#ste||ar on 2017-06-29 — irc logs at irclog.cct.lsu.edu

2017-05-17 13:54 aserio changed the topic of #ste||ar to: STE||AR: Systems Technology, Emergent Parallelism, and Algorithm Research | stellar.cct.lsu.edu | HPX: A cure for performance impaired parallel applications | github.com/STEllAR-GROUP/hpx | Buildbot: http://rostam.cct.lsu.edu/ | Log: http://irclog.cct.lsu.edu/

00:38 zbyerly_ has quit [Remote host closed the connection]

00:38 zbyerly_ has joined #ste||ar

00:41 hkaiser has quit [Read error: Connection reset by peer]

00:41 hkaiser_ has joined #ste||ar

01:06 bikineev has quit [Remote host closed the connection]

01:12 vamatya has quit [Ping timeout: 276 seconds]

01:21 parsa has quit [Quit: Zzzzzzzzzzzz]

01:26 EverYoung has joined #ste||ar

01:30 EverYoung has quit [Ping timeout: 246 seconds]

01:31 parsa has joined #ste||ar

01:38 zbyerly_ has quit [Remote host closed the connection]

01:38 zbyerly_ has joined #ste||ar

01:45 hkaiser_ has quit [Quit: bye]

02:37 parsa has quit [Quit: Zzzzzzzzzzzz]

02:38 zbyerly_ has quit [Remote host closed the connection]

02:38 zbyerly_ has joined #ste||ar

02:48 ajaivgeorge has joined #ste||ar

02:52 ajaivgeorge_ has quit [Ping timeout: 255 seconds]

02:59 K-ballo has quit [Quit: K-ballo]

03:09 vamatya has joined #ste||ar

03:38 zbyerly_ has quit [Remote host closed the connection]

03:38 zbyerly_ has joined #ste||ar

04:06 pree has joined #ste||ar

04:11 ajaivgeorge has quit [Ping timeout: 260 seconds]

04:14 ajaivgeorge has joined #ste||ar

04:38 zbyerly_ has quit [Remote host closed the connection]

04:38 zbyerly_ has joined #ste||ar

04:39 <wash[m]> @zao: thanks, enjoy :)

05:24 pree has quit [Quit: AaBbCc]

05:25 pree has joined #ste||ar

05:38 zbyerly_ has quit [Remote host closed the connection]

05:38 zbyerly_ has joined #ste||ar

06:01 <thundergroudon[m> Can I know how to compile native cuda on Rostam?

06:02 <thundergroudon[m> Currently I am getting unsupported GNU version error

06:02 <thundergroudon[m> and I am unable to link another compiler to nvcc

06:15 <heller> What have you done so far and what exactly do you want to achieve?

06:19 <thundergroudon[m> I want to compile a native cuda application to benchmark

06:19 <thundergroudon[m> I am comparing it against a HPXCL benchmark

06:22 <thundergroudon[m> I can run the application on Bahram/Tycho but I need to build it first :')

06:36 zbyerly_ has quit [Ping timeout: 276 seconds]

06:41 Matombo has joined #ste||ar

06:46 zbyerly_ has joined #ste||ar

06:51 <heller> thundergroudon[m: so what did you try?

07:03 <heller> https://twitter.com/MalwareMinigun/status/880299153493970945

07:03 <heller> jbjnr: ^^

07:06 <thundergroudon[m> heller: I built a .cu file using cmake and it throws up #error -- unsupported GNU version! gcc versions later than 5 are not supported!

07:06 <thundergroudon[m> even though my gcc version is 4.8.5

07:07 <heller> well, then change the compiler?

07:07 <heller> how did you run cmake?

07:07 <heller> which modules do you have loaded?

