[tei-council] oxygen-tei package

[Martin Holmes]

On 14-09-19 08:55 AM, Lou Burnard wrote:
> On 19/09/14 16:39, Peter Stadler wrote:
>> Am 19.09.2014 um 17:26 schrieb Sebastian Rahtz <sebastian.rahtz at IT.OX.AC.UK>:
>>
>>>> The last item on the list made the build during release fail (to my knowledge). I wonder whether this part is (or: can be) maintained by the Council or whether it needs to be maintained by the oxygen people?
>>> the latter, because it all depends on interaction with the main oxygen code base. we can’t diverge from them.
>> Ok, then we treat this part as a black box and simply call oxygen support when our build fails :)
>
> Call me old-fashioned, but this makes all sorts of alarm bells ring in
> my head. Our build process should not be be dependent on a commercial
> product, however wonderful!

It has to be when what we are producing is an add-on for their project.

I think a) the product is wonderful, as you say; b) we should be glad 
they work so closely with us and give us such a large degree of control 
over what goes into our add-on (although we have to stop them from 
putting "Brown" stuff in -- that should be a separate add-on from Brown, 
surely); c) it's our fault that up to now we've left it all up to 
Sebastian and not kept our eye on whether it was easy to build, or what 
conditions needed to be set up to build it properly.

We're going to remedy c) by making it Jenkins job that runs regularly, 
so we see as soon as it breaks and we can fix it.

Cheers,
Martin