<     May 2017     >
Su Mo Tu We Th Fr Sa  
    1  2  3  4  5  6  
 7  8  9 10 11 12 13  
14 15 16 17 18 19 20  
21 22 23 24 25 26 27  
28 29 30 31
00:00 griffinbyatt joined
00:14 PaReeOhNos joined
00:16 jkreeftmeijer joined
00:16 Cohedrin joined
00:17 MarioBranco joined
00:22 __charly__ joined
00:39 willyk joined
00:41 Ilyes512 joined
01:08 kronicdeth joined
01:09 Ven joined
01:17 PaReeOhNos joined
01:27 srxa joined
01:39 tomterl joined
01:42 kronicdeth joined
01:46 __charly__ joined
02:04 __charly__ joined
02:04 elgenie joined
02:06 Ilyes512 joined
02:13 mikebonar joined
02:17 jkreeftmeijer joined
02:18 MarioBranco joined
02:18 junsuijin joined
02:19 PaReeOhNos joined
02:37 MarioBranco joined
02:38 Lokathor joined
02:51 Lokathor joined
03:04 mikebonar joined
03:15 __charly__ joined
03:19 jkreeftmeijer joined
03:22 notdaniel joined
03:23 PaReeOhNos joined
03:42 nahtnam joined
04:09 nhooyr joined
04:16 PaReeOhNos joined
04:17 nhooyr joined
04:20 proteusguy joined
04:35 arquebus joined
04:36 arquebus joined
04:36 arquebus joined
04:37 arquebus joined
04:41 mikebonar joined
04:58 mikebonar joined
05:08 codestorm joined
05:18 PaReeOhNos joined
05:20 jkreeftmeijer joined
05:20 meandi_2 joined
05:30 wsieroci joined
05:34 segmond joined
05:44 griffinbyatt joined
05:55 cemilowski joined
05:58 PSvils joined
06:03 <hexkey[m]> ..hello?
06:06 <hexkey[m]> Right, timzeones. Anyway, if anyone sees this: I'm trying to measure the performance of a few sorting algorithims, and am seeing some weird shit.
06:06 <* hexkey[m]> uploaded an image: mergesort.png (27KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/hHUtDzTMPhGolBBpnFjJWldH>
06:08 <hexkey[m]> mergesort should be around O(nlog) if I'm not mistaken, but something weird happens around 1000 items
06:09 <hexkey[m]> this isn't some one-off thing either, these numbers are averaged across 25 runs.
06:11 elgenie joined
06:13 <hexkey[m]> To be a little more specific: my code is sorting segments of a list of random numbers--starting with the first zero numbers, then first 10, 20, 30, all the way up to 2,490 (I can go up to 10,000 if needed).
06:14 <hexkey[m]> are there extra processses being spawned to compensate or something? that's my guess
06:15 <hexkey[m]> anyway, if you are awake and reading this and have any idea about what the reason for this might be, that'd b a big help.
06:15 Tica2 joined
06:16 <hexkey[m]> I can toss my code up on a paste, share details about my env, screencaps of my cpu usage graph, etc
06:18 <hexkey[m]> this isn't super urgent or anything, but I am just so mystified about how BEAM is seemingly performing magic.
06:19 <asonge> hexkey[m]: won't mind looking in a few min.
06:19 <hexkey[m]> cool
06:21 jkreeftmeijer joined
06:21 PaReeOhNos joined
06:24 refriedchicken joined
06:29 jkreeftmeijer joined
06:35 tuacker joined
06:35 jkreeftmeijer joined
06:39 <hexkey[m]> also, one other thing: I've done somefancy excel magic so I can get new graphs as soon as things finish running, so there's that too
06:44 griffinbyatt joined
06:53 nahtnam joined
06:54 jimmyrcom_ joined
06:56 <asonge> hexkey[m]: got a pastebin?
06:56 <asonge> sorry it took so long, the turkey i thought was defrosted...wasn't.
07:02 <hexkey[m]> right, back. one sec.
07:04 <hexkey[m]> I have a bunch of algs I'm testing, all of which are acting weird interms of time. The mergesort is short enough that I can stick it here.
