--- Log opened Wed Feb 15 00:00:25 2017 | ||
@sukey | Issue #3609 "Linking errors with ATLAS library" closed by vigsterkr - https://github.com/shogun-toolbox/shogun/issues/3609 | 00:04 |
---|---|---|
@sukey | Pull Request #3622 "Add Scala to DockerFile" merged by vigsterkr - https://github.com/shogun-toolbox/shogun/pull/3622 | 00:06 |
@sukey | New Commit "Merge pull request #3622 from abhinavrai44/docker | 00:06 |
@sukey | Add Scala to DockerFile" to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/2fab5f1e8d5a2fd9b6a086a146dee19390d1c284 | 00:06 |
shogun-buildbot | build #88 of trusty - libshogun is complete: Failure [failed test] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun/builds/88 blamelist: Viktor Gal <vigsterkr@gmail.com>, abhinavrai44 <abhinavrai44@gmail.com> | 00:21 |
@sukey | Pull Request #3610 "Java meta examples only re-compile when dependencies change." synchronized by geektoni - https://github.com/shogun-toolbox/shogun/pull/3610 | 00:52 |
-!- suhas2go [uid201652@gateway/web/irccloud.com/x-ikvklnosjejkcxcu] has quit [Quit: Connection closed for inactivity] | 00:56 | |
-!- travis-ci [~travis-ci@ec2-54-80-142-4.compute-1.amazonaws.com] has joined #shogun | 00:56 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201694848 | 00:56 |
-!- travis-ci [~travis-ci@ec2-54-80-142-4.compute-1.amazonaws.com] has left #shogun [] | 00:56 | |
@sukey | Pull Request #3610 "Java meta examples only re-compile when dependencies change." merged by vigsterkr - https://github.com/shogun-toolbox/shogun/pull/3610 | 01:54 |
@sukey | New Commit "Merge pull request #3610 from geektoni/patch-2 | 01:54 |
@sukey | Java meta examples only re-compile when dependencies change." to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/df30e2b4bc4a0d42986df6cda3c74fe24ee4555e | 01:54 |
@sukey | Issue #3626 "Add " opened by vigsterkr - https://github.com/shogun-toolbox/shogun/issues/3626 | 01:59 |
@sukey | Issue #3626 "Add " vigsterkr added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/3626 | 01:59 |
@sukey | Issue #3626 "Add " vigsterkr added label: "Cleanups" - https://github.com/shogun-toolbox/shogun/issues/3626 | 01:59 |
@sukey | Issue #3626 "Add "- https://github.com/shogun-toolbox/shogun/issues/3626 | 01:59 |
@sukey | Issue #3556 "`make build_java_meta_examples` always re-compiles all java examples" closed by vigsterkr - https://github.com/shogun-toolbox/shogun/issues/3556 | 02:00 |
@sukey | Issue #3557 "`make build_csharp_meta_examples` always re-compiles all java examples" closed by vigsterkr - https://github.com/shogun-toolbox/shogun/issues/3557 | 02:00 |
@sukey | Pull Request #3606 "Add CMake to Detect Scala" merged by vigsterkr - https://github.com/shogun-toolbox/shogun/pull/3606 | 02:05 |
@sukey | New Commit "Merge pull request #3606 from abhinavrai44/dev | 02:05 |
@sukey | Add CMake to Detect Scala" to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/44f13ae7bbdc2f195b5d4312d15ed52f819d26c5 | 02:05 |
shogun-buildbot | build #89 of trusty - libshogun is complete: Success [build successful] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun/builds/89 | 02:09 |
@sukey | Pull Request #3593 "Port existing BrayCurtis Distance example to meta" merged by vigsterkr - https://github.com/shogun-toolbox/shogun/pull/3593 | 02:10 |
@sukey | New Commit "Merge pull request #3593 from abhinavrai44/abhi | 02:10 |
@sukey | Port existing BrayCurtis Distance example to meta" to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/308d300461088ead48aabdcbe55c67b1fa30eb63 | 02:10 |
@sukey | New Commit "fix LTO status message | 02:14 |
@sukey | and fix shogun-data" to shogun-toolbox/shogun by vigsterkr: https://github.com/shogun-toolbox/shogun/commit/e8cb1e8efc4bf071b5b8f5276549965ef5172a4d | 02:14 |
shogun-buildbot | build #123 of trusty - libshogun - viennacl is complete: Failure [failed test] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/123 blamelist: Viktor Gal <vigsterkr@gmail.com>, abhinavrai44 <abhinavrai44@gmail.com> | 02:14 |
shogun-buildbot | build #124 of trusty - libshogun - viennacl is complete: Success [build successful] Build details are at http://buildbot.shogun-toolbox.org/builders/trusty%20-%20libshogun%20-%20viennacl/builds/124 | 02:16 |
