[gentoo-dev] Conversion of Emacs virtual packages

[gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

The Emacs team has converted the following virtual packages:

   virtual/emacs
   virtual/flim

to new-style (aka GLEP 37) virtuals. If there are no objections,
I will remove the old virtuals from the profiles one week from now
(22 May 2007). Affected are:

   profiles/base
   profiles/embedded
   profiles/default-bsd/fbsd/6.2/sparc
      (This will be empty then. Shall I remove the virtuals file?)
   profiles/default-darwin

virtual/semi was also removed some time ago, because it was only
provided by a single package (app-emacs/semi) and therefore no longer
needed.

I could not find any official guidelines for keywording of virtual
packages. Current practice seems to be that keywords are simply added
as needed. (Which makes sense: The virtual package installs no files,
so there is nothing that could be tested on a specific architecture.)

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Jakub Moc]

[-- Attachment #1: Type: text/plain, Size: 681 bytes --]

Ulrich Mueller napsal(a):
> The Emacs team has converted the following virtual packages:
> 
>    virtual/emacs
>    virtual/flim
> 
> to new-style (aka GLEP 37) virtuals. If there are no objections,
> I will remove the old virtuals from the profiles one week from now
> (22 May 2007).

Leaving the old-style virtuals in place leaves the new style ones
non-fuctional, so you'd better remove them right now. ;)


-- 
Best regards,

 Jakub Moc
 mailto:jakub@gentoo.org
 GPG signature:
 http://subkeys.pgp.net:11371/pks/lookup?op=get&search=0xCEBA3D9E
 Primary key fingerprint: D2D7 933C 9BA1 C95B 2C95  B30F 8717 D5FD CEBA 3D9E

 ... still no signature   ;)


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 513 bytes --]

On Tue, 15 May 2007 12:44:56 +0200
Ulrich Mueller <ulm@gentoo.org> wrote:
> The Emacs team has converted the following virtual packages:
> 
>    virtual/emacs
>    virtual/flim
> 
> to new-style (aka GLEP 37) virtuals. If there are no objections,
> I will remove the old virtuals from the profiles one week from now
> (22 May 2007).

Wait. Are you saying that currently you have identically named new and
old style virtuals in the tree? Because, uh, that's not a good idea.

-- 
Ciaran McCreesh


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Jakub Moc]

[-- Attachment #1: Type: text/plain, Size: 525 bytes --]

Ulrich Mueller napsal(a):
> The Emacs team has converted the following virtual packages:
> 
>    virtual/emacs
>    virtual/flim
> 
> to new-style (aka GLEP 37) virtuals.

Oh, and naming the new-style virtuals the same as the real thing kills
binpkgs. :/


-- 
Best regards,

 Jakub Moc
 mailto:jakub@gentoo.org
 GPG signature:
 http://subkeys.pgp.net:11371/pks/lookup?op=get&search=0xCEBA3D9E
 Primary key fingerprint: D2D7 933C 9BA1 C95B 2C95  B30F 8717 D5FD CEBA 3D9E

 ... still no signature   ;)


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

>>>>> On Tue, 15 May 2007, Petteri Räty wrote:

>> I could not find any official guidelines for keywording of virtual
>> packages.

> They are keyworded as any other packages. You can't make them stable
> if they have ~arch only dependencies. When one provider is stable,
> they are fine to go stable.

Yes, I think this part is clear.

The point of my question was more if the usual rules apply, i.e.:
keywording and stabilising only by arch teams; wait one month before
the package can go stable. Which seems not to be current practice for
virtual packages. (And makes no sense anyway, IMHO.)

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

>>>>> On Tue, 15 May 2007, Jakub Moc wrote:

> Leaving the old-style virtuals in place leaves the new style ones
> non-fuctional, so you'd better remove them right now. ;)

>>>>> On Tue, 15 May 2007, Ciaran McCreesh wrote:

> Wait. Are you saying that currently you have identically named new and
> old style virtuals in the tree? Because, uh, that's not a good idea.

