[15:56:28] <Amir1>	 halfak: o/ I hope you feel better today :)
[16:11:15] <grrrit-wm>	 (03CR) 10Ladsgroup: [C: 04-1] "I think it may have some problems with LocalSettings.php. Need to check more carefully" [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265676 (https://phabricator.wikimedia.org/T122684) (owner: 10Ladsgroup)
[17:59:23] <Amir1>	 legoktm: hey, do you think it's the proper way to implement arrays in extension.json. line ~60 https://gerrit.wikimedia.org/r/#/c/265676/4/extension.json,cm
[17:59:28] <Amir1>	 ?
[17:59:59] <Amir1>	 someone pointed out to this bug https://phabricator.wikimedia.org/T121378
[18:00:05] <Amir1>	 you commented 
[18:10:51] <legoktm>	 Amir1: yeah, that looks right
[18:11:19] <legoktm>	 but maybe just OresDamagingThresholds: { "soft": 0.5.0, ... } no need to repeat ores-damaging multiple times :)
[18:11:53] <Amir1>	 oh yeah
[18:11:57] <Amir1>	 you're right
[18:12:01] <Amir1>	 I do it
[18:44:09] <pipivoj>	 Hello 
[18:45:43] <pipivoj>	 YuviPanda, what's up with scipy. I can get it to install with pip after these reboots.
[18:49:04] <pipivoj>	 Some "libraries openblass not found", "libraries mkl,vml,guide not found" messages and many more up till "#warning "Using deprecated NumPy API, disable it by " \
[18:49:05] <pipivoj>	       ^
[18:49:05] <pipivoj>	     x86_64-linux-gnu-g++: internal compiler error: Killed (program cc1plus)".
[18:49:22] <pipivoj>	 I googled it but I can't find a solution.
[18:49:43] <pipivoj>	 *googled _them_
[18:57:30] <halfak>	 o/
[18:57:38] <pipivoj>	 o/
[18:58:11] <Amir1>	 o/ halfak :)
[18:58:23] <Amir1>	 I hope you feel better today
[18:58:29] <halfak>	 pipivoj, can you paste the exact error?
[18:58:39] <halfak>	 Amir1, I do.  Not much time to hack though.  Just an hour
[18:58:57] <Amir1>	 it's okay :)
[18:59:03] <pipivoj>	 stderr or stdout?
[18:59:23] <Amir1>	 I made some notes in the work log
[18:59:23] <halfak>	 pipivoj?
[18:59:26] <pipivoj>	 stdout is quite large
[18:59:49] <halfak>	 well, give me the part that is an error :P 
[19:00:13] <Amir1>	 halfak: vagrant patch is merged so people can simply use ORES by "enable role ORES"
[19:00:19] <pipivoj>	 There goes stderr:
[19:00:21] <pipivoj>	 Command "/srv/paws/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-mwanrqr1/scipy/setup.py';exec(compile(getattr(tokenize, 'open', open)(__file__).read().replace('\r\n', '\n'), __file__, 'exec'))" install --record /tmp/pip-2etdf_rv-record/install-record.txt --single-version-externally-managed --compile --install-headers /srv/paws/include/site/python3.4/scipy" failed with error code 1 in /tmp/pip-build-mwanrqr1/scipy
[19:01:01] <pipivoj>	 And this is the last part of stdout:
[19:01:02] <pipivoj>	     In file included from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/ndarraytypes.h:1781:0,
[19:01:02] <pipivoj>	                      from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/ndarrayobject.h:18,
[19:01:02] <pipivoj>	                      from scipy/sparse/sparsetools/sparsetools.h:5,
[19:01:03] <pipivoj>	                      from scipy/sparse/sparsetools/bsr.cxx:4:
[19:01:04] <pipivoj>	     /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/npy_1_7_deprecated_api.h:15:2: warning: #warning "Using deprecated NumPy API, disable it by " "#defining NPY_NO_DEPRECATED_API NPY_1_7_API_VERSION" [-Wcpp]
[19:01:10] <pipivoj>	      #warning "Using deprecated NumPy API, disable it by " \
[19:01:12] <pipivoj>	       ^
[19:01:14] <pipivoj>	     x86_64-linux-gnu-g++: internal compiler error: Killed (program cc1plus)
[19:01:16] <pipivoj>	     Please submit a full bug report,
[19:01:18] <pipivoj>	     with preprocessed source if appropriate.