@sukey | Issue #3605 "port parallel code using pthread"- https://github.com/shogun-toolbox/shogun/issues/3605 | 02:21 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Remote host closed the connection] | 02:48 | |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-dajmozufoyahcrni] has joined #shogun | 02:55 | |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 02:55 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 03:09 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 276 seconds] | 03:14 | |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 03:19 |
-!- travis-ci [~travis-ci@ec2-54-91-71-85.compute-1.amazonaws.com] has joined #shogun | 03:25 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201720442 | 03:25 |
-!- travis-ci [~travis-ci@ec2-54-91-71-85.compute-1.amazonaws.com] has left #shogun [] | 03:25 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 03:41 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 03:45 | |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has joined #shogun | 04:01 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201721411 | 04:01 |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has left #shogun [] | 04:01 | |
-!- tctara_ [~quassel@128.199.61.169] has joined #shogun | 04:06 | |
-!- tctara [~quassel@128.199.61.169] has quit [Ping timeout: 255 seconds] | 04:09 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 04:13 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 268 seconds] | 04:17 | |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has joined #shogun | 04:42 | |
travis-ci | it's Viktor Gal's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201721961 | 04:42 |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has left #shogun [] | 04:42 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 04:43 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 04:48 | |
@sukey | Pull Request #3623 "Add Scala to .travis.yml" closed by abhinavrai44 - https://github.com/shogun-toolbox/shogun/pull/3623 | 04:55 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 05:15 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 05:20 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 05:47 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 05:52 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 06:18 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 255 seconds] | 06:23 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 06:34 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Ping timeout: 240 seconds] | 06:39 | |
-!- praisethemoon [~praisethe@197.3.11.178] has joined #shogun | 09:06 | |
@sukey | Pull Request #3627 "Make meta_examples depend on python generator and language files." opened by geektoni - https://github.com/shogun-toolbox/shogun/pull/3627 | 09:53 |
@sukey | Pull Request #3597 "Add Scala Interface" synchronized by abhinavrai44 - https://github.com/shogun-toolbox/shogun/pull/3597 | 13:13 |
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has joined #shogun | 13:15 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 13:15 | |
@sukey | Pull Request #3597 "Add Scala Interface" synchronized by abhinavrai44 - https://github.com/shogun-toolbox/shogun/pull/3597 | 13:17 |
@sukey | Pull Request #3611 "Enable meta examples when ENABLE_TESTING is OFF." merged by karlnapf - https://github.com/shogun-toolbox/shogun/pull/3611 | 13:20 |
@sukey | New Commit "Merge pull request #3611 from geektoni/patch-3 | 13:20 |
@sukey | Enable meta examples when ENABLE_TESTING is OFF." to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/3130c158861b1999ef7a27546384b4a1c5cac089 | 13:20 |
@sukey | Pull Request #3619 "LinalgRefactor - Enable matrix block elementwise product" merged by karlnapf - https://github.com/shogun-toolbox/shogun/pull/3619 | 13:26 |
@sukey | New Commit "Merge pull request #3619 from OXPHOS/linalg_element_square | 13:26 |
@sukey | LinalgRefactor - Enable matrix block elementwise product" to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/2beda114e7ec42be3be7b15598e025d2bffda0f8 | 13:26 |
@wiking | HeikoS, h0 | 13:37 |
@wiking | y0 | 13:37 |
@HeikoS | wiking: jojo | 13:37 |
@wiking | ok things are rolling :) | 13:37 |
@HeikoS | they are :) | 13:37 |
@HeikoS | fingers crossed about our gsoc application | 13:37 |
@wiking | ah yeah | 13:37 |
@wiking | so 2 things | 13:37 |
@wiking | a) Gunnar | 13:37 |
@wiking | b) joining an umbrella org | 13:38 |
@wiking | and c) i've prepared a draft for the new release | 13:38 |
@wiking | https://github.com/shogun-toolbox/shogun/releases | 13:38 |
@wiking | for the release actually i need to fix 2 things | 13:38 |
@wiking | 1) modular octave force 1 thread | 13:38 |