Wow. Jakub and Ciaran agree, so this must be right. :-)
Removed from profiles, ebuilds will follow.

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Marking virtuals stable

[Petteri Räty]

[-- Attachment #1: Type: text/plain, Size: 911 bytes --]

Ulrich Mueller kirjoitti:
>>>>>> On Tue, 15 May 2007, Petteri Räty wrote:
> 
>>> I could not find any official guidelines for keywording of virtual
>>> packages.
> 
>> They are keyworded as any other packages. You can't make them stable
>> if they have ~arch only dependencies. When one provider is stable,
>> they are fine to go stable.
> 
> Yes, I think this part is clear.
> 
> The point of my question was more if the usual rules apply, i.e.:
> keywording and stabilising only by arch teams; wait one month before
> the package can go stable. Which seems not to be current practice for
> virtual packages. (And makes no sense anyway, IMHO.)
> 
> Ulrich

The month is not set in stone. About who marks them, it's probably best
to get the opinion of the arch teams. I don't think they will object to
normal developers marking them. Arch teams: what do you think?

Regards,
Petteri


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 252 bytes --]

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

>>>>> On Tue, 15 May 2007, Jakub Moc wrote:

>> virtual/emacs
>> virtual/flim

> Oh, and naming the new-style virtuals the same as the real thing
> kills binpkgs. :/

Sorry, I don't get your point here.

The "real thing" is, e.g., emacs-21.4-r12 and the virtual is emacs-21.
How is this not sufficiently different?

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Chris Gianelloni]

On Tue, 2007-05-15 at 12:44 +0200, Ulrich Mueller wrote:
> The Emacs team has converted the following virtual packages:
> 
>    virtual/emacs

You made a new-style virtual with the same name as a package?  This can
cause problems with binary package usage, as you will now end up with
two packages with the same "name" but different versions.  We likely
won't be able to include emacs on Gentoo releases any more because of
this.  Just thought I would let you know.

(Just realizing I need to finish configuring this mail client at
work...)

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Chris Gianelloni]

On Tue, 2007-05-15 at 15:25 +0200, Ulrich Mueller wrote:
> >>>>> On Tue, 15 May 2007, Jakub Moc wrote:
> 
> >> virtual/emacs
> >> virtual/flim
> 
> > Oh, and naming the new-style virtuals the same as the real thing
> > kills binpkgs. :/
> 
> Sorry, I don't get your point here.
> 
> The "real thing" is, e.g., emacs-21.4-r12 and the virtual is emacs-21.
> How is this not sufficiently different?

You end up with this:

/usr/portage/packages/All/emacs-21.4-r12
/usr/portage/packages/All/emacs-21

When you go to merge your virtual, it will always merge the package,
even if you're *trying* to merge an emacs alternative.


-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

>>>>> On Wed, 16 May 2007, Chris Gianelloni wrote:

>> virtual/emacs

> You made a new-style virtual with the same name as a package? This
> can cause problems with binary package usage, as you will now end up
> with two packages with the same "name" but different versions.

There are several others, BTW:

   {virtual,media-libs}/glut
   {virtual,sys-devel}/pmake
   {virtual,dev-db}/mysql

New-style virtuals are just *packages*, or did I get this completely
wrong? So how is this situation different from two packages with the
same name, but in different categories?

> We likely won't be able to include emacs on Gentoo releases any more
> because of this. Just thought I would let you know.

If you believe that this will make users happy, go ahead. :7

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Chris Gianelloni]

On Wed, 2007-05-16 at 19:52 +0200, Ulrich Mueller wrote:
> >>>>> On Wed, 16 May 2007, Chris Gianelloni wrote:
> 
> >> virtual/emacs
> 
> > You made a new-style virtual with the same name as a package? This
> > can cause problems with binary package usage, as you will now end up
> > with two packages with the same "name" but different versions.
> 
> There are several others, BTW:
> 
>    {virtual,media-libs}/glut
>    {virtual,sys-devel}/pmake
>    {virtual,dev-db}/mysql
> 
> New-style virtuals are just *packages*, or did I get this completely
> wrong? So how is this situation different from two packages with the
> same name, but in different categories?

