Re: [sympy] Re: Travis-CI is not keeping up with sympy

2013-03-03 Thread Ondřej Čertík
Hi, I never heard back from the Travis support guys, but it looks like Travis has caught up, so all is good now. Implementing any of the things below would however help anyway. Ondrej On Sat, Feb 23, 2013 at 6:21 PM, Ondřej Čertík wrote: > Hi, > > Thanks everybody for a feedback. I think a few

Re: [sympy] Re: Travis-CI is not keeping up with sympy

2013-02-23 Thread Ondřej Čertík
Hi, Thanks everybody for a feedback. I think a few things we should definitely do now: * measure time of each test and report it in some form (e.g. print tests in ascending order by their execution time at the end...) * Skip a few of the longest running tests on Travis --- I think it should be p

Re: [sympy] Re: Travis-CI is not keeping up with sympy

2013-02-23 Thread Chris Smith
On Sat, Feb 23, 2013 at 2:09 PM, mario wrote: > It seems to me that a possible way to reduce the average time of tests > could be to test only the modules which depend on the modified modules. > I do not know how long it would > take to build the dependency tree; if it takes too much, it could be

[sympy] Re: Travis-CI is not keeping up with sympy

2013-02-23 Thread mario
It seems to me that a possible way to reduce the average time of tests could be to test only the modules which depend on the modified modules. I do not know how long it would take to build the dependency tree; if it takes too much, it could be stored. Making tests when only a leaf of the tree is m