@HeikoS | wiking: looks good | 13:39 |
@HeikoS | wiking: you are on fire ;) | 13:39 |
@wiking | 2) fix all the failing stuff that are caused in the analytics | 13:39 |
@HeikoS | yes I saw the email | 13:39 |
@wiking | build jobs | 13:39 |
@wiking | ah and the last thing | 13:39 |
@wiking | i need to look at the notebooks | 13:39 |
@wiking | ipython | 13:39 |
@wiking | i reckon they are now not being all generated | 13:39 |
@wiking | ah and fuck | 13:39 |
@wiking | the notebook jobs | 13:40 |
@wiking | :< | 13:40 |
@wiking | that actually is required for the PR job as well | 13:40 |
@wiking | fucking old cmake on 7nn.de sonney2k ! | 13:40 |
@wiking | so yeah there are like 5 more tasks | 13:40 |
@wiking | but after that | 13:40 |
@wiking | it's releasable | 13:40 |
@wiking | :DDDDD | 13:40 |
@HeikoS | hehe | 13:41 |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 13:41 |
@HeikoS | wiking: so what about merginng linalg soon? | 13:47 |
@sukey | Pull Request #3627 "Make meta_examples depend on python generator and language files." synchronized by geektoni - https://github.com/shogun-toolbox/shogun/pull/3627 | 13:47 |
@sukey | Pull Request #3627 "Make meta_examples depend on python generator and language files." merged by karlnapf - https://github.com/shogun-toolbox/shogun/pull/3627 | 13:48 |
@sukey | New Commit "Merge pull request #3627 from geektoni/patch-5 | 13:48 |
@sukey | Make meta_examples depend on python generator and language files." to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/618a2989e3a0b0412bd196856f410c9fd93c09bb | 13:48 |
@wiking | yeah that shit should go | 13:48 |
@wiking | but we should get | 13:48 |
@HeikoS | wiking: to get rid of all the cmake fuck | 13:48 |
@HeikoS | for linalg | 13:48 |
@wiking | Pan to fix the multithreading/atomicity stuff done | 13:48 |
@HeikoS | ah yes | 13:48 |
@wiking | i can try rebasing | 13:48 |
@wiking | the feature branch | 13:48 |
@HeikoS | wiking: I mean it is not yet used in develop a lot | 13:48 |
@HeikoS | wiking: yeah good idea | 13:49 |
@wiking | ah and there i think | 13:49 |
@wiking | we should actually use the new feature of cmake 3.1 | 13:49 |
@wiking | https://cmake.org/cmake/help/v3.1/variable/CMAKE_CXX_STANDARD.html | 13:49 |
@wiking | so you can just do this | 13:49 |
@wiking | set(CMAKE_CXX_STANDARD 11) | 13:49 |
@wiking | :> | 13:49 |
@HeikoS | thats cool! | 13:49 |
@HeikoS | wiking: ah man sweet with the python ppa | 13:50 |
@HeikoS | Ill add to the website readme I guess? | 13:50 |
@wiking | yeah and the full restruct of cmake | 13:50 |
@wiking | makes things much much more cleaner | 13:50 |
@wiking | yeah we should | 13:50 |
@wiking | now i'm working on the jupytherhub | 13:50 |
@wiking | :) | 13:50 |
@wiking | that should fucking kill it :) | 13:50 |
@wiking | jupyterhub + aws | 13:50 |
@wiking | and then we blog a bit about it | 13:50 |
@HeikoS | wiking: so the python module is in the existing ppa, the same package | 13:51 |
@HeikoS | wiking: yeah man | 13:51 |
@wiking | yes | 13:51 |
@HeikoS | actually | 13:51 |
@wiking | so just add-apt-repository | 13:51 |
@wiking | and then | 13:51 |
@HeikoS | maybe lea can write that, she wanted to do something | 13:51 |
@wiking | apt-get install python-shogun | 13:51 |
@wiking | yeah that'd be huge help | 13:51 |
@wiking | and | 13:51 |
@wiking | we should put there the credz for AWS | 13:51 |
@wiking | for giving us this | 13:51 |
@HeikoS | yep yep yep | 13:51 |
@wiking | so maybe we could get creds this year as well | 13:51 |
@wiking | because you know | 13:51 |
@HeikoS | yep absolutely | 13:51 |
@wiking | that our fundings are running low in may | 13:51 |
@wiking | and man | 13:51 |
@wiking | we barely used 1k | 13:52 |
@HeikoS | good timing because we need to reapply for credits | 13:52 |
@wiking | but they are gonna be liek | 13:52 |
@wiking | "wtf u asked for 15k | 13:52 |
@wiking | u used 1k | 13:52 |
@wiking | :D | 13:52 |
@HeikoS | man its going well :) | 13:53 |
@HeikoS | just doiscussing with lea the post | 13:53 |
@HeikoS | brb | 13:53 |
@wiking | HeikoS, $1,199.80 | 13:53 |
@wiking | :) | 13:53 |
@wiking | used creds | 13:53 |
lisitsyn | hey | 13:53 |
@wiking | sup? | 13:53 |
lisitsyn | what's going on | 13:54 |
lisitsyn | :) | 13:54 |
@wiking | lisitsyn, idk if you realised but i've just merged the other day commits of yours from 2015 | 13:54 |
@wiking | :) | 13:54 |
lisitsyn | really | 13:54 |
lisitsyn | which ones? | 13:54 |
