Arek (James Potts) wrote:
> Donnie Berkholz wrote:
>>> >=virtual/x11-7 is hiding breakage in ebuilds that are not ported for
>>> modular X.

>> I couldn't agree more, but I was forced to add this rather than allow
>> unported ebuilds to break.

> Hmmm...Looks to me like it would be a great idea to fix the unported
> ebuilds.  Would it be possible to mark virtual/x11-7 as deprecated
> (using enotice/ewarn or similar), in order to get people to port any
> build relying on it to modular X?
> The way I see it, once virtual/x11-7 has been deprecated for a while (6
> months to a year) and most popular packages have been ported to modular
> X, virtual/x11-7 and any packages still relying on it could be given
> Last Rites.

Hmm, I don't think so... There's been a plenty of time to do this when
modular X has been package.masked, the remaining unported stuff didn't
get much further even after it's been unmasked. There's been a
(debatable) repoman check, which has been too annoying so devs nuked it
for themselves, now it's non-fatal warning again (which is mostly being

Soooo - I'd pretty much say until the real breakage is *visible* and
users start to scream - not much will change. Making it visible could
also help us differentiate between used and used stuff. If there's
something unported and you get no bug, then probably noone uses the
thing, nothing depends on it and it can be punted from the tree.

On a side note, this virtual also hides potential bugs in ebuilds that
already have been ported, you can miss dependencies there if you have
already them emerged b/c of the virtual.

Best regards,

 Jakub Moc
 GPG signature:
 Primary key fingerprint: D2D7 933C 9BA1 C95B 2C95  B30F 8717 D5FD CEBA 3D9E

 ... still no signature   ;)

Attachment: signature.asc
Description: OpenPGP digital signature

Reply via email to