It isn't different.  That's the problem.  If you have two packages with
the same name, you have the same problem.

> > We likely won't be able to include emacs on Gentoo releases any more
> > because of this. Just thought I would let you know.
> 
> If you believe that this will make users happy, go ahead. :7

It has nothing to do with making users happy and everything to do with
our inability to work around this issue with the current binary package
support in portage.


-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[William L. Thomson Jr.]

[-- Attachment #1: Type: text/plain, Size: 864 bytes --]

On Wed, 2007-05-16 at 13:07 -0700, Chris Gianelloni wrote:
> On Wed, 2007-05-16 at 19:52 +0200, Ulrich Mueller wrote:
> >
> > New-style virtuals are just *packages*, or did I get this completely
> > wrong? So how is this situation different from two packages with the
> > same name, but in different categories?
> 
> It isn't different.  That's the problem.  If you have two packages with
> the same name, you have the same problem.

On that note I would hope the vim/vi peeps would rename.
app-vim/ant

It's quite annoying when one needs the real ANT ( dev-java/ant ), not
the vim menu plugin so vi can invoke ant or etc.

IMHO app-vim/ant should really be app-vim/vim-ant or something other
than just ant.

Personal pet peeve of some time now, just lacked the occasion to mention
it till now ;)


-- 
William L. Thomson Jr.
Gentoo/Java

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Conversion of Emacs virtual packages

[Ulrich Mueller]

>>>>> On Wed, 16 May 2007, Chris Gianelloni wrote:

> You end up with this:

> /usr/portage/packages/All/emacs-21.4-r12
> /usr/portage/packages/All/emacs-21

> When you go to merge your virtual, it will always merge the package,
> even if you're *trying* to merge an emacs alternative.

This is no longer an issue in (stable) portage-2.1.2.2. I just talked
to zmedico and he confirmed that 2.1.2 avoids collisions.

So, no need to rename virtual/emacs.

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Thilo Bangert]

[-- Attachment #1: Type: text/plain, Size: 449 bytes --]

> > It isn't different.  That's the problem.  If you have two packages
> > with the same name, you have the same problem.
>
> On that note I would hope the vim/vi peeps would rename.
> app-vim/ant

and app-vim/sudo

> IMHO app-vim/ant should really be app-vim/vim-ant or something other
> than just ant.

or app-vim/sudo-syntax and app-vim/ant-syntax as there already are a 
number of ebuilds following that scheme...

regards
Thilo

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Jakub Moc]

[-- Attachment #1: Type: text/plain, Size: 301 bytes --]

Thilo Bangert napsal(a):
>>> It isn't different.  That's the problem.  If you have two packages
>>> with the same name, you have the same problem.
>> On that note I would hope the vim/vi peeps would rename.
>> app-vim/ant
> 
> and app-vim/sudo

and app-xemacs/emerge, grrrr


-- 
jakub


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[gentoo-dev] Re: Packages with same name was -> Conversion of Emacs virtual packages

[Ali Polatel]

Thilo Bangert <bangert@gentoo.org> yazdı:
> > > It isn't different.  That's the problem.  If you have two packages
> > > with the same name, you have the same problem.
> >
> > On that note I would hope the vim/vi peeps would rename.
> > app-vim/ant
> 
> and app-vim/sudo
> 
  which is soon to be dead...


-- 
Ali Polatel (hawking) <polatel@gmail.com>
http://hawking.nonlogic.org/
gpg: 0x8E724EDC fp: DBC2 2BC7 95B8 6D6C 8BC3  37EC CA00 CFC1 8E72 4EDC
()  ascii ribbon campaign - against html mail
/\
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Carsten Lohrke]

[-- Attachment #1: Type: text/plain, Size: 283 bytes --]

While I always was for uniq package names, tree-wide, renaming doesn't solve 
anything. Gentoo's binary packages are fundamentally broken. You can't have 
two binary packages of the same ebuild differing e.g. in use flags, 
architecture, toolchain, etc. pp. either.


