On Fri, Mar 26, 2010 at 3:20 AM, Brice Figureau < [email protected]> wrote:
> On Fri, 2010-03-26 at 01:23 -0700, Markus Roberts wrote: > > Now with breakdown on the failing spec tests (which we're focusing on > > for the next few days). > > > > testing (Branching off master) succeeded: > > [snipped] > > markus:ticket/0.25.x/3366 (bool settings fix) succeeded: > > * 25f6edf Fix for #3366 - --tags '' treated as boolean 'true' > > 7770 examples, 24 failures, 41 pending > > 1 more in unit/type/file.rb (1 total) > > 21 more in unit/type/package.rb (21 total) > > 651 tests, 5351 assertions, 68 failures, 29 errors > > 1 fewer in /ral/providers/user.rb (2 total) > > brice:tickets/0.25.x/3396 (Faster event propagation) succeeded: > > * 7c3cd8c Fix inefficient SimpleGraph#matching_edge > > 7771 examples, 28 failures, 41 pending > > 4 more in unit/simple_graph.rb (4 total) > > This is not good. Looks like we are missing Luke's glue event patch, or > there is something at play here. > That may well be; one of the problems I encountered when reordering the patches that had been rebased onto testing was that they had absorbed fragments of each other (what we've been calling the stray pixel problem, by analogy to what happens when you shuffle parts of picture in an image manipulation program) that can prevent them from merging properly later. > 651 tests, 5347 assertions, 71 failures, 30 errors > > 4 more in /other/transactions.rb (16 total) > > 1 fewer in /ral/providers/cron/crontab.rb (0 total) > > 1 more in /ral/providers/user.rb (3 total) > > I'll check those. I might have missed those tests when I did the patch > on 0.25. > But I think their root cause is the same as for the 4 failing specs. > > BTW, is there a way to distinguish between test failing (ie assertions > are false) and test erroring? > The later would be a strong indicator the code merged is missing > something (ie merge issue for instance). > That would be a good refinement. I just got this basic statistics extraction going yesterday with much grumbling; one of the things I noted was that the unit test output seemed designed to resist parsing by all but the uglyist of regular expressions. But yeah, breaking them down by failure/error would be a logical next step. I'm taking today off, so I'll try to do it tomorow. -- Markus -- You received this message because you are subscribed to the Google Groups "Puppet Developers" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/puppet-dev?hl=en.
