03:50 < favre49> zoq: just wanted to give you a reminder to take a look at ny proposed NEAT interface whenever you're free

11:29 < zoq> favre49: Ahh, thanks, just responded.

13:47 < favre49> rcurtin: I've recently reinstalled ubuntu, and hence mlpack. There are a couple maybe pedantic issues i have

13:47 < favre49> with the website

13:48 < favre49> Maybe it's just me, but it's weird to me that the installation instructions are under a "tutorials" section

13:49 < favre49> It's not as intuitive as having a dedicated link to installation instructions from the main page

13:51 < favre49> Also, copy pasting the install command from the "Dependencies" section in http://mlpack.org/doc/stable/doxygen/build.html doesn't work since there are newlines at the end of the line

13:51 < favre49> (No clue if this can be fixed, just something that i found somewhat annoying)

13:53 < favre49> Also, there are multiple scroll bars on the screen. This may be something only I find weird though.

14:04 < favre49> It occurred to me that the install command is more annoying to bash users, since it treats newlines as enter keypresses. Another reason to use zsh :)

14:09 < zoq> favre49: gmanlan is working on an update (https://github.com/mlpack/mlpack/issues/1892) which I think will address some of the issues you mentioned

15:28 < jeffin143> Zoq : I have used boost::string_view and for that included <boost/utility/string_view.hpp> , but Travis cl is failing throwing error that no file found

15:28 < jeffin143> Compiles fine In My local system*

15:28 < jeffin143> What should I do to debug it

15:29 < lozhnikov> jeffin143: string_view is a part of boost since 1.60. Travis uses boost 1.58

15:29 < jeffin143> Ohh ok*

15:30 < jeffin143> The I should just ignore it

15:30 < jeffin143> Then*

15:31 < lozhnikov> jeffin143: Actually string_view is a pretty small class. So, I think we could backport it to src/mlpack/core/boost_backport

15:32 < jeffin143> Ok I will look into it*

15:32 < jeffin143> And once done will ping you

15:32 < lozhnikov> jeffin143: But it's better to discuss with Ryan and Marcus before backporting.

15:32 < jeffin143> Ok*

15:32 < jeffin143> Sure

15:33 < zoq> If it's small, sure backporting is an option.

17:35 < jeffin143> Lozhnikov : added some commits to pr1814, would like to have your comments on it, also I am unable to resolve Travis failure .

17:36 < jeffin143> Added string_view.hpp class to src/core/boost_backport , and also update cmake * , but it still throws an error.

17:41 < lozhnikov> jeffin143: Looks like string_view.hpp includes string_view_fwd.hpp. And you didn't backport this file. Actually, maybe string_view_fwd.hpp isn't needed at all. We need to check that.

17:48 < lozhnikov> jeffin143: I think this line is incorrect: "#if BOOST_VERSION <= 105800". string_view appeared in boost 1.61 (I've just checked).

18:04 < jeffin143> Ok*

18:04 < jeffin143> So I have to If boost_version <= 106100..?

18:06 < lozhnikov> jeffin143: I guess < 106100 or <= 106000

18:07 < jeffin143> Ok

22:45 < rcurtin> lozhnikov: jeffin143: yeah, backporting string_view is probably the right call here.

22:47 < rcurtin> also, I hope that including string_view doesn't also include tons of other headers. if so, there are a few options, like how we split data/load.hpp and data/load_impl.hpp and used explicit template specialization

23:01 < jenkins-mlpack2> Project docker mlpack nightly build build #329: ABORTED in 2 days 8 hr: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/329/