sb0 changed the topic of #m-labs to: ARTIQ, Migen, MiSoC, Mixxeo & other M-Labs projects :: fka #milkymist :: Logs http://irclog.whitequark.org/m-labs
<whitequark> sb0: hm you are right
<whitequark> the py35 is taken from --python 3.5
FabM has quit [Ping timeout: 252 seconds]
FabM has joined #m-labs
Mon_ has joined #m-labs
Mon_ is now known as Guest52775
ylamarre has joined #m-labs
<whitequark> sb0: tbh I think you fucked up with transition to 3.5
<whitequark> I can't use Debian's packages for ARTIQ because it does not have pip3.5 and llvmlite_artiq does not compile properly with conda
<whitequark> and I have no idea why
<whitequark> OSError: /home/whitequark/.miniconda/envs/py35/lib/python3.5/site-packages/llvmlite_artiq/binding/libllvmlite.so: undefined symbol: _ZN4llvm3sys22getDefaultTargetTripleB5cxx11Ev
<whitequark> I've compiled this shit just now, what is its problem
Guest52775 has quit [Quit: This computer has gone to sleep]
Mon_1 has joined #m-labs
<whitequark> ... oh
<whitequark> nevermind, gcc decided they want to change their C++ ABI again
Mon_1 has quit [Quit: This computer has gone to sleep]
Mon_ has joined #m-labs
Mon_ is now known as Guest67397
Guest67397 has quit [Client Quit]
Mon_1 has joined #m-labs
Mon_1 has quit [Client Quit]
Mon_1 has joined #m-labs
<sb0> whitequark, i guess the debian packages will come. and anaconda should be 3.5 by default by the end of the month
<sb0> whitequark, are there further problems or just this gcc one?
Mon_1 has quit [Quit: This computer has gone to sleep]
Mon_1 has joined #m-labs
Mon_1 has quit [Quit: This computer has gone to sleep]
Mon_1 has joined #m-labs
<cr1901_modern> sb0: I'm looking over ARTIQ now. I am on standby if you wish to chat.
Mon_1 has quit [Quit: This computer has gone to sleep]
ylamarre has quit [Quit: ylamarre]
<GitHub128> [misoc] sbourdeauducq pushed 2 new commits to new: http://git.io/vCVmJ
<GitHub128> misoc/new 93a615a Sebastien Bourdeauducq: Merge branch 'new' of github.com:m-labs/misoc into new
<GitHub128> misoc/new ecc4c57 Sebastien Bourdeauducq: integration/builder: fix building for SoCSDRAM-based targets when SDRAM is disabled...
kmehall has quit [Remote host closed the connection]
kmehall has joined #m-labs
fengling has joined #m-labs
Mon_ has joined #m-labs
Mon_ is now known as Guest19159
Guest19159 has quit [Client Quit]
fengling has quit [Ping timeout: 245 seconds]
Mon_ has joined #m-labs
Mon_ is now known as Guest35368
Guest35368 has quit [Client Quit]
Mon_1 has joined #m-labs
Mon_1 has quit [Client Quit]
Mon_1 has joined #m-labs
<sb0> cr1901_modern, do you have any questions? just ask...
Mon_1 has quit [Quit: This computer has gone to sleep]
fengling has joined #m-labs
Mon_ has joined #m-labs
Mon_ is now known as Guest2596
Guest2596 has quit [Quit: This computer has gone to sleep]
fengling has quit [Ping timeout: 245 seconds]
fengling has joined #m-labs
Mon_1 has joined #m-labs
Mon_1 has quit [Quit: This computer has gone to sleep]
Mon_1 has joined #m-labs
Mon_1 has quit [Quit: This computer has gone to sleep]
Mon_ has joined #m-labs
Mon_ is now known as Guest62407
Guest62407 has quit [Quit: This computer has gone to sleep]
Mon_1 has joined #m-labs
<GitHub151> [artiq] sbourdeauducq pushed 4 new commits to master: http://git.io/vCwUg
<GitHub151> artiq/master 9f04af6 Sebastien Bourdeauducq: scheduler: raise logging severity of errors
<GitHub151> artiq/master da159f1 Sebastien Bourdeauducq: gui/log: ensure black/white colors
<GitHub151> artiq/master 86e201c Sebastien Bourdeauducq: master/log: parse names, better filtering, cleanup
<sb0> whitequark, while you're messing
<sb0> oops
<sb0> sorry
<sb0> whitequark, while you're dealing with travis/packaging, can you split the artiq package into 1) the python part + the flashing script 2) bitstream/binaries for kc705/qc1 3) kc705/qc2 4) pipistrello?