07:08 <thundergroudon[m> heller: The point being, I think nvcc is using a different compiler

07:08 <heller> it uses whatever you tell it to

07:08 <thundergroudon[m> and I can change that using something as mentioned in https://stackoverflow.com/questions/6622454/cuda-incompatible-with-my-gcc-version

07:09 <heller> usually looking up gcc in $PATH

07:09 <thundergroudon[m> But I can't link it because I don't have sudo access

07:09 <thundergroudon[m> but even when my compiler is 4.8.5, it still throws up the same error

07:12 <heller> tell me the output of the following commands: 1) module list 2) gcc --version 3) echo -n "#include <cuda.h>\nint main(){}\n" > test.cu && nvcc test.cu

07:12 <heller> then tell me how you ran cmake

07:14 <heller> and 4) g++ --version

07:15 <thundergroudon[m> This doesn't throw up that error

07:16 <thundergroudon[m> Currently Loaded Modulefiles:

07:16 <thundergroudon[m> 1) cmake/3.7.2 2) boost/1.56.0

07:16 <thundergroudon[m> gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-11)

07:16 <thundergroudon[m> g++ version is the same

07:18 <thundergroudon[m> The problem should righlty be with the build process using cmake

07:18 Matombo has quit [Remote host closed the connection]

07:19 <thundergroudon[m> Let me re-check this

07:30 EverYoung has joined #ste||ar

07:33 ajaivgeorge has quit [Read error: Connection reset by peer]

07:34 ajaivgeorge has joined #ste||ar

07:35 EverYoung has quit [Ping timeout: 246 seconds]

07:35 david_pfander has joined #ste||ar

07:47 zbyerly_ has quit [Remote host closed the connection]

07:47 zbyerly_ has joined #ste||ar

08:05 zbyerly_ has quit [Ping timeout: 276 seconds]

08:20 <Vir> david_pfander: don't be afraid of mixing the AVX and SSE abis, if you want a 4-wide datapar<int> and don't want to use fixed_size<4> then sse it is

08:20 <david_pfander> Vir: so it will generate 128bit VEX instruction?

08:20 <Vir> david_pfander: note that the ABI tag does not restrict the instruction selection/coding. It uses ternary (VEX-coded) instructions for the sse ABI if AVX is available

08:20 <Vir> yes

08:20 <david_pfander> Vir: perfect

08:21 <Vir> david_pfander: a library has very limited influence on instruction coding...

08:22 Matombo has joined #ste||ar

08:23 <Vir> david_pfander: consider that there are two, almost orthogonal choices: choosing the target ISA and choosing the way datapar objects are passed on function calls. The latter is. more or less, what the ABI tag is for

08:23 <Vir> david_pfander: the target ISA is chosen by the compiler (i.e. the flags)

08:28 Matombo has quit [Ping timeout: 260 seconds]

09:12 <zao> I took a look into my CUDA environments at our clusters, a lot of our env is stuff like setenv("CUDA_CC", 'nvcc -ccbin="g++"')

09:14 <zao> (also CUDA_CXX)

09:15 pree has quit [Ping timeout: 268 seconds]

09:22 bikineev has joined #ste||ar

09:32 EverYoung has joined #ste||ar

09:37 EverYoung has quit [Ping timeout: 276 seconds]

09:39 bikineev has quit [Remote host closed the connection]

09:47 ajaivgeorge has quit [Read error: Connection reset by peer]

09:47 ajaivgeorge has joined #ste||ar

09:56 <github> [hpx] StellarBot pushed 1 new commit to gh-pages: https://git.io/vQlzF

09:56 <github> hpx/gh-pages f92f780 StellarBot: Updating docs

10:05 <heller> zao: hard to tell what's going on without having the information i asked for

10:05 <heller> the cmake step is missing

10:05 <heller> thundergroudon[m: ^^

10:06 <zao> indeedily

10:08 Matombo has joined #ste||ar

10:09 hkaiser has joined #ste||ar

10:11 <github> [hpx] hkaiser pushed 12 new commits to master: https://git.io/vQl2j

10:11 <github> hpx/master c99b1e2 : initial import of valarray serializer

10:11 <github> hpx/master 5293df3 : added unit test

10:11 <github> hpx/master e8dc283 : modified copyright to use correct developer name

10:12 <heller> hkaiser: good morning

10:12 <heller> hkaiser: https://twitter.com/MalwareMinigun/status/880299153493970945

10:13 <zao> Easy to be fast if you don't care about correctness :P

10:14 <heller> I am sure that test is just splitting out random numbers as timing results

10:15 <heller> we clearly mess with that

10:16 <hkaiser> heh

10:17 <hkaiser> the unseq is not implemented and falls back to par only anyways

10:18 <hkaiser> heller: this is the guy who implements the parallel algorithms for the MS compiler

10:18 <heller> I know ;)

10:19 <hkaiser> frankly, I was expecting for us to horribly loose there

10:19 <heller> me too

10:19 <heller> I wouldn't have expected that

10:20 <heller> I mean, a team of a few vs. microsoft, how crazy is that?

