13:23 <aserio> hkaiser: hey

13:23 <hkaiser> aserio: hey

13:23 <hkaiser> g'morning

13:23 <aserio> see pm please :)

13:27 <hkaiser> parsa: yt?

13:27 <parsa> i am

13:27 <hkaiser> at CCT?

13:27 <parsa> yes sir, at the CCT

13:28 <hkaiser> see pm, pls

13:37 <mdiers_> did the special hpx/config/boost/compiler/intel.hpp disappear?

13:40 <hkaiser> mdiers_: I think so, yes

13:41 <hkaiser> mdiers_: the whole config module has moved, but your code should still compile - you might have to delete your hpx build directory, however

13:43 <mdiers_> yes, unfortunately there is an old default entry in cmake/templates/config_defines.hpp.in

13:57 <hkaiser> huh?

13:57 <hkaiser> mdiers_: yes, you're right

13:57 <hkaiser> why does it build for us?

13:58 <hkaiser> ...because we don't test with intel :/

13:58 <hkaiser> please remove that from the template file

13:59 <hkaiser> mdiers_: ^^

13:59 <mdiers_> remove or change to the original boost path?

13:59 <hkaiser> remove the lines alltogether

13:59 <hkaiser> these lines: https://github.com/STEllAR-GROUP/hpx/blob/master/cmake/templates/config_defines.hpp.in#L11-L14

14:00 <hkaiser> darn, this breaks our release for Intel compilers

14:00 <K-ballo> that's odd... doesn't it mean we were broken for years?

14:01 <hkaiser> K-ballo: no this file was removed by introducing the config module

14:01 <K-ballo> ah, so it wasn't just a module transformation

14:02 <mdiers_> hkaiser: Okay, I'll test. i'm trying the new intel release to see if c++17 is finally better supported.

14:02 <hkaiser> ok, thanks

14:03 <K-ballo> it's not part of the release, acc\ording to github

14:10 <hkaiser> K-ballo: ahh good

14:16 <aserio> hkaiser: headed to CCT

14:16 <hkaiser> aserio: ok

14:35 <parsa> simbergm: ping

14:58 <diehlpk> jbjnr, Will you join the operation bell meeting?

14:58 <diehlpk> hkaiser, ?

15:14 <mdiers_> hkaiser: removing the three lines helped. to the intel compiler: success, that for the first time after two years our new code base didn't crash the compiler.

15:14 <hkaiser> mdiers_: cool, care to create a PR?

15:14 <hkaiser> diehlpk: here

15:15 <diehlpk> Do you want to join the operation bell meeting

15:17 <mdiers_> hkaiser: Yes, tomorrow, go home now.

15:17 <hkaiser> mdiers_: sure, no rush

15:17 <hkaiser> and thanks for checking

15:17 <hkaiser> diehlpk: sorry, I don't feel too well

15:18 <diehlpk> hkaiser, Ok, sorry to hear

15:20 <mdiers_> hkaiser: not for that, thanks a lot for the incredible library. ;-)

15:52 <diehlpk> K-ballo, We have some application for GSoD from Argentina :)

15:55 <K-ballo> really?? who?

15:57 <diehlpk> K-ballo, see pm

17:15 <hkaiser> aserio: yt?

18:05 <aserio> hkaiser: yes

18:05 <hkaiser> aserio: I forgot what I wanted to ask :/

18:06 <aserio> hkaiser: lol, were you planning on joining the Phylanx meeting today?

18:06 <hkaiser> aserio: I'd rather not - do you think I'm really required to join?

18:06 <hkaiser> do you need me there?