07:04 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:04:37.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/kXRaVurLASpyLFakYVuteEDk>
07:04 <hexkey[m]> close enough
07:04 <micmus> one thing to remember is that you have lists and not arrays
07:05 <asonge> yeah, list operations are going to be linear
07:05 <asonge> all the time
07:05 jkreeftmeijer joined
07:06 <micmus> there are usually algorithms that are better suited for immutable languages
07:06 <micmus> or light modifications of regular ones
07:07 <hexkey[m]> Mergesort should work pretty nice here though. There's no random access required, for one thing.
07:07 <hexkey[m]> the quicksport I wrote for comparision, on the other hand...
07:08 <hexkey[m]> And it does perform well. I'm just wondering why the execution time levels out at ~1000 items.
07:09 <hexkey[m]> I'm assuming it's some multiprocessing thing in the BEAM for now.
07:10 <micmus> no, definitely not. Unless you explicitly start a process, there will be only one.
07:10 <micmus> How do you measure time?
07:11 <micmus> I'd generally recommend https://github.com/PragTob/benchee for running benchmarks
07:11 <hexkey[m]> one sec
07:11 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:11:22.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/kdZoDvqwtPIKpGgzOdMCBXAG>
07:12 <hexkey[m]> There's some extra stuff in other modules, but this is the main attraction I suppose
07:12 <hexkey[m]> wait
07:12 <hexkey[m]> it cut off some
07:12 <micmus> guard with length is going to be pretty bad
07:12 <micmus> since it has to traverse the list
07:13 <micmus> pattern match on the list to check length, should be much, much better
07:13 josevalim joined
07:13 <hexkey[m]> where am I using a guard?
07:14 <hexkey[m]> also, lemme wuick try to paste that again
07:14 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:14:19.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/XNzfxFtKGLvHTQYHvzriuLyo>
07:14 <hexkey[m]> that looks better
07:15 jcspencer joined
07:15 <micmus> hexkey[m]: in merge_sort, you have when length(list) <= 1
07:15 <hexkey[m]> do: list
07:16 <hexkey[m]> right?
07:16 cemilowski joined
07:16 <hexkey[m]> here's the sort itself again:
07:16 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:16:53.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/zPUGnYDelbCOLDMDeYlHQOGg>
07:17 <hexkey[m]> It's pretty much the only implementation that comes up when googling 'elixir mergesort'
07:19 ejpcmac joined
07:22 PaReeOhNos joined
07:23 <micmus> hexkey[m]: if you add def merge_sort([]), do: []; def merge_sort([_] = list), do: list
07:23 <micmus> it should improve perf significantly
07:24 jkreeftmeijer joined
07:25 <hexkey[m]> I get that first line, but what exactly is going on in the second?
07:25 <micmus> you match on a single element list
07:25 <sevenseacat> sorting a one-item list
07:25 <micmus> you could do def merge_sort([elem]), do: [elem]
07:26 <micmus> but that would allocate the list again, assigning with `= list` allows you to return the same value
07:27 codestorm joined
07:27 danguita joined
07:28 codestorm joined
07:28 <hexkey[m]> I really was not expecting guards to be such a slowdown here
07:28 <hexkey[m]> I'll give this a spin and see what happens
07:30 <hexkey[m]> alrighty, done
07:30 <hexkey[m]> one sec while I screenshot some graphs
07:32 <hexkey[m]> got it
07:32 <hexkey[m]> before:
07:32 <* hexkey[m]> uploaded an image: ms.png (34KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/hjBrvAwvlfpKuOFNyfeedKXZ>
07:32 <hexkey[m]> and after:
07:33 <* hexkey[m]> uploaded an image: ms-again.png (31KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/rSkuSqhYMEaQwovXIvbxRPiz>
07:33 <hexkey[m]> definite improvment
07:34 <micmus> It looks so strange, it flattens out...
07:35 <hexkey[m]> That's the part I'm trying to figure out
07:35 <hexkey[m]> It's so consistent too.
07:35 <hexkey[m]> This is averaged over 25 runs
07:35 <josevalim> why not use :lists.sort or Enum.sort which are merge_sort iirc?
07:36 <hexkey[m]> I could, but I also want to figure this out
07:37 <hexkey[m]> It's quite the spooky mystery
07:37 <micmus> I would first test if it's really sorted at the end :P
07:38 <hexkey[m]> I think I do that, one sec
07:38 <hexkey[m]> I'll check
07:38 <hexkey[m]> okay, so I do check...