lisitsyn | oh | 13:54 |
lisitsyn | oh god IT WAS 2015 | 13:54 |
lisitsyn | crazyiest shietz | 13:55 |
@wiking | :> | 13:55 |
@wiking | but yeah gut stuff | 13:55 |
@wiking | man | 13:55 |
@wiking | wanted to ask you | 13:55 |
@wiking | but now i cannot recall wtf | 13:56 |
lisitsyn | haha | 13:56 |
lisitsyn | wiking: so what are you thoughts on bazel or stuff like that? | 13:56 |
@wiking | mmm good shit | 13:56 |
@wiking | but man | 13:56 |
@wiking | i dont wanna do it | 13:56 |
@wiking | you know how shitty it is | 13:57 |
@wiking | ... | 13:57 |
@wiking | fuck | 13:57 |
lisitsyn | yeah it's like shitload of work | 13:57 |
@wiking | no fucking way | 13:57 |
@wiking | actually now | 13:57 |
@wiking | cmake is in a very very good state | 13:57 |
@wiking | it's almost totally devided | 13:57 |
@wiking | into subprojects | 13:57 |
lisitsyn | yeah true | 13:57 |
@wiking | libshogun, python | 13:57 |
lisitsyn | no need to bother | 13:57 |
@wiking | everything | 13:57 |
@wiking | i removed a lot of shit from the main cmakelists.txt | 13:57 |
@wiking | (the one in the root) | 13:57 |
lisitsyn | one other thought I have | 13:57 |
lisitsyn | we may store our basic stuff in protos | 13:57 |
@wiking | now you can actually build separately modular interfaces from libshogun | 13:58 |
lisitsyn | I mean it may be some logic over protobufs | 13:58 |
@wiking | yeah | 13:58 |
@wiking | i'm totally ++ for that | 13:58 |
lisitsyn | this way people can introspect anything | 13:58 |
@wiking | makes things very clean | 13:58 |
@wiking | and yeah exactly | 13:58 |
lisitsyn | ok glad you like it | 13:58 |
@wiking | you dont care about for swig interface at that time | 13:58 |
@wiking | you just get the generated code | 13:58 |
@wiking | and fuck all | 13:58 |
@wiking | you use that directly | 13:58 |
@wiking | there are couple of things i head in mind actuall | 13:58 |
@wiking | y | 13:59 |
lisitsyn | are you talking about protobuf interfaces? | 13:59 |
lisitsyn | that's even better | 13:59 |
@wiking | yes | 13:59 |
@wiking | i would totally use | 13:59 |
@wiking | .proto | 13:59 |
@wiking | for defining interface | 13:59 |
@wiking | why not? | 13:59 |
lisitsyn | yeah sure | 13:59 |
@wiking | then you can even rpc | 13:59 |
lisitsyn | it is good as well | 13:59 |
@wiking | and then all that shit | 13:59 |
@wiking | is totally transparent for you | 13:59 |
@wiking | is it called locally | 13:59 |
@wiking | is it called on the moon | 13:59 |
@wiking | you dont fucking care | 13:59 |
lisitsyn | not always good but | 13:59 |
lisitsyn | still | 13:59 |
lisitsyn | but again not only that | 13:59 |
lisitsyn | but | 14:00 |
lisitsyn | another thing | 14:00 |
lisitsyn | we may store everything in protos | 14:00 |
@wiking | but what do you mean by storing? | 14:00 |
lisitsyn | this way you may easily hack any classifier | 14:00 |
lisitsyn | well | 14:00 |
lisitsyn | say some state of classifier | 14:00 |
lisitsyn | weights etc | 14:00 |
lisitsyn | make it exposable | 14:00 |
@wiking | ah and u use proto for that? | 14:00 |
lisitsyn | yes | 14:00 |
@wiking | sure | 14:00 |
@wiking | that sounds very reasonable for me | 14:00 |
@wiking | :) | 14:00 |
lisitsyn | so it is like some code operating on protobufs | 14:01 |
@wiking | yeah | 14:01 |
lisitsyn | sounds like google | 14:01 |
@wiking | let's have like | 14:01 |
@wiking | JIT for proto | 14:01 |
@wiking | right? | 14:01 |
@wiking | :D | 14:01 |
lisitsyn | haha | 14:01 |
lisitsyn | why | 14:01 |
@wiking | i mean you can totally do that btw | 14:01 |
@wiking | nativeJIT + protoc | 14:01 |
@wiking | and fuck all | 14:01 |
@wiking | :) | 14:01 |
lisitsyn | not sure what are advantages | 14:01 |
lisitsyn | just to speed it up? | 14:01 |
@wiking | ha well | 14:02 |
@wiking | if you do autodiff | 14:02 |
@sukey | Pull Request #3608 "Clean up KNN" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3608 | 14:02 |
@wiking | or anything where you actually need some sort of codegen | 14:02 |
@wiking | right? | 14:02 |
@HeikoS | re | 14:02 |
@wiking | in runtime | 14:02 |
-!- praisethemoon [~praisethe@197.3.11.178] has quit [Ping timeout: 268 seconds] | 14:02 | |
lisitsyn | ahh ok | 14:02 |
@wiking | btw so i was wondering if we wanna look at this YOLO project https://arrow.apache.org/ | 14:03 |
@wiking | for interoperable memory storage | 14:03 |
lisitsyn | I am not sure this should go into shogun | 14:04 |
lisitsyn | I mean there are dozens of them | 14:04 |
