On Wed, Jun 24, 2020 at 11:52:28AM +0200, Thomas Deutschmann wrote:
> On 2020-06-20 21:24, Aaron Bauman wrote:
> > Thomas, unfortunately, I am shocked at your choice of words here. I
> > think it is reasonable that any developer would understand a lack
> > of forward momentum in removing Py2 only packages only drives
> > stagnation.
> > 
> > If you have a more effective method to doing so, I am open to
> > suggestions.
> Like I am shocked about your recent actions:
> Remember what you did in January. I thought it became clear that next
> time you will share your list before just masking stuff to avoid things
> which happened then.

Developers have many tools and *hopefully* the organic ability to determine
which packages are impacted. Especially given previous threads on this
very ML with pleas from other Python team members to assist in cleaning
things up in a deliberate manner.

This is why the QA team interceded... because a couple of individuals
screamed loudly for no reason. Fix it and move on.

> In the beginning of this month you just decided to disband graphics
> project. On your own. Please tell me what gave you the authority to just
> do that? You didn't even share your plan before executing it on any
> mailing list. Something that should be common sense, if not even necessary.
> The whole action was so destructive that you couldn't evenb just undo it
> because you also deleted stuff on Wiki.

I will not apologize for doing something that others have lacked the
intestinal fortitude to do.

> Like multiple people have already shown you, many packages from that
> list are not even blocking Py3 transition.

This isn't about transitioning to Py3... it is about removing Py2.

> Let me tell you what a mask will cause:
> A mask is destructive and requires user interaction. Therefore a mask
> isn't something to play with, "Oh, let's test if someone will
> complain... it's just a mask, we can just unmask in case...".

Is that why you assume I masked these things?

> No, imagine there are people out there using Gentoo in production and
> not as playground. These people maybe have automated build systems which
> are creating systems/images (do you know Dockers for example?). Whenever
> you mask something and that package is referenced in configuration, you
> will break that build.
> That's not funny if this is happening for no real reason.

I know you use Gentoo in production, but does this mean we (Gentoo)
can't move forward because *you* want to use something that is EOL and
dying? What if you used a major distro that removed Py2 support already?
Why complain here? There are other ways to safely run your tooling with
Py2 if you so choose.

> > re: net-mail/offlineimap... there are alternatives.
> I think you don't really know that tool. It's an industry standard.
> Sure, there are already successors (however, not in Gentoo). But the
> package itself is still working and actively maintained and when you
> will use it in production you usually have extended/adjusted the tool
> for your environment using the plugin system the tool provides. That's
> not something you will be able to replace with something new in 5 minutes.

You continuously speak condescendingly to me. I am truly starting to
regret my nomination for you on both the security project and for
council. Do you speak to others this way simply because you don't agree
with them?

Yes, I use net-mail/offlineimap... I know how it works. No, I really
hope that a tool which has not been maintained in many years is an
"industry standard" 

Yes, there are successors in Gentoo.

> And I repeat myself: Especially not when there is no need to do that
> because because the package itself is working fine and there is absolute
> no reason to get rid of it.

Take Patrick's approach and move it to an overlay if you want it that

> Last but not least: Gentoo is about choices. It's not your job to decide
> what people should use. Sure, if you maintained a package and will stop
> using it so it will become maintainer-needed and masked for removal at
> some point because it's outdated, vulnerable and/or not working anymore,
> that's OK. But if someone else will pick up this package... and
> offlineimap in Gentoo is working and up-to-date.

Are you implying that because "Gentoo is about choice" that we never
remove an ebuild, interpreter, compiler, etc? Let ::gentoo grow in size
forever to appease the few?


Attachment: signature.asc
Description: PGP signature

Reply via email to