--- Log opened Mon Aug 29 00:00:20 2011 | ||
--- Day changed Mon Aug 29 2011 | ||
blackburn | 55k$ per year?! the default salary in ohloh.net calculator rocks | 00:00 |
---|---|---|
blackburn | sonney2k: lets sell shogun :D | 00:00 |
@sonney2k | heh | 00:00 |
blackburn | currently my salary is 3k/year ahaha | 00:02 |
blackburn | sonney2k: make in examples currently requires doc for .class files | 00:05 |
blackburn | in java | 00:06 |
@sonney2k | what? | 00:06 |
blackburn | documentation for "MatrixTest.class" (file "descriptions/modular/MatrixTest.txt") missing | 00:08 |
blackburn | documentation for "MatrixTest.java" (file "descriptions/modular/MatrixTest.txt") missing | 00:08 |
@sonney2k | erase the class files before doing make doc or fix the generate_documented.sh to exclude .exe / .class files | 00:12 |
@sonney2k | matrix/vector test should be removed btw | 00:15 |
CIA-3 | shogun: Sergey Lisitsyn master * r2b3ca25 / (2 files): Remove Matrix/VectorTest from java modular examples - http://git.io/pEV2Zw | 00:24 |
CIA-3 | shogun: Sergey Lisitsyn master * r28ce82a / (6 files): More doc for examples - http://git.io/pTtd-w | 00:37 |
CIA-3 | shogun: Sergey Lisitsyn master * rfb4ddae / (20 files): Kernel examples doc - http://git.io/zFOd6Q | 00:58 |
-!- in3xes [~in3xes@180.149.49.227] has joined #shogun | 01:16 | |
shogun-buildbot | build #182 of libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/libshogun/builds/182 blamelist: blackburn91@gmail.com | 01:34 |
serialhex | HA!!! blackburn GOT BLAMED!!!!! IT'S UR FAULT!!!!!! | 01:50 |
serialhex | btw, is sonney2k around still?? | 01:50 |
serialhex | hmm, asleep probably, i'll e-mail ya the problem i'm getting (same problem, 2 files) | 01:51 |
blackburn | serialhex: definitely not my fault ;) | 01:53 |
serialhex | no?? shogun-buildbot says different: "blamelist: blackburn91@gmail.com" | 01:53 |
blackburn | yeah but take a look on log | 01:54 |
blackburn | compiler got segfault hah | 01:54 |
serialhex | ooh, lovely! | 01:55 |
serialhex | you just got blamed b/c you were the last to commit, and with a bunch of textfiles | 01:55 |
blackburn | yeah | 01:56 |
serialhex | i would think it would blame everyone who commited between the time it last built & the time it compiled | 01:57 |
serialhex | hey, do you have python modular compiled & installed?? | 01:59 |
serialhex | ^ blackburn | 01:59 |
blackburn | serialhex: I could compile now | 01:59 |
serialhex | well i want to test something, if it's not too much of a pain for you | 02:00 |
serialhex | though the ruby would be better | 02:00 |
shogun-buildbot | build #163 of octave_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/octave_modular/builds/163 blamelist: blackburn91@gmail.com | 02:00 |
serialhex | cause then i could give you the code thats bein a pain for me | 02:00 |
blackburn | so what to install? | 02:01 |
blackburn | ruby-modular? | 02:01 |
shogun-buildbot | build #147 of ruby_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/ruby_modular/builds/147 blamelist: blackburn91@gmail.com | 02:01 |
blackburn | buildbot goes crazy | 02:02 |
blackburn | serialhex: got python installed, what to do? | 02:03 |
serialhex | see if distribution_histogram_modular.py in examples/undoc.../pyth_mod works | 02:04 |
serialhex | (yes, ui'm lazy) | 02:04 |
serialhex | when i run the ruby version iget a wonky error, and i'm not sure if it's just ruby or if it's shogun in general | 02:06 |
blackburn | serialhex: yeah it works | 02:08 |
blackburn | what is the error you get? | 02:08 |
serialhex | ...i'll gist it | 02:09 |
serialhex | https://gist.github.com/1fe53c321c001132e6c4 | 02:10 |