Carsten

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Mike Kelly]

On Thu, 17 May 2007 00:37:23 +0200
Thilo Bangert <bangert@gentoo.org> wrote:

> > > It isn't different.  That's the problem.  If you have two packages
> > > with the same name, you have the same problem.
> >
> > On that note I would hope the vim/vi peeps would rename.
> > app-vim/ant
> 
> and app-vim/sudo

That's getting the axe in a few weeks.

> > IMHO app-vim/ant should really be app-vim/vim-ant or something other
> > than just ant.
> 
> or app-vim/sudo-syntax and app-vim/ant-syntax as there already are a 
> number of ebuilds following that scheme...

Well, sudo and ant aren't syntax plugins, so that wouldn't make any
sense. Also, we're keeping the same names that the upstream script
writers use, just as we do everywhere else in Gentoo. The whole point
of having category names is so that we can have two packages w/ the
same name and not have issues.

-- 
Mike Kelly
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Georgi Georgiev]

Quoting Mike Kelly <pioto@gentoo.org>:

> On Thu, 17 May 2007 00:37:23 +0200
> Thilo Bangert <bangert@gentoo.org> wrote:
  ...

>> > IMHO app-vim/ant should really be app-vim/vim-ant or something other
>> > than just ant.
>>
>> or app-vim/sudo-syntax and app-vim/ant-syntax as there already are a
>> number of ebuilds following that scheme...
>
> Well, sudo and ant aren't syntax plugins, so that wouldn't make any
> sense. Also, we're keeping the same names that the upstream script
> writers use, just as we do everywhere else in Gentoo. The whole point
> of having category names is so that we can have two packages w/ the
> same name and not have issues.

All this is so familiar... where have I heard that before?

http://thread.gmane.org/gmane.linux.gentoo.devel/27770/focus=27838

The thread is pretty long, but some of the issues of the current  
thread have been covered in length there.

-- 
Georgi

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

--
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 279 bytes --]

On Wed, 16 May 2007 17:23:58 -0400
"William L. Thomson Jr." <wltjr@gentoo.org> wrote:
> On that note I would hope the vim/vi peeps would rename.
> app-vim/ant

Policy says to go with upstream's naming. This is, after all, why we
have categories.

-- 
Ciaran McCreesh


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name

[Ulrich Mueller]

>>>>> On Thu, 17 May 2007, Jakub Moc wrote:

>>> On that note I would hope the vim/vi peeps would rename.
>>> app-vim/ant
>> 
>> and app-vim/sudo

> and app-xemacs/emerge, grrrr

But (X)Emacs was first. It had "emerge" in 1992 already. :P

   ;; LCD Archive Entry:
   ;; emerge|Dale R. Worley|drw@math.mit.edu
   ;; |File merge
   ;; |92-12-11|version 5 gamma|~/packages/emerge.el.Z
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name

[Raúl Porcel]

Blame drobbins :P

-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[William L. Thomson Jr.]

[-- Attachment #1: Type: text/plain, Size: 511 bytes --]

On Thu, 2007-05-17 at 08:49 +0100, Ciaran McCreesh wrote:
> On Wed, 16 May 2007 17:23:58 -0400
> "William L. Thomson Jr." <wltjr@gentoo.org> wrote:
> > On that note I would hope the vim/vi peeps would rename.
> > app-vim/ant
> 
> Policy says to go with upstream's naming. This is, after all, why we
> have categories.

Sure, and along those lines upstream seems to call it ant_menu or
ant_menu.vim :)

http://www.vim.org/scripts/script.php?script_id=155

-- 
William L. Thomson Jr.
Gentoo/Java

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 409 bytes --]

On Thu, 17 May 2007 10:59:09 -0400
"William L. Thomson Jr." <wltjr@gentoo.org> wrote:
> > Policy says to go with upstream's naming. This is, after all, why we
> > have categories.
> 
> Sure, and along those lines upstream seems to call it ant_menu or
> ant_menu.vim :)
> 
> http://www.vim.org/scripts/script.php?script_id=155

Not for the versions in the tree they don't.

