Re: 'master' 'make check' lcov code coverage report

On Mon, Nov 3, 2014 at 12:38 PM, Maarten Hoes <hoes.maarten@gmail.com> wrote:
> Looking briefly at the differences (or lack thereof) in the reports of both
> 11-Apr-2014 and today, I cannot help but wonder if running the report as
> often as weekly would be beneficial at this point. Doing it manually once
> every few months seems often enough for now ? Of course, I would be glad to
> be wrong here.
> ;)

It's a catch-22 problem... if this is run rarely, it is on the mind of
people rarely, it is unlikely to attract the attention of new blood
and even less likely to kkep their attention.
If it run say weekly... they there is a 'instant' gratification
associated with putting some effort to improve the coverage in some
part of the code
that in turn often motivate to do even more...

Surely much of the coverage could be tacckled as easy hacks.. but
positive feedback is very important, to motivate new contributor and
show them the result of their good work in a tangible way.

> >
> If bourne/korn shell is considered acceptable, Im willing to try to hack up
> a script that runs a git pull/lcov report and scp uploads the results
> somewhere.

Sure bash is fine.. and a properly scripted bash script can be tuck
under almost any automation system we can think of.
for the record buidbot's client are pure bash :-)

Norbert