10:21 <hkaiser> right

10:21 <github> [hpx] hkaiser force-pushed fixing_2722 from 40a0f95 to b0c5c98: https://git.io/vQla9

10:21 <github> hpx/fixing_2722 b0c5c98 Hartmut Kaiser: Enable lcos::channel<T>::register_as...

10:21 ajaivgeorge has quit [Ping timeout: 276 seconds]

10:22 hkaiser has quit [Read error: Connection reset by peer]

10:22 Matombo has quit [Remote host closed the connection]

10:22 hkaiser has joined #ste||ar

10:30 ajaivgeorge has joined #ste||ar

10:32 bikineev has joined #ste||ar

10:38 Matombo has joined #ste||ar

10:44 Matombo has quit [Ping timeout: 260 seconds]

10:53 pree has joined #ste||ar

11:00 bikineev has quit [Ping timeout: 240 seconds]

11:03 <heller> https://akrzemi1.wordpress.com/2017/06/28/compile-time-string-concatenation/ <-- this is interesting

11:03 <heller> we could get rid of all our std::string stuff for the action/component/plugin registration

11:12 bikineev has joined #ste||ar

11:16 Matombo has joined #ste||ar

11:37 <zao> Bah, was going to use HPX to stress-test my CPU but the code is broken :(

11:37 <zao> https://gist.github.com/zao/d985bb9bebe542ad1bd0ee5d5a83e91b

11:38 <zao> Eugh, that's in Boost code...

11:38 Matombo has quit [Remote host closed the connection]

11:41 <hkaiser> zao: configure with HPX_WITH_CXX14=On, that works around the issue

11:41 <zao> Ick.

11:42 <hkaiser> this is gcc 7.x, right?

11:42 <zao> gcc version 7.1.1 20170622 (Red Hat 7.1.1-3) (GCC)

11:42 <hkaiser> nod

11:43 <hkaiser> this gets configured for c++17 by default which breaks the boost code

11:43 <zao> On Fedora 26 Beta as part of trying to tease out microcode bugs in my CPU :D

11:43 <zao> I'm guessing upstream Boost hasn't fixed this yet?

11:43 <hkaiser> lol

11:43 <hkaiser> I doubt they know about it

11:45 <zao> ./boost/exception/exception.hpp:491:13: internal compiler error: Segmentation fault

11:45 <hkaiser> hah!

11:45 <hkaiser> I have not seen that one

11:45 <zao> hkaiser: GCC is not at fault here, amusingly enough.

11:45 <hkaiser> how so?

11:45 <zao> CPU bug.

11:46 <hkaiser> isn't an ICE always the compiler's fault?

11:46 <hkaiser> ahh, read about it

11:46 <hkaiser> skylake+ HT

11:46 <zao> Load down the cores in a HT pair with userspace code while interrupts are serviced and cores may hang and segfaults occur.

11:46 <zao> Ryzen, but same kind of µcode madness.

11:47 <zao> Even more fun, the machine a meter to my right is a vulnerable Skylake :)

11:51 <zao> /home/zao/slask/stellar/hpx/hpx/util/jenkins_hash.hpp:195:19: warning: this statement may fall through [-Wimplicit-fallthrough=]

11:51 <zao> I like these new warnings, you have to mitigate them with comments that say that you're intentionally falling through :)

11:52 <hkaiser> zao: I think we have a fix for c++17 for this

11:52 <zao> (or [[fallthrough]], I guess)

11:52 <hkaiser> yes

11:53 <hkaiser> https://github.com/STEllAR-GROUP/hpx/blob/master/hpx/util/jenkins_hash.hpp#L196

11:53 <zao> Ah, explains why I see it with the forced C++14 build I guess.

11:53 <hkaiser> nod

11:53 <zao> There's supposedly a GNU variation for earlier standards.

