<     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 _2_7  
28 29 30 31
00:01 pbj joined
00:08 wpcarro joined
00:09 jkreeftmeijer joined
00:20 jkreeftmeijer joined
00:21 <patientplatypus> ok
00:21 <patientplatypus> dumb question
00:21 <patientplatypus> on the standard hello world that mix phoenix.new gives
00:21 <patientplatypus> where and or how can i navigate to app.html?
00:21 <patientplatypus> i dont see this in the routes
00:21 <patientplatypus> like i must be dumb
00:27 wsieroci joined
00:30 jkreeftmeijer joined
00:31 jhack joined
00:31 meh` joined
00:33 <Ankhers> app.html is a layout. You can find it in the templates/layout folder.
00:33 <patientplatypus> ok
00:34 <patientplatypus> i guess where im getting stuck is that if i navigate to that page
00:34 <Ankhers> You have this in websites because most pages look alike side from the main content.
00:34 wpcarro joined
00:34 <Ankhers> There is no app.html page though.
00:34 <patientplatypus> so i made a page that navigates to there
00:35 <patientplatypus> through the routes as it appeared to be html
00:35 <patientplatypus> and it worked but this line
00:35 <patientplatypus> <%= render @view_module, @view_template, assigns %>
00:35 <patientplatypus> threw an error - no @view_module
00:35 <patientplatypus> which is like, of course not, there isnt one in there
00:36 <Ankhers> Yes, because that line is attempting to render the specific information for that page that you are wanting to load.
00:38 <patientplatypus> im not sure i follow - is the app.html.eex file being called from somewhere when the hello world file loads? it seems that instead that index.html.eex is being rendered
00:39 <Ankhers> app.html has a bunch of common information in it that you have across a bunch of pages. Things like which CSS file(s) to load, and the overall design of your application. Your index.html file and other page specific files will house just the information for that page.
00:40 <patientplatypus> ahhh
00:40 <patientplatypus> so its an example page that is not intended to be functional, but just provide naming conventions on how to link to things like your static css file for example
00:41 jkreeftmeijer joined
00:42 pbj joined
00:43 <Ankhers> No. That layout file is called for every page load. It will house things like your navigation bar and whatever else is common through all of your pages.
00:43 <Ankhers> That way you do not have to duplicate everything across all pages.
00:45 <patientplatypus> what is the file one parent above it that calls it
00:45 <patientplatypus> ?
00:46 <Ankhers> I don't understand what you are asking.
00:48 gvaughn joined
00:52 jkreeftmeijer joined
01:02 jkreeftmeijer joined
01:03 gvaughn joined
01:09 <Ankhers> It is just something in Phoenix that calls it. It is set automatically, but you can have different layouts for different actions or controllers.
01:12 jkreeftmeijer joined
01:13 mika_ joined
01:15 wpcarro joined
01:18 frigidcode joined
01:42 irclogger_com joined
01:42 Topic for
01:43 wpcarro joined
02:00 wpcarro joined
02:13 jkreeftmeijer joined
02:13 bambanx joined
02:18 <patientplatypus> hi anyone still awake in here
02:20 merlin__ joined
02:23 jkreeftmeijer joined
02:27 wpcarro joined
02:32 pbj joined
02:34 jkreeftmeijer joined
02:36 <hexkey[m]> Yep. And I have some questions if anyone can halp
02:38 wpcarro joined
02:38 <hexkey[m]> I'm trying to time a function and am having a hard time getting good accuracy .it seems like the Time module might be...broken?
02:38 mika_ joined
02:39 <hexkey[m]> ...Or numbers maybe, because all my results usually are 0.0
02:40 <hexkey[m]> No matter how many times I've run this example code, I've never gotten a time besides 0:
02:40 <hexkey[m]> http://stackoverflow.com/questions/29668635/how-can-we-easily-time-function-calls-in-elixir
02:41 <hexkey[m]> Any chance someone could try it themself and confirm or deny that it behaves that way?
02:44 <hexkey[m]> It's getting fairly frustrating
02:44 jkreeftmeijer joined
02:55 jkreeftmeijer joined
02:58 wpcarro joined
03:05 jkreeftmeijer joined
03:09 wpcarro joined
03:15 jkreeftmeijer joined
03:26 jkreeftmeijer joined
03:27 fmcgeough joined
03:29 <asonge> hexkey[m]: :timer.tc is the standard way to go, and there's a good benchmark library that does some statistical runs
03:30 <asonge> hexkey[m]: but timing is rather unpredictable because the VM is so scheduled. it makes individual runs, particularly on a noisy developer laptop, highly variable, but it's also what makes performance stable when the machine is under heavy load
03:30 wpcarro joined
03:31 <hexkey[m]> yeah, I have noticed that timings of things here are extremely consistent
03:32 <asonge> like, a process might finish its reductions in your benchmark, and then another scheduler picks it up because of work-stealing
03:34 <asonge> in a lot of "big" erlang system benchmarks, you will notice that this helps keep that 99.9% latency figure very low while under load...but it makes the "best case scenario" slightly slower. but wasting resources while you have a lot to spare is not *that* bad of a performance property
03:34 <hexkey[m]> Makes sense. But I still can't replicate the results in the example.
03:34 <asonge> what results, exactly?
03:35 <patientplatypus> hi
03:35 <patientplatypus> im getting the following error
03:35 <asonge> hexkey[m]: if you can gist/pastebin something i could look at that you are testing, i'll see if i can reproduce/help
03:36 <patientplatypus> could not compile dependency :phoenix_ecto, "mix compile" failed. You can recompile this dependency with "mix deps.compile phoenix_ecto", update it with "mix deps.update phoenix_ecto" or clean it with "mix deps.clean phoenix_ecto"
03:36 wpcarro joined
03:36 <hexkey[m]> Sure, in a sec. in the stackoverflow page I linked, the example benchmark is for (fn -> 123456*654321 end).
03:36 <patientplatypus> im not sure why
03:36 <hexkey[m]> My results have consistently been zero.
03:36 <patientplatypus> i added the comeonin: dependency and it seemed like it went in fine enough
03:37 jkreeftmeijer joined
03:38 <asonge> hexkey[m]: i'm getting 3.0e-6 here, but the return value of :timer.tc is in microseconds (1 1millionth), and if you're formatting that badly, it'll round to 0.
03:38 <asonge> hexkey[m]: try just leaving off the division and using the microseconds?
03:39 <hexkey[m]> I have it in milliseconds right now, I'll try micro
03:39 <hexkey[m]> cls
03:39 <hexkey[m]> whups wrong place
03:40 <asonge> patientplatypus: mix deps.compile phoenix_ecto didn't work?
03:40 <patientplatypus> no i get the same error
03:40 <asonge> is there any other output that you can pastebin?
03:41 <patientplatypus> would you mimd taking a look at a repo?
03:41 <hexkey[m]> I'm still getting zero.
03:41 <asonge> patientplatypus: i don't mind, no
03:42 <asonge> hexkey[m]: what OS are you on? this might be a thing where you're on an OS that doesn't provide an accurate timing facility.
03:42 <asonge> (for some reason)
03:42 <hexkey[m]> Win10 is what I'm on at the moment
03:42 <* asonge> pulls up the erlang module docs to remind himself where to get this info...