serialhex | maybe it's soemthing silly i'm missing in the translation between python & ruby.... | 02:10 |
serialhex | ^ blackburn | 02:10 |
shogun-buildbot | build #174 of cmdline_static is complete: Exception [exception git] Build details are at http://www.shogun-toolbox.org:5347/builders/cmdline_static/builds/174 blamelist: blackburn91@gmail.com | 02:11 |
shogun-buildbot | build #152 of python_modular is complete: Exception [exception git] Build details are at http://www.shogun-toolbox.org:5347/builders/python_modular/builds/152 blamelist: sonne@debian.org | 02:11 |
blackburn | serialhex: yeah I know what the error is | 02:13 |
serialhex | ??? please tell!!! | 02:13 |
blackburn | serialhex: check if your load_dna removes '\n' char | 02:14 |
blackburn | if yes so it could be typemap error | 02:14 |
serialhex | hmm, ok | 02:15 |
blackburn | serialhex: there are 120 lines in fm_test_dna.dat | 02:17 |
blackburn | err | 02:17 |
blackburn | 92 in train_dna | 02:17 |
serialhex | hmm, it was joining them nto an array, each line a new element in the array | 02:17 |
blackburn | so hist[0]=92 | 02:17 |
blackburn | is number of '\n's | 02:18 |
blackburn | does it remove \n char? | 02:18 |
serialhex | yes, it did, but then i changed it to put the entire file into the array, or just use the entire file, nothing :( | 02:18 |
blackburn | serialhex: try to cut lines | 02:20 |
serialhex | hmm... something wonky is going on... | 02:20 |
blackburn | e.g. take half of all of them | 02:20 |
serialhex | oh, derr.... | 02:20 |
blackburn | if it is still failing - I suspect only typemap | 02:20 |
blackburn | serialhex: so how? | 02:23 |
serialhex | i forgot, the version of modshogun.rb i'm using is *not* being loaded, so nothing i do in it will fix anything, but i'm going to copy it over and see if i can fix the problem... | 02:24 |
blackburn | serialhex: but it is ruby load | 02:25 |
blackburn | not modshogun.rb related | 02:25 |
serialhex | well the modshogun.rb has the load_whatever functions in it | 02:25 |
blackburn | ahhhh | 02:26 |
blackburn | I see | 02:26 |
blackburn | okay | 02:26 |
blackburn | serialhex: just make a try with some not loaded data but ACTG | 02:26 |
serialhex | hmm, ok | 02:26 |
blackburn | you can be sure in this way there are no '\n' | 02:26 |
serialhex | hmm... with or w/o the \n i get a new error with this: traindna = "ACTG" | 02:28 |
serialhex | distribution_histogram_modular.rb:16:in `set_features': Wrong arguments for overloaded method 'StringCharFeatures.set_features'. | 02:29 |
serialhex | ^ is trhe new error :P | 02:29 |
blackburn | serialhex: try array of "ACTG" "TCGA" | 02:29 |
blackburn | [] or so don't know how it is in ruby | 02:29 |
blackburn | :) | 02:29 |
serialhex | nope, same error as before with traindna = ["ACTG\n", "TGCA\n"] | 02:30 |
blackburn | wrong number?? | 02:30 |
blackburn | hmm | 02:30 |
serialhex | also w/o the newlines it dosnt work... | 02:32 |
serialhex | so i'm thinking the typemap... | 02:32 |
blackburn | serialhex: so [] things doesn't work anyhow? | 02:32 |
serialhex | nope, apparently not | 02:33 |
blackburn | serialhex: so what returns load_dna ? | 02:34 |
blackburn | [] too | 02:35 |
blackburn | serialhex: okay sorry but it is 4-35 there already :) I have to sleep | 02:35 |
serialhex | ["GAGACGGACCGTATGGGCAGGAT", | 02:35 |
serialhex | "GCGCATATTGTAGAGTATGGAGG", | 02:35 |
serialhex | ... and so on | 02:35 |
blackburn | hmm wait | 02:35 |
serialhex | ok, np... i'll e-mail sonney2k & sploving | 02:36 |
blackburn | so how can ["ACTG", "TACG"] doesn't | 02:36 |
blackburn | work | 02:36 |
blackburn | but ^ do? | 02:36 |