lisitsyn | we'd better make it interoperable with all that stuff | 14:04 |
@wiking | yteah thats' what i meant | 14:04 |
lisitsyn | no idea how though | 14:04 |
@wiking | yeah that is a good question | 14:05 |
@wiking | i mean now i'm looking into thrill | 14:05 |
@wiking | and wondering how we could use like making batch models with shogun | 14:06 |
@wiking | and then aggregate for example | 14:06 |
@wiking | ah and one more thing i had in mind | 14:06 |
@wiking | remember the other day | 14:06 |
@wiking | i was wishing i would have reflection in c++ | 14:06 |
@wiking | so now that whenever in the distance future | 14:06 |
@wiking | we manage to get this modularization done | 14:07 |
@wiking | it'd be cool to have a tool as well | 14:07 |
@wiking | that would help you to create pipelines | 14:07 |
@wiking | of these modules | 14:07 |
@HeikoS | wiking: first step, being able to serialize our models :) | 14:07 |
lisitsyn | what reflection do you need? | 14:07 |
@wiking | where the execution is actually a just a config file | 14:07 |
@wiking | not code | 14:07 |
@wiking | so wha ti meant is say | 14:08 |
@HeikoS | wiking: you got a link to that openmp patch you mentioned earlier, cannot find it somehow | 14:08 |
@wiking | - CSVFileread | 14:09 |
@wiking | - input: "file.csv" | 14:09 |
@wiking | - Label | 14:09 |
@wiking | - labels: input[0] | 14:09 |
@wiking | etc etc. | 14:09 |
@HeikoS | ah got it | 14:09 |
@wiking | oh cool | 14:09 |
@wiking | lisitsyn, see what i mean? | 14:10 |
@sukey | Wiki page: GSoC_2017_project_fundamental_usual_suspects edited on shogun-toolbox/shogun by karlnapf | 14:10 |
lisitsyn | wiking: yeah | 14:10 |
@wiking | but for that | 14:10 |
@wiking | i need a way to do reflection right? | 14:11 |
lisitsyn | ahh | 14:11 |
lisitsyn | ok | 14:11 |
@wiking | i mean with our sgserialization (class_list.cpp) | 14:11 |
@wiking | we have our class factory | 14:11 |
@wiking | so that's cool | 14:11 |
@wiking | but then | 14:11 |
@wiking | how do i know that i can call Label.set_labels(...) | 14:12 |
@wiking | see what i mean? | 14:12 |
@wiking | without reflection | 14:12 |
@wiking | or fuckthis shit and jsut do this in a language that has reflection? :D | 14:12 |
@sukey | Wiki page: GSoC_2017_detox edited on shogun-toolbox/shogun by karlnapf | 14:12 |
@wiking | i mean the great stuff with tags | 14:12 |
@wiking | that at least with that | 14:13 |
@wiking | doing something like | 14:13 |
@wiking | - CSVM: | 14:13 |
@wiking | - param1: value | 14:13 |
@wiking | - param2: value | 14:13 |
@wiking | - param3: value | 14:13 |
@wiking | would be possible easily | 14:13 |
@wiking | right? | 14:13 |
@wiking | as that is just translated into | 14:13 |
@wiking | obj.set(key, value) | 14:13 |
@wiking | but when you wanna call a function with the right args | 14:14 |
@wiking | things gets messy | 14:14 |
@wiking | but yeah i dont see the reason why couldn't we have like a scripting language or something | 14:15 |
@wiking | where we could create these pipelines | 14:15 |
@wiking | because this would help easy prototyping | 14:16 |
@wiking | and actually | 14:16 |
@wiking | the good stuff with your idea of protos | 14:16 |
@wiking | that you could use those interfaces to define these pipelines :) | 14:16 |
@sukey | Issue #3628 "Surpress "unused variable" warnings in compiled meta examples" opened by karlnapf - https://github.com/shogun-toolbox/shogun/issues/3628 | 14:19 |
@sukey | Issue #3628 "Surpress "unused variable" warnings in compiled meta examples" karlnapf added label: "Cleanups" - https://github.com/shogun-toolbox/shogun/issues/3628 | 14:19 |
@sukey | Issue #3628 "Surpress "unused variable" warnings in compiled meta examples" karlnapf added label: "entrance" - https://github.com/shogun-toolbox/shogun/issues/3628 | 14:19 |
-!- abhinav [~abhinav@103.211.52.154] has joined #shogun | 14:19 | |
@sukey | Issue #3628 "Surpress "unused variable" warnings in compiled meta examples"- https://github.com/shogun-toolbox/shogun/issues/3628 | 14:20 |
-!- HeikoS [~heiko@host-92-0-178-129.as43234.net] has quit [Ping timeout: 260 seconds] | 14:23 | |
-!- abhinav [~abhinav@103.211.52.154] has quit [Quit: Leaving] | 14:24 | |
lisitsyn | wiking: ok got it | 14:25 |
@wiking | lisitsyn, and what's your opinion? :) | 14:47 |
@sukey | Pull Request #3620 "Add time comparison for KNN_KDTREE" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3620 | 15:08 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 15:13 | |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has joined #shogun | 16:10 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 16:10 | |
