[oe] [RFC] duplicate recipes

Frans Meulenbroeks fransmeulenbroeks at gmail.com
Wed Mar 10 15:44:48 UTC 2010


Time for a summary I guess:

original proposal from me:
Probably the best scenario is to remove all versions which
- are not the latest version
- are not  pinned with  PREFERRED_VERSION  in conf/distro/*
- do not have the largest DEFAULT_PREFERENCE

opinions:
Koen: see glib proposal
Mickey: move out of sight or own repo would have my support
Philip: no as we do not know who is using the older versions
Graeme: expressed that there are more important tasks to be done

and in the glib-2.0 cleanup thread:
koen: suggests to clean up after improving the recipes and prefers to
move recipe to obsolete instead of delete it
mickey: move out of sight
graham: comes with more offending case, no opinion on the proposal given
Marcin (on the gcc stuff: "If someone wants one then we have SCM for it."
[note from FM: not sure if this a specific remark on the gcc debian
patches or a generic statement]
khem: moving to recipes/obsolete may not be so bad
Denys: +1 (we already have recipes/obsolete in BBMASK, reduces parse
time, git-mv does preserve history)

and in the recipes/${distro} thread there were also a few comments on this.
Most of these are from the same persons as mentioned above. Only
addition that is related to this thread is:
otavio: in favour of removing (faster parsing)

Seems most people are in favour of remving out of sight:

Counting head's for that variant (assuming that people wo stated that
removal is the best choice are not against moving to an obsolete dir).
in favour: frans, mickey, khem, denys, otavio and perhaps marcin
against: koen, philip
I've ocunted graeme as neutral and graham as no opinion given.

If you feel I misrepresented your position or if I missed your
position, feel free to follow up.

Best regards, frans

(ps: i expect that this concerns some 2000 recipes (out of 8000))




More information about the Openembedded-devel mailing list