[oe] [meta-python][PATCH 3/4] python-msgpack: add recipe

[Bruce Ashfield]

On Thu, Feb 19, 2015 at 12:21 PM, Martin Jansa <martin.jansa at gmail.com> wrote:
> On Thu, Feb 19, 2015 at 12:09:09PM -0500, Bruce Ashfield wrote:
>> On Wed, Feb 18, 2015 at 6:03 PM, Ben Shelton <ben.shelton at ni.com> wrote:
>> > python-msgpack implements the MessagePack binary serialization format
>> > for Python.
>> >
>> > Copied from commit 91ed2353847d4121e3dd8ad10ca0c24194ac9765 in
>> > meta-cloud-services/meta-openstack and updated to version 0.4.5.
>>
>> Can you elaborate on what is the use case for these recipes being cherry
>> picked ?
>
> It was discussed on IRC yesterday, the use case is simple. Some project
> needs couple python-* recipes and they happen to be available in
> meta-openstack now.. so instead of adding dependency on whole
> meta-openstack which has unwanted side-effects and a lot of stuff they
> don't care about they move 4 recipes to meta-python.

Is there an IRC log someone can point me at ? IRC is an incredibly bad
medium for looking up an old discussion. Sorry for not seeing it!

Can someone also elaborate on 'unwanted side effects' ? Sounds like
a bug, and something that can be fixed. Yanking recipes, not emailing
the originators and citing bugs/issues without details .. really isn't a
great way to foster collaboration and generally improve the ecosystem.

Again .. that's all that I'm saying. There's no harm in taking the extra
time to document things a little bit, and offer explanations.

That's a valid reason for a move, but what I'm really asking for is
that to be in
the commit messages, and a description of the user. Having the floating
set of recipes without a way to know how they are used and tested is
going to cause a giant mess.

>
>> If they are simply being moved "because they are python' .. that is annoying,
>> and causes pain for dependent layers that have specific version dependencies.
>
> They are moved before different projects need them and meta-python is
> best project-agnostic layer for python recipes.

No debate there. But again, we should simply log the reason in the original
commit. There's no reason to lose that information to IRC and everyone's
fading memory.

>
>> I'm going to have to pin nearly all versions in the openstack layers and keep
>> duplicate recipes around, so since I'm suffering that overhead, it would be nice
>> if we could elaborate on the technical gain from doing this.
>
> You don't need to pin it if your layer has higher priority.

Yah. Paul mentioned that. I'll do that for now, since there will be
times that the
openstack variants are newer than the ones in meta-python, and that will
handle both use cases.

Cheers,

Bruce

>
>> > Signed-off-by: Ben Shelton <ben.shelton at ni.com>
>> > ---
>> >  .../recipes-devtools/python/python-msgpack_0.4.5.bb     | 17 +++++++++++++++++
>> >  1 file changed, 17 insertions(+)
>> >  create mode 100644 meta-python/recipes-devtools/python/python-msgpack_0.4.5.bb
>> >
>> > diff --git a/meta-python/recipes-devtools/python/python-msgpack_0.4.5.bb b/meta-python/recipes-devtools/python/python-msgpack_0.4.5.bb
>> > new file mode 100644
>> > index 0000000..206b03d
>> > --- /dev/null
>> > +++ b/meta-python/recipes-devtools/python/python-msgpack_0.4.5.bb
>> > @@ -0,0 +1,17 @@
>> > +DESCRIPTION = "MessagePack (de)serializer"
>> > +HOMEPAGE = "https://pypi.python.org/pypi/msgpack-python/"
>> > +SECTION = "devel/python"
>> > +LICENSE = "Apache-2.0"
>> > +LIC_FILES_CHKSUM = "file://COPYING;md5=cd9523181d9d4fbf7ffca52eaa2a5751"
>> > +
>> > +PR = "r0"
>> > +SRCNAME = "msgpack-python"
>> > +
>> > +SRC_URI = "https://pypi.python.org/packages/source/m/${SRCNAME}/${SRCNAME}-${PV}.tar.gz"
>> > +
>> > +SRC_URI[md5sum] = "3b82bc542d5599896695512e7c32f42d"
>> > +SRC_URI[sha256sum] = "a07cd6615a6bf38cfa2f010b121c7e77b74a3e7b971ef3e475c3d33308014cbb"
>> > +
>> > +S = "${WORKDIR}/${SRCNAME}-${PV}"
>> > +
>> > +inherit setuptools
>> > --
>> > 2.3.0
>> >
>> > --
>> > _______________________________________________
>> > Openembedded-devel mailing list
>> > Openembedded-devel at lists.openembedded.org
>> > http://lists.openembedded.org/mailman/listinfo/openembedded-devel
>>
>>
>>
>> --
>> "Thou shalt not follow the NULL pointer, for chaos and madness await
>> thee at its end"
>> --
>> _______________________________________________
>> Openembedded-devel mailing list
>> Openembedded-devel at lists.openembedded.org
>> http://lists.openembedded.org/mailman/listinfo/openembedded-devel
>
> --
> Martin 'JaMa' Jansa     jabber: Martin.Jansa at gmail.com
>
> --
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel at lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-devel
>



-- 
"Thou shalt not follow the NULL pointer, for chaos and madness await
thee at its end"