07:39 willyk joined
07:39 <hexkey[m]> but with the exact same sorter I use in the first place
07:39 <hexkey[m]> :|
07:39 <hexkey[m]> but to be fair
07:39 <hexkey[m]> I did check it manually several times earlier
07:39 <hexkey[m]> and there should be no good reason to suddenly change
07:40 <micmus> You could try testing with x = my_sort_alg; ^x = Enum.sort(x)
07:41 <hexkey[m]> here's my test after modinfying that one part:
07:41 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:41:23.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/HmSTnrcenzDZMhoavsSbpfNi>
07:41 <hexkey[m]> now to run it
07:42 <hexkey[m]> ran it, no exceptions.
07:44 <micmus> hexkey[m]: have you tried benching the stdlib algorithm and comparing?
07:45 <hexkey[m]> I can just swap it out and try right now, I suppose
07:45 griffinbyatt joined
07:46 <hexkey[m]> there we go
07:46 <hexkey[m]> ``def merge_sort(list) do
07:46 <hexkey[m]> :lists.sort(list)
07:46 <hexkey[m]> end
07:46 <hexkey[m]> ``def merge_sort(list) do
07:46 <hexkey[m]> :lists.sort(list)
07:46 <hexkey[m]> end``
07:46 <hexkey[m]> better
07:46 <hexkey[m]> now to run it
07:47 <hexkey[m]> okay now things are just getting weird
07:47 <hexkey[m]> before:
07:48 <* hexkey[m]> uploaded an image: ms-again.png (31KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/OtpMcbQcHHIUKfSIuUSEGVWf>
07:48 <hexkey[m]> aaaaand after:
07:48 <* hexkey[m]> uploaded an image: ms-wat.png (27KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/vJRJgvWWjjyvtCZPmfZEPqvn>
07:49 <asonge> hexkey[m]: how do you time your code?
07:49 <asonge> i assume x axis is number of items in the list?
07:49 <nox> hexkey[m]: How could length() not hurt perfs?
07:50 <hexkey[m]> yes. Y axis is time in milliseconds
07:50 <nox> hexkey[m]: It takes linear time proportional to the list length.
07:50 <hexkey[m]> here
07:50 <hexkey[m]> *here's the timing section:
07:50 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-21_07:50:56.txt <https://matrix.org/_matrix/media/v1/download/matrix.org/WKHyXyOcdmxkrqQLImyLpYfG>
07:50 <asonge> if you're converting :os.timestamp incorrectly, it might show up with that kind of graph
07:51 <asonge> are you on windows?
07:52 <hexkey[m]> I was unitl I discovered that erlang windows sucks at time
07:52 <hexkey[m]> It's currently running in xubuntu
07:52 <hexkey[m]> in a VM
07:52 <asonge> try :timer.tc instead?
07:53 Guest67 joined
07:57 <hexkey[m]> I'm not entirely sure how to do this with my current setup
07:57 <hexkey[m]> ``{elapsed, sorted} = :timer.tc(fn -> apply(sorter, list_segment) end, [])``
07:57 <hexkey[m]> something vaugely like this I know
07:58 <hexkey[m]> wait I think I got it
07:59 <hexkey[m]> ``{elapsed, sorted} = :timer.tc(fn(srt, lst) -> apply(srt, lst) end, [sorter, list_segment])``
07:59 <hexkey[m]> looks right
08:00 <hexkey[m]> timer.tc returns millisecs, right?
08:00 <asonge> yeah
08:00 <asonge> i wonder if nanoseconds just might be some kind of guess.
08:01 <micmus> it should be enough to do :timer.tc(sorter, list_segment) - use it as you would apply
08:01 <hexkey[m]> I think timer.tc/1 only takes in a function
08:02 <hexkey[m]> also I kinda need to use apply with the way I've set it up
08:02 <micmus> http://erlang.org/doc/man/timer.html#tc-2
08:03 <hexkey[m]> oboy errors
08:05 <hexkey[m]> apparently I'm calling my sorter with no arguments
08:08 <hexkey[m]> okay, got that to work
08:08 <hexkey[m]> the number's I'm getting have lot less percision
08:08 <hexkey[m]> guessing it's because I'm no longer fetching nanoseconds
08:09 <hexkey[m]> I don't know if they're valid or not
08:09 <hexkey[m]> but I'm about to update the graph
08:09 <hexkey[m]> before picture is same as above
08:10 <hexkey[m]> aaaaaaaand:
08:10 <* hexkey[m]> uploaded an image: ms-5.png (29KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/eLDgNvcXnRRpnRiIKKdFhNAs>
08:11 <asonge> can you try making the counts jump by 50 items at a time? wonder if the scale is making things weird.
