I am a faithfull reader of GWN.
I have had my lib-bonobo bonobobobobo me continually. (it still does becuase
something out there depends on the old one, but that is besides the point)
and my apache stopped working because of glep11 related occurences.
granted, glep11 will keep things working in the future, but my php and
squirrelmail have not caught up that major change that happened to apache... on
roughly the day the last GWN came out.
So I ask that major changes be announced extra early.. even before going in
unstable, so people have time to prepare.
Maybe this bug depends on there only being 2 use flags unstable and stable...
and my missusing ~x86 as my everyday machine... but that is the kind of testing
I'm willing to do. I'm willing to submit an occasional bug report.. but not all
Maybe there is a way to say "temporarily unstable" meaning that I'm willing to
go to ~x86, and I dont want to downgrade back to stable on next emerge, but I'll
wait for stable to catch up to me again.....
anyway, that would be cool... harness the masses while keeping them happy.
Steps to Reproduce:
This surely is a candidate for policy.
Marking this as "LATER" because... well... the whole point of this bug just screams later all over the place :) So we'll try to inform the users later when big changes happen.
Ideas for implementation...
(reopening because I've been bitten by system changes again, and I have new ideas)
the portage snapshot could contain a list of fixes/changes in the line up. a diff of the incomming one vs the old one could be emailed to root.
So, i'm saying include some documentation in with the portage hierarchy... that contains future information. Each package maintainer should be responsible for noting planned changes..
marking RESOLVED LATER again won't bug me, but this seems like a reasonable method to get it going...
Security updates could also be emailed, and notes given by the ebuild could also be emailed (rather than have to be grepped from the portage logs)
I don't think portage is suitable for informing yet. There are patches going in that record the einfo/ewarning's that get displayed when you update a package (which puts considerable amount of interesting information on your screen) so that you can read all this after emerge's.
I'm marking as later again (for the GWN stuff part :) That doesn't mean we don't look at this bug anymore, comments always propagate to us.
Closing this bug as per Josh's request in gentoo-doc's ML.