[19:01:20] <pipivoj>	     See <file:///usr/share/doc/gcc-4.9/README.Bugs> for instructions.
[19:01:22] <halfak>	 Amir1, great!  So that will start up a devserver?  And how do we deploy a new version of ORES?
[19:01:22] <pipivoj>	     error: Command "x86_64-linux-gnu-g++ -pthread -DNDEBUG -g -fwrapv -O2 -Wall -g -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -fPIC -D__STDC_FORMAT_MACROS=1 -Iscipy/sparse/sparsetools -I/srv/paws/lib/python3.4/site-packages/numpy/core/include -I/usr/include/python3.4m -I/srv/paws/include/python3.4m -c scipy/sparse/sparsetools/bsr.cxx -o build/temp.linux-x86_64-3.4/scipy/sparse/sparsetools/bsr.o" failed with exit sta
[19:01:27] <pipivoj>	 tus 4
[19:01:29] <pipivoj>	   
[19:01:35] <halfak>	 pipivoj, ahhh.  Paste service like pastebin.com please :P 
[19:01:42] <pipivoj>	 ok
[19:02:09] <Amir1>	 it uses ores.wmflabs.org
[19:02:28] <Amir1>	 (the testwiki by default)
[19:03:28] <halfak>	 Amir1, oH!  That'll work. 
[19:04:19] <Amir1>	 we can change everything in LocalSettings.php
[19:07:06] <Amir1>	 halfak: btw. It's break between semesters, I've nothing to do except ORES
[19:07:56] <halfak>	 This'll help with the KDD writing.  IME, it's easier to write when it is my primary creative outlet. 
[19:08:42] <pipivoj>	 halfak: http://pastebin.ca/3348621
[19:08:43] <Amir1>	 yeah, exactly
[19:08:59] <Amir1>	 I'm planning to get the KDD article to somewhere
[19:09:55] <halfak>	 pipivoj, not seeing the error.  It seems like the end is cut off. 
[19:10:15] <halfak>	 If you could give me /just the error/  -- not the compiling output or warnings -- that would be most helpful. 
[19:10:39] <pipivoj>	 Well I saved the stderr to a different file and already posted it here. Here, I'll paste it again
[19:10:52] <pipivoj>	 Command "/srv/paws/bin/python3 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-mwanrqr1/scipy/setup.py';exec(compile(getattr(tokenize, 'open', open)(__file__).read().replace('\r\n', '\n'), __file__, 'exec'))" install --record /tmp/pip-2etdf_rv-record/install-record.txt --single-version-externally-managed --compile --install-headers /srv/paws/include/site/python3.4/scipy" failed with error code 1 in /tmp/pip-build-mwanrqr1/scipy
[19:11:48] <halfak>	 ^ that's not enough.  That line just says that there was an error. 
[19:11:55] <pipivoj>	 Hm
[19:12:04] <pipivoj>	 How do I get more?
[19:12:28] <pipivoj>	 I did "pip install scipy > stdout.txt 2> stderr.txt"
[19:13:15] <halfak>	 pipivoj, something is wrong with your pasting.  I don't see the error you pasted into the chat in your pastebin.ca
[19:16:01] <pipivoj>	 well it's the first thing I thought it was error: line 124
[19:16:25] <pipivoj>	 it's not openblass it was openblas :(
[19:17:16] <halfak>	 124 has "warning: ‘elsize’ may be used uninitialized in this function"  
[19:18:06] <halfak>	 Woops.  No, it has "Block: zptsv"
[19:18:57] <halfak>	 Could you try copy-pasting the last 100 lines of output that appear in the terminal?
