[Bioperl-l] BioPerl and Travis-CI

[Fields, Christopher J]

On Jul 27, 2012, at 10:03 AM, Peter Cock wrote:

> On Fri, Jul 27, 2012 at 3:58 PM, Fields, Christopher J
> <cjfields at illinois.edu> wrote:
>> On Jul 27, 2012, at 9:47 AM, Peter Cock wrote:
>> 
>>> On Thu, Jul 26, 2012 at 4:22 PM, Peter Cock <p.j.a.cock at googlemail.com> wrote:
>>>> 
>>>> That's done now - except for the circular dependencies, and GD,
>>>> which might be easy to solve if anyone knows what the error
>>>> means - see commit message here:
>>>> https://github.com/peterjc/bioperl-live/commit/905441ac09939be3368c14de38d04486c7e9849a
>>> 
>>> Re: https://twitter.com/cjfields/status/228861370454638592
>>> Not sure why you got GD to work when something very similar
>>> had failed for me. Oh well - job done :)
>> 
>> It was the lack of gdlib-config in the libgd2-xpm package, you need
>> libgd2-xpm-dev.  One of the fun things about Debian packaging.
> 
> Ah - I should have guessed that.
> 
>>>> Would a single clean commit of the (current) .travis.yml file be
>>>> preferable to the current series of commits? And you you want
>>>> a pull request, or would you just merge/cherry-pick manually?
>>> 
>>> Given all the churn between our revisions, personally I'd opt for
>>> a single clean commit to bioperl/master - but your call.
>>> 
>>> Peter
>> 
>> Yep, about to merge it over.  It's working now, just need to
>> whitelist master instead of travis after the merge.
> 
> I'd removed the whitelist altogether here:
> https://github.com/peterjc/bioperl-live/commit/96dc5866f4406179353909c72d812623341c8fbd
> 
> My thinking was BioPerl seems to have multiple feature branches
> under the official repo, so they should get tested too. You'd be
> in a better position than me to judge what would work best for
> BioPerl here.
> 
> Peter

We'll keep it to master for now.  It's pretty easy to add branches as needed, and I didn't want to expand to all the potentially stale branches unless explicitly set (we need to triage all those at some point).

chris