Me too! :) I was hoping tests.dups would do it. Dawid promised me he will one day find the time to do it! :)
In the meanwhile, Mike pointed me to repeatLuceneTest.py under luceneutil. I've yet been successful to run it, but I think it's just a matter of configuration. Maybe if you have a shellscript that's generic enough, it's worth putting under dev-tools? Shai On Fri, May 10, 2013 at 6:06 PM, Robert Muir <[email protected]> wrote: > I'm still hoping for the elusive tests.beast or whatever that gives each > run a whole different seed and acts like the shellscripts that run 'ant > test -Dtestcase' over and over :) > > > On Fri, May 10, 2013 at 10:59 AM, Shai Erera <[email protected]> wrote: > >> Thanks for the tip. If it's just me, I won't make the change. I just find >> it annoying to have to type it everytime, and think that this is probably >> what you intend to achieve when running w/ tests.iters -- search for a seed >> that trips the tests. >> >> If there are no objections, I'll make the trivial change later. >> >> Shai >> >> >> On Fri, May 10, 2013 at 5:55 PM, Robert Muir <[email protected]> wrote: >> >>> I dont have any strong opinion really, i just want to mention on the >>> side you can also change your 'personal defaults' by editing >>> build.properties in your home directory too. >>> >>> >>> On Fri, May 10, 2013 at 9:27 AM, Shai Erera <[email protected]> wrote: >>> >>>> Because I think it makes more sense as a default? Usually when I run >>>> with tests.iters, I just search for a seed that fails. When I find it, I go >>>> fix the bug and run again. >>>> >>>> But that's just me. That's why I asked if others think it makes sense >>>> as a default too. There is no functionality change, just different default. >>>> >>>> Shai >>>> On May 10, 2013 3:33 PM, "Robert Muir" <[email protected]> wrote: >>>> >>>>> why change it? >>>>> >>>>> On Fri, May 10, 2013 at 1:20 AM, Shai Erera <[email protected]> wrote: >>>>> >>>>>> Hi >>>>>> >>>>>> Can we default tests.failfast to true? I don't know how common it is >>>>>> to run with -Dtests.iters=N, not specify -Dtests.maxfailures AND wanting >>>>>> to >>>>>> see all N tests finish. >>>>>> >>>>>> At any rate, if someone wishes to, he can always specify >>>>>> -Dtests.failfast=false. It now depends what is the default behavior we >>>>>> want >>>>>> to have. >>>>>> >>>>>> Any thoughts? >>>>>> >>>>>> Shai >>>>>> >>>>> >>>>> >>> >> >