08:12 <hexkey[m]> can do
08:12 <hexkey[m]> one sec
08:13 <hexkey[m]> it'll still go from 0 to 2490,but I can't guarentee what kind of rounding will be done there
08:14 <hexkey[m]> it should go under instead of over
08:14 <hexkey[m]> but there are enough data points that it prob doesnt matter
08:15 <* hexkey[m]> uploaded an image: ms-6.png (16KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/RvJnUhzUhjBGlQpGseKUguis>
08:15 <hexkey[m]> ...
08:15 <hexkey[m]> lemme check what I entered
08:16 <hexkey[m]> yep, that was it
08:16 <hexkey[m]> Imma try 20
08:18 <* hexkey[m]> uploaded an image: ms-7.png (27KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/oeFuomxGATfOngXFDtWjTerm>
08:18 <asonge> something seems wrong with this.
08:18 <hexkey[m]> gonna check a thing quick
08:21 jcspencer joined
08:21 <hexkey[m]> Okay, so what I think is going on is that timer.tc is returning numbers small enough that they get rounded downtozero
08:22 codestorm joined
08:22 <hexkey[m]> why at a certain point, I have noidea
08:22 <asonge> the vm just might not be giving very accurate numbers
08:23 <hexkey[m]> It just seems so strangely consistent though
08:24 <hexkey[m]> I could try running on an actual linux device, but my dual boot may or may not be in a working/sane state
08:24 <asonge> well, if it randomly returns 0 answers and rounds stuff down in the small cases, that explains the rise
08:24 wsieroci joined
08:24 <asonge> the linearity might show up on a larger scale.
08:24 <hexkey[m]> I can go up to 10,000
08:24 <asonge> i usually see sort stats for 10, 100, 1000, 10k, 100k, 1m, etc.
08:24 <asonge> something log-ish.
08:26 PaReeOhNos joined
08:26 <hexkey[m]> gonna try 0~1000 in steps of 20
08:29 <* hexkey[m]> uploaded an image: ms-8.png (24KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/lviwRvbRJaybbryopkyhfTqU>
08:30 <hexkey[m]> okay, so switching to 10 as the interval gave this:
08:30 <* hexkey[m]> uploaded an image: ms-9.png (28KB) <https://matrix.org/_matrix/media/v1/download/matrix.org/IotjVpbkDyltfabkzoSyAXvI>
08:31 <hexkey[m]> the dropoff is expected though
08:31 <hexkey[m]> I didn't ajducst the bounds of th cahrt yet
08:31 <hexkey[m]> hm.
08:31 <micmus> try running multiple iterations instead of just one. This should increase the overall time and make it more reliable
08:31 <hexkey[m]> well I kinda gotta sleep right now.
08:32 <hexkey[m]> I already do.
08:32 <hexkey[m]> this is the average of 25 runs
08:33 <hexkey[m]> it sorts from lists of size x to size y in steps of z, and repeats w intervals
08:33 <micmus> I'm waiting on a delayed flight, so I have lots of time :P
08:34 <hexkey[m]> I am barely awake though :/
08:34 <hexkey[m]> wait, idea:
08:34 <hexkey[m]> I can stick all the code up somewhere if you wanna look at it
08:35 <hexkey[m]> cant really pastebin it, multiple files
08:35 <asonge> gist supports multiple files
08:36 <hexkey[m]> that'll work
08:39 willykaram joined
08:41 <hexkey[m]> okay, done:
08:42 <hexkey[m]> https://gist.github.com/anonymous/af550ab4b4c11caf350733dc24cf1610
08:42 <hexkey[m]> It's a bit messy, but it works
08:42 <hexkey[m]> anyway I really need to sleep
08:43 <hexkey[m]> I'll be back here in ~10-ish hours
08:43 <hexkey[m]> one last thing
08:45 <hexkey[m]> if you run it, you'll want to use the 'overwrite-all' flag I included
08:46 <hexkey[m]> okay, now I'm gong to bed.