[19:19:02] <halfak>	 pipivoj, ^
[19:19:49] <pipivoj>	 http://pastebin.ca/3348621 on line 124 is       libraries openblas not found in ['/srv/paws/lib', '/usr/local/lib', '/usr/lib', '/usr/lib/x86_64-linux-gnu']
[19:19:52] <pipivoj>	 ok I will
[19:20:37] <pipivoj>	     #warning "Using deprecated NumPy API, disable it by " \
[19:20:37] <pipivoj>	       ^
[19:20:37] <pipivoj>	     In file included from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/ndarrayobject.h:27:0,
[19:20:37] <pipivoj>	                      from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/arrayobject.h:4,
[19:20:38] <pipivoj>	                      from scipy/sparse/csgraph/_min_spanning_tree.c:242:
[19:20:42] <pipivoj>	     /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/__multiarray_api.h:1634:1: warning: �_import_array� defined but not used [-Wunused-function]
[19:20:45] <pipivoj>	      _import_array(void)
[19:20:47] <pipivoj>	      ^
[19:20:49] <pipivoj>	     x86_64-linux-gnu-gcc -pthread -shared -Wl,-O1 -Wl,-Bsymbolic-functions -Wl,-z,relro -Wl,-z,relro -g -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 build/temp.linux-x86_64-3.4/scipy/sparse/csgraph/_min_spanning_tree.o -Lbuild/temp.linux-x86_64-3.4 -o build/lib.linux-x86_64-3.4/scipy/sparse/csgraph/_min_spanning_tree.cpython-34m.so
[19:20:54] <pipivoj>	     building 'scipy.sparse.csgraph._reordering' extension
[19:20:56] <pipivoj>	     compiling C sources
[19:20:58] <pipivoj>	     C compiler: x86_64-linux-gnu-gcc -pthread -DNDEBUG -g -fwrapv -O2 -Wall -Wstrict-prototypes -g -fstack-protector-strong -Wformat -Werror=format-security -D_FORTIFY_SOURCE=2 -fPIC
[19:21:01] <pipivoj>	     
[19:21:05] <pipivoj>	     compile options: '-I/srv/paws/lib/python3.4/site-packages/numpy/core/include -I/srv/paws/lib/python3.4/site-packages/numpy/core/include -I/usr/include/python3.4m -I/srv/paws/include/python3.4m -c'
[19:21:08] <pipivoj>	     x86_64-linux-gnu-gcc: scipy/sparse/csgraph/_reordering.c
[19:21:12] <pipivoj>	     In file included from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/ndarraytypes.h:1781:0,
[19:21:15] <pipivoj>	                      from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/ndarrayobject.h:18,
[19:21:18] <pipivoj>	                      from /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/arrayobject.h:4,
[19:21:21] <pipivoj>	                      from scipy/sparse/csgraph/_reordering.c:242:
[19:21:23] <pipivoj>	     /srv/paws/lib/python3.4/site-packages/numpy/core/include/numpy/npy_1_7_deprecated_api.h:15:2: warning: #warning "Using deprecated NumPy API, disable it by " "#defining NPY_NO_DEPRECATED_API NPY_1_7_API_
[19:21:44] <halfak>	 Please don't paste in IRC
[19:22:11] <halfak>	 (1) I can't read it because the formatting is all ruined and (2) it fills our logs with nonsense. 
[19:31:09] <pipivoj>	 ok sorry
[19:32:55] <pipivoj>	 http://pastebin.ca/3348739
[19:39:05] <pipivoj>	 Sorry for delay, my kid was ready for bath so I had to bathe her.
[19:46:49] <halfak>	 pipivoj, again the last line ends with "[?25h"
[19:46:56] <halfak>	 It doesn't look like the full error. 
[19:47:09] <halfak>	 Line 119
[19:47:59] <pipivoj>	 Ok I'll try install it again and see what I get.