serialhex | it dosnt | 02:36 |
serialhex | none of them work | 02:36 |
blackburn | serialhex: I mean when you use load_dna it have different error about alphabet? | 02:36 |
blackburn | but if shorter ACTG version it says about wrong num of argz | 02:37 |
serialhex | nope, when it's not in an array i get a different error, but when it's in an array i get the: [ERROR] ALPHABET too small to contain all symbols in histogram | 02:37 |
blackburn | can't understand the difference between load_dna result and [] | 02:37 |
serialhex | with or without newlines | 02:37 |
serialhex | load_dna result is a longer array | 02:38 |
blackburn | serialhex: so will ["ACTG", "TACG"] produce alphabet error? | 02:38 |
serialhex | yes | 02:38 |
serialhex | and so will ["ACTG\n", "TGCA\n"] | 02:38 |
blackburn | serialhex: how many histogram elements there are with "ACTG" ? | 02:39 |
blackburn | 5? | 02:39 |
serialhex | 2 it says | 02:39 |
blackburn | no | 02:40 |
blackburn | hist[] = | 02:40 |
blackburn | how many records of this form ^ | 02:40 |
blackburn | ? | 02:40 |
serialhex | hist[0]=2 | 02:40 |
serialhex | hist[65]=2 | 02:40 |
serialhex | hist[67]=2 | 02:40 |
serialhex | hist[71]=2 | 02:40 |
serialhex | hist[84]=2 | 02:40 |
blackburn | 5 | 02:40 |
blackburn | okay | 02:40 |
blackburn | and with ACTG\n? | 02:40 |
serialhex | hist[0]=2 | 02:41 |
serialhex | hist[10]=2 | 02:41 |
serialhex | hist[65]=2 | 02:41 |
serialhex | hist[67]=2 | 02:41 |
serialhex | hist[71]=2 | 02:41 |
serialhex | hist[84]=2 | 02:41 |
blackburn | serialhex: I have a fix but you probably have to recompile | 02:41 |
serialhex | ok... i should prob do that anyway with the latest :P | 02:42 |
blackburn | serialhex: can you please gist script you used for ^? | 02:42 |
serialhex | yes | 02:42 |
serialhex | https://gist.github.com/1fe53c321c001132e6c4 | 02:43 |
serialhex | thats the latest v of the file & error | 02:44 |
blackburn | compiling | 02:45 |
blackburn | serialhex: the problem is for some *STRANGE* reason baozeng copies len+1 of string chars | 02:46 |
blackburn | in "in" typemap | 02:46 |
serialhex | hmm, ok | 02:46 |
blackburn | serialhex: so ACTG is something 6567718400 | 02:47 |
serialhex | ok, sounds like a fun error :P | 02:47 |
shogun-buildbot | build #165 of octave_static is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/octave_static/builds/165 blamelist: blackburn91@gmail.com | 02:56 |
blackburn | serialhex: got wrong number of args.. | 03:03 |
serialhex | hmm... really? | 03:03 |
blackburn | yeah | 03:03 |
blackburn | strange | 03:03 |
serialhex | yes, it is :-/ | 03:03 |
blackburn | serialhex: w/o \n produces wrong number of args | 03:03 |
blackburn | w/ - alphabet error | 03:04 |
serialhex | well, get to bed, i'll e-mail sonney2k & sploving about it and maybe we can come up with something? | 03:04 |
serialhex | huh... | 03:04 |
serialhex | i dunno :( | 03:04 |
blackburn | serialhex: let they know that string takes *one* char more than needed | 03:06 |
blackburn | it is pretty certain | 03:06 |
serialhex | ok, will do | 03:06 |
blackburn | but no fix yet | 03:06 |
blackburn | okay 5-06 time to sleep ahah | 03:07 |
blackburn | see you | 03:07 |
-!- blackburn [~blackburn@109.226.89.85] has quit [Quit: Leaving.] | 03:08 | |
shogun-buildbot | build #151 of r_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/r_modular/builds/151 blamelist: sonne@debian.org | 03:13 |
-!- in3xes [~in3xes@180.149.49.227] has quit [Remote host closed the connection] | 03:23 | |
shogun-buildbot | build #154 of csharp_modular is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org:5347/builders/csharp_modular/builds/154 blamelist: sonne@debian.org | 03:40 |