Mon_1 has quit [Quit: Leaving]
<sb0> and those could be built in parallel
Mon_ has joined #m-labs
Mon_ is now known as Guest88328
Guest88328 has quit [Quit: This computer has gone to sleep]
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#505 (master - ea5fbc9 : Sebastien Bourdeauducq): The build passed.
travis-ci has left #m-labs [#m-labs]
Mon_ has joined #m-labs
Mon_ is now known as Guest53923
omega_red has joined #m-labs
Guest53923 has quit [Quit: This computer has gone to sleep]
omega_red has quit [Read error: Connection reset by peer]
fengling has quit [Ping timeout: 245 seconds]
<whitequark> sb0: I think so, yes
<sb0> ysionneau, what happens when a new parameter (now dataset) that does not match the search string is added?
<sb0> filtering by string/regexp is rather easy, but custom filter functions seem to be a pain, though
<whitequark> sb0, did you build llvmdev-or1k?
<sb0> whitequark, the conda package? I didn't
<whitequark> hm, that one is built properly though
<whitequark> why does llvmlite-artiq eventually depend on libLLVMLTO.so... there are no .so in llvmdev-or1k
<whitequark> in llvmlite-artiq/build.sh: PATH=/usr/local/llvm-or1k/bin:$PATH $PYTHON setup.py install
<whitequark> what the fuck?
<GitHub182> [artiq] whitequark pushed 4 new commits to master: http://git.io/vCwAT
<GitHub182> artiq/master fa9cbab whitequark: conda: fix README.
<GitHub182> artiq/master 760ecb5 whitequark: conda: add lit package.
<GitHub182> artiq/master ca3a0fd whitequark: conda: llvmlite-artiq: do not use LLVM from /usr/local/llvm-or1k.
<whitequark> still doesn't work...
<whitequark> oh ffs, I also have to rebuild llvmdev-or1k in conda
Mon_ has joined #m-labs
Mon_ is now known as Guest59187
Guest59187 has quit [Quit: This computer has gone to sleep]
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#506 (master - 6b9cc5a : whitequark): The build has errored.
travis-ci has left #m-labs [#m-labs]
<whitequark> sb0: how do you say "ISO100 viewport" in chinese?
ylamarre has joined #m-labs
<sb0> they have a taobao store
<sb0> I ordered from them this week so I'll soon know roughly about the quality
<whitequark> ohh, 观察窗
<sb0> their catalogue is incomplete, they didn't list the KF50/25 adapter they actually had and I ordered...
<whitequark> taobao doesn't seem to have ISO100 aka DN100KF viewports
<whitequark> only DN100CF
Mon_ has joined #m-labs
Mon_ is now known as Guest8041
<sb0> whitequark, that would surprise me. the chinese are good at machining this stuff. what is hard to find is more complicated devices like scroll pumps, turbopumps, etc.
<sb0> they need more "foreign experience absorption", as they call it...
<whitequark> heh
<whitequark> turbopumps are hard to make...
<whitequark> did you know, RU does not have a /single/ rebalancing stand for turbopumps?
<whitequark> or at least that's what an org which repairs those told me
Guest8041 has quit [Quit: This computer has gone to sleep]
<sb0> how are russian turbopumps made then?
<whitequark> I mean, any stands that would work with imported turbopumps, Edwards etc
<whitequark> russian turbopumps are a whole different beast
<whitequark> I don't think they are made in "under 100kg" category at all
<sb0> isn't it just a price problem? the edwards stuff is luxury
<whitequark> I wouldn't really say it is? if you have a plant, Edwards turbopumps are practically disposable
<whitequark> which is why you get so many on ebay
<sb0> also, I saw a smaller (~20kg?) russian turbopump at Alek's institute a couple years ago...
Mon_1 has joined #m-labs
<whitequark> hm
<whitequark> interesting
<GitHub16> [artiq] whitequark pushed 2 new commits to master: http://git.io/vCrn8
<GitHub16> artiq/master d033745 whitequark: conda: llvmdev-or1k: bump.
