Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Sergei Trofimovich]

On Sun, 11 Oct 2015 09:23:20 +0200
Michał Górny  wrote:

> Contributors are repeatedly adding base.eclass uses, so we should
> finally make the deprecation formal, even at the cost of adding warnings
> for some frequently used eclasses.
> 
> Signed-off-by: Michał Górny 
> ---
>  pym/repoman/checks/ebuilds/checks.py | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/pym/repoman/checks/ebuilds/checks.py 
> b/pym/repoman/checks/ebuilds/checks.py
> index ae6d96e..a00d518 100644
> --- a/pym/repoman/checks/ebuilds/checks.py
> +++ b/pym/repoman/checks/ebuilds/checks.py
> @@ -409,6 +409,7 @@ class InheritDeprecated(LineCheck):
>  
>   # deprecated eclass : new eclass (False if no new eclass)
>   deprecated_eclasses = {
> + "base": False,
>   "bash-completion": "bash-completion-r1",
>   "boost-utils": False,
>   "distutils": "distutils-r1",
> -- 
> 2.6.1

I would also suggest:

- adding a @DEPRECATED documentation stanza
   right into base.eclass with a migration strategy link
- and a ewarn/eqawarn call to flag all offenders and prevent
   new ebuilds to creep dependency back

-- 

  Sergei


signature.asc
Description: PGP signature

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Michał Górny]

Dnia 2015-10-11, o godz. 13:24:43
Zac Medico  napisał(a):

> On 10/11/2015 12:23 PM, Zac Medico wrote:
> > On 10/11/2015 09:51 AM, Alexander Berntsen wrote:
> >> On 11/10/15 18:45, Michał Górny wrote:
> >>> Nope, don't have any. Though AFAICS there's all-over-the-place 
> >>> consensus that it's deprecated, and most of the developers complain
> >>>  when they see it.
> >> If you can show me the consensus, I'll ACK it. Alternatively anyone
> >> else on the team that have observed this consensus may ACK it.
> > 
> > I've found a tracker bug here:
> > 
> > https://bugs.gentoo.org/show_bug.cgi?id=497022
> > 
> 
> The patch is consistent with the QA team's policy discussed in bug
> 497022, it looks good to me.

Pushed.

-- 
Best regards,
Michał Górny



pgpeq_FJK77kS.pgp
Description: OpenPGP digital signature

[gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Michał Górny]

Contributors are repeatedly adding base.eclass uses, so we should
finally make the deprecation formal, even at the cost of adding warnings
for some frequently used eclasses.

Signed-off-by: Michał Górny 
---
 pym/repoman/checks/ebuilds/checks.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/pym/repoman/checks/ebuilds/checks.py 
b/pym/repoman/checks/ebuilds/checks.py
index ae6d96e..a00d518 100644
--- a/pym/repoman/checks/ebuilds/checks.py
+++ b/pym/repoman/checks/ebuilds/checks.py
@@ -409,6 +409,7 @@ class InheritDeprecated(LineCheck):
 
# deprecated eclass : new eclass (False if no new eclass)
deprecated_eclasses = {
+   "base": False,
"bash-completion": "bash-completion-r1",
"boost-utils": False,
"distutils": "distutils-r1",
-- 
2.6.1

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Michał Górny]

Dnia 2015-10-11, o godz. 18:16:44
Alexander Berntsen  napisał(a):

> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA512
> 
> I don't even remember the last time this was discussed. Could you link
> to some discussion or a decision on this?

Nope, don't have any. Though AFAICS there's all-over-the-place
consensus that it's deprecated, and most of the developers complain
when they see it.

> Also, this patch won't actually apply to HEAD since you are basing it
> on one of your patches that aren't in yet. Try to avoid that in the
> future.

If I based it on HEAD, it wouldn't apply with the other patch on ;-P.

-- 
Best regards,
Michał Górny



pgp8WF3lA6zbY.pgp
Description: OpenPGP digital signature

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Zac Medico]

On 10/11/2015 09:51 AM, Alexander Berntsen wrote:
> On 11/10/15 18:45, Michał Górny wrote:
>> Nope, don't have any. Though AFAICS there's all-over-the-place 
>> consensus that it's deprecated, and most of the developers complain
>>  when they see it.
> If you can show me the consensus, I'll ACK it. Alternatively anyone
> else on the team that have observed this consensus may ACK it.

I've found a tracker bug here:

https://bugs.gentoo.org/show_bug.cgi?id=497022
-- 
Thanks,
Zac

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Alexander Berntsen]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512

I don't even remember the last time this was discussed. Could you link
to some discussion or a decision on this?

Also, this patch won't actually apply to HEAD since you are basing it
on one of your patches that aren't in yet. Try to avoid that in the
future.
- -- 
Alexander
berna...@gentoo.org
https://secure.plaimi.net/~alexander
-BEGIN PGP SIGNATURE-
Version: GnuPG v2
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=oiR6
-END PGP SIGNATURE-

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Alexander Berntsen]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512

On 11/10/15 18:45, Michał Górny wrote:
> Nope, don't have any. Though AFAICS there's all-over-the-place 
> consensus that it's deprecated, and most of the developers complain
>  when they see it.
If you can show me the consensus, I'll ACK it. Alternatively anyone
else on the team that have observed this consensus may ACK it.

I'm sorry, but I've been so busy that I'm a bit out of the loop with
Gentoo. I'm not doing this to be annoying, but it is effectively
(probably) all my fault.

> If I based it on HEAD, it wouldn't apply with the other patch on 
> ;-P.
Yeah, but if we for some reason we decided to not apply that one, but
do apply this one, it would be a bit of a hassle. And they are not
related enough to be in a patch series, so... But this is nitpicking
and a non-issue since I ACKed the first patch. :]

- -- 
Alexander
berna...@gentoo.org
https://secure.plaimi.net/~alexander
-BEGIN PGP SIGNATURE-
Version: GnuPG v2
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=W+yH
-END PGP SIGNATURE-

Re: [gentoo-portage-dev] [PATCH] repoman: Finally deprecate base.eclass

[Zac Medico]

On 10/11/2015 12:23 PM, Zac Medico wrote:
> On 10/11/2015 09:51 AM, Alexander Berntsen wrote:
>> On 11/10/15 18:45, Michał Górny wrote:
>>> Nope, don't have any. Though AFAICS there's all-over-the-place 
>>> consensus that it's deprecated, and most of the developers complain
>>>  when they see it.
>> If you can show me the consensus, I'll ACK it. Alternatively anyone
>> else on the team that have observed this consensus may ACK it.
> 
> I've found a tracker bug here:
> 
> https://bugs.gentoo.org/show_bug.cgi?id=497022
> 

The patch is consistent with the QA team's policy discussed in bug
497022, it looks good to me.
-- 
Thanks,
Zac