shogun-buildbot | build #166 of octave_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/octave_static/builds/166 | 03:48 |
shogun-buildbot | build #164 of octave_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/octave_modular/builds/164 | 04:47 |
@sonney2k | serialhex, seems like the NULL is in the string | 05:59 |
serialhex | ahh, ok | 05:59 |
@sonney2k | guess the fix is trivial | 05:59 |
@sonney2k | just not use len+1 in the typemap but len | 06:00 |
serialhex | ok cool | 06:00 |
CIA-3 | shogun: Soeren Sonnenburg master * r50aac3e / src/interfaces/ruby_modular/swig_typemaps.i : don't use len+1 for the string but len - http://git.io/ET0xow | 06:02 |
@sonney2k | this ^ should help | 06:02 |
serialhex | ok, i'll pull & try | 06:03 |
shogun-buildbot | build #183 of libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/libshogun/builds/183 | 06:09 |
shogun-buildbot | build #163 of python_static is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/python_static/builds/163 | 06:15 |
@sonney2k | serialhex, did you make some progress on examples? | 06:17 |
serialhex | i've got a bunch... i think like10-20 or something? a lot of them just won't work though | 06:18 |
serialhex | ...like the preprocesor examples, i can't find the modules :-/ | 06:18 |
@sonney2k | serialhex, what happens? | 06:20 |
serialhex | well, like preprocessor_hessianlocallylinearembedding_modular.rb, i can't find a Modshogun::Hessian* anything.... | 06:21 |
@sonney2k | serialhex, then you don't have atlas/blas/lapack installed | 06:21 |
serialhex | ahh, ok | 06:22 |
serialhex | where is that in debian? just apt-get atlas?? | 06:22 |
@sonney2k | do apt-get install libblas-dev liblapack-dev libatlas-base-dev | 06:22 |
@sonney2k | then recompile and then it will work | 06:22 |
serialhex | ok, recompiling... my machine is slow so i'll know later in the am | 06:25 |
@sonney2k | serialhex, btw we had the same issue w/ '\0' being in the string in other $LANG | 06:27 |
@sonney2k | so I am pretty sure that this fixes | 06:27 |
@sonney2k | it | 06:27 |
serialhex | ahh, ok... | 06:27 |
serialhex | i'll let you know tomorrow | 06:27 |
serialhex | or later today :P | 06:27 |
@sonney2k | the script you had first should work - the one where you add '\n' won't | 06:27 |
serialhex | ok | 06:27 |
shogun-buildbot | build #155 of csharp_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/csharp_modular/builds/155 | 06:32 |
shogun-buildbot | build #150 of ruby_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/ruby_modular/builds/150 | 06:32 |
shogun-buildbot | build #152 of r_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/r_modular/builds/152 | 06:33 |
shogun-buildbot | build #153 of java_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/java_modular/builds/153 | 06:34 |
shogun-buildbot | build #153 of lua_modular is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org:5347/builders/lua_modular/builds/153 | 06:34 |
-!- heiko [~heiko@134.91.52.61] has joined #shogun | 12:42 | |
-!- heiko [~heiko@134.91.52.61] has quit [Ping timeout: 258 seconds] | 13:02 | |
-!- heiko [~heiko@134.91.52.61] has joined #shogun | 13:13 | |
@sonney2k | heiko, hi! | 13:52 |
heiko | sonney2k, hi | 13:52 |
heiko | dont worry I will finish the migration soon | 13:52 |
@sonney2k | how is it going? | 13:52 |
heiko | fine here, what about you? | 13:53 |
@sonney2k | ok, so soon means hours or days? | 13:53 |
heiko | It is already working, I am trying out some more examples to eliminate some possible bugs | 13:53 |