03:44 <patientplatypus> here asonge https://github.com/patientplatypus/phoenixdepsnocompile
03:44 <asonge> hexkey[m]: can you paste what the result of :erlang.system_time() is?
03:45 <patientplatypus> i was supposed to get a return of an error if the password was more than 20 characters but that didnt seem to work either
03:46 <hexkey[m]> In iex or my project? also, sorry for unintentionally double-teaming you
03:46 <asonge> hexkey[m]: iex
03:46 <asonge> (either is fine)
03:46 <hexkey[m]> > 1529987894000640
03:47 <hexkey[m]> also, just found this on an erlang mailing list:
03:47 jkreeftmeijer joined
03:47 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-07_03:47:46.txt - https://matrix.org/_matrix/media/v1/download/matrix.org/ZdZbrPBvdaibViVccdYHwAid
03:47 <asonge> hrm, seems to be fast enough.
03:48 <hexkey[m]> http://erlang.org/pipermail/erlang-questions/2011-May/058908.html
03:48 <asonge> oh, that's definitely changed since OTP 19.0
03:48 <hexkey[m]> ok yeah just noticed the timestamp on the email
03:48 <asonge> like, last year the native erlang time stuff changed a *lot*
03:48 <asonge> 2 years ago, and that would've still been right :)
03:49 <asonge> the fact that you get 0 is weird
03:49 <asonge> i wonder if windows is using a timestamp caching trick somewhere
03:49 <asonge> for performance reasons
03:50 <hexkey[m]> I'm going to try just storing start/stop values from system_time
03:50 <hexkey[m]> see what happens
03:50 <asonge> hexkey[m]: actually, 0 should be impossible. this is some kind of bug, i think.
03:51 <asonge> i just looked up the source of :timer.tc and it uses :erlang.monotonic_time/0
03:51 <asonge> oh, you know what. it's then converting that.
03:51 <hexkey[m]> rounding?
03:51 <asonge> hexkey[m]: https://github.com/erlang/otp/blob/master/lib/stdlib/src/timer.erl#L168
03:52 <hexkey[m]> hmmm...
03:52 <asonge> so if it's 0, it's less than a microsecond.
03:54 <hexkey[m]> so it might work if I call monotonic_time() myself? that's what it's sounding like
03:54 <asonge> oh yeah, you can easily convert that tc to elixir
03:55 <patientplatypus> hi asonge
03:56 <asonge> oh wait, i just got the compile error
03:56 <patientplatypus> ah
03:57 <asonge> well, phoenix_ecto, phoenix, and comeonin are somewhat out of date
03:57 <patientplatypus> hmmm
03:57 <asonge> i wonder if this might be some kind of version mismatch but of some kind
03:57 <patientplatypus> how do i update all my deps
03:57 <patientplatypus> ?
03:58 <asonge> run `mix hex.outdated` to see
03:58 <patientplatypus> is this possible
03:58 jkreeftmeijer joined
03:58 <asonge> patientplatypus: oh, you need to bump the version requirements in mix.exs
03:58 <patientplatypus> ah? how do i do that
03:59 <asonge> yeah, some of these APIs are declared compatible that aren't.
03:59 <asonge> patientplatypus: look in deps() in mix.exs
04:00 WhiskyRyan joined
04:00 <asonge> they have the ~> stuff, bump those numbers up.
04:00 <asonge> ~> means 2.x, but not 3.x, for example
04:00 <asonge> the hex.outdated stuff will tell you what's farther out of date
04:01 <hexkey[m]> brb, trying this on by Xubuntu VM
04:02 <asonge> this stuff should run on windows or bash for windows just fine.
04:03 <patientplatypus> im sorry im still getting the same error although that was a good catchc
04:03 <asonge> well, after you bump the numbers, you have to deps.update and the mix.lock will change with the new versions
04:03 <asonge> phoenix_ecto isn't compatible with the version of ecto that's installed.
04:04 <asonge> ecto 1.x is really really old
04:04 r00tl3ss joined
04:05 <asonge> phoenix_ecto is trying to use a field that's no longer in ecto.
04:05 <asonge> which is throwing a compile error
04:07 Shados joined
04:07 mika_ joined
04:08 <patientplatypus> um
04:08 <patientplatypus> how do i install the latest version of ecto then
04:08 <patientplatypus> ?
04:08 <patientplatypus> i thought i had it
04:08 jkreeftmeijer joined
04:08 <patientplatypus> ive cleaned out all dependencies and run
04:09 <patientplatypus> mix deps.unlock --all
04:09 <patientplatypus> mix deps.get all
04:10 <patientplatypus> *mix deps.get --all
04:10 <patientplatypus> sorry
04:10 <patientplatypus> it was 1) mix deps.unlock --all
04:10 <patientplatypus> 2) mix deps.clean --all
04:10 <patientplatypus> 2) mix deps.get --all
04:10 <patientplatypus> and that didnt fix the compile error
04:17 <asonge> when you run mix hex.outdated, what pops up?
04:18 <asonge> is it the same kind of thing? i'm sure the mix.lock stuff changed.
04:19 jkreeftmeijer joined
04:19 <patientplatypus> hold on one sec
04:25 <patientplatypus> thanks that fixed the problem on to the next error ;_;
04:29 jkreeftmeijer joined
04:29 ssswitch joined
04:35 jkreeftmeijer joined
04:42 Ioyrie joined
04:45 jkreeftmeijer joined
04:49 <asonge> np. now all the documentation should make more sense, btw.
04:54 gvaughn joined
04:56 jkreeftmeijer joined
04:59 wpcarro joined
05:07 jkreeftmeijer joined
05:12 wpcarro joined
05:14 gvaughn joined
05:17 jkreeftmeijer joined
05:26 yourname joined
05:28 jkreeftmeijer joined
05:28 <hexkey[m]> oh. my. god.
05:28 <hexkey[m]> I was right.
05:29 <asonge> hexkey[m]: what was it?
05:30 <hexkey[m]> asonge[m]: Windows has been fucking my results up the whole time.
05:30 <asonge> been doing weird time things?
05:30 <asonge> time isn't real anyway
05:30 <* asonge> waves hand
05:31 WhiskyRyan joined
05:31 <hexkey[m]> one sec
05:34 <patientplatypus> step 1
05:34 <patientplatypus> uninstall windows
05:34 <patientplatypus> step 2
05:34 <patientplatypus> DO ANYTHING ELSE
05:35 <* hexkey[m]> uploaded an image: 2017-05-07 00_32_20-.png (6KB) - https://matrix.org/_matrix/media/v1/download/matrix.org/qnPEHboBXpekWJJUPrbtceTN
05:35 <hexkey[m]> The carefull eye may notice a difference
05:36 <asonge> hexkey[m]: were you getting 0 from subtracting the raw calls to :erlang.monotonic_time()?
05:37 wsieroci joined
05:37 <hexkey[m]> let's see how pasting this goes:
05:38 <* hexkey[m]> sent a long message: hexkey[m]_2017-05-07_05:38:01.txt - https://matrix.org/_matrix/media/v1/download/matrix.org/nFCQFkAGneIooRkQYXGxruVb
05:38 <hexkey[m]> close enough
05:38 jkreeftmeijer joined
05:38 <asonge> *that* returns 0?