@wiking | lol | 16:22 |
@wiking | https://github.com/search?utf8=%E2%9C%93&q=remove+password&type=Commits&ref=searchresults | 16:22 |
@wiking | :>> | 16:22 |
@sukey | Wiki page: Google-Summer-of-Code-2017-Projects edited on shogun-toolbox/shogun by karlnapf | 16:23 |
@sukey | Wiki page: GSoC_2017_detox edited on shogun-toolbox/shogun by karlnapf | 16:24 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit created on shogun-toolbox/shogun by karlnapf | 16:25 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 16:25 |
CaBa_ | wiking: yeah, since this was on reddits especially the page 1 hits get commented on their commits ^^ | 16:26 |
CaBa_ | 16:26 | |
@wiking | :) | 16:26 |
@sukey | Issue #3612 "Stratified Cross Validation with Combined Kernels using Custom Kernels"- https://github.com/shogun-toolbox/shogun/issues/3612 | 16:27 |
CaBa_ | wiking: "I'm a bit disappointed now that putting 'protected' in front of the password doesn't protect it ;)" | 16:27 |
@wiking | :>>>>>>>>>>>> | 16:27 |
@wiking | HeikoS, btw i think | 16:27 |
@wiking | we should try re-revert | 16:27 |
@wiking | the commit for the multithreaded xval | 16:27 |
@wiking | :D | 16:27 |
@HeikoS | wiking: hehe :) | 16:27 |
@HeikoS | yeah | 16:27 |
@HeikoS | do it! :) | 16:27 |
@HeikoS | or hsall I? | 16:27 |
@HeikoS | wiking: btw had a cool idea for another gsoc project | 16:28 |
@HeikoS | potentially a joint one where everyone has to participate | 16:28 |
@wiking | self-driving cars? :D | 16:28 |
@HeikoS | wiking: The low hanging fruit (TM) | 16:28 |
@HeikoS | will put it in wiki later today | 16:28 |
@wiking | cool | 16:28 |
@HeikoS | wiking: btw for release | 16:29 |
@HeikoS | I guess we put all issues in the 5.1 milestone to 5.2 | 16:29 |
@wiking | mmmmmmmm | 16:29 |
@wiking | i would do a 6.0 | 16:29 |
@wiking | or here goes a scary question | 16:29 |
@wiking | when do we add c++11 required? | 16:29 |
@HeikoS | linalg merge | 16:29 |
@wiking | and that wold be 5.2? | 16:29 |
@wiking | *would | 16:29 |
@HeikoS | yeah | 16:29 |
@wiking | no | 16:29 |
@wiking | then go 6.0 | 16:30 |
@HeikoS | also good | 16:30 |
@wiking | because that's a big change | 16:30 |
@wiking | different libc++ dependency | 16:30 |
@wiking | etc | 16:30 |
@HeikoS | I guess once we merge linalg, tags, and bigtest | 16:30 |
@HeikoS | then we can do it for that | 16:30 |
@HeikoS | I want to do these 3 before gsoc starts | 16:30 |
@HeikoS | and maybe even bsd | 16:30 |
@HeikoS | then 6.0 would be bsd | 16:30 |
@wiking | mmm we can do | 16:30 |
@wiking | 6.0 with c++11 | 16:30 |
@wiking | 7.0 with bsd | 16:30 |
@wiking | no problem | 16:30 |
@wiking | :D | 16:30 |
@wiking | come on | 16:30 |
@HeikoS | also good :) | 16:30 |
@HeikoS | I know | 16:30 |
@wiking | chrome is like | 16:30 |
@wiking | what | 16:30 |
@wiking | 4 years old | 16:30 |
@wiking | or 7 | 16:30 |
@wiking | and it's 5x. now | 16:31 |
@wiking | :) | 16:31 |
@HeikoS | hehe | 16:31 |
@wiking | lol | 16:31 |
@wiking | The browser was first publicly released on September 2, 2008 | 16:31 |
@wiking | so 9 years | 16:31 |
@wiking | so yeah i would maybe force more major ++ | 16:31 |
@wiking | wouldn't hurt | 16:31 |
@HeikoS | yeah good from my sidfe | 16:32 |
@wiking | cool | 16:32 |
@wiking | so who and when | 16:32 |
@wiking | rebases | 16:32 |
@wiking | linalg feature branch | 16:32 |
@HeikoS | I can rebase linalg | 16:32 |
@wiking | with current develop? | 16:32 |
@wiking | can u? | 16:32 |
@wiking | becaue then | 16:32 |
@HeikoS | I can, just not today maybe | 16:32 |
@wiking | after you fixed the rebase stuff | 16:33 |
@wiking | add on top | 16:33 |
@HeikoS | icml deadline next week aaaaah | 16:33 |
@wiking | a commit | 16:33 |
@wiking | that requires c++11 | 16:33 |
@wiking | the way i said | 16:33 |
@wiking | and drop all that fucking c++11 feature tester in the cmake dir | 16:33 |
@wiking | :D | 16:33 |
@wiking | or just rebase | 16:33 |
@wiking | and ping me | 16:33 |
@HeikoS | wiking: shall we do that in linalg branch? | 16:33 |
@wiking | and i can help | 16:33 |
@HeikoS | yeah | 16:33 |
@HeikoS | ok | 16:33 |
@wiking | yeyey | 16:33 |
@wiking | because linalg is fully c++11 | 16:33 |
@wiking | anyways no? | 16:33 |
@HeikoS | wiking: shall I way for the lock thing? | 16:33 |
@wiking | well rebase | 16:34 |
@wiking | and then we'll see | 16:34 |
