One thought I had during the day would be to write and submit aspec file with no %files, no prep, build, etc.
Whose sole purpose is to generate a package with a single dependency to point to the new package. Followed by a submission, vote/approval and such. At least until the real obsoletion stuff is ready. Opinions? Comments? +1/-1s? Other ideas? Doug. Doug Leavitt wrote: > Thanks Shawn, and my apologies, I should have been more clear. > I was asking more w.r.t. /contrib & jucr process. I am aware > of the current physical steps TBD. > > We have in /contrib at least 1 case where a project delivered > into /contrib in December has now been superceded by the same > deliverable in /release. > > We need to instigate the sw-porters /contrib process to > republish the said package(a) such that python-pysqlite2 > in /contrib directs people to install SUNWpysqlite > instead. > > So what is the jucr and/or /contrib process to do this? > The web page: > http://opensolaris.org/os/community/sw-porters/contributing/jucrprocess/ > only says: > 'In Work' > aka yet to be written. > > > doug. > > Shawn Walker wrote: >> Doug Leavitt wrote: >>> I just found out that snv_111 delivered python sqlite as: >>> SUNWpysqlite >>> >>> The package also currently exists in /contrib as: >>> python-pysqlite2. It was delivered there last December. >>> >>> What procedure do we use to obsolete a package that >>> has been promoted (re-ported) to a release repo? >>> >>> We now have at least 1 test case. >> >> Because pkg(5) does not yet support obsoletion or renames directly, so >> the best way to obsolete a package is to publish a new version of the >> old package (python-pysqlite2) that depends on the new package >> (SUNWpysqlite) but that is otherwise empty (you may wish to keep the >> description and summary). >> >> That will cause clients upgrading to the new version to get the new >> package, and remove all of the old package's files. >> >> In the next few months, official rename/obsoletion support will be added. >> >> Cheers, > _______________________________________________ > sw-porters-discuss mailing list > sw-porters-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/sw-porters-discuss