05:38 <hexkey[m]> on Windows.
05:38 <asonge> this is some kind of bug, if i'm not mistaken.
05:39 <asonge> monotonic_time() is never supposed to return the same value twice as a safety guaratee, iirc
05:39 <hexkey[m]> At least I have one OS that knows what time is
05:40 <asonge> the word monotonic means that even if the OS provides the same time, erlang is supposed to increase it by 1.
05:40 <asonge> no matter what.
05:41 <asonge> can you give me the output of :erlang.system_info(:os_monotonic_time_source)?
05:41 <hexkey[m]> I'm gonna see what the other impl w/ :timer gets
05:41 wpcarro joined
05:42 <asonge> oh wait, that's only for "strictly monotonic"
05:42 <asonge> okay, it *can* return the same value
05:42 <hexkey[m]> (on my linux box this time?)
05:42 <asonge> on windows
05:42 <asonge> i'm just curious as hell now
05:43 <hexkey[m]> > [function: :GetTickCount64, resolution: 100, extended: :no, parallel: :yes,
05:43 <hexkey[m]> time: 24675375104]
05:43 <asonge> ahhhhhhhhh
05:43 <asonge> you have a very low resolution clock.
05:43 <hexkey[m]> :/
05:44 nhooyr joined
05:44 WhiskyRyan joined
05:44 <hexkey[m]> ...and my VM has a high enough one. Neat.
05:45 <hexkey[m]> I had suspected a hardware thing, really
05:45 <asonge> obviously your vm is faking it.
05:45 <asonge> (time is always fake though)
05:46 <asonge> yeah, i'm reading this right. the resolution on this clock source is 10ms. there is probably a higher resolution clock source for windows, but erlang's using this one.
05:46 <hexkey[m]> hm.
05:46 <hexkey[m]> trying it in the VM just cuz...
05:46 <asonge> erlang only guarantees timer accuracy to like, a few milliseconds, so if this function is cheapest, that might be best.
05:46 <hexkey[m]> ...aaaand...
05:47 <hexkey[m]> ...damn son.
05:47 <hexkey[m]> > [function: :clock_gettime, clock_id: :CLOCK_MONOTONIC, resolution: 1000000000,
05:47 <hexkey[m]> extended: :no, parallel: :yes, time: 6444348116872]
05:47 <asonge> yeah, os x has an extra 0.
05:48 <asonge> ew, high resolution timers on windows are CPU-dependent. "it could be the number of clock cycles" ooph.
05:48 r00tl3ss joined
05:48 <asonge> yeah, i wouldn't want to figure out how to code a high resolution timer on windows.
05:49 <asonge> on linux, it'd just be a system call.
05:49 <hexkey[m]> also: any reason the result list has more items on the VM?
05:49 jkreeftmeijer joined
05:49 <asonge> slight difference in erlang version?
05:50 refriedchicken joined
05:51 <asonge> hexkey[m]: ah, from the manual: "{clock_id, ClockId} This tuple only exists if Function can be used with different clocks. ClockId corresponds to the clock identifier used when calling Function."
05:51 <hexkey[m]> windows isn't telling me the clock id
05:51 <hexkey[m]> which I kind of would like to know
05:51 <hexkey[m]> This is both of them side by side:
05:51 <hexkey[m]> [function: :clock_gettime, clock_id: :CLOCK_MONOTONIC, resolution: 1000000000, extended: :no, parallel: :yes, time: 6444348116872]
05:51 <hexkey[m]> [function: :GetTickCount64, resolution: 100, extended: :no, parallel: :yes, time: 24675375104]
05:51 <hexkey[m]> one sec
05:51 <asonge> so, there's nothing else to "choose" as the clock for the clock function
05:52 <hexkey[m]> spose I'm stuck with it on Windows then?
05:53 <asonge> contribute a patch that uses a high resolution timer?
05:53 justelex joined
05:53 <hexkey[m]> If I can figure out how to, sure
05:54 <hexkey[m]> also I'm just now discovering that erl -v doesnt seem to work on windows
05:54 <hexkey[m]> it brings up eshell right away
05:55 <asonge> windows stuff isn't a high priority for most of the core erlang contributors, but beam development is increasing. erlang's getting more popular, along with elixir growing beam usage by others. many patches have come from the elixir community now
05:55 <hexkey[m]> wait
05:55 <hexkey[m]> elixir -v does work
05:55 <asonge> the erl command on windows is kinda...weird. elixir stuff mostly works the same, but the elixir/iex/etc commands are all wrapper shell scripts, and so is the erl command
05:56 <hexkey[m]> Found a difference between the installs.
05:56 <hexkey[m]> xubuntu:
05:56 <hexkey[m]> > Erlang/OTP 19 [erts-8.3] [source-d5c06c6] [64-bit] [async-threads:10] [hipe] [kernel-poll:false]
05:56 <asonge> it's not the installs, though. i found it in the manual. that tuple only appears if you can select another clock for that function on that OS.
05:57 <hexkey[m]> Right. but I found one interesting other difference.
05:57 cschneid_ joined
05:58 hbraun joined
05:58 <hexkey[m]> (and here's the windows one):
05:58 <hexkey[m]> > Erlang/OTP 19 [erts-8.2] [64-bit] [smp:8:8] [async-threads:10]
05:59 <hexkey[m]> Windows is under-hipeing itself!
05:59 <asonge> lol
05:59 <hexkey[m]> (there's a correct version of that joke in there somewhere)
05:59 jkreeftmeijer joined
06:00 <hexkey[m]> I'd enable it on windows (I think I've done it before), but w/o the extra time percision, It's not going to help me rn
06:01 griffinbyatt joined
06:02 <hexkey[m]> Would you suspect that a baremetal linux install would have any different resolution?
06:03 <hexkey[m]> (In my case it's be Xubuntu for both, if that matters)
06:03 DarkArcher207 joined
06:08 wpcarro joined
06:10 jkreeftmeijer joined
06:10 <hexkey[m]> I'd assume not if it's hardware only, but still..
06:10 <asonge> hexkey[m]: it's entirely possible that the vm uses a high resolution timer in order to emulate the hardware that linux needs to talk to
06:11 gvaughn joined
06:11 <asonge> it's just a lot more bookkeeping to do.
06:12 <hexkey[m]> I just tried the same thing with Process.sleep(333) and got 334263952.
06:13 <hexkey[m]> and on windows: 351232
06:14 <asonge> msdn said that accuracy of the getticks function is 10-16ms.
06:16 <hexkey[m]> any way I can specify the unit? (:msec) aint it
06:16 <hexkey[m]> think i got it
06:17 <hexkey[m]> (:millisecond)
06:17 <hexkey[m]> much better
06:17 <hexkey[m]> now I can be platform-agnostic!
06:18 <hexkey[m]> It'll still be fundementally broken on Windows, but hey, Interoperabillity! :DDDD
06:20 jkreeftmeijer joined
06:22 <asonge> hexkey[m]: you know how to read erlang docs?
06:23 <asonge> http://erlang.org/doc/apps/erts/time_correction.html is the time stuff, and the erlang module itself has most of the raw functions
06:24 <hexkey[m]> I remember some things about Time Warp from Learn You Some Erlang
06:25 <asonge> ah, well, there you go.
