[Bioperl-l] Network tests overhaul

Spiros Denaxas spiros at lokku.com
Wed Jun 20 11:49:48 EDT 2007


Yep, they are not all done. Some still need to be ported over, doing
some here and there at home. However, the recent email Sendu sent, the
one about abstracting the setup of testing is actually something i was
thinking myself so it might be a better way to tackle the problem. For
once it would save us from duplicating the same 30 lines of code
across all tests.

As far as network tests are involved, ive always been an avid hater of
them. I believe they only bring more troubles than what they
contribute due to the diversity of setups people have. My way of
tackling them was always to group all the tests that required live
access into one file and then forcibly just run that - iff needed and
not by default. Like i said, thats just my opinion, ive been bitten by
them one time too many.

Spiros

On 6/18/07, Chris Fields <cjfields at uiuc.edu> wrote:
>
> On Jun 18, 2007, at 11:34 AM, Sendu Bala wrote:
>
> > Chris Fields wrote:
> >> Couldn't you enable BIOPERLDEBUG, disable network access, then
> >> iterate through tests checking for those which fail or skip?
> >
> > Yes, good idea, though my dev machine is also my email/webserver so
> > I'd rather come up with an alternate solution than one involving
> > 'disable network access'.
> >
> > Still, that's what I'll probably end up doing. Cheers!
> >
> >
> > Oh, Chris, Spiros, how goes the Test::More conversion? I might want
> > to wait for you to finish, or join in? If you're not going to have
> > time to do any more in the next few weeks, can you please update
> > http://www.bioperl.org/wiki/TestMoreProgress removing your name (or
> > in the opposite case, add your name in)? Its not quite clear to me
> > which tests are assigned to whom. Can someone clarify what the
> > markings mean?
> >
> > Cheers,
> > Sendu.
>
> Not sure how far along spiros is; I handed it over after I finished
> up to the 'Q' tests.  In general the ones marked out have been
> converted over, ones with names next to them have been claimed.  If
> you need help I'll prob. start back up again to finish them off; we
> just need to divy them up.
>
> chris
>


More information about the Bioperl-l mailing list