08:46 Uniaika joined
08:46 <hexkey[m]> seeya~
08:46 griffinbyatt joined
08:47 <hexkey[m]> one other last thing: might wanna send me a private message if you find anything so that it doesnt get lost in all the scrolling
08:48 <hexkey[m]> okay, now that should be it
08:51 stephen_m joined
08:52 PaReeOhNos joined
08:54 indyone joined
08:58 squallstter joined
09:16 codestorm joined
09:20 wsieroci joined
09:27 lexmag joined
09:28 jkreeftmeijer joined
09:36 Tica2 joined
09:47 griffinbyatt joined
09:48 codestorm joined
09:57 Uniaika joined
10:00 dustinfarris joined
10:04 junsuijin joined
10:07 danguita joined
10:13 codestorm joined
10:26 Ven joined
10:30 jkreeftmeijer joined
10:31 PaReeOhNos joined
10:32 srxa joined
10:40 craigp joined
10:47 griffinbyatt joined
10:48 Ven_ joined
10:53 <rs`> I ported a Ruby script to Elixir, which was called every 10 minutes via cronjob. Is there an idiomatic way to do something like this in Elixir/OTP? Or do I just continue using cronjobs?
10:55 <Nicd-> depends on what kind of guarantees you want. the simplest is to have a GenServer call itself with a timer
10:55 <asonge> rs`: there is a library called quantum that takes crontab strings and will run tasks in a process
10:56 <asonge> you can just add the entries in your config and also query status of the jobs
10:56 <asonge> (no system cron needed)
10:57 <rs`> ok thanks, I will have a look at this
10:59 <asonge> the github readme has info not in the docs
11:02 cemilowski joined
11:07 codestorm joined
11:10 m00dy joined
11:11 notdaniel joined
11:15 m00dy joined
11:25 m00dy joined
11:28 m00dy_ joined
11:28 craigp joined
11:33 mbwe joined
11:34 Gasher joined
11:37 mikebonar joined
11:40 PaReeOhNos joined
11:45 Tica2_ joined
11:46 PaReeOhNos joined
11:48 gvaughn joined
11:48 <OliverMT> rs`: are you running on aws?
11:48 griffinbyatt joined
11:49 <OliverMT> I'm loving the cloudwatch crontab events
11:49 <OliverMT> move the responsibility of sending events outside your app
11:50 <Nicd-> Process.send_after best
11:50 <OliverMT> depends how complex your stuff is
11:50 <OliverMT> and once you have a cluster of nodes and need to do stuff like making sure your job runs on one node only
11:50 <OliverMT> bleh
11:50 <OliverMT> thats why I went aws lambda/cloudwatch
11:51 <Nicd-> 1 node best
11:51 <Nicd-> if you have too many users for 1 node, you need to limit registrations
11:58 dustinfarris joined
12:00 cemilowski joined
12:01 codestorm joined
12:05 srxa_ joined
12:12 jkreeftmeijer joined
12:19 Cloudflare joined
12:23 jkreeftmeijer joined
12:29 jkreeftmeijer joined
12:40 jkreeftmeijer joined
12:44 Tica2 joined
12:45 wsieroci joined
12:48 Tica2 joined
12:49 griffinbyatt joined
12:51 jkreeftmeijer joined
12:55 codestorm joined
12:56 Guest67 joined
12:57 jkreeftmeijer joined
13:00 wong2 joined
13:05 PaReeOhNos joined
13:06 Tica2 joined
13:07 jkreeftmeijer joined
13:09 Uniaika joined
13:10 Uniaika joined
13:12 Jam666 joined
13:12 Jam666 joined
13:17 josevalim_ joined
13:18 jkreeftmeijer joined
13:24 jkreeftmeijer joined
13:28 srxa joined
13:31 dimitarvp joined
13:35 jkreeftmeijer joined
13:36 gvaughn joined
13:36 ssswitch joined
13:37 wsieroci joined
13:40 dustinfarris joined
13:40 josevalim joined
13:45 jkreeftmeijer joined
13:49 Tica2 joined
13:50 codestorm joined
13:50 griffinbyatt joined
13:51 jkreeftmeijer joined
13:57 kau joined
13:59 tuacker1 joined
14:01 srxa_ joined
14:02 jkreeftmeijer joined
14:02 MarioBranco joined
14:09 PaReeOhNos joined