[19:49:28] <halfak>	 http://socio-technologist.blogspot.com/2016/01/notes-on-writing-wikipedia-vandalism.html
[19:49:31] <halfak>	 Amir1, ^ 
[19:49:41] <halfak>	 Will want to do better for the paper, but those are some thoughts. 
[19:51:07] <Amir1>	 thanks halfak 
[19:51:16] <Amir1>	 blogpost is blocked in Iran
[19:51:19] <halfak>	 So, it turns out that "recall" has a name, but I'm not quite sure what TP+FR/ALL would be called.  
[19:51:23] <halfak>	 Amir1, lolwat
[19:51:29] * halfak emails a PDF
[19:51:32] <Amir1>	 it may take some time until I get a workaround
[19:51:35] <Amir1>	 thanks
[19:51:36] <Amir1>	 :)
[19:52:24] <halfak>	 Amir1, I think I want to spend some time with the thresholds I proposed. 
[19:52:54] <pipivoj>	 halfak: FR?
[19:52:54] <halfak>	 I think that ClueBot doesn't measure "False-positive rate" or "Recall" the way I expect. 
[19:53:07] <halfak>	 Sorry that was supposed to be FP
[19:53:11] <halfak>	 False-positives
[19:53:40] <halfak>	 Review-proportion = all true predictions / all obs. when recall is fixed at 95%. 
[19:54:11] <halfak>	 We might drop that recall value.  We can, after all, put some review burden on people and their watchlists. 
[19:54:55] <halfak>	 Maybe we should do coloring in ScoredRevisions based on these recall bands. 
[19:55:03] <halfak>	 E.g. 95% is light yellow
[19:55:27] <halfak>	 Meh.  That's not as intuitive 
[19:55:36] <halfak>	 I need to think about that more. 
[19:56:29] <halfak>	 95% light yellow, 75% orange, 50% red.  Around 50%, we should be getting to the things that can be auto-reverted by bots. 
[19:57:11] <Amir1>	 we can examine with different thresholds
[19:57:17] <Amir1>	 and see what happens
[19:57:40] <Amir1>	 halfak: I already got some people who can help with reviewing the new datasets for fa.wp
[19:57:51] <Amir1>	 tell me when you deployed a new changeset
[19:57:53] <halfak>	 Amir1, +1.  So one could say, "It's OK if I only find 50% of vandalism, but I want all of the stuff I review to be vandalism"
[19:58:08] <halfak>	 Amir1, a new campaign?
[19:58:09] <halfak>	 Sure!  
[19:58:34] <halfak>	 Email sent
[19:58:38] <Amir1>	 you told me that the number is not enough
[19:59:20] <halfak>	 +1, unless we do some re-weighting of the input set. 
[19:59:24] <halfak>	 I need to explore that. 
[19:59:38] <halfak>	 Right now, none of the items in fawiki's test set are predicted true. 
[19:59:48] <pipivoj>	 halfak it looks it's the same stuff but without the "[?25h"
[20:00:26] <pipivoj>	 Dunno if I should redirect again to a text file.
[20:00:37] <Amir1>	 halfak: regarding the threshold, that's my main reason to implement "soft" and "hard" thresholds for the extension
[20:01:53] <halfak>	 "internal compiler error: Killed (program cc1plus)"  This makes me wonder if it was sent an interrupt. 
[20:02:06] <halfak>	 Amir1, +1  Seems to work well for scored revisions. 
[20:03:11] <halfak>	 pipivoj, looks like you might be running out of memory 
[20:03:15] <halfak>	 http://stackoverflow.com/questions/19595944/trouble-installing-scipy-in-virtualenv-on-a-amazon-ec2-linux-micro-instance
[20:03:33] <halfak>	 YuviPanda, could comment better.  I'm not sure how much memory is available to a PAWS instance. 
[20:03:47] <halfak>	 But you shouldn't need to install scipy now that I think of it. 
[20:03:57] <halfak>	 It should already be installed for revscoring to work. 