11:54 <hkaiser> interesting, don't know anything about this

11:54 <zao> https://developers.redhat.com/blog/2017/03/10/wimplicit-fallthrough-in-gcc-7/

11:54 <hkaiser> ok, I'll add that

11:56 <hkaiser> or simply add -Wimplicit-fallthrough ;)

11:56 <zao> Hehe, that's also a "solution" :)

12:00 <hkaiser> zao: we've been living without this warning for ages, so why have it now ;)

12:00 <mcopik> hkaiser: are you going to add gcc C++14 solution or should I do it?

12:00 <zao> hkaiser: We've been living with the problem of unintentional fallthrough for ages too :)

12:01 <hkaiser> mcopik: feel free to do that - even if I think we should just test for -Wimplicit-fallthrough=0 being available in C++14 and below

12:01 <hkaiser> since when is [[gnu:fallthrough]] (or the __attribute__() version) available?

12:02 <hkaiser> we would need to add another feature test...

12:03 <zao> Is this in user-facing code or library internal build?

12:03 <mcopik> hkaiser: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=7652

12:03 <mcopik> 7.0

12:03 <zao> If the latter, I'm fine with just muting it.

12:03 <hkaiser> nod

12:03 <hkaiser> mcopik: let's just mute the warning for C++14 and below

12:04 <mcopik> ok

12:05 <hkaiser> if(NOT HPX_WITH_CXX17) hpx_add_compile_flag_if_available(-Wimplicit-fallthrough=0 LANGUAGES CXX) endif()

12:06 <mcopik> ach, so we already have a facility for trying compilation flags

12:07 <hkaiser> mcopik: even better: let the HPX_FALLTHROUGH macro expand to a comment: /* fallthrough */

12:08 <hkaiser> that keeps it explicit and shuts up the compiler

12:08 <mcopik> right, GCC should be happy with that

12:08 <hkaiser> it doesn't hurt th eothers either

12:09 <mcopik> hkaiser: btw, I noticed an interesting thing while I've been tracking down a CMake bug

12:09 <hkaiser> ok?

12:09 <mcopik> in the build directory, we create a share dir with share/hpx symlink to share/hpx-${HPX_VERSION}

12:09 <mcopik> the problem is that both directories are empty

12:09 <zao> hkaiser: Comment doesn't seem to work.

12:10 <zao> I can never recall my phases, but aren't those stripped earlier?

12:10 <hkaiser> the article says it should :/

12:10 <hkaiser> does it work if you replace the macro in the jenkins file with a comment?

12:11 <mcopik> furthermore, in the install directory we have share/share-${HPX_VERSION} with docs inside

12:11 <zao> Hrm, no.

12:11 K-ballo has joined #ste||ar

12:11 <hkaiser> mcopik: ok

12:11 <mcopik> I'm wondering if it's not a leftover after a disabled Python part? https://github.com/STEllAR-GROUP/hpx/blob/master/CMakeLists.txt#L1753

12:12 <hkaiser> what would you suggest?

12:12 <mcopik> and here is the symlink: https://github.com/STEllAR-GROUP/hpx/blob/master/CMakeLists.txt#L1877

12:12 <mcopik> I don't understand what is the purpose of this directory inside the build

12:12 <hkaiser> those directories shouldn't be empty

12:12 <hkaiser> they have the docs, if built

12:13 <mcopik> so what is the purpose of the symlink in the build dir?

12:13 <mcopik> a symlink to hpx-version makes sense if you want to install multiple builds

12:14 <hkaiser> mcopik: the idea is that you can use hpx without installing, so the install structure is mirrored in the build dir

12:14 <zao> I'm guessing that however we build the fallthrough level is higher than 3? odd.

12:14 <mcopik> zao: that's the default for gcc, 3

12:14 <hkaiser> zao wrong text inside the comment?

12:14 <mcopik> hkaiser: ok, so I'll simply fix the issue of share not being created

12:14 <mcopik> zao: https://developers.redhat.com/blog/2017/03/10/wimplicit-fallthrough-in-gcc-7/

12:14 <hkaiser> mcopik: I guess we don't need the symlink in the build dir

12:15 <mcopik> try one of these, 2 matches a regex but 3 gives a list of accepted comments

12:15 <zao> I have the exact spelling of "/* FALLTHRU */", but it seems that a trailing semicolon screws it over.