@HeikoS | yeah true | 16:34 |
@wiking | when pan comes up with the multithreading | 16:34 |
@HeikoS | wiking: I think Pan is almost done with porting all the calls we had | 16:34 |
@wiking | cool | 16:34 |
@wiking | let's just have that thread-safe | 16:34 |
@wiking | and we are good | 16:34 |
@wiking | but i reckon | 16:34 |
@HeikoS | kool! | 16:34 |
@wiking | rebasing is not gonna be straightforward | 16:34 |
@wiking | :))) | 16:34 |
@HeikoS | dont konw | 16:34 |
@wiking | lisitsyn, buuuuuuuujaaaaaaaaa | 16:35 |
@HeikoS | I mean we didnt change too much in there | 16:35 |
@HeikoS | more like adding/removing | 16:35 |
@HeikoS | so lets see | 16:35 |
@wiking | lisitsyn, with tags, how do you imagine api documentation? :D | 16:35 |
@HeikoS | I can give it a go | 16:35 |
@HeikoS | wiking: discussed that with him acutally | 16:35 |
@HeikoS | but I forgot :) | 16:35 |
@HeikoS | we had like a two layer thing | 16:35 |
@HeikoS | core API docs | 16:35 |
@HeikoS | and then one can have things for plugins documented by example | 16:35 |
lisitsyn | uhm | 16:37 |
lisitsyn | its dynamic | 16:37 |
@wiking | yeah i know | 16:37 |
@HeikoS | shogun goes python | 16:37 |
@wiking | but man | 16:37 |
@HeikoS | no more member variables | 16:37 |
@wiking | we need a way to document it | 16:37 |
@wiking | :) | 16:37 |
@HeikoS | dynamic docs | 16:37 |
@HeikoS | haha | 16:37 |
@wiking | i mean what | 16:37 |
@wiking | you | 16:38 |
@wiking | say | 16:38 |
@wiking | try any fucking combination | 16:38 |
@wiking | ? | 16:38 |
@wiking | and see what sticks? | 16:38 |
@wiking | :D | 16:38 |
@HeikoS | I think first step is the SWIG interface docs for the core API | 16:38 |
@HeikoS | that should be really polished | 16:38 |
@HeikoS | and thats aside any algorithm implementations | 16:39 |
@HeikoS | doxygen | 16:39 |
@HeikoS | static | 16:39 |
@HeikoS | and then for the implementations, we can either document the possibilities by example (thats minimal thing) | 16:39 |
@HeikoS | or the thing can tell you what it does | 16:39 |
@HeikoS | at runtime | 16:39 |
@HeikoS | lisitsyn: thats kind of what we had in mind last time right? | 16:40 |
@wiking | yeah but still | 16:40 |
@HeikoS | wiking: dont like it? :D I would have guessed that | 16:40 |
@wiking | nono | 16:40 |
@wiking | what i mean is | 16:40 |
@wiking | i dont fucking care of swig now | 16:40 |
@wiking | api docs for it | 16:40 |
@wiking | how do you create documentation | 16:41 |
@wiking | for the c api? | 16:41 |
@wiking | i mean dont tell me that you dont need it | 16:41 |
@wiking | because then wtf? | 16:41 |
@wiking | how do you write code for it? | 16:41 |
@HeikoS | yeah true | 16:41 |
@HeikoS | nono you need it | 16:41 |
@HeikoS | all the stuff that is not exposed via swig but that you might want to use still, right? | 16:42 |
@wiking | no i mean even the tags | 16:42 |
@wiking | how do you know | 16:42 |
@wiking | in the c++ | 16:42 |
@wiking | which tag is valid? | 16:42 |
@HeikoS | which members are there? | 16:42 |
@wiking | i mean vlid you will know | 16:42 |
@wiking | by trial and error i guess | 16:42 |
@wiking | but isn't there a better way? :D | 16:42 |
@HeikoS | with the tag system we have, you cannot know that as they are generated at runtime | 16:42 |
@wiking | and then what | 16:43 |
@HeikoS | I mean we can have methods that give you the tags | 16:43 |
@wiking | how do you know what are the params | 16:43 |
@wiking | for CSVM? | 16:43 |
@HeikoS | but not sure how to generate docs | 16:43 |
@HeikoS | how is that problem solved in python docs? | 16:44 |
@wiking | mmm define them in the docstrings | 16:45 |
@wiking | afaik | 16:45 |
@wiking | but then again | 16:45 |
@wiking | you can just do that | 16:45 |
@wiking | for vars that are func args | 16:45 |
@wiking | dunno if you can actually define dynamically added | 16:45 |
@wiking | member args | 16:45 |
@wiking | and btw | 16:45 |
@wiking | how the hell you validate | 16:45 |
@wiking | that you have set all the vars | 16:45 |
@wiking | for the model | 16:45 |
@wiking | you'll have a validate tags? :d | 16:45 |
@wiking | and even better i might add | 16:46 |
@HeikoS | wiking: the model will default value all things of course | 16:46 |
@wiking | how do you enforce in this case the type? | 16:46 |
@HeikoS | and then you can change if wanted | 16:46 |
@wiking | HeikoS, and then if you try to set | 16:46 |
@wiking | set("C1", "adfg") | 16:46 |
@wiking | then what happens? | 16:46 |