[20:09:56] * pipivoj quitting - will be back
[20:10:01] <halfak>	 o/ pip
[20:10:29] <halfak>	 So, I don't see a good way to balance the obs.  I could use the same strategy that we use in SVC or we could just adjust how we think of probability. 
[20:10:35] <halfak>	 What do you think Amir1?
[20:11:07] <halfak>	 If suddenly 50% probability corresponded to a much higher precision, would that be a big problem?
[20:12:02] <Amir1>	 I don't think we should give percentage to people 
[20:12:05] <halfak>	 Either way, it seems that it would be most desirable if tools like the ORES extension and ScoredRevisions could read a machine-readable endpoint to find out where to set its thresholds. 
[20:12:21] <halfak>	 So that we can safely change the underlying scores. 
[20:12:28] <YuviPanda>	 halfak: hmm, installing scipy yourself into PAWS is going to be painful
[20:12:36] <YuviPanda>	 I should just install sklearn and scipy into it myself shouldn't I
[20:12:42] <Amir1>	 we should find a way to balance that and just let users decide on an intuitive "sensitivity meter" 
[20:12:46] <halfak>	 YuviPanda, +1
[20:12:48] <YuviPanda>	 I don't have revscoring installed into it yet
[20:12:59] <YuviPanda>	 so I guess I'll just do that now
[20:13:03] * YuviPanda goes to find that ticket
[20:13:10] <halfak>	 Amir1, so, the only reason I can see to balance it is if we *are* going to show it to people. 
[20:13:28] <Amir1>	 show percentage?
[20:13:32] <halfak>	 Yeah. 
[20:13:35] <Amir1>	 oh no, it's really bad idea
[20:13:49] <halfak>	 So leave it unbalanced for now?
[20:13:55] <Amir1>	 yeah
[20:13:58] <Amir1>	 that's my idea
[20:14:07] <halfak>	 This means that anything that is relying on our percentage to be consistent is going to be sad. 
[20:14:22] <halfak>	 E.g. if you still have that bot on fawiki, the score threshold is going to change. 
[20:14:27] <halfak>	 You might even set it at 50%
[20:15:25] <Amir1>	 we should let programmers that use our API know about changes
[20:15:32] <Amir1>	 maybe by sending email to AI-l
[20:15:50] <halfak>	 Yeah.   I'm thinking that this will be a rough switch even if we give lots of notice.  We need a more stable output. 
[20:15:52] <halfak>	 Hmmm. 
[20:16:32] <YuviPanda>	 https://phabricator.wikimedia.org/T120317#1933458
[20:16:33] <YuviPanda>	 hmm
[20:16:36] <YuviPanda>	 so
[20:16:38] <YuviPanda>	 I should install
[20:16:43] <YuviPanda>	 'revscoring' and 'wikiclass' modules
[20:16:55] <Amir1>	 here is my idea: we need to have some developers that built upon our tools and then we have users that they have no idea about how ORES works, 
[20:17:07] <Amir1>	 the users of tools should not get any percentage from ORES
[20:17:21] <halfak>	 Amir1, +1 to that. 
[20:17:39] <halfak>	 But we might want them to get a recall or precision -- just not the "probability" field. 
[20:17:40] <Amir1>	 they should just get "colors" or radio buttons of "soft and hard" threshold in their preferences 
[20:18:11] <Amir1>	 we should give "precision", "recall" to devs
[20:18:13] <Amir1>	 not users
[20:18:29] <halfak>	 Amir1, users could think effectively about these values, I think. 
[20:19:00] <Amir1>	 mostly they don't know about precision or recall
[20:19:04] <halfak>	 E.g. for a given score you can say (This has a X% chance of actually being vandalism) 
[20:19:07] <Amir1>	 I talked to most of them in my wiki
[20:19:21] <halfak>	 Right now, our probability doesn't really do that, but it could. 
[20:20:01] <Amir1>	 I see
[20:20:13] <Amir1>	 in that case we might need scaling
[20:20:28] <Amir1>	 normal scaling 
[20:20:35] <Amir1>	 turning these numbers to z scores
[20:20:36] <halfak>	 So, i think that us *exactly* what we are getting with the Gradient Boost model. 