heiko | One thing sucks: I forgot my Laptop power cable at home and my remaining time is only 30mins | 13:54 |
heiko | I think I will send a pull request tomorrow | 13:54 |
@sonney2k | it would be optimal if we could try to work on the migrations this evening | 13:54 |
@sonney2k | uh | 13:54 |
heiko | then we will have to add the migraion methods to each class that has a changed parameter | 13:54 |
@sonney2k | yes - I would have wished to start that tonight if you are that far already | 13:54 |
@sonney2k | we have only 2 days before the intended release data so time is pressing a bit now | 13:55 |
heiko | yes true | 13:55 |
heiko | mmh, I have some urgent stuff to do this evening, I will move out of my flat on friday and nothing is done yet | 13:56 |
heiko | Lets do it like this | 13:56 |
@sonney2k | so if you have an example that shows how to use migrations and your migrations don't break anything I would like to merge even sooner | 13:56 |
heiko | I will go home now and work there offline | 13:57 |
@sonney2k | ok | 13:57 |
heiko | then I will put a "pull"-able patch together | 13:57 |
heiko | and come to university this evening to send it | 13:57 |
@sonney2k | or internet cafe or so | 13:57 |
heiko | yes, but its not far away from here | 13:57 |
@sonney2k | k | 13:58 |
heiko | I think it should work pretty fine | 13:58 |
heiko | at least for numerical stuff | 13:58 |
@sonney2k | would be great - we have everything running now http://shogun-toolbox.org:5347/waterfall | 13:58 |
@sonney2k | (all examples work) | 13:58 |
heiko | the most important parts are the actual migrate methods, which have to be written by hand for each changed parameter | 13:58 |
heiko | cool | 13:59 |
@sonney2k | and documentations is also updated http://shogun-toolbox.org/doc/en/latest/ | 13:59 |
@sonney2k | heiko, yeah but we didn't do *huge* changes | 13:59 |
@sonney2k | mostly additions | 13:59 |
@sonney2k | or sgvector replacement | 13:59 |
heiko | yes, I am trying that out in the moment | 13:59 |
@sonney2k | so if you have an example on how to ignore new things / sgvector stuff then we should be able to get the old test suite to work | 14:00 |
heiko | the map stuff works and also recursively replacing parameters using the migrate method | 14:00 |
heiko | ok, I will do that now and then send a patch this evening | 14:01 |
@sonney2k | ok | 14:01 |
@sonney2k | thanks | 14:01 |
heiko | I will be here working on it tomorrow before noon also | 14:01 |
heiko | Ok then, see you this evening | 14:01 |
@sonney2k | so there is a chance we can do it | 14:01 |
@sonney2k | cu | 14:01 |
-!- heiko [~heiko@134.91.52.61] has quit [Ping timeout: 258 seconds] | 14:07 | |
-!- alesis-novik [~alesis@188.74.87.206] has joined #shogun | 16:52 | |
-!- blackburn [~blackburn@188.168.2.146] has joined #shogun | 17:38 | |
CIA-3 | shogun: Soeren Sonnenburg master * race079d / (28 files): add further c# examples - http://git.io/WkwKjA | 17:50 |
@sonney2k | blackburn, heiko said that he will be back w/ a patch tonight :) | 17:52 |
blackburn | sonney2k: nice | 17:53 |
@sonney2k | in the evening already | 17:53 |
blackburn | any other news? | 17:53 |
@sonney2k | so we might give the transitions a shot | 17:53 |
@sonney2k | more c# examples | 17:53 |
@sonney2k | from daniel | 17:53 |
blackburn | I'm pretty busy today with all the moving to samara from my home | 17:53 |
@sonney2k | when are you leaving? | 17:53 |
blackburn | already in samar | 17:54 |
blackburn | a | 17:54 |
@sonney2k | ahh | 17:54 |
blackburn | so now unpacking aha | 17:54 |
blackburn | :D | 17:54 |
@sonney2k | gzip -d | 17:54 |
blackburn | will do something with examples or so this night | 17:55 |