-- 
Ciaran McCreesh


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[William L. Thomson Jr.]

[-- Attachment #1: Type: text/plain, Size: 816 bytes --]

On Thu, 2007-05-17 at 16:03 +0100, Ciaran McCreesh wrote:
> On Thu, 17 May 2007 10:59:09 -0400
> "William L. Thomson Jr." <wltjr@gentoo.org> wrote:
> > > Policy says to go with upstream's naming. This is, after all, why we
> > > have categories.
> > 
> > Sure, and along those lines upstream seems to call it ant_menu or
> > ant_menu.vim :)
> > 
> > http://www.vim.org/scripts/script.php?script_id=155
> 
> Not for the versions in the tree they don't.

You mean 0.5.3 released 2003-12-28. IMHO that borderlines a stale
package or one that should be punted. Or at min bumped since there has
been two releases since then. With the latest one being 0.5.5
2006-11-19.

So is that app really used? Does anyone care about? Sure looks like not
in both cases.

-- 
William L. Thomson Jr.
Gentoo/Java

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Ciaran McCreesh]

[-- Attachment #1: Type: text/plain, Size: 1089 bytes --]

On Thu, 17 May 2007 12:48:11 -0400
"William L. Thomson Jr." <wltjr@gentoo.org> wrote:
> > > http://www.vim.org/scripts/script.php?script_id=155
> > 
> > Not for the versions in the tree they don't.
> 
> You mean 0.5.3 released 2003-12-28. IMHO that borderlines a stale
> package or one that should be punted. Or at min bumped since there has
> been two releases since then. With the latest one being 0.5.5
> 2006-11-19.

Eh, perhaps the vim herd could take a look at that -- maybe someone
should file a bug about it. Although perhaps the version in the tree
works just fine, which is why no-one's noticed that there's an update.

> So is that app really used? Does anyone care about? Sure looks like
> not in both cases.

'Twas added to the tree at user request. Given that Java's basically a
dead language and only being used for legacy applications now, it's
probably not necessary any more. But then, keeping at least one of the
colliding app-vim/ packages around is probably a good idea if only to
remind people why categories exist...

-- 
Ciaran McCreesh


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Josh Sled]

On Thu, May 17, 2007 12:53 pm, Ciaran McCreesh wrote:
> 'Twas added to the tree at user request. Given that Java's basically a
> dead language and only being used for legacy applications now, it's

I'm having a hard time trying to figure out how you justify calling Java a
dead language.

-- 
...jsled


-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Vlastimil Babka]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Josh Sled wrote:
> On Thu, May 17, 2007 12:53 pm, Ciaran McCreesh wrote:
>> 'Twas added to the tree at user request. Given that Java's basically a
>> dead language and only being used for legacy applications now, it's
> 
> I'm having a hard time trying to figure out how you justify calling Java a
> dead language.
> 

It's not C++ nor Ruby.

Anyway, offtopic :)

- --
Vlastimil Babka (Caster)
Gentoo/Java
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGTI0ztbrAj05h3oQRAiSHAJ9/3FzS1r+rnfxTXsqN0aOJh3fveQCfW2TV
4JREonUvieuIMHkFMcAFI5o=
=JB4J
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Packages with same name was -> Conversion of Emacs virtual packages

[Chris Gianelloni]

[-- Attachment #1: Type: text/plain, Size: 763 bytes --]

On Thu, 2007-05-17 at 13:07 -0400, Josh Sled wrote:
> On Thu, May 17, 2007 12:53 pm, Ciaran McCreesh wrote:
> > 'Twas added to the tree at user request. Given that Java's basically a
> > dead language and only being used for legacy applications now, it's
> 
> I'm having a hard time trying to figure out how you justify calling Java a
> dead language.

