02:47 < rcurtin> cult-: hmm, do you mean the training loop? since it's boosting that may be hard---iteration i+1 depends on the results of iteration i

04:50 < niteya> rcutrn: I checked that already and they are different ,I a tried a bit more trial and error and the header files for pca_test.cpp(main) dont give the error , it is only when i include range_search_main.cpp that i get the error

04:51 < niteya> so it seems that the problem lies with that file or its includes ( mainly rs_model.hpp and rs_model_impl.hpp) as the error is caught in rs_model_impl.hpp

11:11 < zoq> niteya: Do you have an mlpack fork that we could use to take a quick look?

13:01 < rcurtin> niteya: maube there is some non-template method declared somewhere that needs to be marked inline

13:01 < rcurtin> maybe*

14:49 < niteya> rcurtin: https://github.com/niteya-shah/mlpack this is my fork with minimum code that generates the error

15:11 < rcurtin> niteya: did you try looking for a non-templated non-inlined function implementation in a .hpp file?

15:11 < rcurtin> any non-templated function that's implemented in a header file needs to be marked inline

15:16 < ShikharJ_> rcurtin: Are you there?

15:17 < ShikharJ_> For some reason, Atharva's PR (https://github.com/mlpack/mlpack/pull/1594) is timing out on one of the tests, even after re-running the job. Do you think it maybe because of a bug?

15:26 < niteya> Thanks a lot , that fixed the error .

15:29 < niteya> Is there any advantage to function definitions in hpp files ( Everything book I have read till now speaks of code bloat and less reusability )

15:29 < niteya> Every*

16:00 < rcurtin> niteya: any templated function has to be in a header file, so we don't have much choice there

16:01 < rcurtin> sometimes we include non-templated functions in an _impl.hpp file, just so that we don't have three files (a .hpp, .cpp, and _impl.hpp for templated functions)

16:01 < rcurtin> but, as you have found out, also I think that sometimes we forget the 'inline' marker :)

16:02 < rcurtin> ShikharJ_: that's strange. let's try building it again

16:02 < rcurtin> I hit "restart job", let's see what it does...

16:06 < ShikharJ_> rcurtin: I wouldn't bet too much on its completion, it has stopped at the exact time of 35 mins in the past five runs that I initiated.

16:22 < rcurtin> ShikharJ__: oh, didn't realize you already did that. that is quite strange, I have not seen anything like that before

16:23 < rcurtin> have you tried building locally with the same flags as the failing build? (-DDEBUG=ON -DPROFILE=OFF -DBUILD_PYTHON_BINDINGS=OFF)

16:43 < ShikharJ_> Hmm, I'll try that out. Thanks.

17:27 < rcurtin> hmm, I didn't realize that with the Jenkins build inaccessible to non-logged-in users, the badge on the main github page doesn't work

17:27 < rcurtin> I guess I will have to debug that Jenkins issue then

20:07 < zoq> rcurtin: or we use travis to show the build status for the master branch

20:14 < rcurtin> that could work too, but the Jenkins thing is annoying enough that I need to solve it anyway

20:14 < rcurtin> I'm getting a bit tired of not being able to click on anything until I log in, and the login times out after a couple hours

20:16 < zoq> yeah, I doubled checked the settings some time ago; looks correct to me

21:54 < rcurtin> I think I'll first try just updating all the Jenkins plugins and seeing if that helps

21:55 < rcurtin> (probably that will break things even worse, but hey, worth a shot)

21:56 < zoq> okay, fingers crossed

22:10 < rcurtin> oh nice, it worked! that was unexpected

22:15 < zoq> nice!

23:31 < travis-ci> mlpack/mlpack#5739 (master - adbfe10 : Ryan Curtin): The build passed.

23:31 < travis-ci> Change view : https://github.com/mlpack/mlpack/compare/376c17e11d27...adbfe102a23b

23:31 < travis-ci> Build details : https://travis-ci.org/mlpack/mlpack/builds/476587164