@sonney2k | blackburn, hopefully you have some time tonight - maybe we can get the first few migration related things to work | 17:55 |
@sonney2k | if things run smoothly - then we are good | 17:55 |
@sonney2k | blackburn, btw octave is broken on the buildbot (still) so tests related to it might be a bit unreliable... | 17:55 |
CIA-3 | shogun: Soeren Sonnenburg master * rfc63be5 / src/CONTRIBUTIONS : mention daniel in contributions - http://git.io/74eg_Q | 18:03 |
blackburn | sonney2k: why octave is broken? | 19:34 |
-!- blackburn [~blackburn@188.168.2.146] has quit [Ping timeout: 258 seconds] | 19:41 | |
-!- blackburn [~blackburn@188.168.5.154] has joined #shogun | 19:43 | |
@sonney2k | debian unstabe mess | 20:20 |
-!- sonney2k [~shogun@7nn.de] has left #shogun ["Leaving"] | 20:31 | |
-!- sonney2k [~shogun@7nn.de] has joined #shogun | 20:31 | |
-!- mode/#shogun [+o sonney2k] by ChanServ | 20:31 | |
-!- heiko [~heiko@134.91.52.61] has joined #shogun | 20:45 | |
-!- blackburn [~blackburn@188.168.5.154] has quit [Ping timeout: 258 seconds] | 21:12 | |
heiko | sonney2k, I just made my last "test"-example work, I will put together an example with a SGVector transition now | 21:20 |
heiko | argh found another error | 21:31 |
@sonney2k | kk | 21:33 |
heiko | ok scalars work now | 21:37 |
heiko | mmh sgvectors still make problems | 21:37 |
heiko | remeber that "treated in the same way as scalar variables" thing? | 21:38 |
-!- blackburn [~blackburn@188.168.5.129] has joined #shogun | 21:43 | |
CIA-3 | shogun: Sergey Lisitsyn master * ra4d19ca / (6 files): Added few examples descriptions - http://git.io/E-ON_g | 21:46 |
@sonney2k | heiko, I don't understand... | 21:51 |
heiko | i got problems with these new container types | 21:52 |
heiko | CT_SGVECTOR etc | 21:52 |
heiko | but I just thought of a way how to handle it | 21:52 |
@sonney2k | ok | 21:52 |
heiko | it will do it for SGVector for now | 21:52 |
heiko | then you can have a look before i do the other ones | 21:52 |
heiko | this is getting more and more complicated | 21:52 |
heiko | but at least: scalars work :) | 21:52 |
heiko | type changes, name changes, no changes - everything | 21:53 |
blackburn | I have bad connection, what's up here? | 21:53 |
blackburn | sonney2k: any help I can do? | 21:53 |
-!- alesis-novik [~alesis@188.74.87.206] has quit [Remote host closed the connection] | 22:09 | |
-!- blackburn [~blackburn@188.168.5.129] has quit [Read error: Operation timed out] | 22:21 | |
heiko | sonney2k, sorry the sgvector migration leads to a bunch of new problems | 22:26 |
heiko | i will not finish this today | 22:26 |
@sonney2k | what problems? | 22:26 |
heiko | copying Tparameter instances (with their data) is complicated with these types | 22:27 |
heiko | every time, an SGVector has to be allocated, along with its data | 22:27 |
heiko | and this has to be done when migrating | 22:28 |
heiko | I am currently changing the migration method to be able to use already allocated data | 22:28 |
heiko | sorry that this takes so long, but the stuff is even more complicated than the model selection stuff | 22:29 |
heiko | sonney2k, I gotta go now, my girlfriend is killing me already, we have to clear up our flat. I will be back tomorrow morning and continue | 22:30 |
@sonney2k | hmmhh seems like the problem is more w/ this SGVector stuff | 22:30 |
heiko | yes | 22:30 |
heiko | normaly it is easy to create a new TParameter instance | 22:30 |
heiko | simply copy the m_parameter memory | 22:30 |
@sonney2k | if we 'just' had double* stuff it would work I guess | 22:30 |
heiko | but with SGVector this is not possible like this | 22:31 |