It's called baiting.  He's purposefully calling Java dead in a
conversation with someone from the Java team in an attempt to bring on
an emotional response rather than technical.  In case you're wondering,
you feel for it.  *grin*

-- 
Chris Gianelloni
Release Engineering Strategic Lead
Alpha/AMD64/x86 Architecture Teams
Games Developer/Council Member/Foundation Trustee
Gentoo Foundation

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

[gentoo-dev] Re: Marking virtuals stable

[Christian Faulhammer]

[-- Attachment #1: Type: text/plain, Size: 411 bytes --]

Petteri Räty <betelgeuse@gentoo.org>:

> The month is not set in stone. About who marks them, it's probably
> best to get the opinion of the arch teams. I don't think they will
> object to normal developers marking them. Arch teams: what do you
> think?

 Speaking for x86/amd64 and Emacs...I am ok with that. :)

V-Li

-- 
http://www.gentoo.org/
http://www.faulhammer.org/
http://www.gnupg.org/

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: [gentoo-dev] Re: Marking virtuals stable

[Ulrich Mueller]

>>>>> On Wed, 23 May 2007, Christian Faulhammer wrote:

> Petteri Räty <betelgeuse@gentoo.org>:
>> Ulrich Mueller kirjoitti:
>>> The point of my question was more if the usual rules apply, i.e.:
>>> keywording and stabilising only by arch teams; wait one month
>>> before the package can go stable.

>> The month is not set in stone. About who marks them, it's probably
>> best to get the opinion of the arch teams. I don't think they will
>> object to normal developers marking them. Arch teams: what do you
>> think?

>  Speaking for x86/amd64 and Emacs...I am ok with that. :)

So, only this reply.

May I conclude that nobody objects to the above?

Ulrich
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Marking virtuals stable

[Michael Cummings]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Ulrich Mueller wrote:
> So, only this reply.
> 
> May I conclude that nobody objects to the above?
> 
> Ulrich

Wearing only my perl team hat, it would seem to lowly me that if a
virtual points to packages foo and bar, and both foo and bar were tested
and marked stable by the arch's previously, that its silly to then wait
for them to mark the virtual stable as well, since at least in my
perception the only function of that virtual is to say use one of these
packages - which have already been marked stable.

/me hopes some arch brains step in, like weeve in particular, who is
usually far more eloquent at defending an arch's position

- --

- -----o()o----------------------------------------------
Michael Cummings   |    #gentoo-dev, #gentoo-perl
Gentoo Perl Dev    |    on irc.freenode.net
Gentoo/SPARC
Gentoo/AMD64
GPG: 0543 6FA3 5F82 3A76 3BF7  8323 AB5C ED4E 9E7F 4E2E
- -----o()o----------------------------------------------
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.4 (GNU/Linux)

iD8DBQFGXpVDq1ztTp5/Ti4RAn3fAJ93fLU/G2QOB5p6jeGQst4lnyXEWgCfeuPg
kQDUKObUWKYcIWndm6zGm6U=
=beGq
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Marking virtuals stable

[Jorge Manuel B. S. Vicetto]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Michael Cummings wrote:
> Ulrich Mueller wrote:
>> So, only this reply.
> 
>> May I conclude that nobody objects to the above?
> 
>> Ulrich
> 
> Wearing only my perl team hat, it would seem to lowly me that if a
> virtual points to packages foo and bar, and both foo and bar were tested
> and marked stable by the arch's previously, that its silly to then wait
> for them to mark the virtual stable as well, since at least in my
> perception the only function of that virtual is to say use one of these
> packages - which have already been marked stable.
> 
> /me hopes some arch brains step in, like weeve in particular, who is
> usually far more eloquent at defending an arch's position
> 

Michael,

for a virtual pointing to packages foo and bar, only one of them needs
to be stable before the virtual can be marked as stable, right?
So your above comment should read "if a virtual points to packages foo
and bar, and [either foo or bar was] tested and marked stable by the
arch's previously, that its silly to then wait for them to mark the
virtual stable as well", right?