14:09 tuacker joined
14:10 MarioBranco joined
14:12 jkreeftmeijer joined
14:12 MarioBranco joined
14:13 fernandomm joined
14:14 PaReeOhNos joined
14:15 justelex joined
14:18 jkreeftmeijer joined
14:19 Lyubo1 joined
14:25 willyk joined
14:28 jkreeftmeijer joined
14:31 gvaughn joined
14:32 Tica2 joined
14:39 jkreeftmeijer joined
14:44 codestorm joined
14:45 jkreeftmeijer joined
14:50 superherointj joined
14:51 cemilowski joined
14:54 jcspencer joined
14:55 jkreeftmeijer joined
14:59 gvaughn joined
15:01 willykaram joined
15:02 griffinbyatt joined
15:05 jkreeftmeijer joined
15:08 meh` joined
15:11 jkreeftmeijer joined
15:12 PragTob joined
15:14 <PragTob> Hello my Elixir/Erlang friends - I've come to you for guidance on how I can make an Elixir library usable from Erlang and It's been suggested to ping tristan__ or alisdair :) https://elixirforum.com/t/how-to-make-an-elixir-library-available-to-erlang-users/5172
15:14 <PragTob> any help greatly appreciated :)
15:16 __charly__ joined
15:21 <ciawal> https://www.microsoft.com/en-us/research/blog/p-programming-language-asynchrony/
15:22 jkreeftmeijer joined
15:22 tuacker1 joined
15:26 <PragTob> found rebar3_elixir_compile which seems to do what I want it to :) https://github.com/barrel-db/rebar3_elixir_compile
15:26 PaReeOhNos joined
15:27 Ilyes512_ joined
15:28 <Ilyes512_> hey guys
15:28 <Ilyes512_> is there a way to IEx.pry Phoenix.Controller? I know you shouldn't change vendors but still...?
15:29 <Ilyes512_> i think what i am missing now is a "force" recompile
15:31 <Ilyes512_> nvm ... i needed mix deps.compile
15:32 bitmod_ joined
15:32 jkreeftmeijer joined
15:33 srxa joined
15:38 codestorm joined
15:38 jkreeftmeijer joined
15:45 Gasher joined
15:48 jkreeftmeijer joined
15:52 kronicdeth joined
15:54 griffinbyatt joined
15:59 jkreeftmeijer joined
15:59 Ven joined
16:00 verto joined
16:02 m00dy_ joined
16:05 jkreeftmeijer joined
16:05 Tica2 joined
16:10 danguita joined
16:11 dorei joined
16:13 Tica2 joined
16:15 jkreeftmeijer joined
16:21 willykaram joined
16:22 elgenie joined
16:24 m00dy joined
16:25 jkreeftmeijer joined
16:25 ambrosia joined
16:26 Tica2 joined
16:31 jkreeftmeijer joined
16:32 codestorm joined
16:36 PaReeOhNos joined
16:37 PaReeOhNos joined
16:38 returntrip joined
16:39 Tica2 joined
16:42 jkreeftmeijer joined
16:47 Ven joined
16:49 craigp joined
16:50 cemilowski joined
16:51 __charly__ joined
16:52 jkreeftmeijer joined
16:56 jcspencer joined
16:56 gvaughn joined
17:01 awea joined
17:03 jkreeftmeijer joined
17:03 <awea> Hi there! I go some git related problems with edeliver to build an upgrade. When it come to checkout the last changes it fail with a "please, commit your changes". I'm missing something on my build server?
17:03 snowowl joined
17:04 <snowowl> hi, i'm trying to get phoenix up and running but i'm having some issues with using the mix phoenix.new proj
17:05 <snowowl> if i use that, npm doesn't like the fact that deps/phoenix dir and file don't exist
17:05 <snowowl> so i have to modify the package.json and remove them to get it working
17:05 m00dy joined
17:06 <snowowl> i'm not sure if those are supposed to be empty files to be used later, or what the deal is?
17:07 Ven_ joined
17:08 <snowowl> second, after installing hex, ecto still doesn't want to run mix ecto.create
17:08 <snowowl> complains about something revolving around mix deps.get
17:08 jkreeftmeijer joined
17:09 refriedchicken joined
17:09 <snowowl> so i'm really wondering, what to do about this?
