[Sugar-devel] GCI and our build environment

Walter Bender walter.bender at gmail.com
Sat Nov 12 16:10:47 EST 2016


+1 to Quozl's suggestion to drop the actual collaboration wrapper code into
the repo. We can consider adding gitsubmodule  support to 0.112.

-walter

On Sat, Nov 12, 2016 at 4:01 PM, Utkarsh Tiwari <iamutkarshtiwari at gmail.com>
wrote:

>
> ---------- Forwarded message ----------
> From: James Cameron <quozl at laptop.org>
> Date: Tue, Nov 8, 2016 at 8:58 AM
> Subject: Re: [Sugar-devel] Sugar-Build
> To: "Nguyen, Lam" <nguyenla at dickinson.edu>
>
>
> G'day,
>
> Browse is a critical part of Sugar, so to exclude it from
> sugar-build does not seem to be a good long term solution.
>
> Many fixes have been proposed, but none have gained community support.
> Perhaps because the fixes have been wrong, and have not solved the
> underlying problem.
>
> Please confirm by your own research my theory below.
>
> The problem was caused by a commit to the Browse repository that
> introduced a Git Submodule (fb62edc), without a corresponding commit
> to sugar-build or sugar-toolkit-gtk3 to support submodules and
> recursive clones.
>
> If my theory is true, an alternative is to revert the commit to
> Browse, removing the submodule and replacing it with a copy of the
> files from the collabwrapper repository.
>
> You might test the theory by:
>
> 1.  cloning the Browse repository and reverting the commit,
>
> 2.  editing your module.json to use your clone of Browse,
>
> 3.  re-running sugar-build from a fresh clone.
>
> As for myself, I build Sugar as Fedora or Debian packages and apply
> these to my systems, rather than use sugar-build.  sugar-build is not
> used when Sugar is a primary desktop environment; it is only a
> development aid for people who do not need to make packages.
>
> Disclosure: I'm paid by One Laptop per Child.
>
> On Tue, Nov 08, 2016 at 01:52:55AM +0000, Nguyen, Lam wrote:
> > Hi Sugar developers,
> >
> > The issue that Walter brought up had been discussed before. The fix for
> this
> > issue can be found here [1]http://www.mail-archive.com/
> > sugar-devel at lists.sugarlabs.org/msg42561.html
> > I was thinking of changing the installation instruction on [2]
> > developer.sugarlabs.org to include this fix as a common issue. What I
> wondered
> > is that should we permanently remove the lines that caused the error
> from the
> > file /build/module.json? Would there be any consequence to permanently
> removing
> > these lines?
> >
> > {
> > >> "name": "browse",
> > >> "repo": "[3]https://github.com/sugarlabs/browse-activity.git";,
> > >> "clean_stamp": 1
> > >> },
> >
> > Best,
> > Lam
> >
> >     On Oct 30, 2016, at 9:20 AM, Walter Bender <[4]
> walter.bender at gmail.com>
> >     wrote:
> >
> >     ---------- Forwarded message ----------
> >     From: Ridhwanul Haque <[5]rhemon19 at gmail.com>
> >     Date: Sun, Oct 30, 2016 at 6:30 AM
> >     Subject: Sugar-Build
> >     To: Walter Bender <[6]walter.bender at gmail.com>
> >
> >     Hello,
> >     I am trying to set up the sugar development environment in fedora
> following
> >     this link: [7]https://developer.sugarlabs
> .org/dev-environment.md.html#run
> >     On build i got and error, but i still tried running but got an error
> as
> >     well. I have shared part of the error with the image.
> >
> >     <on build.png><on run.png>
> >     In the website, it said to check the relay chat to share problems
> but I
> >     wasn't sure if I should be taking this problem there so I thought I
> should
> >     first share it with you. Thanks and sorry in advance.
> >
> >     --
> >     Walter Bender
> >     Sugar Labs
> >     [8]http://www.sugarlabs.org
> >     [9]
> >     _______________________________________________
> >     Sugar-devel mailing list
> >     [10]Sugar-devel at lists.sugarlabs.org
> >     http://lists.sugarlabs.org/listinfo/sugar-devel
> >
> > References:
> >
> > [1] http://www.mail-archive.com/sugar-devel@lists.
> sugarlabs.org/msg42561.html
> > [2] http://developer.sugarlabs.org/
> > [3] https://github.com/sugarlabs/browse-activity.git%22
> > [4] mailto:walter.bender at gmail.com
> > [5] mailto:rhemon19 at gmail.com
> > [6] mailto:walter.bender at gmail.com
> > [7] https://developer.sugarlabs.org/dev-environment.md.html#run
> > [8] http://www.sugarlabs.org/
> > [9] http://www.sugarlabs.org/
> > [10] mailto:Sugar-devel at lists.sugarlabs.org
>
> > _______________________________________________
> > Sugar-devel mailing list
> > Sugar-devel at lists.sugarlabs.org
> > http://lists.sugarlabs.org/listinfo/sugar-devel
>
>
> --
> James Cameron
> http://quozl.netrek.org/
>
>
>
> --
> Regards,
> Ütkarsh Tiwari
>
>
>
> On Sun, Nov 13, 2016 at 12:46 AM, Walter Bender <walter.bender at gmail.com>
> wrote:
>
>> What with Google Code In beginning in two weeks, we are about to get
>> bombarded with questions from people struggling with the build system. It
>> has recently been pointed out that building browse is causing problems
>> again. I imagine there may be other problems as well, for example, when
>> building on non-Fedora-based systems.
>>
>> It would be really helpful if you could take the time to do some testing
>> of the build system on various platforms and post recommendations in the
>> wiki for any necessary workarounds. (We can also use process to generate
>> some GCI tasks to improve the robustness of the build system.)
>>
>> Would be good to post results here. We can summarize the results in
>> developer.sl.o once we accumulate some data.
>>
>> Thanks.
>>
>> -walter
>>
>> --
>> Walter Bender
>> Sugar Labs
>> http://www.sugarlabs.org
>> <http://www.sugarlabs.org>
>>
>> _______________________________________________
>> Sugar-devel mailing list
>> Sugar-devel at lists.sugarlabs.org
>> http://lists.sugarlabs.org/listinfo/sugar-devel
>>
>>
>
>
> --
> Regards,
> Ütkarsh Tiwari
>
>
>
>


-- 
Walter Bender
Sugar Labs
http://www.sugarlabs.org
<http://www.sugarlabs.org>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/sugar-devel/attachments/20161112/c42928b6/attachment-0001.html>


More information about the Sugar-devel mailing list