18:06 <aserio> hkaiser: not if you do not feel up to it

18:07 <aserio> I just though I would ask

18:07 <hkaiser> then I'd stay away

18:07 <aserio> I have not heard anything further from Chris

18:07 <hkaiser> sure

18:07 <hkaiser> he feels bad

18:08 <aserio> he probably feels a bit stressed

18:08 <hkaiser> right

18:08 <hkaiser> that's what I meant

18:08 <aserio> hkaiser: let me know if you need help interpreting the excel doc I sent you

18:08 <hkaiser> aserio: no worries, it's what we need

18:08 <hkaiser> see pm, pls

19:09 <Yorlik> Which HPX facilities could I use to ease the development of a full remote console? I just finished a first remote REPL and can issue simple server command through a very basic client using nng. Before doing any fancy web based / web socket stuff I'd like to know if there are any ways how hpx might help me? Especially I'd like to maintain a buffer of logging messages I could query from my client. Any ideas?

19:10 <Yorlik> I imagin a command like !LASTLOG N or sth which would send me the last N log messages or so

19:10 <Yorlik> I'm not yet using any hpx logging stuff - I'm just getting started doing this.

19:11 <Yorlik> Just stumbled on this when I enabled remote Lua commands and ofc - I don't get any Lua console output

19:12 <Yorlik> In the moment its just simple ping-pong - client asks and server answers.

19:33 <parsa> can someone help me understand what "template<class... Ts> overloaded(Ts...) -> overloaded<Ts...>;" in the example on this page is https://en.cppreference.com/w/cpp/utility/variant/visit ?

19:34 <zao> parsa: I believe it's some sort of "deduction helper"?

19:35 <zao> "deduction guide"

19:35 <zao> https://en.cppreference.com/w/cpp/language/class_template_argument_deduction

19:35 <parsa> zao: i've never encountered one before. thanks a bunch!

19:52 <Yorlik> IS there any good example for hpx logging? I find the docs a little bit sparse on that.

20:20 <Yorlik> Isn't an hpx app supposed to automagically read hpx.ini ? I think mine doesn't

20:20 <bita> hkaiser, I am working on dilated subtensor. I have seen in subtensor that sometime page is the third argument (row,column,page), does it happen really often? Should I change the order(from page,row,column) in my try(something)assign functions?

20:24 <hkaiser> Yorlik: it should, yah, see the docs, this is described

20:24 <hkaiser> bita: no

20:24 <Yorlik> I tried to put some simple logging into it, but it refuses to work

20:24 <Yorlik> Environment variables work nicely though

20:25 <hkaiser> the index order is alsways page. row, column, except for the functions that expect 2 indices, there its always row, page

20:25 <Yorlik> So - logginmg is functional - it just ignores my hpx.ini

20:25 <hkaiser> interesting

20:25 <hkaiser> sounds like a bug

20:25 <hkaiser> but you should be able to pass a ini file on the command line

20:25 <hkaiser> --hpx:config=<ini>

20:26 <bita> thanks

20:26 <hkaiser> that at least lets us verify whether it's reading the files or something else

20:26 <Yorlik> hkaiser: that actually led to a crash

20:26 <hkaiser> lol

20:26 <hkaiser> nice

20:27 <Yorlik> VC runtime library calling abort()

20:27 <hkaiser> Yorlik: please create a ticket with your ini file attached, I'll try to have a look

20:27 <Yorlik> OK

20:27 <hkaiser> thanks

20:27 <aserio> hkaiser: please see pm

20:35 <Yorlik> hkaiser: https://github.com/STEllAR-GROUP/hpx/issues/3890

20:36 <hkaiser> Yorlik: use --hpx:config=<ini>, not --hpx:ini

20:36 <Yorlik> Woops?

20:37 <Yorlik> OK - that seems to work

20:37 <Yorlik> Still - why is it not reading it in the first place?

20:37 <Yorlik> I mean - without the CL option

20:42 <Yorlik> However: Thanks for resolving the blooper. Though I'd consider a crash something that shouldn'tz happen - rather "Wrong Option - aborting" or sth

20:57 <K-ballo> yeah, crash means bug

20:58 <K-ballo> > Steps to Reproduce the Problem

20:58 <K-ballo> > 1. Debug build of my app

20:58 <K-ballo> well then.............................

21:00 <diehlpk_work> hkaiser, see pm

21:07 <zao> K-ballo: Can't reproduce, not a bug :P