What about code coverage? If I understand it correctly the tests take to long on Travis to include code coverage (just an impression I got while browsing Issues, no idea how it works in practice). Maybe there is something nifty built in on Azure Pipelines that can be used for that? Or if it just a matter of CPU time, maybe one can run the code coverage tests on Azure and the other tests on Travis? I really would like if it was possible to get a codecov.io or something up and running.
BR Oscar Den onsdag 30 januari 2019 kl. 19:35:49 UTC+1 skrev Aaron Meurer: > > > > On Wed, Jan 30, 2019 at 8:32 AM Matt Cooper <[email protected] > <javascript:>> wrote: > >> Thanks Isuru. If you see any issues again, I'd like to dig into it. Could >> have been transient, or could be a browser-specific problem we need to >> address. >> >> Also, I'm interested to know more about the complexity of our YAML >> syntax. (I don't disagree, I'm mainly curious what specifically you tripped >> over.) >> > > You can see from https://github.com/sympy/sympy/pull/15297/commits how I > struggled with the syntax, for instance, trying to figure out how to make > the conditions work. > > My biggest gripe is that the documentation doesn't have a real search > feature. There are two search boxes on > https://docs.microsoft.com/en-us/azure/devops/pipelines/get-started/index?view=azdevops. > > One searches only the titles, and the other searches all of Azure DevOps. I > ended up downloading the PDF documentation and using that because it was > the only way I could get a reasonable search. Even then, I had a hard time > with the documentation because of lack of concrete examples. I remember > searching for a long time until I figured out how to make a build an > allowed failure. > > My biggest hurdle was working with environment variables, and the fact > that they are not persisted across scripts. This means that things like > source activate for conda/virtualenv do not work. A lot of time was wasted > from me not knowing this fact. > > Aaron Meurer > > >> On Wednesday, January 30, 2019 at 10:17:15 AM UTC-5, Isuru Fernando wrote: >>> >>> The azure build log link didn't work for me when Aaron sent it 2 days >>> ago and now it opens at the correct line just fine. >>> >>> Isuru >>> >>> On Wed, Jan 30, 2019 at 7:34 AM Matt Cooper <[email protected]> wrote: >>> >>>> (Azure Pipelines PM chiming in.) >>>> >>>> Aaron, can you send me a screenshot what you see when you click that >>>> log link? For me it opens directly to the line, as it's intended to do. >>>> I'm >>>> using Firefox, and most of the team uses Chrome - it's possible we have a >>>> bug though. >>>> >>>> With no additional clicks, here's what I see: >>>> >>>> On Tuesday, January 29, 2019 at 1:59:25 PM UTC-5, Aaron Meurer wrote: >>>>> >>>>> The question is if its worth it, given the other issues Azure has. >>>>> Keep in mind that we have to maintain whatever we set up. >>>>> >>>>> For benchmarks, I don't know if any public CI is appropriate. You >>>>> really need a dedicated machine so you can get consistent runtimes. >>>>> Code quality tests are done as part of the test suite already >>>>> (test_code_quality.py). >>>>> >>>>> I'm not opposed to the idea of using some other CI for other tests. I >>>>> think it should be driven by a real need, though, not just saying we >>>>> should use the resource because it is there. There are other options >>>>> too, such as CircleCI and AppVeyor. >>>>> >>>>> Aaron Meurer >>>>> >>>>> On Tue, Jan 29, 2019 at 3:09 AM S.Y. Lee <[email protected]> wrote: >>>>> > >>>>> > There had been some issues with azure pipelines with conda >>>>> environment setups, and I would also agree that some features are not >>>>> better than travis. >>>>> > >>>>> > However, I have a question that, if they are offering free resources >>>>> for open source projects, >>>>> > we may allocate some of the tests to azure, which can either reduce >>>>> overall build time, >>>>> > or making some space for running benchmarks or code quality tests. >>>>> > >>>>> > Would there be any reason making this idea naive? >>>>> > >>>>> > -- >>>>> > You received this message because you are subscribed to the Google >>>>> Groups "sympy" group. >>>>> > To unsubscribe from this group and stop receiving emails from it, >>>>> send an email to [email protected]. >>>>> > To post to this group, send email to [email protected]. >>>>> > Visit this group at https://groups.google.com/group/sympy. >>>>> > To view this discussion on the web visit >>>>> https://groups.google.com/d/msgid/sympy/3bd7d189-b67e-457f-ac3b-994d3c5deda0%40googlegroups.com. >>>>> >>>>> >>>>> > For more options, visit https://groups.google.com/d/optout. >>>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "sympy" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to [email protected]. >>>> To post to this group, send email to [email protected]. >>>> Visit this group at https://groups.google.com/group/sympy. >>>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/sympy/702492e8-8233-417a-aeba-a9cde5aeb668%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/sympy/702492e8-8233-417a-aeba-a9cde5aeb668%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> For more options, visit https://groups.google.com/d/optout. >>>> >>> -- >> You received this message because you are subscribed to the Google Groups >> "sympy" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> To post to this group, send email to [email protected] <javascript:> >> . >> Visit this group at https://groups.google.com/group/sympy. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/sympy/7b98fac0-f9c8-413d-9885-c3eb33aafc25%40googlegroups.com >> >> <https://groups.google.com/d/msgid/sympy/7b98fac0-f9c8-413d-9885-c3eb33aafc25%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> For more options, visit https://groups.google.com/d/optout. >> > -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/sympy. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/03917a20-d077-411f-a27e-c994cc1519db%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