06:26 <hexkey[m]> It's a good thing I don't have to write this on windows.
06:26 <hexkey[m]> Also did you say earlier that you thought the VM was faking the accuracy?
06:27 jkreeftmeijer joined
06:32 <asonge> it isn't faking the accuracy
06:33 <asonge> it's keeping track in a giant number, but that number is getting recalculated with very little accuracy
06:33 <hexkey[m]> Good.
06:33 <asonge> precision and accuracy being different and all
06:34 wpcarro joined
06:34 <hexkey[m]> that'd be the diff in resolution?
06:35 <hexkey[m]> and one billion being different from ten.
06:35 <asonge> pretty much
06:35 <asonge> it was 100, iirc
06:35 <asonge> .1 is tenths, .01 is hundredths, .001 is thousandths (milliseconds)
06:37 jkreeftmeijer joined
06:38 <hexkey[m]> Just tried checking, and:
06:38 <hexkey[m]> > iex(16)> :erlang.monotonic_time(:microsecond)
06:38 <hexkey[m]> -576460670582752
06:39 <hexkey[m]> that's a new one
06:39 <hexkey[m]> (on the linux vm)
06:40 <hexkey[m]> huh. :os.system_time counts up
06:40 <hexkey[m]> > iex(25)> :os.system_time
06:40 <hexkey[m]> 1494139184350159376
06:40 borgtu joined
06:40 <hexkey[m]> something to sort out later.
06:41 <hexkey[m]> for now I need to either sleep or faint, whichever comes first.
06:41 <asonge> it's kinda described in the docs. but basically, the value of monotonic time is just an internal detail, all you know is that you get an integer.
06:42 <hexkey[m]> Choosing an negative one is still a weird decision
06:44 <hexkey[m]> I probably gotta sleep for now. Thanks for your help!
06:46 wpcarro joined
06:48 jkreeftmeijer joined
06:54 wpcarro joined
06:58 jkreeftmeijer joined
07:01 refriedchicken joined
07:05 griffinbyatt joined
07:06 elgenie joined
07:07 gvaughn joined
07:09 jkreeftmeijer joined
07:20 jkreeftmeijer joined
07:23 tuacker joined
07:25 josevalim joined
07:25 wpcarro joined
07:30 jkreeftmeijer joined
07:41 jkreeftmeijer joined
07:51 r00tl3ss joined
07:51 jkreeftmeijer joined
07:57 jkreeftmeijer joined
08:03 gvaughn joined
08:05 r00tl3ss joined
08:06 griffinbyatt joined
08:07 jkreeftmeijer joined
08:18 jkreeftmeijer joined
08:18 junsuijin joined
08:23 Ioyrie joined
08:28 jkreeftmeijer joined
08:34 jkreeftmeijer joined
08:39 icanhazbroccoli joined
08:39 gvaughn joined
08:43 jkreeftmeijer joined
08:46 superj3y joined
08:49 travis-ci joined
08:49 <travis-ci> elixir-lang/elixir#14426 (master - 5ab8c6e : José Valim): The build was broken.
08:49 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/f6f4bad18ae4...5ab8c6e93ad6
08:49 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/229622600
08:49 travis-ci left
08:50 PaReeOhNos joined
08:54 jkreeftmeijer joined
08:57 marr joined
09:00 gvaughn joined
09:01 gmcabrita joined
09:02 squall joined
09:05 jkreeftmeijer joined
09:15 jkreeftmeijer joined
09:17 asabil joined
09:21 jkreeftmeijer joined
09:25 travis-ci joined
09:25 <travis-ci> elixir-lang/elixir#14429 (master - ed58508 : José Valim): The build is still failing.
09:25 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/14e69109696b...ed58508dfef5
09:25 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/229628985
09:25 travis-ci left
09:28 Siyo joined
09:28 Siyo joined
09:30 r00tl3ss joined
09:32 jkreeftmeijer joined
09:35 lexmag joined
09:35 asabil joined
09:39 asabil joined
09:39 wpcarro joined
09:42 jkreeftmeijer joined
09:47 ejpcmac joined
09:48 Ioyrie joined
09:52 jkreeftmeijer joined
09:53 stephen_m joined
09:55 asabil joined
09:56 gvaughn joined
10:02 elgenie joined
10:04 jkreeftmeijer joined
10:04 junsuijin joined
10:07 griffinbyatt joined
10:08 asabil joined
10:11 PaReeOhNos joined
10:13 wsieroci joined
10:14 PaReeOhNos joined
10:15 jkreeftmeijer joined
10:15 asabil joined
10:19 asabil joined
10:25 jkreeftmeijer joined
10:26 nbg joined
10:31 jkreeftmeijer joined
10:36 r00tl3ss joined
10:43 dch_ joined
10:44 wsieroci joined
10:45 mika_ joined
10:47 Guest67 joined
10:50 iamvery joined
10:52 gvaughn joined
10:55 r00tl3ss joined
10:59 iamvery joined
11:00 squallstter joined
11:03 eddd joined
11:08 griffinbyatt joined
11:12 josevalim joined
11:21 pbj joined
11:24 r00tl3ss joined
11:24 wsieroci joined
11:31 treaki joined
11:33 tuacker1 joined
11:34 tuacker1 joined
11:38 r00tl3ss joined
11:48 gvaughn joined
11:51 pbj joined
11:51 wpcarro joined
11:54 elgenie joined
11:59 justelex joined
12:00 eddd joined
12:05 PaReeOhNos joined
12:09 griffinbyatt joined
12:17 tuacker joined
12:29 wsieroci joined
12:34 Guest67 joined
12:38 ZippoWeb joined
12:41 pbj joined
12:44 gvaughn joined
12:47 r00tl3ss joined
12:49 yourname joined
12:55 elgenie joined
12:55 yourname joined
12:57 dimitarvp joined
12:57 jkreeftmeijer joined
12:58 markw_ joined
13:00 <dimitarvp> hexkey[m]: did you manage to fix your problem with timing a function?
13:00 <dimitarvp> if not, could you provide a gist?
13:00 jimboy joined
13:01 eddd joined
13:10 griffinbyatt joined
13:12 pbj joined
13:20 DarkArcher207 joined
13:24 jkreeftmeijer joined
13:30 <Ankhers> Does anyone know how to guarantee a port is closed after your program finishes running? https://hexdocs.pm/elixir/Port.html#module-zombie-processes doesn't seem to work for me.
13:34 jkreeftmeijer joined
13:44 jkreeftmeijer joined
13:48 jkreeftmeijer joined
13:52 jkreeftmeijer joined
13:53 pbj joined
13:55 lexmag_ joined
13:58 tuacker1 joined
13:59 nbg joined
14:02 eddd joined
14:02 jkreeftmeijer joined
14:03 wpcarro joined
14:14 asabil joined
14:16 jkreeftmeijer joined
14:18 gvaughn joined
14:20 r00tl3ss joined
14:21 jkreeftmeijer joined
14:26 asabil joined
14:27 jkreeftmeijer joined
14:29 <benwilson512> @Ankhers it's up to the other program to terminate itself
14:29 <benwilson512> there's nothing the BEAM can do to force that
14:29 wsieroci_ joined
14:30 PaReeOhNos joined
14:30 leons joined
14:33 jhack joined
14:35 Guest67 joined
14:35 <Ankhers> That's fair. Would you by chance know of any reasons why the zombie processes part didn't help?