[20:21:03] <halfak>	 We've been living with crazy scaling issues with SVCs
[20:21:24] <halfak>	 Moving away from support vectors should make these probability estimates a lot saner. 
[20:22:35] <Amir1>	 +1
[20:22:39] <Amir1>	 let's give it a try
[20:23:44] <halfak>	 OK.  We'll be able to test out in staging first either way. 
[20:24:40] * halfak continues work on editquality models. 
[20:24:49] <halfak>	 Will load up an fawiki campaign after this is done. 
[20:25:53] <Amir1>	 thanks
[20:25:54] <Amir1>	 :)
[20:27:55] <halfak>	 Our goodfaith classifier in fawiki kicks ass. 
[20:28:25] <Amir1>	 yay
[20:28:34] <Amir1>	 we have lots of people working on it
[20:28:46] <Amir1>	 ORES has lots of fans in fa.wp
[20:28:47] <halfak>	 At 50%, we get 93% recall and 99.9% precision
[20:29:16] <halfak>	 The damaging classifier isn't as good.  Maybe the vandal bot should target the goodfaith model instead :) 
[20:30:45] <pipivoj>	 Had to restart my modem.
[20:31:50] <pipivoj>	 Halfak I had to install scipy before - before the reboots - because I couldn't use sklearn, it said it needed scipy
[20:31:57] <pipivoj>	 The same again
[20:32:39] <halfak>	 Looks like YuviPanda is going to work on it. 
[20:32:56] <halfak>	 He'll get the dependencies for revscoring and that includes sklearn 
[20:33:05] <YuviPanda>	 yeah
[20:33:06] <YuviPanda>	 so
[20:33:17] <YuviPanda>	 that's the revscoring and editquality pip packages right?
[20:33:30] <YuviPanda>	 pipivoj: if I install it in the base everyone'll get it automatically
[20:33:43] <pipivoj>	 that would be great
[20:34:17] <YuviPanda>	 halfak: can you confirm it's just the revscoring and editquality pip packages? :D sorry if you already did and I missed it
[20:34:24] <pipivoj>	 Then my notebooks wouldn't complain if reboot happened.
[20:34:25] <halfak>	 YuviPanda, +1 for pip
[20:34:54] <YuviPanda>	 ok
[20:34:57] <YuviPanda>	 let me do
[20:36:25] <YuviPanda>	 ok so
[20:36:26] <YuviPanda>	 I'm installing
[20:36:43] <YuviPanda>	 revscoring editquality wikiclass
[20:36:49] <YuviPanda>	 pipivoj: would you like any other libraries?
[20:36:51] <YuviPanda>	 sklearn-nn?
[20:36:57] <pipivoj>	 yes
[20:37:07] <halfak>	 Thanks YuviPanda :)  
[20:37:09] <pipivoj>	 its name is sklearn-neuralnetwork
[20:38:02] <halfak>	 YuviPanda, if you have a minute when you are done, could you check out the staging and web nodes.  They are running the "uwsgi" service rather than the "uwsgi-ores-web" service. 
[20:38:15] <halfak>	 This messes up our deploy scripts and it seemed weird enough to have you take a look. 
[20:38:26] <halfak>	 I'm worried puppet is doing something weird. 
[20:38:37] <YuviPanda>	 halfak: I responded yesterday
[20:38:43] <halfak>	 Oh!  I missed it. 
[20:38:44] <YuviPanda>	 you're right re: reboot
[20:38:53] <YuviPanda>	 I killed the uwsgi service and started uwsgi-ores-web
[20:39:01] <YuviPanda>	 and since we had the lb it was fine to do
[20:39:06] <YuviPanda>	 haven't on staging
[20:39:11] <YuviPanda>	 I need to fix puppet to not do that anymore
[20:39:21] <halfak>	 Great.  OK.  I'll get the task made. 