<GitHub16> artiq/master da4a6a2 whitequark: conda: llvmlite-artiq: test by importing binding.
fengling has joined #m-labs
Mon_1 has quit [Quit: This computer has gone to sleep]
<GitHub24> [artiq] sbourdeauducq pushed 3 new commits to master: http://git.io/vCrBA
<GitHub24> artiq/master f855544 Sebastien Bourdeauducq: gui: dataset filtering with QSortFilterProxyModel
<GitHub24> artiq/master f3b3bf3 Sebastien Bourdeauducq: gui: log filtering
<GitHub24> artiq/master 744ae4c Sebastien Bourdeauducq: Merge branch 'master' of github.com:m-labs/artiq
Mon_ has joined #m-labs
Mon_ is now known as Guest24400
<GitHub153> [artiq] sbourdeauducq pushed 1 new commit to master: http://git.io/vCruo
<GitHub153> artiq/master 49acd10 Sebastien Bourdeauducq: gui: fix data update in DictSyncModel/ListSyncModel
<GitHub123> [migen] sbourdeauducq pushed 1 new commit to new: http://git.io/vCrgf
<GitHub123> migen/new 48d22a7 Sebastien Bourdeauducq: genlib/fifo: width_or_layout -> width
travis-ci has joined #m-labs
<travis-ci> m-labs/migen#128 (new - 48d22a7 : Sebastien Bourdeauducq): The build passed.
travis-ci has left #m-labs [#m-labs]
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#507 (master - da4a6a2 : whitequark): The build failed.
travis-ci has left #m-labs [#m-labs]
fengling has quit [Ping timeout: 245 seconds]
<GitHub32> [misoc] sbourdeauducq pushed 1 new commit to new: http://git.io/vCrVf
<GitHub32> misoc/new 884faed Sebastien Bourdeauducq: integration/builder: escape backslash in makefile defines
<GitHub32> [artiq] whitequark pushed 3 new commits to new-py2llvm: http://git.io/vCr6T
<GitHub32> artiq/new-py2llvm e9adfd6 whitequark: Merge branch 'master' into new-py2llvm
<GitHub32> artiq/new-py2llvm 5d64df8 whitequark: transforms.artiq_ir_generator: fix devirtualized call annotation.
<GitHub32> artiq/new-py2llvm b91ffa1 whitequark: ir: fix default argument fiasco.
<whitequark> "error: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version `GLIBCXX_3.4.21' not found (required by /home/travis/miniconda/envs/py35/lib/python3.5/site-packages/llvmlite_artiq/binding/libllvmlite.so)"
<whitequark> jesus fuck I hate glibc
<whitequark> sb0: I would like you to buy two DigitalOcean VMs, with 32 and 64-bit Ubuntu LTS installed
<whitequark> they're like $10/month
<whitequark> and this will allow anyone in m-labs to easily build conda packages without bothering to set up this particular environment, on reasonably fast hardware
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#508 (master - 744ae4c : Sebastien Bourdeauducq): The build has errored.
travis-ci has left #m-labs [#m-labs]
<whitequark> if conda actually cared about decent isolation, they would ship their own copy of glibc
<whitequark> or even just musl
<whitequark> sb0: tangential suggestion, I want to split conda/ from the ARTIQ repo
<whitequark> specifically so that updates to that subtree do not show up in the main repo. Then I could make it so that build number is incremented and linux-32 and linux-64 packages are built, all with a single script that ssh's into the VMs and kicks off the build and uploads it
<whitequark> I suggest a m-labs/conda-recipes repo
<whitequark> actually, hang on. I have an even better idea.
<whitequark> travis is 64-bit Ubuntu LTS, right? this means we can build linux-64 and linux-32 binaries right there. i.e. look at the diff, see which packages were changed, rebuild and push.
<whitequark> then we can use Appveyor CI to do the same thing for Windows, some day.