14:35 <Ankhers> benwilson512: ^^
14:36 bitmod joined
14:38 wsieroci joined
14:41 <benwilson512> Ankhers: well, what exactly are you doing from that section?
14:41 <benwilson512> you've got a wrapper script?
14:42 <Ankhers> Yes. I'm calling the wrapper script, I can see that it is called using ps. When my tests finish though, the wrapper and my program are still running.
14:43 <benwilson512> how are you spawning it? same as is written?
14:43 <Ankhers> Let me quickly gist what I have. But yes.
14:44 squallstter joined
14:44 <Ankhers> https://gist.github.com/ankhers/3717943dcb439cdea597d87c44b13ad5 -- I tried with and without the spawn_link/1. I was told to look at https://github.com/hexpm/hex/blob/master/test/support/hexpm.ex#L56-L99 for how to do something like this.
14:46 cemilowski joined
14:46 <benwilson512> hm so
14:46 <benwilson512> the first thing I would do
14:46 <benwilson512> is in your wrapper script add something after the `while` loop that indicates that it knows the stdin / stdout is closed
14:46 <benwilson512> so for example
14:48 <benwilson512> https://www.irccloud.com/pastebin/j3K7zeZO/
14:48 <benwilson512> note there on line 7
14:48 <benwilson512> so after you run your elixir stuff
14:48 <benwilson512> see if you have a ~/foo file
14:48 <benwilson512> if you don't, for some reason the script thinks that stdin / stdout is still alive
14:49 <benwilson512> from there
14:49 <benwilson512> I'd do a similar echo after line 2
14:49 <benwilson512> to make sure you actually START the while loop, maybe something about how the other executable is started is actually blocking
14:49 <benwilson512> I'm not really an sh expert
14:51 fschuindt joined
14:51 PaReeOhNos joined
14:53 <Ankhers> benwilson512: Thanks. I'll do that.
14:56 elgenie joined
15:01 jimboy joined
15:03 pbj joined
15:03 eddd joined
15:07 meh` joined
15:08 wpcarro joined
15:12 gvaughn joined
15:13 ssswitch joined
15:13 gvaughn_ joined
15:18 wsieroci joined
15:23 lexmag joined
15:31 pbj joined
15:33 asabil joined
15:35 wpcarro joined
15:36 PaReeOhNos joined
15:40 gvaughn joined
15:41 kochev joined
15:48 PaReeOhNos joined
15:52 patientplatypus joined
15:52 <patientplatypus> hi
15:53 <markw_> Hi patientplatypus
15:53 <patientplatypus> so....
15:53 <patientplatypus> im getting an error -
15:54 <patientplatypus> in a phoenix tutorial im working on
15:54 <patientplatypus> ERROR 42703 (undefined_column): column v0.url does not exist
15:54 <patientplatypus> in this def
15:54 <patientplatypus> def index(conn, _params) do
15:54 <patientplatypus> 7 videos = Repo.all(Video)
15:54 <patientplatypus> 8 render(conn, "index.html", videos: videos)
15:54 <patientplatypus> 9 end
15:55 <patientplatypus> it highlights line 7
15:55 <Nicd-> have you run all migrations?
15:55 <patientplatypus> (this lives in controllers/video_controller.ex)
15:55 <Nicd-> it seems the column url was not created into the database
15:55 <patientplatypus> whats the command for outputting whats currently living in the psql database?
15:56 <patientplatypus> also, i am not sure why it would not be getting created, so that's annoyinng
15:56 <Nicd-> you can check with psql if you are familiar with it
15:57 <Nicd-> \d tablename
15:57 WhiskyRyan joined
15:57 <Nicd-> or use something like pgadmin
15:58 asabil joined
15:58 <patientplatypus> whats the terminal command? psql \d videos?
15:58 <patientplatypus> *'psql \d videos'?
15:58 <Nicd-> no
15:59 <Nicd-> psql -W dbname username
15:59 <Nicd-> then inside psql \d tablename
16:04 <patientplatypus> would someone be willing to look at a repo?
16:05 <patientplatypus> im still getting a broken psql and i just cant figure it out
16:05 <patientplatypus> https://github.com/patientplatypus/tableissuephoenix
16:06 <Nicd-> did you check that the column exists?
16:06 <patientplatypus> it doesnt but it should
16:07 <patientplatypus> im running this mix phoenix.gen.html Video videos user_id:references:users \
16:07 <patientplatypus> url:string title:string description:text
16:07 <patientplatypus> and it says that url doesnt exist
16:07 <patientplatypus> straight copy/paste from the programming phoenix book
16:07 rkazak joined
16:07 <patientplatypus> so....confused
16:07 <patientplatypus> also after i did ecto.migrate
16:07 <Nicd-> and you ran that migration and not a previous version of the same migration?
16:08 <patientplatypus> ive been trying to make it work again, so ive been running it over again after minor changes, HOWEVER, I do delete the migration file and repopulate when i run
16:09 wsieroci joined
16:09 wpcarro joined
16:09 <Nicd-> you clear the whole db?
16:10 <patientplatypus> i just delete the migrations file in my file tree - how do i clear the whole db? im unfamiliar with that command
16:10 <patientplatypus> i would think it would just overwrite it anyway
16:10 <Nicd-> it works for me
16:10 <Nicd-> it won't overwrite the b
16:10 <Nicd-> db*
16:11 <patientplatypus> how do i clear the db then?
16:11 <Nicd-> the db will contain a migrations table and will contain info that it has already run the migrations
16:11 <patientplatypus> like whats the psql command for that
16:12 <Nicd-> it's easier to drop the whole database
16:12 <patientplatypus> ok so mix ecto.drop
16:12 <patientplatypus> one sec
16:12 <Nicd-> but you can do this also in psql: `drop schema public cascade; create schema public;`
16:13 meandi_2 joined
16:16 <bitmod> Anyone here deployed an elixir/phoenix app to digital ocean (using docker)?
16:23 <patientplatypus> thanks for the help Nicd I got it up again
16:23 WhiskyRyan joined
16:24 wpcarro joined
16:26 cemilowski joined
16:28 jesseshieh joined
16:33 wsieroci joined
16:34 refriedchicken joined
16:42 rodolfojcj joined
16:48 Uniaika joined
16:50 pbj joined
16:52 nishithshah2211 joined
16:54 Uniaika joined
16:58 elgenie joined
16:58 mika_ joined
17:01 wpcarro joined
17:01 nhooyr joined
17:05 trias joined
17:05 eddd joined
17:07 __0xAX joined
17:17 miwa_ joined
17:20 wpcarro joined
17:29 wpcarro joined
17:30 pbj joined
17:31 NeverDie joined
17:33 craigp joined
17:34 <Nicd-> hey Uniaika, did you watch hockey tonight? :D
17:35 <Uniaika> Nicd-: check your queries
17:37 Guest67 joined
17:49 southpawlar joined
17:54 sunaku joined
17:56 WhiskyRyan joined
17:58 <bitmod> Nicd-, does this docker image work? https://hub.docker.com/r/marcelocg/phoenix/
17:59 <Nicd-> I don't know, I don't use docker
17:59 <bitmod> ok
17:59 elgenie joined
18:00 mbwe joined
18:02 Tidur joined
18:06 eddd joined
18:10 wpcarro joined
18:16 tomaz_b joined
18:17 icanhazbroccoli joined
18:21 justelex joined
18:23 <asonge> bitmod: define "work"
18:23 <asonge> bitmod: bitwalker's docker images are pretty solid in general, just btw.
