On Sun, Feb 16, 2014 at 8:29 PM, Matt Turner <[email protected]> wrote: > On Fri, Feb 14, 2014 at 11:24 AM, Eric Anholt <[email protected]> wrote: >> Dylan Baker <[email protected]> writes: >> >>> This replaces the console spewer with a simpler console reporting >>> mechanism inspired by the output of ninja. This reduces code, remove all >>> singleton instances and speeds up piglit runs. There is a drawback, the >>> output is much more terse than the previous implementation, giving only >>> the following output: >>> [<finished>/<total>] Running Test(s): <test number> >>> [16008/16011] Running Test(s): 16007 16008 16009 16010 >> >> I don't like this and don't want to see it land. Seeing pass-fail >> scrolling is really imporatnt to me, and saves me more time than you're >> proposing to save with this patch. > > Wouldn't a running total of pass/fail/crash/skip accomplish the same > thing? I.e., pass stays at zero and fail zooms up would be pretty > effective at showing that your run is botched.
Yes. There was a bit of a crossed-wires situation with these... Take a look at my sequence of patches which addresses various people's concerns. I'm going to commit the R-b'd bits soonish. I think that Dylan and I are on the same page now, at least :) And hopefully everyone else is too... -ilia _______________________________________________ Piglit mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/piglit