12:15 <mcopik> hkaiser: and symlink in the install dir? do we need that?

12:16 <mcopik> ha, we forgot about the semicolon

12:16 <hkaiser> the symlink in install is supposed to link to the latest version

12:16 <hkaiser> mcopik: the semicolon should be outside of the picture

12:16 <hkaiser> don't think it matters

12:17 <zao> In any way, the spellings in the document work, as long as you don't have a statement after them, like a semicolon.

12:17 <mcopik> hkaiser: and there's no symlink. I think there's is a mistake there, I'll fix that by moving symlink from build -> install

12:17 <zao> The macro way never works.

12:18 <zao> (that is, HPX_FALLTHROUGH; vs. HPX_FALLTHROUGH)

12:18 <hkaiser> mcopik: ok, but only during installation, pls

12:24 <hkaiser> zao: so disable the warning?

12:26 <zao> Won't do much good if it ends up in end-user code anyway.

12:35 EverYoung has joined #ste||ar

12:37 bikineev has quit [Ping timeout: 276 seconds]

12:39 EverYoung has quit [Ping timeout: 246 seconds]

12:44 ajaivgeorge has quit [Ping timeout: 260 seconds]

12:53 bikineev has joined #ste||ar

12:58 bikineev has quit [Ping timeout: 255 seconds]

12:59 <github> [hpx] ct-clmsn closed pull request #2715: initial import of icu string serializer (master...icu) https://git.io/vQ3OK

13:02 <github> [hpx] Element-126 opened pull request #2724: Fix examples in Action documentation (master...master) https://git.io/vQl7X

13:02 bikineev has joined #ste||ar

13:07 bikineev has quit [Ping timeout: 260 seconds]

13:07 ajaivgeorge has joined #ste||ar

13:09 eschnett has quit [Quit: eschnett]

13:21 <github> [hpx] ajaivgeorge opened pull request #2725: Implemented segmented algorithms extending existing parallel algorithms (master...master) https://git.io/vQldV

13:36 EverYoung has joined #ste||ar

13:41 EverYoung has quit [Ping timeout: 276 seconds]

13:42 pree has quit [Quit: AaBbCc]

13:42 pree has joined #ste||ar

13:46 david_pfander1 has joined #ste||ar

13:46 david_pfander1 has quit [Remote host closed the connection]

13:47 eschnett has joined #ste||ar

13:53 aserio has joined #ste||ar

13:55 zbyerly_ has joined #ste||ar

13:57 <aserio> david_pfander, heller, hkaiser, jbjnr, wash[m]: Operation Bell Meeting

14:00 <heller> aserio: can't attend today, sorry

14:09 vamatya has quit [Ping timeout: 260 seconds]

14:09 <aserio> heller: working hard on your thesis?

14:10 <heller> aserio: yes

14:10 <heller> more or less

14:10 <heller> aserio: also this: https://twitter.com/MalwareMinigun/status/880299153493970945

14:15 <aserio> heller: :0

14:16 <aserio> Did he contact you or did you stumble across this?

14:16 <heller> aserio: best thing, as far as I am aware, none of us here was involved in this ;)

14:16 <heller> I was made aware of it by someone else

14:16 <aserio> wow!

14:16 <aserio> its like it works :p

14:18 <K-ballo> Billy keeps looking at our stuff

14:18 <K-ballo> (aserio: for context, Billy is working on parallel algorithms for the MSVC standard library)

14:19 zbyerly_ has quit [Ping timeout: 276 seconds]

14:19 <K-ballo> *implementation

14:21 <aserio> K-ballo: that's awesome

14:33 hkaiser has quit [Quit: bye]

14:38 EverYoung has joined #ste||ar

14:41 jakemp has joined #ste||ar

14:42 EverYoung has quit [Ping timeout: 258 seconds]

14:47 ajaivgeorge_ has joined #ste||ar

14:48 ajaivgeorge has quit [Ping timeout: 276 seconds]

14:53 zbyerly_ has joined #ste||ar

14:56 aserio has quit [Ping timeout: 276 seconds]

15:02 aserio has joined #ste||ar

15:14 ajaivgeorge_ has quit [Ping timeout: 255 seconds]

15:20 ajaivgeorge_ has joined #ste||ar

15:21 aserio has quit [Ping timeout: 240 seconds]

15:40 <heller> Damn

15:41 <heller> We failed :(

15:42 <heller> No reviews

15:43 <heller> jbjnr: have you heard from your folks?