@sonney2k | but can't you do the same thing with the SGVector stuff? | 22:31 |
@sonney2k | you just say a=b | 22:31 |
heiko | no, the m_parameter field contains a vector | 22:31 |
heiko | this has to be copied | 22:31 |
heiko | but also the ->vector field | 22:31 |
@sonney2k | the vector has to be copied too? | 22:31 |
@sonney2k | not just the ptr? | 22:31 |
heiko | currently, yes | 22:32 |
heiko | but i will change this | 22:32 |
heiko | also, a=b only works if the types are clear | 22:32 |
@sonney2k | ptr should be sufficient | 22:32 |
@sonney2k | a=b is the same like memccpy in our case | 22:32 |
@sonney2k | do you copy the double* ptr's or do you copy memory for that too? | 22:33 |
heiko | memory too | 22:33 |
heiko | but i will change it | 22:33 |
@sonney2k | ok that definitely needs to be changed | 22:34 |
heiko | when a parameter does not change from one version to next one | 22:34 |
heiko | yes | 22:34 |
@sonney2k | doesn't make sense to copy memory around | 22:34 |
heiko | work in progress... | 22:34 |
heiko | it is there because of this: | 22:35 |
heiko | when a migration is performed, a type may change or something else may happen with the data | 22:35 |
heiko | so that is cannot be used, but has to be converted somehow | 22:35 |
heiko | in this case, migrate returns a new TParameter instance (with allocated memory) with the converted data | 22:35 |
heiko | and this data has to be deleted afterwards (because of multiple migration steps) | 22:36 |
heiko | i did this as a first step | 22:36 |
heiko | and then noticed that sometimes, data does not get changed | 22:36 |
@sonney2k | I would say one should assume that things that are migrated from have to stay constant | 22:36 |
heiko | in this case, no copy has to be returned, but also this stuff must not be deleted afterwords | 22:36 |
@sonney2k | and the things we migrate to might have to be clones | 22:37 |
heiko | yes, thats how i want to do it | 22:37 |
heiko | but just realised it | 22:37 |
@sonney2k | ok | 22:37 |
@sonney2k | yeah it seems really much more involved than I ever imagined | 22:37 |
heiko | yes, its complicated stuff | 22:38 |
@sonney2k | what is important now is that we get something to work good enough for the initial migration | 22:38 |
heiko | yes | 22:38 |
@sonney2k | we can polish this later on I would say | 22:38 |
@sonney2k | current migrations aren't too difficult - so it shouldn't be too hard | 22:38 |
heiko | yes, but the infrastructure has to be capable of more complex migrations so that we avoid doing all this stuff all over again in the next release | 22:39 |
@sonney2k | that is true | 22:39 |
heiko | but is has not to be implemented yet | 22:39 |
heiko | sonney2k, I will go for it tomorrow and hopefully it will go then, I have the feeling that "it will work soon" for quite some time now | 22:40 |
@sonney2k | yeah that's not goo | 22:40 |
@sonney2k | d | 22:40 |
@sonney2k | ok | 22:40 |
@sonney2k | thanks! | 22:40 |
@sonney2k | apologies to your gf | 22:40 |
heiko | :) thanks | 22:41 |
heiko | so cu | 22:41 |
@sonney2k | cu | 22:42 |
-!- heiko [~heiko@134.91.52.61] has quit [Ping timeout: 258 seconds] | 22:45 | |
-!- blackburn [~blackburn@188.168.4.41] has joined #shogun | 23:25 | |
-!- blackburn1 [~blackburn@188.168.4.53] has joined #shogun | 23:29 | |
-!- blackburn [~blackburn@188.168.4.41] has quit [Ping timeout: 245 seconds] | 23:29 | |
blackburn1 | have read the logs | 23:58 |
blackburn1 | didn't understand what's up with serialization :) | 23:58 |
--- Log closed Tue Aug 30 00:00:37 2011 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!