Re: Travis failures

A thing to be aware of for the future - the SC build-test workflow depends
on quite a few external components to work correctly. The versions that we
fetch can be somewhat loose, depending on what happens to be installed on
the travis system (which changes regularly), and current versions available
via pip/brew/apt/etc. The python requirements for testing have been firmed
up, but there's still lots of slop in exactly what we have installed on a
system for any given build-test run. If there's an unexplainable build
failure, the first thing to check is whether any component versions have
changed between the first failing build and the last successful build -
this has historically caused LOTS of breakages. You should be able to see
details at the beginning of the raw log, where we install dependencies.

- Scott C


On Tue, Apr 17, 2018 at 11:16 AM <[email protected]> wrote:

> Hi all,
>
> Just an update - the issues with qpm/pip/Travis have been resolved,
> everything is running smoothly now!
>
> -Brian
>
> On Mon, Apr 16, 2018 at 4:34 PM, brian heim <[email protected]> wrote:
>
>> > on a side note, wouldn't it be great to reduce travis build log size?
>> > right now there are 3000+ lines of code posted just for the tests..
>>
>> > I'd be happy to help but I'd need some pointers as where to start
>> looking ;)
>>
>> What specifically would you like to reduce? For qpm, you will want to
>> look at the qpm project (github.com/scztt/qpm). The verbosity increased
>> lately as we were trying to change the way tests were run and were hitting
>> some hard to debug problems on Travis. I think the rest of the output is
>> manageable. A more important issue is to get qpm on our Appveyor builds
>> (see: appveyor docs on .appveyor.yml in main repo), and get it running with
>> the real test suite as opposed to the deprecated CommonTests (see:
>> https://github.com/supercollider/supercollider/issues/3132).
>>
>> Please start a separate thread if you want to discuss any of that.
>>
>> -Brian
>>
>> On Mon, Apr 16, 2018 at 4:14 PM, <yvan.sc-gM/[email protected]> wrote:
>>
>>> On 04/16/2018 08:41 PM, [email protected] wrote:
>>> > Travis builds are currently failing due to a bug that popped up in qpm.
>>> > A patch update in one of the subpackages is causing a crash. Scott and
>>> I
>>> > are working on fixing this, it should be resolved soon.
>>>
>>> thanks for tackling that guys =)
>>>
>>> on a side note, wouldn't it be great to reduce travis build log size?
>>> right now there are 3000+ lines of code posted just for the tests..
>>>
>>> I'd be happy to help but I'd need some pointers as where to start
>>> looking ;)
>>>
>>> cheers,
>>> y
>>>
>>> --
>>> http://yvanvolochine.com
>>> http://s3pt3ntrion.com
>>> http://www.facebook.com/pages/Septentrion/357546467715618
>>> http://soundcloud.com/yvanvolochine
>>> http://soundcloud.com/elgusanorojo
>>> http://vimeo.com/yv
>>>
>>> _______________________________________________
>>> sc-dev mailing list
>>>
>>> info (subscription, etc.):
>>> http://www.birmingham.ac.uk/facilities/ea-studios/research/supercollider/mailinglist.aspx
>>> archive: http://www.listarc.bham.ac.uk/marchives/sc-dev/
>>> search: http://www.listarc.bham.ac.uk/lists/sc-dev/search/
>>>
>>
>>
>