Release buildstream / plugins 1.95.4 as 2.0
Hi Ben,

On Sat, 2022-11-05 at 15:25 +0000, Benjamin Schubert wrote:
> Hey everyone,
>
[...]
> -0
>
> I believe https://github.com/apache/buildstream/issues/1787 should be
> a blocker for this release, as it would otherwise negatively impact
> the first experience with it.

As you will notice I've currently closed this issue as I believe you're
not up to speed on the intentional behavioral change, which is to
ensure that after quitting a build, that any built elements (failed or
successful) continue to be pushed, but no new elements will be queued
to be built.

Also, I disagree with raising the bar this high for the initial 2.0
release: an initial 2.0 release is not necessarily the same as a "first
experience" with BuildStream 2.

Right now we've been struggling to cut through the red tape and get
used to the voting process and end up with a reasonable cadence for
subsequent releases (which is proving tiresome and more difficult than
expected); ensuring our release channel works and striking while the
iron is hot is a top priority for the project in my view at this time.

Right now we don't have any known issues which render the tool unusable
in any way, there may be some minor inconveniences or lack of polish in
some places, but nothing we cannot fix in a subsequent bugfix release
in an API stable fashion.

Cheers,
    -Tristan

Reply via email to