[20:39:28] <YuviPanda>	 yeah +1
[20:39:44] <YuviPanda>	 not too hard, just ensure that the 'uwsgi' service is dead everywhere
[20:40:11] <halfak>	 +1
[20:40:41] <YuviPanda>	 pipivoj: am doing the install / builds now
[20:41:06] <pipivoj>	 great
[20:41:32] <halfak>	 https://phabricator.wikimedia.org/T124621
[20:41:35] <halfak>	 YuviPanda, ^
[20:41:52] <YuviPanda>	 thanks halfak
[20:42:06] <YuviPanda>	 we're using assignment in the labs team to track some stuff, so am leaving this unassigned
[20:42:18] <YuviPanda>	 plus I hope to convince madhuvishy to do this instead of me :D
[20:42:41] <halfak>	 I'm happy to bug madhu about it.  Could be a fun problem for schana too. 
[20:42:49] <halfak>	 I've got to run now though. 
[20:42:56] <halfak>	 Gotta "go outside" and "be a human"
[20:42:57] <halfak>	 Psssh
[20:42:58] <halfak>	 o/
[20:43:40] <YuviPanda>	 I should probably do those things too eh
[20:46:29] <pipivoj>	 Have to reset my modem again. Brb.
[20:47:18] <YuviPanda>	 installation is chugging along
[20:47:23] <YuviPanda>	 one of the negatives of having a low powered CPU
[20:47:28] <YuviPanda>	 is that docker builds take forever
[20:52:21] <pipivoj>	 Back :)
[20:57:19] <YuviPanda>	 cool
[20:57:30] <YuviPanda>	 pipivoj: it's still building, going to be at least 30mins maybe more
[20:59:21] <pipivoj>	 Ok, I was just wondering what happened cause it was faster with my "user" install.
[21:00:26] <pipivoj>	 Ah, ofcourse you are building it I was merely installing it. :)
[21:00:45] <YuviPanda>	 pipivoj: yeah, this is rebuilding the entire docker image
[21:00:55] <YuviPanda>	 and since it isn't in my cache anymore it's rebuilding *all the libraries*
[21:40:34] <grrrit-wm>	 (03PS5) 10Ladsgroup: Let people choose a threshold [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265676 (https://phabricator.wikimedia.org/T122684) 
[21:40:50] <grrrit-wm>	 (03CR) 10jenkins-bot: [V: 04-1] Let people choose a threshold [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265676 (https://phabricator.wikimedia.org/T122684) (owner: 10Ladsgroup)
[21:46:45] <grrrit-wm>	 (03CR) 10Hoo man: "All other changes look good to me." (031 comment) [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265944 (https://phabricator.wikimedia.org/T124443) (owner: 10Ladsgroup)
[22:23:06] <grrrit-wm>	 (03PS6) 10Ladsgroup: Let people choose a threshold [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265676 (https://phabricator.wikimedia.org/T122684) 
[22:55:51] <grrrit-wm>	 (03CR) 10Ladsgroup: Some minor improvements to the database schema (031 comment) [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265944 (https://phabricator.wikimedia.org/T124443) (owner: 10Ladsgroup)
[23:06:59] <grrrit-wm>	 (03CR) 10Ladsgroup: [C: 032] Update authors += ladsgroup [extensions/ORES] - 10https://gerrit.wikimedia.org/r/264607 (owner: 10Awight)
[23:07:48] <Amir1>	 :D
[23:07:57] <grrrit-wm>	 (03Merged) 10jenkins-bot: Update authors += ladsgroup [extensions/ORES] - 10https://gerrit.wikimedia.org/r/264607 (owner: 10Awight)
[23:15:06] <grrrit-wm>	 (03CR) 10Hoo man: "re" (031 comment) [extensions/ORES] - 10https://gerrit.wikimedia.org/r/265944 (https://phabricator.wikimedia.org/T124443) (owner: 10Ladsgroup)
[23:44:51] <pipivoj>	 YuviPanda, still no sign of scipy. Nevermind, will try tomorrow. Bye