17:11 <mloy> snowowl: paste your specific error messages please, or we can't help
17:13 <snowowl> https://paste.pound-python.org/show/GlD6lp7WfgKWY1h4yngJ/
17:14 <snowowl> mloy: there they are, I just tried it with the --no-brunch option and it worked
17:19 <snowowl> seems like using brunch has an npm issue?
17:19 jkreeftmeijer joined
17:19 <mloy> snowowl: and i take it you did run mix deps.get?
17:19 <manukall> snowowl: did you run mix deps.get?
17:19 <manukall> :-)
17:20 <mloy> hehe
17:21 patientplatypus joined
17:22 <snowowl> mloy: not manually
17:23 <snowowl> mloy: i thought it was running that when i did mix phonix.new proj
17:23 <snowowl> i'll try
17:24 Tica2 joined
17:24 Ven_ joined
17:24 <mloy> snowowl: it prompts "Fetch and install dependencies? [Yn]" after running `mix phoenix.new proj`
17:24 <mloy> you may have accidentally not accepted
17:24 <snowowl> i did
17:25 <mloy> ah, hmm
17:25 <snowowl> running it with mix deps.get manually seems to work
17:25 <mloy> good :)
17:26 <snowowl> yep that was the issue, everythings running swimmingly
17:26 <snowowl> thanks mloy and manukall
17:26 codestorm joined
17:27 ejpcmac joined
17:27 <mloy> glad to help! happy hacking
17:28 wsieroci joined
17:29 jkreeftmeijer joined
17:34 PaReeOhNos joined
17:34 segmond joined
17:35 jkreeftmeijer joined
17:37 Ven_ joined
17:46 jkreeftmeijer joined
17:49 __charly__ joined
17:56 jkreeftmeijer joined
18:00 kronicdeth joined
18:06 jkreeftmeijer joined
18:12 jkreeftmeijer joined
18:12 srxa joined
18:13 kronicdeth joined
18:20 codestorm joined
18:21 m00dy joined
18:22 elgenie joined
18:22 m00dy_ joined
18:22 jkreeftmeijer joined
18:24 willyk joined
18:27 codestorm joined
18:33 jkreeftmeijer joined
18:35 srxa_ joined
18:38 jkreeftmeijer joined
18:44 gvaughn joined
18:46 m00dy joined
18:49 jkreeftmeijer joined
18:50 Cohedrin joined
18:50 __charly__ joined
18:52 PaReeOhNos joined
18:59 jkreeftmeijer joined
18:59 dignifiedquire joined
19:02 <Ilyes512> hmm i think i might have found an error in Phoenix.Controller.accepts/2
19:03 <Ilyes512> anyone here that got time to confirm my findings?
19:05 jkreeftmeijer joined
19:06 <Nicd-> Ilyes512: what kind of error?
19:06 <Ilyes512> it does not except a valid Accept value as a type
19:07 <Ilyes512> for instance
19:08 <Ilyes512> I got this in my pipeling: "plug :accepts ~w(xml)"
19:08 marr joined
19:08 <Ilyes512> When I send a request with the value of the Accept header set as "xml/*". It does not recognize it as xml
19:09 <Ilyes512> that is because of Phoenix.Controller.find_format()
19:09 <Ilyes512> I think the fallowing is missing at the moment: defp find_format(exts, accepted) when is_binary(exts), do: exts in accepted
19:09 <Nicd-> isn't xml "application/xml" or "text/xml"?
19:10 <Ilyes512> yes true but when you send "something/*" it set something as the extension
19:10 <Ilyes512> defp find_format(exts, accepted) when is_binary(exts), do: exts in accepted
19:11 <Ilyes512> oeps
19:11 <Ilyes512> https://github.com/phoenixframework/phoenix/blob/master/lib/phoenix/controller.ex#L1140-L1143
19:11 <Ilyes512> parse_header_accept passes a binary in my case but find_format expects a list of binaries
19:13 <Ilyes512> Here is the code that returns xml in the case of "xml/*" https://github.com/phoenixframework/phoenix/blob/master/lib/phoenix/controller.ex#L1137
19:14 <Ilyes512> I think this should work, but I have been looking at this now for mor then an hour. I guess I have to create issue to explain this better.
