Hi Guys,

I'm playing with github today. And now I'm a bit confused if this is
really the right way.

What I did:
$ git clone [email protected]:redsleeve-linux/el7.git
-> this created an empty repository

in here I cloned a couple of upstream repos:
$ cd el7
$ git clone https://git.centos.org/git/rpms/atlas.git
$ cd atlas
$ git checkout c7

now I added these artifacts to our repo:
$ cd ..
$ git add atlas/SPECS atlas/SOURCES
$ git commit -m "something"
$ git push

Next I went for committing my patch:
$ cd atlas
$ git checkout -b redsleeve
$ patch -p1 < /path/to/patch_file
$ git commit -a -m "patched"
$ git format-patch c7
$ cd ..
$ git add altas/0001-patched.patch
$ git commit -m "patched atlas"
$ git push

so far so good, I now have a tree with upstream git repos and my patches
to that. I can also go to the sub directories (atlas in this case) and
do get_sources.sh and do rpmbuild to make a SRPM.

However I'm a bit at a loss if this makes sense to anybody else. To try
that, I came up with this:
$ git clone [email protected]:redsleeve-linux/el7.git
$ cd el7
$ rm -rf atlas
$ git clone https://git.centos.org/git/rpms/atlas.git
$ cd atlas
$ git checkout c7
$ cd ..
$ git checkout -- atlas

This has (just for one package) the same layout as I have.

This all feels a bit awkward. Does anybody have a better Idea on how to
(easily) work with this?

Jacco



On 18-11-16 11:38, Bjarne Saltbæk wrote:
> Hi Gordan/Jacco.
>
>
>
> On 13-11-2016 18:19, Gordan Bobic wrote:
>> On 13/11/16 16:55, Bjarne Saltbæk wrote:
>>> Hi Jacco.
>>>
>>> What is your ideas about Github?
>>>
>>> As you can see I tried putting my RSEL spec files into Github, but I
>>> think it is going to be a big mess. You will need to download all spec
>>> files + all patches + all changes each time you want to build/work on
>>> just one package.
>> This is a fair point, but there is a practical issue, specifical that 
>> of creating 2000 or so repositories.
>>
>> In reality, even back in the day of RSEL 6.0 only about 120 or so spec 
>> file changes were needed, and only a fraction of those needed 
>> additional patches. This is a reasonably practical amount to keep in a 
>> single repository, or at least more practical than having a separate 
>> repository for every package that needs changes. Additionally, it 
>> makes it not obvious how to deal with cases where a package used to 
>> need a patch but no longer does. With a single repository we simply 
>> remove the spec file and the patch. In the case of one repository per 
>> package, do we delete the whole repository?
> I have scripted an export of my patches/changes from my git server and 
> uploaded them to github. It is mostly correct but there are some 
> Work-In-Progress patches which are not completed. Also there are some 
> script extract errors (nobody is perfect :D).
>
> BR,
> Bjarne
>
> _______________________________________________
> users mailing list
> [email protected]
> https://lists.redsleeve.org/mailman/listinfo/users



_______________________________________________
users mailing list
[email protected]
https://lists.redsleeve.org/mailman/listinfo/users

Reply via email to