[gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Alec Warner]

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

As suggested by SpanKY in the bug report itself but never done until now
I'll gladly start up the discussion.

The bug report[1] is a request from the embedded team to change how
DEPEND and REDEPEND are assigned in an ebuild.  At present the
documentation[2] states that RDEPEND if not specified will be set to
DEPEND.  Apparently the embedded team has issue with that, mostly that
it pollutes exactly what RDEPEND is there for, making their ROOT much
larger than it has to be.

The actual fix to the bug is a minor one, a small check to Repoman to
make sure ebuilds have both DEPEND and RDEPEND set; and to warn if they
are not set.

However the use of DEPEND and RDEPEND in the manner that they are
requesting is a large change and thus was brought here for discussion.

[1]http://bugs.gentoo.org/show_bug.cgi?id=79337
[2]http://www.gentoo.org/proj/en/devrel/handbook/handbook.xml?part=2&chap=1#doc_chap5
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org
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=lwur
-----END PGP SIGNATURE-----
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Mike Frysinger]

On Sunday 29 May 2005 01:48 am, Alec Warner wrote:
> The actual fix to the bug is a minor one, a small check to Repoman to
> make sure ebuilds have both DEPEND and RDEPEND set; and to warn if they
> are not set.
>
> However the use of DEPEND and RDEPEND in the manner that they are
> requesting is a large change and thus was brought here for discussion.

i'm against this ... the current behavior is the logical default imho
-mike
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Brian Harring]

[-- Attachment #1: Type: text/plain, Size: 945 bytes --]

On Sun, May 29, 2005 at 02:32:45AM -0400, Mike Frysinger wrote:
> On Sunday 29 May 2005 01:48 am, Alec Warner wrote:
> > The actual fix to the bug is a minor one, a small check to Repoman to
> > make sure ebuilds have both DEPEND and RDEPEND set; and to warn if they
> > are not set.
> >
> > However the use of DEPEND and RDEPEND in the manner that they are
> > requesting is a large change and thus was brought here for discussion.
> 
> i'm against this ... the current behavior is the logical default imho
What of eclasses?  Bug 58819 comes to mind...

Ebuilds get the RDEPEND="${REPEND-${DEPEND}}" default, but 
eclasses don't.  Would rather there was a standard the whole way 
through personally, even if eclasses are a bit more complex 
in enforcing the RDEPEND="${RDEPEND-${DEPEND}}" behaviour.  That 
said, I'm not much for the auto-setting of RDEPEND :)

So... extend the ebuild behaviour to eclasses, or...
~harring

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

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Mike Frysinger]

On Monday 30 May 2005 08:51 pm, Brian Harring wrote:
> On Sun, May 29, 2005 at 02:32:45AM -0400, Mike Frysinger wrote:
> > On Sunday 29 May 2005 01:48 am, Alec Warner wrote:
> > > The actual fix to the bug is a minor one, a small check to Repoman to
> > > make sure ebuilds have both DEPEND and RDEPEND set; and to warn if they
> > > are not set.
> > >
> > > However the use of DEPEND and RDEPEND in the manner that they are
> > > requesting is a large change and thus was brought here for discussion.
> >
> > i'm against this ... the current behavior is the logical default imho
>
> What of eclasses?  Bug 58819 comes to mind...

i'd be for 'fixing' the eclass behavior ... making it work the same way as 
ebuilds
-mike
-- 
gentoo-dev@gentoo.org mailing list

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Jason Stubbs]

[-- Attachment #1: Type: text/plain, Size: 1646 bytes --]

On Tuesday 31 May 2005 09:55, Mike Frysinger wrote:
> On Monday 30 May 2005 08:51 pm, Brian Harring wrote:
> > On Sun, May 29, 2005 at 02:32:45AM -0400, Mike Frysinger wrote:
> > > On Sunday 29 May 2005 01:48 am, Alec Warner wrote:
> > > > The actual fix to the bug is a minor one, a small check to Repoman to
> > > > make sure ebuilds have both DEPEND and RDEPEND set; and to warn if
> > > > they are not set.
> > > >
> > > > However the use of DEPEND and RDEPEND in the manner that they are
> > > > requesting is a large change and thus was brought here for
> > > > discussion.
> > >
> > > i'm against this ... the current behavior is the logical default imho
> >
> > What of eclasses?  Bug 58819 comes to mind...
>
> i'd be for 'fixing' the eclass behavior ... making it work the same way as
> ebuilds

I'd be for having DEPEND required to be set manually. ;)

But seeing that it would be a huge task and there aren't the resources or 
support to do it at this time, and as I feel standards (even when they're 
wrong :P ) are most important of all, I'd agree to making RDEPEND default to 
DEPEND for eclasses too.