@wiking | is it exception because not the right type? | 16:46 |
@HeikoS | set is type safe at runtime | 16:47 |
@HeikoS | yes | 16:47 |
@wiking | or is it set to a string? | 16:47 |
@wiking | ? | 16:47 |
@HeikoS | nono | 16:47 |
@HeikoS | first | 16:47 |
@HeikoS | because | 16:47 |
@HeikoS | C1 is set in constructor | 16:47 |
@HeikoS | so the type is fixed then | 16:47 |
CaBa_ | you two with your same nicklength... really tough to read through your conversations O.O | 17:01 |
-!- CaBa_ [~Diu7saig@lounge.imp.fu-berlin.de] has quit [Quit: Changing server] | 17:06 | |
-!- CaBa [~Diu7saig@unaffiliated/caba] has joined #shogun | 17:06 | |
lisitsyn | HeikoS: wiking well | 17:10 |
lisitsyn | we can extract it from code | 17:10 |
lisitsyn | not a big deal | 17:10 |
lisitsyn | if you're talking about easiest way | 17:11 |
lisitsyn | no idea yet | 17:11 |
lisitsyn | I am not a big fan of docs | 17:11 |
lisitsyn | in python you try things at runtime | 17:11 |
@wiking | and then what | 17:23 |
@wiking | we encourage to use cling? :) | 17:23 |
-!- travis-ci [~travis-ci@ec2-54-92-224-73.compute-1.amazonaws.com] has joined #shogun | 17:30 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201852987 | 17:30 |
-!- travis-ci [~travis-ci@ec2-54-92-224-73.compute-1.amazonaws.com] has left #shogun [] | 17:30 | |
@HeikoS | I am actually not a big fan of docs either, apart from a very stable set of base API docs | 17:33 |
@wiking | and then what | 17:37 |
@wiking | trial and erro? | 17:37 |
@wiking | i mean come on guys | 17:38 |
@wiking | lets be a bit more serious | 17:38 |
@wiking | ;0 | 17:38 |
@wiking | anyhow | 17:40 |
@wiking | i need to sleep | 17:40 |
@wiking | :) | 17:40 |
@wiking | ttyl | 17:40 |
@wiking | one more thing | 17:41 |
@wiking | can we move the site (website) somewhere else? | 17:41 |
@wiking | otherwise this story with the cookbook | 17:41 |
@wiking | is a mess | 17:41 |
@HeikoS | wiking: move somewhere else as in server? | 17:43 |
@HeikoS | wiking: I mean yeah whatever helps | 17:43 |
@HeikoS | wiking: lets discuss this doc thing again | 17:43 |
@HeikoS | maybe its a good one for the hackathon | 17:43 |
-!- mikeling [uid89706@gateway/web/irccloud.com/x-dajmozufoyahcrni] has quit [Quit: Connection closed for inactivity] | 19:09 | |
@sukey | Issue #1415 "Static interfaces support needed in cmake" closed by karlnapf - https://github.com/shogun-toolbox/shogun/issues/1415 | 19:53 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 19:57 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 19:59 |
@HeikoS | lisitsyn: | 20:00 |
@HeikoS | jo! | 20:00 |
@HeikoS | still around? | 20:00 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 20:01 |
@HeikoS | https://github.com/shogun-toolbox/shogun/wiki/GSoC_2017_low_hanging_fruit | 20:01 |
@HeikoS | feel free to add things | 20:01 |
@sukey | Wiki page: GSoC_2017_low_hanging_fruit edited on shogun-toolbox/shogun by karlnapf | 20:02 |
@sukey | Wiki page: Getting-involved edited on shogun-toolbox/shogun by karlnapf | 20:05 |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Remote host closed the connection] | 20:05 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 20:08 | |
@sukey | New Commit "bump to latest version" to shogun-toolbox/shogun by karlnapf: https://github.com/shogun-toolbox/shogun/commit/eceda2dc3a5ed163e834e3ca2f70d038f2a99f9b | 20:11 |
@sukey | Wiki page: Getting-involved edited on shogun-toolbox/shogun by lgoetz | 20:13 |
@sukey | Wiki page: Getting-involved edited on shogun-toolbox/shogun by lgoetz | 20:14 |
@sukey | Wiki page: Getting-involved edited on shogun-toolbox/shogun by lgoetz | 20:14 |
lisitsyn | HeikoS: yes | 20:17 |
lisitsyn | back | 20:17 |
-!- HeikoS [~heiko@untrust-out.swc.ucl.ac.uk] has quit [Ping timeout: 240 seconds] | 20:28 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has quit [Remote host closed the connection] | 20:40 | |
-!- goksinen [~goksinen@rrcs-50-75-193-138.nyc.biz.rr.com] has joined #shogun | 20:42 | |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has joined #shogun | 21:01 | |
travis-ci | it's Heiko Strathmann's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: https://travis-ci.org/shogun-toolbox/shogun/builds/201982436 | 21:01 |
-!- travis-ci [~travis-ci@ec2-54-90-247-33.compute-1.amazonaws.com] has left #shogun [] | 21:01 | |
CaBa | wiking: ping | 21:58 |
--- Log closed Thu Feb 16 00:00:27 2017 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!