Guest24400 has quit [Quit: This computer has gone to sleep]
travis-ci has joined #m-labs
<travis-ci> m-labs/artiq#509 (master - 49acd10 : Sebastien Bourdeauducq): The build failed.
travis-ci has left #m-labs [#m-labs]
Mon_ has joined #m-labs
Mon_ is now known as Guest87507
Guest87507 has quit [Client Quit]
fengling has joined #m-labs
Mon_1 has joined #m-labs
Mon_1 has quit [Client Quit]
cr1901_modern has quit [Read error: Connection reset by peer]
<sb0> whitequark, you should be able to create repositories under m-labs I think
<whitequark> insufficient permission
<sb0> ah right, i thought that option was enabled but somehow it wasnt
<sb0> can you try again?
<whitequark> yeah that works
<sb0> whitequark, do you have any silicone diffusion pump oil?
<whitequark> no. I don't do diffusion pumps
fengling has quit [Ping timeout: 245 seconds]
cr1901_modern has joined #m-labs
<sb0> it sounds like you need to be lucky to find a turbopump on ebay that a) is working b) includes a controller c) is relatively inexpensive ...
<whitequark> hmm
<whitequark> if your budget is $1k that is easy
<whitequark> half that is going to be more tricky
<sb0> Please note you are purchasing only the pump pictured, no controller is included
<whitequark> the controller is #120
<whitequark> alternatively this one is in much better condition
<whitequark> note that if you get an Edwards EXT pump, I have lots of spare oil for it
<sb0> are those controllers compatible with the pump?
<whitequark> yes
<whitequark> I have that specific pump and controller
<whitequark> also compatible with EXDC80
<whitequark> well, all of these are
<sb0> I have *never* won anything on ebay, which frustrated me, and I now stay away from it
<whitequark> I won things on ebay plenty of times
<whitequark> I have five weldable high-voltage feedthroughs for $5 each, two working HPLC pumps for $20 both...
<whitequark> a $200 for an almost functional Edwards pump
<whitequark> though I also lost plenty of times. someone sniped a microscope stage right from under my nose in the literal last second
<whitequark> anyhow, I believe any of the above will work for you
<whitequark> they should be re-oiled and potentially cleaned, all of which I am equipped to do
<whitequark> are you going to get one?
<whitequark> (if not, I was going to sell most of the oil to someone else)
<sb0> hmm, maybe in a few months
<whitequark> alright
<GitHub89> [conda-recipes] whitequark created master (+1 new commit): https://github.com/m-labs/conda-recipes/commit/a14eebf1a99e
<GitHub89> conda-recipes/master a14eebf whitequark: Initial commit.
<whitequark> sb0: while we're at it
<whitequark> what is the purpose and nature of main and dev channels?
<sb0> not sure about that... ysionneau used those
<ysionneau> the idea was that the current artiq development could be followed by using the dev channel
<ysionneau> where each commit was being bundled and uploaded to this channel
<ysionneau> and when there is a stable release, it can be uploaded to main
<ysionneau> so that end users who only want stable releases can just follow the main channel
<ysionneau> so far indeed only the dev channel was really used
<whitequark> I think it makes sense to put most packages onto main channel and let those who want to use dev add both
<GitHub132> [conda-recipes] whitequark pushed 1 new commit to master: https://github.com/m-labs/conda-recipes/commit/cbed9dd7918b923e9cf916037eaa95bcedb9b4a3
<GitHub132> conda-recipes/master cbed9dd whitequark: Move packages under conda/.
<GitHub93> [conda-recipes] whitequark pushed 1 new commit to master: https://github.com/m-labs/conda-recipes/commit/e92259a3339ecb7235e30084d049df10e5087313
<GitHub93> conda-recipes/master e92259a whitequark: Add .travis.yml.
<sb0> whitequark, so you don't need those VMs, right?
<whitequark> sb0: assuming I can get travis to work, no
<GitHub127> [conda-recipes] whitequark pushed 1 new commit to master: https://github.com/m-labs/conda-recipes/commit/64ebde56b72e67ad6ac06226ce8b5ec55782c07e
<GitHub127> conda-recipes/master 64ebde5 whitequark: lit: bump.
<whitequark> "We STRONGLY recommend installing the x86_64 version of Miniconda3 on an x86_64 system." do not tell me what to do
fengling has joined #m-labs
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#2 (master - e92259a : whitequark): The build was canceled.
travis-ci has left #m-labs [#m-labs]
<whitequark> ugh, and they don't even skip it with -b
<GitHub57> [conda-recipes] whitequark force-pushed master from 64ebde5 to 709376f: https://github.com/m-labs/conda-recipes/commits/master
<GitHub57> conda-recipes/master 5bd84f8 whitequark: Add .travis.yml.