- --
Regards,

Jorge Vicetto (jmbsvicetto) - jmbsvicetto at gentoo dot org
Gentoo- forums / Userrel / Proctors
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.4 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGXs9kcAWygvVEyAIRAnoUAJ4iQc4qhyn8Yehuvs2w5AHknU2crgCfVvCx
PWibZvOya/nyGDZDi72rwLs=
=YAH2
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Marking virtuals stable

[Jeroen Roovers]

On Thu, 31 May 2007 05:28:35 -0400
Michael Cummings <mcummings@gentoo.org> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Ulrich Mueller wrote:
> > So, only this reply.
> > 
> > May I conclude that nobody objects to the above?

I think marking virtuals is OK. If you cannot mark them because some
DEPENDs have not been marked (stable) for some arch, you couldn't do it
anyway (while at the same time getting past repoman) and would have to
file a keywording bug. I think I should probably review this stance at
the earliest when virtuals threaten to become more than containers for
DEPENDs.

> Wearing only my perl team hat, it would seem to lowly me that if a
> virtual points to packages foo and bar, and both foo and bar were
> tested and marked stable by the arch's previously, that its silly to
> then wait for them to mark the virtual stable as well, since at least
> in my perception the only function of that virtual is to say use one
> of these packages - which have already been marked stable.

I have seen many Perl virtuals go straight to stable and haven't ever
experienced any adverse effects. :)

> /me hopes some arch brains step in, like weeve in particular, who is
> usually far more eloquent at defending an arch's position

Oh sorry. :)


Kind regards,
     JeR
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Marking virtuals stable

[Graham Murray]

"Jorge Manuel B. S. Vicetto" <jmbsvicetto@gentoo.org> writes:

> for a virtual pointing to packages foo and bar, only one of them needs
> to be stable before the virtual can be marked as stable, right?
> So your above comment should read "if a virtual points to packages foo
> and bar, and [either foo or bar was] tested and marked stable by the
> arch's previously, that its silly to then wait for them to mark the
> virtual stable as well", right?

At first sight what you say sounds right, but further thought shows that
both foo and bar would have to be marked stable before the virtual could
be.

Take the instance that the appropriate version of foo is marked stable
but that for bar is still in ~arch. If someone has foo installed then
upgrading the virtual will pull in the new (stable) foo and all is
well. However if someone else has bar installed but not foo, then the
upgrade to the virtual will not cause bar to be upgraded (as it is still
masked ~arch) but will cause the upgraded foo to be installed (as a new
package) to satisfy the virtual. Or have I (as a mere user)
misunderstood the concepts of virtuals?
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Re: Marking virtuals stable

[Michael Cummings]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Jeroen Roovers wrote:
> I have seen many Perl virtuals go straight to stable and haven't ever
> experienced any adverse effects. :)

well, that's the idea :) But like I think it was Graham said in another
subthread of this, perl team's virtuals only go to stable if both (in
our case, there are usually only 2 possible sources for fulfillment)
deps are already stable, so really all you're doing is updating a
pointer reference. um. can i be more confusing? let's find out!

> 
>> /me hopes some arch brains step in, like weeve in particular, who is
>> usually far more eloquent at defending an arch's position
> 
> Oh sorry. :)

bah - that was me failing to finish the sentence that in my head
concluded as "than me." That and softserve machines aside, weeve and
gustovoz are pretty (in a positive sense) vocal about arch related
items, so they were the ones that came to mind when i threw in the
comment about arch's commenting.

Double bah. I think i've dug a good hole here. Let me get in it.

- --

- -----o()o----------------------------------------------
Michael Cummings   |    #gentoo-dev, #gentoo-perl
Gentoo Perl Dev    |    on irc.freenode.net
Gentoo/SPARC
Gentoo/AMD64
GPG: 0543 6FA3 5F82 3A76 3BF7  8323 AB5C ED4E 9E7F 4E2E
- -----o()o----------------------------------------------
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.4 (GNU/Linux)

iD8DBQFGXvLHq1ztTp5/Ti4RAtf/AKCLWpCWcsD+m8njHSdfWltt+owQ1gCfaV2P
S4o2MhHNJFs2gv2oN6yms/A=
=sj0h
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list