18:24 <asonge> that docker file looks like a dev-focused docker file. a production one wouldn't have the phoenix mix archive in it or anything like that
18:28 <bitmod> asonge: this is what i get: https://paste.ubuntu.com/24532115/
18:28 <bitmod> asonge: ah so does that command create a working application?
18:30 <asonge> bitmod: are you trying to have a docker-based dev environment or a docker-based deploy? because those are 2 entirely different kinds of things, and what you're running there is creating a new app.
18:31 <bitmod> asonge: basically i want to start a phoenix app (new) on my DO droplet
18:32 <asonge> bitmod: so, usually, you take an existing app that's outside of docker and you run it in a docker container to dev on *or* make an erlang release (via distillery) and put it in a docker container to deploy it to production
18:33 <asonge> you can make the release itself inside of a docker container you make for that task, but like...your use case is just weird.
18:33 <asonge> particularly since your command auto-deletes the phoenix app when the application ends, and it's also not running in the background.
18:33 <bitmod> asonge: is there a post that goes over deploying a phoenix app to production?
18:34 wpcarro joined
18:34 <asonge> bitmod: a few, and i've used docker to push to production as well as other things
18:35 <bitmod> asonge: also as an aside, what do you think of https://nanobox.io/? Do you think its benefits outway the costs?
18:37 <asonge> dunno, some of that "depends". deploying to production is kinda in a weird space where it's so trivially easy to just shell script it in most cases, that no one's made a strategy that works for a super-wide range of use-cases (if that makes sense)
18:38 <asonge> the reason i like to use docker is that docker is probably the best way to have deployable artifacts that can run on any provider.
18:39 <asonge> bitmod: i will say that erlang/elixir deal a lot better with load and are a lot more efficient, and so auto-scaling is not quite as common as just getting a bigger server.
18:39 apotry joined
18:41 wpcarro joined
18:43 <bitmod> asonge: maybe i'll stick with docker then
18:43 squallstter joined
18:44 <bitmod> asonge: also would you suggest this (http://www.phoenixframework.org/v0.6.2/docs/deployment) as a starting point for deploying a new phoenix app to DO?
18:44 <bitmod> i just want to get used to the process
18:44 <bitmod> right now it's all very abstract and nebulous, not really sure where i stand
18:45 <asonge> bitmod: i'd use the newest version of those docs if anything.
18:45 <bitmod> asonge: also do you have the url for a post on using phoenix with docker?
18:45 <asonge> bitmod: http://www.phoenixframework.org/v1.2.0/docs/deployment
18:45 Uniaika joined
18:45 <bitmod> asonge: ah thanks, didn't realize that was an old version
18:46 <asonge> bitmod: no url/post, but the hardest part is actually getting a release running. if you can run a release locally using distillery (which has instructions for setting up everything you need), the "docker" bits are very thin.
18:46 <asonge> does that make sense?
18:46 <asonge> and the author of distllery does maintain some docker images
18:46 srxa joined
18:47 NeverDie joined
18:47 <asonge> bitmod: here's one, https://hub.docker.com/r/bitwalker/alpine-elixir-phoenix/
18:48 <asonge> bitmod: see readme here: https://github.com/bitwalker/alpine-elixir
18:48 patientplatypus joined
18:48 <patientplatypus> hi
18:48 <bitmod> asonge: could you briefly described how docker comes into using distillary, just so i know im on the right path?
18:49 <patientplatypus> is there a good resource/tutorial out there on building an ember front end and a phoenix backend? ive looked around and most of the tutorials are out of date.
18:49 sunaku joined
18:50 <bitmod> asonge: oh yeah, i saw that image before. Is that a production or dev image?
18:50 <tristan__> if you build a target system from the release with erts included you just need a base docker image that can run that erts
18:50 <asonge> bitmod: basically, the path is that you mount some dir on your local filesystem to a docker container that you run to build the release. then you build a new docker image that copies that release file (docker actually auto-extracts tar.gz files with the copy and add commands)
18:51 wpcarro joined
18:51 <asonge> and yes, you should make sure that the erlang runtime system (the OTP versions) match really exactly if you don't build an "embedded" release (saves space not to)
18:51 <tristan__> meaning nothing special installed (except maybe openssl), just compatible libc. so if you want to run on alpine you have to build with muscl
18:52 <patientplatypus> bueller
18:52 wsieroci_ joined
18:52 <bitmod> tristan__: sorry what is muscl?
18:53 <bitmod> also is there a post on running phoenix on alpine?
18:54 <tristan__> sorry, musl. the C lib used by alpine
18:55 kochev joined
18:55 <asonge> bitmod: there's nothing special about phoenix, btw. you can just use the word "elixir" only about 80% of the time, unless it's in the phoenix API
18:55 <asonge> bitmod: but don't worry about running a release on alpine if you build it on another alpine image (if that makes sense)
18:57 <bitmod> asonge: er not exactly...
18:58 trias joined
18:58 <asonge> bitmod: so, you have a docker image that you have around, that its only job is (when you run it) to mount a directory, build your application and then store the release .tar.gz file in the mounted directory.
18:59 <asonge> when that happens, you have a .tar.gz file that is on your host system that can run in alpine
18:59 <asonge> then you run another docker command to copy the release and make an image with your app ready to go with 1 command
19:00 <asonge> that's what's in bitwalker's readme on bitwalker/alpine-elixir. the tar extraction part.
19:01 elgenie joined
19:02 <asonge> in a few months, there's going to be a new kind of dockerfile directive for multi-stage images that allows you to separate build and production steps, and it drops the build artifacts from the image (leaving just production stuff)
19:02 <asonge> but right now, you have to do this 2-step
19:03 <bitmod> asonge: ah ok, i think i've got it: 1) use distillery to compile(?) your elixir app to binaries (or .tar.gz), 2) unpack the application and run it
19:04 Uniaika joined
19:06 eddd joined
19:10 zv joined
19:12 <patientplatypus> hi
19:12 <patientplatypus> is there any particular reason someone would use ember in the front end and phoenix in the back as opposed to just doing the whole thing in phoenix?
19:13 <patientplatypus> and also are there any decent tutorials on ember/phoenix that are recent enough not to have errors on deprecated packages
19:13 <Nicd-> if they wanted to make an SPA?
19:13 <patientplatypus> phoenix doesnt do spa?
19:13 <Nicd-> Phoenix runs on the backend
19:13 wsieroci joined
19:13 <Nicd-> you need a frontend to do SPA
19:14 <patientplatypus> do you really?
19:14 <patientplatypus> like as a necessity?