15:46 aserio has joined #ste||ar

15:52 bikineev has joined #ste||ar

15:52 <heller> aserio: :(

15:53 <aserio> I know

15:53 <heller> why are there no reviews?

15:53 <aserio> Perhaps we will get access to reviews later?

15:53 <heller> would be great

15:54 <heller> or this is just a hoax?

15:54 <aserio> Though it might be one of those political things

15:54 <aserio> idk

15:54 <aserio> I suppose I could google him

15:54 <heller> would be disappointing

15:54 <heller> no, the name of the chair is correct

15:57 <aserio> Yea, seems to be legit

16:06 <heller> so what should we do now?

16:08 parsa has joined #ste||ar

16:09 vamatya has joined #ste||ar

16:12 akheir has quit [Remote host closed the connection]

16:18 <aserio> well you write a dissertation

16:18 <aserio> We need to figure out where to publish the paper

16:18 <aserio> maybe late breaking at SC

16:18 <aserio> and we get in gear for next year

16:20 <aserio> I think if we can get the performance under control we will be golden

16:24 bikineev has quit [Ping timeout: 240 seconds]

16:30 david_pfander has quit [Ping timeout: 260 seconds]

16:34 parsa has quit [Quit: Zzzzzzzzzzzz]

16:38 parsa has joined #ste||ar

16:39 EverYoung has joined #ste||ar

16:40 <heller> aserio: we won't know without reviews

16:40 parsa has quit [Client Quit]

16:41 parsa has joined #ste||ar

16:43 parsa has quit [Client Quit]

16:43 EverYoung has quit [Ping timeout: 240 seconds]

16:50 parsa has joined #ste||ar

16:54 aserio has quit [Ping timeout: 246 seconds]

17:01 parsa has quit [Quit: Zzzzzzzzzzzz]

17:11 pree has quit [Ping timeout: 276 seconds]

17:16 parsa has joined #ste||ar

17:45 Matombo has joined #ste||ar

18:00 hkaiser has joined #ste||ar

18:00 aserio has joined #ste||ar

18:03 <heller> hkaiser: so, we failed

18:08 <hkaiser> nod

18:09 <hkaiser> let's resubmit the paper asap to some conference

18:09 <hkaiser> and get started now for next year

18:09 <heller> Well, reviews would be great

18:10 <hkaiser> you will not get any

18:10 <heller> I'm thinking scala, but we'd sell ourselves under value

18:10 <heller> Why not?

18:10 <hkaiser> no, I'd suggest a real conference

18:10 <hkaiser> not a workshop

18:10 <heller> Do you have anything in mind?

18:11 <hkaiser> not sure what's coming up next

18:12 <heller> Calling it a day. Drying my tears. I'll ask for reviews tomorrow

18:13 <hkaiser> k

18:14 <diehlpk_work> It is strange not to get reviews

18:28 <heller> hkaiser: any comments from Dominic? I haven't heard anything if my suggestions fixed anything

18:30 <hkaiser> he said that he still can't run things and would like to know how you actually built everything

18:30 <hkaiser> can you give him the details (if you haven't done so yet)?

18:30 <hkaiser> modules, git hashes, cmake options, etc.

18:31 parsa has quit [Quit: Zzzzzzzzzzzz]

18:31 parsa has joined #ste||ar

18:36 parsa has quit [Ping timeout: 276 seconds]

18:38 bikineev has joined #ste||ar

18:41 EverYoung has joined #ste||ar

18:42 <mcopik> hkaiser: I'll force my student to use clang-format

18:42 <mcopik> sorry about that, I didn't pressure him to use it before sending a PR

18:43 parsa has joined #ste||ar

18:43 <heller> hkaiser: I gave him the hashes I use. Didn't tell how I built everything. Made one more suggestion, was awaiting a "yay or nay" response to further try to reproduce his problems. I really want to know why it fails for him and not for me... So yeah, will try again tomorrow...

