[gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 427 bytes --]

Hi,

Now that we have proper components for keywording and stabilization,
the old keywords are redundant.  Nevertheless, some people still set
them.  I would like to propose two solutions going forward.  Either:

1. We kill both keywords, and just rely on components, or

2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
appropriate.

Which would you prefer?

-- 
Best regards,
Michał Górny


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 618 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Joonas Niilola]

[-- Attachment #1.1: Type: text/plain, Size: 657 bytes --]


On 4/11/20 6:33 PM, Michał Górny wrote:
> Hi,
>
> Now that we have proper components for keywording and stabilization,
> the old keywords are redundant.  Nevertheless, some people still set
> them.  I would like to propose two solutions going forward.  Either:
>
> 1. We kill both keywords, and just rely on components, or
>
> 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> appropriate.
>
> Which would you prefer?
>
Less noise is better, so I vote for 1.

Wasn't aware KEYWORDREQ and STABLEREQ were useless, thats why I always
set them. Will it break any commonly used search scripts / settings?

-- juippis



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 642 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Michael Orlitzky]

On 4/11/20 11:33 AM, Michał Górny wrote:
> Hi,
> 
> Now that we have proper components for keywording and stabilization,
> the old keywords are redundant.  Nevertheless, some people still set
> them.  I would like to propose two solutions going forward.  Either:
> 
> 1. We kill both keywords, and just rely on components, or
> 

I've been setting them just in case someone has a workflow/automation
involving the keywords that hasn't been updated in ten years. If you
kill the keywords, I wouldn't have to worry about that, so +1.

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[James Le Cuirot]

[-- Attachment #1: Type: text/plain, Size: 866 bytes --]

On Sat, 11 Apr 2020 18:38:27 +0300
Joonas Niilola <juippis@gentoo.org> wrote:

> On 4/11/20 6:33 PM, Michał Górny wrote:
> > Hi,
> >
> > Now that we have proper components for keywording and stabilization,
> > the old keywords are redundant.  Nevertheless, some people still set
> > them.  I would like to propose two solutions going forward.  Either:
> >
> > 1. We kill both keywords, and just rely on components, or
> >
> > 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> > appropriate.
> >
> > Which would you prefer?
> >  
> Less noise is better, so I vote for 1.
> 
> Wasn't aware KEYWORDREQ and STABLEREQ were useless, thats why I always
> set them. Will it break any commonly used search scripts / settings?

Me neither, must have missed that memo. Go for 1.

-- 
James Le Cuirot (chewi)
Gentoo Linux Developer

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Kent Fredric]

[-- Attachment #1: Type: text/plain, Size: 501 bytes --]

On Sat, 11 Apr 2020 11:39:21 -0400
Michael Orlitzky <mjo@gentoo.org> wrote:

> I've been setting them just in case someone has a workflow/automation
> involving the keywords that hasn't been updated in ten years. If you
> kill the keywords, I wouldn't have to worry about that, so +1.

And that's pretty much the same thing I've been doing.

I imagine if these keywords go, then people trying to rely on them in
clients of various descriptions will start hard-erroring when they use
them.

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Thomas Deutschmann]

[-- Attachment #1.1: Type: text/plain, Size: 719 bytes --]

On 2020-04-11 17:33, Michał Górny wrote:
> 1. We kill both keywords, and just rely on components, or
> 
> 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> appropriate.

I think you cannot kill it.

Yes, we have a component for stabilization/keywording, but we also do
stabilization from security bugs which don't have such a component and
some tools must be able to filter.

Just checking CC list is not an option because in theory, you can CC
architectures when you are just requesting some input from them.

So I would tend to #2. It doesn't really hurt, does it?


-- 
Regards,
Thomas Deutschmann / Gentoo Linux Developer
C4DD 695F A713 8F24 2AA1 5638 5849 7EE5 1D5D 74A5


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 618 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Mike Gilbert]

On Sat, Apr 11, 2020 at 12:28 PM Thomas Deutschmann <whissi@gentoo.org> wrote:
>
> On 2020-04-11 17:33, Michał Górny wrote:
> > 1. We kill both keywords, and just rely on components, or
> >
> > 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> > appropriate.
>
> I think you cannot kill it.
>
> Yes, we have a component for stabilization/keywording, but we also do
> stabilization from security bugs which don't have such a component and
> some tools must be able to filter.

Someone proposed that we change the security bug workflow to include a
separate stable request bug, which would resolve this.

That would also help in cases where we are stabilizing
security-unsupported archs at the same time.

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[Michał Górny]

[-- Attachment #1: Type: text/plain, Size: 800 bytes --]

On Sat, 2020-04-11 at 18:28 +0200, Thomas Deutschmann wrote:
> On 2020-04-11 17:33, Michał Górny wrote:
> > 1. We kill both keywords, and just rely on components, or
> > 
> > 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> > appropriate.
> 
> I think you cannot kill it.
> 
> Yes, we have a component for stabilization/keywording, but we also do
> stabilization from security bugs which don't have such a component and
> some tools must be able to filter.
> 
> Just checking CC list is not an option because in theory, you can CC
> architectures when you are just requesting some input from them.

Except that no tools actually filter based on keywords.  If you add
packages and CC arches, the thing becomes a stablereq.

-- 
Best regards,
Michał Górny


[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 618 bytes --]

Re: [gentoo-dev] [RFC] KEYWORDREQ and STABLEREQ keywords

[David Seifert]

On Sat, 2020-04-11 at 12:53 -0400, Mike Gilbert wrote:
> On Sat, Apr 11, 2020 at 12:28 PM Thomas Deutschmann <whissi@gentoo.org> wrote:
> > On 2020-04-11 17:33, Michał Górny wrote:
> > > 1. We kill both keywords, and just rely on components, or
> > > 
> > > 2. I make NATTkA automatically add KEYWORDREQ or STABLEREQ where
> > > appropriate.
> > 
> > I think you cannot kill it.
> > 
> > Yes, we have a component for stabilization/keywording, but we also do
> > stabilization from security bugs which don't have such a component and
> > some tools must be able to filter.
> 
> Someone proposed that we change the security bug workflow to include a
> separate stable request bug, which would resolve this.
> 
> That would also help in cases where we are stabilizing
> security-unsupported archs at the same time.
> 

I agree, this sounds like the better solution, and separating concerns.