19:14 <Nicd-> err
19:14 <Nicd-> you need JavaScript to run in the browser
19:15 <Nicd-> to do a single-page app
19:15 <Nicd-> you can render HTML in Phoenix and send it to the browser. that's just a normal website then
19:16 <patientplatypus> so i can have a subpart of the webpage that gets routed new information but its main layout doesnt change, so the whole page wouldnt reload right? or am i not thinking this through correctly
19:16 iamvery_ joined
19:16 <patientplatypus> so if i have a page that contains a messaging app and a weather app, and when new information comes in the router sends it to the messaging component the whole page wouldnt refresh no?
19:17 <patientplatypus> anyway
19:17 <patientplatypus> do you have a good resource for learning a stack of ember and phoenix together?
19:17 <patientplatypus> front/back
19:19 <Nicd-> yes that is the main advantage of single-page apps
19:19 <Nicd-> I don't know any. I would read how to write an Ember application that connects to a hypothetical backend. then connect it to Phoenix
19:19 <tristan__> and yes you need frontend logic for that
19:19 <Nicd-> the methods for connecting (AJAX and JSON) are the same anyway
19:26 josevalim joined
19:30 Ioyrie joined
19:34 gvaughn_ joined
19:36 wpcarro joined
19:36 eddd joined
19:37 <Lokathor> guys
19:37 <Lokathor> i just started reading about Elixir this week
19:38 <Lokathor> and I'm pretty impressed
19:38 <Uniaika> Lokathor: welcome aboard
19:38 <Lokathor> I used Erlang a while back, just enough to write a mini-IRC bot, but it had a supervisor and stuff
19:39 <Uniaika> :)
19:39 <Lokathor> Elixir feels like a fresh coat of paint on an industrial grade fridge that was just chipping on the outside
19:41 sepow joined
19:42 rodolfojcj joined
19:44 gvaughn_ joined
19:46 zvrk joined
19:48 northfurr joined
19:49 <ivan> I would really like to be able to put a @doc on a defp to avoid having to convert things from comments to @doc or the reverse
19:49 rodolfojcj joined
19:49 <ivan> especially given that defp in Elixir doesn't mean private so much as "not called from any other module or test"
19:49 <Uniaika> ivan: so… public documentation on a function that's not going to appear in the documentation?
19:49 <ivan> sure
19:50 <Uniaika> ex_doc is going to have a hard time understanding to which function it belongs if it's a private one
19:50 patientplatypus joined
19:57 ejpcmac left
19:59 PaReeOhNos joined
20:01 elgenie joined
20:18 gvaughn_ joined
20:20 <Lokathor> the IDE might still be able to just parse the file itself and show the doc in a tooltip, even if it's private, I guess
20:25 griffinbyatt joined
20:37 pbj joined
20:44 PaReeOhNos joined
20:53 srxa joined
20:58 <OliverMT> is there a syntax to bind a match in the middle of a map?
20:58 <OliverMT> %{data = data: %{"url" => url}} this doesnt work
20:59 <asonge> OliverMT: %{data: data=%{"url"=>url}}
20:59 <asonge> OliverMT: that syntax would try to match data to the "pair", which is weird.
20:59 <OliverMT> ah ofc :D
20:59 <OliverMT> thanks asonge
20:59 <OliverMT> do you code in vscode these days asonge ?
20:59 <OliverMT> its become pretty good
20:59 <asonge> yeah
21:00 <OliverMT> orders of magnitude faster than atom at least
21:00 <asonge> you had me switch for a day and i never looked back
21:00 <OliverMT> do you also have this annoying thing where the do-end filler gives you defoverride instead?
21:00 <OliverMT> it came in last update to vscode-elixir
21:00 <asonge> i actually patched the plugin so that if you use vscode to run mix build, the "problems" will actually link you off to the right line of code
21:00 northfurr joined
21:00 <asonge> oh yeah, the selection on that is bad.
21:00 <asonge> that needs to be fixed somehow
21:00 <OliverMT> it had the same thing in atom
21:01 <OliverMT> then got fixed, but I dont think intellisense orders stuff the same way
21:01 <OliverMT> or even expose it in the api
21:01 <OliverMT> I use mix build in a terminal at the bottom
21:01 <asonge> you still have to map cmd+b in the task runners, but you can now reference the task filters that i registered
21:01 <hexkey[m]> does vscode have a handy keyboard shortcut for pulling up docs?
21:02 <ciawal> vscode has elixir support?
21:03 akeating joined
21:03 <Uniaika> ciawal: not built-in but there's at least 1 plugin
21:04 <OliverMT> ciawal: vscode-elixir
21:04 akeating joined
21:04 <OliverMT> alchemist backed thing almost identical to atom
21:04 srxa joined
21:05 <asonge> hexkey[m]: here's the features, it kinda comes up the intellisense, but i don't know how to make it show up large or anything: https://github.com/fr1zle/vscode-elixir
21:06 <asonge> OliverMT: look at the new .vscode/tasks.json in the readme for the plugin, with that, you can just hit cmd+B
21:06 <asonge> to run the build
21:07 <ciawal> might try it out
21:10 <OliverMT> I am sorting a list of tuples {action, l} by doing
21:10 <OliverMT> |> Enum.sort(fn {_action, %{organization: %{id: oid1}, inserted_at: i1}}, {_action, %{organization: %{id: oid2}, inserted_at: i2}} -> {oid1, i1} < {oid2, i2} end)
21:10 <OliverMT> is there something inherently wrong there?
21:11 <OliverMT> i am getting a match error where it shows me two items that should clearly match that
21:11 <OliverMT> both are 2tuples, both have organization.id and inserted_at keys in the 2nd item
21:11 <asonge> and the second item is a map?
21:11 <asonge> oh, _action
21:11 <ciawal> wouldn't Enun.sort_by read nicer
21:11 <asonge> needs to be _action1, _action2
21:12 <OliverMT> ah, ofc it re-matches even if the variable is unbound
21:12 <asonge> yeah, it's not quite "unbound", it just silences warnings
21:12 <asonge> you can actually still use it >_>
21:12 <asonge> _ is truly unbound though
21:12 <asonge> you can reuse _
21:12 <OliverMT> ah ok
21:12 <OliverMT> I almost have a style guide disallowing _
21:13 <OliverMT> unless its incredibly trivial
21:13 <OliverMT> it costs so little to say what you're ignoring
21:13 <asonge> yeah, it doesn't cost anything, really
21:13 <OliverMT> I meant effort wise
21:13 <OliverMT> code wise I am expecting 100% elimination :p
21:13 <asonge> i meant vm-wise
21:14 <asonge> i think it might get cleaned up later in the compiler
21:14 <asonge> like, if you didn't use a variable, it'll emit a warning, and then clean up unused variables in a later stage of compilation
21:15 <asonge> but if you reuse _foo, it won't clean up _foo in that later stage
21:15 <asonge> this is nice if you're keeping a variable around for debugging reasons, but have some kind of macro that compiles out a debug statement
21:16 <asonge> if you built with "prod" on, you'd get the warning during a production build that it was unused, so you prefix the variable with _ anyway
21:17 eddd joined
21:18 zvrk joined
21:20 <asonge> hexkey[m]: if you hit ctrl+space multiple times in vscode intellisense, you toggle between docs and autocomplete menu
21:20 <* asonge> just figured that out
21:20 <hexkey[m]> nice. I'll give it a look sometime.