19:15 jkreeftmeijer joined
19:15 <hexkey[m]> hey asonge: you here? did you find anything out from looking at my code?
19:16 gvaughn joined
19:17 Talltree joined
19:22 patientplatypus joined
19:25 __charly__ joined
19:25 jkreeftmeijer joined
19:31 jkreeftmeijer joined
19:33 squallstter joined
19:34 Cohedrin joined
19:36 <Ilyes512> is it possible to only run 1 describe group within a test file?
19:39 <Ilyes512> hmm running the whole test was slow (0.6 seconds :P). But still for future reference is it possible to only run a single describe group (I know you can run a single test if you provide the line number)
19:40 nd__ joined
19:41 jkreeftmeijer joined
19:43 karmajunkie joined
19:45 <OliverMT> Ilyes512: try doing yourtest.exs:123
19:45 <OliverMT> where 123 is the line number the describe starts
19:45 <OliverMT> oh, didnt read your paranthesis :P
19:46 m00dy joined
19:46 <Ilyes512> well i havent tried setting it on the describe instead of a function
19:47 <Ilyes512> nope it will only run the 1 test after that line
19:47 cemilowski joined
19:52 jkreeftmeijer joined
19:58 jkreeftmeijer joined
20:03 jkreeftmeijer joined
20:04 wsieroci joined
20:06 m00dy joined
20:09 jcspencer joined
20:12 esmiurium joined
20:13 jkreeftmeijer joined
20:15 notdaniel joined
20:16 willyk joined
20:19 jkreeftmeijer joined
20:29 jkreeftmeijer joined
20:32 jerel joined
20:33 codestorm joined
20:37 nbg joined
20:40 jkreeftmeijer joined
20:40 PaReeOhNos joined
20:45 jkreeftmeijer joined
20:51 PSvils joined
20:51 gvaughn joined
20:53 kronicdeth joined
20:56 jkreeftmeijer joined
20:58 cemilowski joined
20:58 srxa joined
21:06 jkreeftmeijer joined
21:09 ur5us joined
21:12 jkreeftmeijer joined
21:16 jmiven joined
21:22 jkreeftmeijer joined
21:26 craigger joined
21:26 kronicdeth joined
21:29 jkreeftmeijer joined
21:30 srxa joined
21:30 Ilyes512 joined
21:34 m00dy joined
21:35 PaReeOhNos joined
21:36 bitmod joined
21:37 pufuddled joined
21:38 srxa_ joined
21:39 srxa joined
21:39 jkreeftmeijer joined
21:40 <OliverMT> whats the name of "..." operator you can use on lists in ecto queries
21:40 <OliverMT> I thought it was spread operator
21:49 cschneid_ joined
21:49 <Ankhers> I can't find any "name" within the code, but I did see it referenced as the spread operator in this thread -- https://groups.google.com/d/msg/elixir-ecto/5teiP0Rld6M/QplFMmx6AwAJ
21:50 jkreeftmeijer joined
21:52 <hexkey[m]> so I think I solved my weird graph problem
21:53 <hexkey[m]> the one where I seemingly defied the laws of computer science
21:53 <hexkey[m]> at least I think I solved it
21:53 <hexkey[m]> dunno how
21:55 jkreeftmeijer joined
21:59 srxa_ joined
22:06 jkreeftmeijer joined
22:16 jkreeftmeijer joined
22:22 jkreeftmeijer joined
22:32 jkreeftmeijer joined
22:42 isubasti joined
22:42 Ilyes512 joined
22:42 jkreeftmeijer joined
22:44 __charly__ joined
22:48 codestorm joined
22:48 jkreeftmeijer joined
22:49 isubasti_ joined
22:58 jkreeftmeijer joined
23:03 patientplatypus joined
23:05 jcspencer joined
23:08 dustinfarris joined
23:09 jkreeftmeijer joined
23:11 adgtl joined
23:11 cemilowski joined
23:15 jkreeftmeijer joined
23:15 Antiarc joined
23:20 PaReeOhNos joined
23:23 m00dy joined
23:25 jkreeftmeijer joined
23:29 kronicdeth joined
23:35 jkreeftmeijer joined
23:41 jkreeftmeijer joined
23:49 Ioyrie joined
23:51 jkreeftmeijer joined
23:58 jkreeftmeijer joined