Perhaps, further down the track we'd be able to work out something with the 
build farm thingy; check for linkage and warn if things specified in RDEPEND 
aren't linked against and build up a whitelist from it... Perhaps repoman 
will become smart enough to detect exactly which RDEPENDs are being defaulted 
to what and where and then provide a warning and a resolution...

Anyway, not much point in increasing an already overflowing workload at this 
point in time.

Regards,
Jason Stubbs

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

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Georgi Georgiev]

[-- Attachment #1: Type: text/plain, Size: 820 bytes --]

maillog: 01/06/2005-23:25:00(+0900): Jason Stubbs types
>...
> Perhaps, further down the track we'd be able to work out something with the 
> build farm thingy; check for linkage and warn if things specified in RDEPEND 
> aren't linked against and build up a whitelist from it... Perhaps repoman 
> will become smart enough to detect exactly which RDEPENDs are being defaulted 
> to what and where and then provide a warning and a resolution...

I don't get it. Linking is not the only reason for a package to RDEPEND
on another one.

Or were you talking about something completely different?

-- 
*)   Georgi Georgiev   *) "Problem solving under linux has never       *)
(*    chutz@gg3.net    (* been the circus that it is under AIX." (By   (*
*)  +81(90)2877-8845   *) Pete Ehlke in comp.unix.aix)                 *)

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

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Jason Stubbs]

[-- Attachment #1: Type: text/plain, Size: 738 bytes --]

On Thursday 02 June 2005 00:21, Georgi Georgiev wrote:
> maillog: 01/06/2005-23:25:00(+0900): Jason Stubbs types
>
> >...
> > Perhaps, further down the track we'd be able to work out something with
> > the build farm thingy; check for linkage and warn if things specified in
> > RDEPEND aren't linked against and build up a whitelist from it... Perhaps
> > repoman will become smart enough to detect exactly which RDEPENDs are
> > being defaulted to what and where and then provide a warning and a
> > resolution...
>
> I don't get it. Linking is not the only reason for a package to RDEPEND
> on another one.
>
> Or were you talking about something completely different?

Nope. That's why I mentioned a whitelist.

Regards,
Jason Stubbs

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

Re: [gentoo-dev] Bugzilla Bug 79337 make repoman complain if DEPEND and RDEPEND are not set.

[Brian Harring]

[-- Attachment #1: Type: text/plain, Size: 1603 bytes --]

On Wed, Jun 01, 2005 at 11:25:00PM +0900, Jason Stubbs wrote:
> I'd be for having RDEPEND required to be set manually. ;)
As would I, actually...

Granted it's a useful convenience, but it also makes nailing the deps 
down much harder.  Personally down the line, I'd like to see packages 
that require compilation actually stating the virtual/gcc dep, and 
RDEPEND="${RDEPEND-${DEPEND}}" kind of screws with that.


> But seeing that it would be a huge task and there aren't the resources or 
> support to do it at this time,
Question is which is preferable.

Changing half the tree is a pita granted and not something to be done 
drop of the hat, but that doesn't mean can't decide to change how 
things are done, and work towards it gradually.

Writing out a helper script wouldn't be too hard, nor would a script 
that does the actual changes- just lift it from ebuild.sh (RDEPEND and 
E_RDEPEND are kept seperate till post sourcing).

> Anyway, not much point in increasing an already overflowing workload at this 
> point in time.

I'm mainly interested if people agree with the convenience feature 
being worthwhile to keep; I don't think so, but I also occasionally 
have strange ideas :)

Again, note, if people did agree rdepend=${rdepend-${depend}} was 
evil, it's not a massive treewide commit to change it; just would 
require gradually adding explicit rdepend into ebuilds till it's 
done, then ixnaying the convenience feature.  Same type of changes 
gradually rolled out for use and has's verbosity (making them no 
longer echo the result)...
~harring

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