<GitHub57> conda-recipes/master 709376f whitequark: lit: bump.
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#3 (master - 64ebde5 : whitequark): The build was canceled.
travis-ci has left #m-labs [#m-labs]
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#4 (master - 709376f : whitequark): The build has errored.
travis-ci has left #m-labs [#m-labs]
<GitHub54> [conda-recipes] whitequark force-pushed master from 709376f to a32f9bc: https://github.com/m-labs/conda-recipes/commits/master
<GitHub54> conda-recipes/master 60bf01f whitequark: Add .travis.yml.
<GitHub54> conda-recipes/master a32f9bc whitequark: lit: bump.
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#5 (master - a32f9bc : whitequark): The build passed.
travis-ci has left #m-labs [#m-labs]
<whitequark> that..... worked?!
<whitequark> ok cool it, like, almost worked. amazing
<GitHub173> [conda-recipes] whitequark force-pushed master from a32f9bc to efe6403: https://github.com/m-labs/conda-recipes/commits/master
<GitHub173> conda-recipes/master e00d0af whitequark: Add .travis.yml.
<GitHub173> conda-recipes/master efe6403 whitequark: lit: bump.
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#6 (master - efe6403 : whitequark): The build passed.
travis-ci has left #m-labs [#m-labs]
<GitHub34> [conda-recipes] whitequark force-pushed master from efe6403 to 2a99bf1: https://github.com/m-labs/conda-recipes/commits/master
<GitHub34> conda-recipes/master 2a99bf1 whitequark: Add .travis.yml.
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#7 (master - 2a99bf1 : whitequark): The build passed.
travis-ci has left #m-labs [#m-labs]
<GitHub162> [conda-recipes] whitequark pushed 1 new commit to master: https://github.com/m-labs/conda-recipes/commit/9f67ed8d51476c5ff236278d66be05a1b6d078cd
<GitHub162> conda-recipes/master 9f67ed8 whitequark: lit: bump.
travis-ci has joined #m-labs
<travis-ci> m-labs/conda-recipes#8 (master - 9f67ed8 : whitequark): The build passed.
travis-ci has left #m-labs [#m-labs]
<whitequark> amazing. it works. I was fully prepared to spend the next 20 hours debugging it
<whitequark> sb0: do you think we should rebuild everything using this method to ensure that it works on all packages?
<sb0> sounds good. i'm just worried that when building heavy packages (eg llvm) we run out of time on travis.
<whitequark> I think they lifted the limit in the container-based builds, presumably because they consume less resources
<whitequark> actually, you know what, I think I can do the same with Appveyor CI now
<whitequark> it should not be much harder
kyak has quit [*.net *.split]
larsc has quit [*.net *.split]
jaeckel has quit [*.net *.split]
[florian] has quit [*.net *.split]
tmbinc__ has quit [*.net *.split]
fengling has quit [Ping timeout: 245 seconds]
jaeckel has joined #m-labs
[florian] has joined #m-labs
tmbinc__ has joined #m-labs
jaeckel has quit [*.net *.split]
[florian] has quit [*.net *.split]
tmbinc__ has quit [*.net *.split]
kyak has joined #m-labs
kyak has joined #m-labs
jaeckel has joined #m-labs
[florian] has joined #m-labs
tmbinc__ has joined #m-labs
larsc has joined #m-labs
fengling has joined #m-labs
ylamarre has quit [Quit: ylamarre]
fengling has quit [Ping timeout: 245 seconds]
fengling has joined #m-labs
<whitequark> fucking anaconda garbage
<whitequark> [warning] It appears you are already logged in from host appvyr-win20122
<whitequark> Would you like to continue [Y|n]:
<whitequark> I don't care, I don't want to be made to care, and why the fuck is there an interactive prompt
<GitHub47> [conda-recipes] whitequark tagged appveyor at 2a99bf1: https://github.com/m-labs/conda-recipes/commits/appveyor
fengling has quit [Ping timeout: 245 seconds]
ylamarre has joined #m-labs
fengling has joined #m-labs
fengling has quit [Ping timeout: 245 seconds]
fengling has joined #m-labs
fengling has quit [Ping timeout: 245 seconds]