18:45 EverYoung has quit [Ping timeout: 240 seconds]

18:45 bikineev has quit [Ping timeout: 276 seconds]

19:23 <jbjnr> heller: we were involved. That felipe chap supplied the sort algorithm and I spent ages getting it into hpx. or do you mean we were not involved with the tts he made?

19:24 <jbjnr> heller: failed at wat? reviews for what? my folks what?

19:24 <heller> jbjnr: we're not involved in the tests he performed

19:24 <jbjnr> ok

19:24 <heller> jbjnr: GB, check your mails

19:26 <heller> jbjnr: I mean, of course we're involved since we provided the stuff he tested. But he did on his own motivation, not with us pushing him

19:26 <jbjnr> GB. Pity. But not a surprise. we knew we had no chance

19:27 <heller> Well, we hoped we had one

19:27 <jbjnr> only a few months' worth of work wasted.

19:27 <heller> Reviews would still be nice

19:28 <hkaiser> mcopik: no worries

19:28 <hkaiser> jbjnr: nothing wasted

19:30 <mcopik> it's not a usual situation where no reviews are sent with a rejection, is it?

19:31 <hkaiser> mcopik: and sorry for the flurry of comments

19:31 <hkaiser> some are really minor in importance

19:31 <mcopik> hkaiser: no, please don't apologize. I feel bad because I hasn't spotted it

19:32 <mcopik> and I really appreciate that you spent so much time to make a detailed review

19:32 <hkaiser> mcopik: might have been easier if the PR was split up into one per algorithm... but I guess it willbe fine this way as well

19:32 <mcopik> I'll send my student an email explaining some things you mentioned and hinting how we can improve that

19:32 <hkaiser> cool, thanks

19:32 <hkaiser> feel free to keep me in the loop

19:33 <mcopik> hkaiser: I'll do. what do you think about 5 variants of find algorithm (standard, if, if_not, first_of, end)? one PR or split?

19:36 jgoncal_ has joined #ste||ar

19:36 jgoncal_ has quit [Client Quit]

19:38 <hkaiser> mcopik : separate might be good - makes reviews easier

19:38 <hkaiser> gtg

19:38 hkaiser has quit [Quit: bye]

19:46 <K-ballo> at least a bunch of them ought to be the same

19:47 <K-ballo> find, find_if, find_if_not

19:48 <mcopik> K-ballo: yes, it's the same implementation but with different function objects

19:48 <K-ballo> nod, I'd threat those 3 as one

19:50 bikineev has joined #ste||ar

19:51 zbyerly_ has quit [Ping timeout: 276 seconds]

20:01 zbyerly_ has joined #ste||ar

20:08 zbyerly_ has quit [Ping timeout: 276 seconds]

20:18 akheir has joined #ste||ar

20:18 hkaiser has joined #ste||ar

20:20 hkaiser has quit [Read error: Connection reset by peer]

20:20 hkaiser has joined #ste||ar

20:23 bikineev_ has joined #ste||ar

20:26 bikineev has quit [Ping timeout: 255 seconds]

20:31 parsa has quit [Quit: Zzzzzzzzzzzz]

20:37 parsa has joined #ste||ar

20:54 aserio has quit [Ping timeout: 246 seconds]

20:58 erl has joined #ste||ar

21:11 mbremer has joined #ste||ar

21:19 aserio has joined #ste||ar

21:20 erl has left #ste||ar [#ste||ar]

21:20 michoski has joined #ste||ar

21:27 bikineev_ has quit [Ping timeout: 240 seconds]

21:28 jakemp has quit [Ping timeout: 240 seconds]

21:29 parsa has quit [Quit: Zzzzzzzzzzzz]

21:44 Matombo has quit [Remote host closed the connection]

21:47 Matombo has joined #ste||ar

21:53 aserio has quit [Quit: aserio]

21:56 bikineev has joined #ste||ar

22:10 parsa has joined #ste||ar

22:56 <hkaiser> ajaivgeorge_: yt?

23:21 bikineev has quit [Remote host closed the connection]

23:22 mbremer has quit [Ping timeout: 260 seconds]

23:24 michoski has quit [Ping timeout: 260 seconds]

23:43 Matombo has quit [Remote host closed the connection]

23:55 eschnett has quit [Quit: eschnett]