21:21 akeating joined
21:24 <patientplatypus> i have an ember.js question
21:25 <patientplatypus> in ember it appears that {{outlet}}, if it is at the end of the page does not render anything, but if it is at the top of the page it renders any js file that has an {{outlet}} call above it. Is this the case?
21:26 <Ankhers> patientplatypus: That may be better suited to a different channel.
21:26 tuacker joined
21:26 <patientplatypus> ah?
21:26 pbj joined
21:26 <ciawal> a channel about ember.js, for example
21:27 squallstter joined
21:27 <patientplatypus> currently that channel has one person in it and that is me
21:27 <Ankhers> This channel is specific to Elixir. While people here may know ember.js, you may get a better / faster response from a ember.js channel, or at the very least, a javascript one.
21:27 <patientplatypus> ah
21:27 akeating joined
21:27 <Ankhers> Did you go to @emberjs?
21:27 refriedchicken joined
21:27 <Ankhers> That is what their website suggests for IRC.
21:27 <Uniaika> more like #EmberJS, no?
21:27 <Ankhers> Doesn't matter.
21:28 <Ankhers> channel names are case insensitive I believe.
21:28 <patientplatypus> sorry thanks
21:28 <Uniaika> 23:28:05 [Users #@emberjs]
21:28 <Uniaika> 23:28:05 [@patientplatypus] [ Uniaika]
21:28 <Uniaika> Ankhers: sorry, I wasn't pointing out the case (^_^)
21:28 <Ankhers> Oh. Yes. I did mean #emberjs... Sorry for the stupidity :P
21:29 <Uniaika> no, it's not stupidity
21:29 <Uniaika> don't be sorry
21:31 asabil joined
21:32 akeating joined
21:46 hakunin joined
21:47 hakunin joined
21:48 griffinbyatt joined
21:50 PaReeOhNos joined
21:52 ssswitch joined
21:58 northfurr joined
22:03 elgenie joined
22:03 sunaku joined
22:06 wpcarro joined
22:08 patientplatypus joined
22:08 <patientplatypus> hi
22:09 <patientplatypus> im interested if anyone has a good resource on how to do a simple call to a third party api (get request ie ajax) in phoenix
22:09 southpawlar joined
22:11 Talltree_ joined
22:11 Quintasan joined
22:11 Quintasan joined
22:12 bambanx joined
22:12 sulphur28 joined
22:12 <patientplatypus> anyone here?
22:12 tonyc__ joined
22:12 <bambanx> i am me
22:13 dj_goku_ joined
22:13 <hexkey[m]> I am, but I'm busy trying to get something to work atm
22:15 refriedchicken joined
22:16 srxa_ joined
22:16 pbj joined
22:17 <asonge> patientplatypus: check hex.pm for an existing library for what you want to do, if not, there are libraries for doing http requests like httpoison or httpotion
22:18 <asonge> hackney is good, but is straight erlang and some elixir folks struggle with using erlang libs early on.
22:18 <OliverMT> HTTPoison is simple and easy
22:19 <OliverMT> its a thin thin wrapper around hackney
22:19 <OliverMT> with saner (easier at least) syntax
22:21 TinkerTyper_ joined
22:21 <patientplatypus> thanks --- do you guys have any phoenix examples for the slow kids in the class?
22:22 <patientplatypus> just looking for a get request, console response pattern to get me started
22:29 wpcarro joined
22:33 cschneid_ joined
22:35 <Radar> patientplatypus: I would suggest checking out httpotion or httpoison for doing that sort of thing
22:35 <Radar> I don't understand what you mean by "console response pattern"
22:42 leons joined
22:42 arpunk joined
22:42 HashNuke joined
22:42 Lex[m]1 joined
22:42 jaydoane joined
22:42 adulteratedjedi joined
22:42 johanw joined
22:42 spawnthink[m] joined
22:42 CARAM__ joined
22:45 CARAM__ joined
22:47 zealsham_ joined
22:48 cschneid_ joined
22:48 <zealsham_> is jose valim the same guy from ruby community?
22:48 nhooyr joined
22:49 <Radar> yes
22:50 <zealsham_> the f, how did he create such an elegant knowledge without former c.s background
22:50 <zealsham_> elegant language i meant
22:58 proteusguy joined
23:00 nhooyr joined
23:00 cschneid_ joined
23:01 PaReeOhNos joined
23:04 elgenie joined
23:12 strykerkkd joined
23:23 mika_ joined
23:23 <bitmod> anyone here?
23:24 <PeregrinePDX> There are people around.
23:24 elgenie joined
23:25 wpcarro joined
23:27 <bitmod> PeregrinePDX: ah great, just wondering why i might be getting this error: https://paste.ubuntu.com/24533573/
23:28 <bitmod> i'm trying to run a container on a DO droplet, but can't connect to postgres for some reason (it is installed)
23:29 <PeregrinePDX> Well, I don't know much elixir but I suspect it's one of 1) Your postgres isn't set to listen on port 5432 on localhost. 2) Your pg_hba.conf isn't set to allow whatever user you're trying to connect as to use a password to connect. Or 3) you didn't specify a user or password
23:31 <patientplatypus> hi
23:31 <patientplatypus> are there any good tutorials on how to get audio streaming working on phoenix?
23:35 srxa joined
23:35 wpcarro joined
23:42 dimitarvp` joined
23:42 <bitmod> PeregrinePDX: i don't think it's 1) because the default port is 5432, 2) i haven't altered, that should work too, could be 3), but my config/prod.secret.exs matches my postgres instance
23:45 <PeregrinePDX> can you connect to postgres using psql with the same user, password, host and port?
23:45 __charly__ joined
23:47 <bitmod> PeregrinePDX: i can connect with "sudo -i -u postgres"
23:48 <PeregrinePDX> Well that's the postgres user. Is that the user you're trying to use in your elixir app?
23:48 sunaku joined
23:48 <bitmod> PeregrinePDX: what's strange is that there's no password prompt, which i assume means there is no password, and in my config/prod.exs the password should be "postgres", so i tried changing the password for "postgres" user on postgresql to "postgres", but it made no difference
23:48 <bitmod> PeregrinePDX: yeah, except with the password "postgres"
23:49 <PeregrinePDX> Ok so you do sudo -i -u postgres.
23:49 <bitmod> tried changing it with \password and ALTER USER ..., not sure if either worked
23:49 <bitmod> PeregrinePDX: that gets me into postgres@machine
23:49 <PeregrinePDX> what is the command line you use to connect to postgres? psql ....
23:49 <bitmod> then i run psql and i get access to postgres shell
23:50 <bitmod> so i can run \d, create databases, etc
23:50 <bitmod> does this mean the password wasn't changed?
23:50 <PeregrinePDX> That connects using a unix domain socket. Not via tcp.
23:50 <PeregrinePDX> try psql -h localhost -p 5432 -U postgres -W
23:51 <PeregrinePDX> that says to connect using a hostname of localhost port 5432 with USer postgres and to prompt for a password
23:51 <bitmod> PeregrinePDX: ok that gives me a password prompt, "postgres" works and i can log in
23:52 <bitmod> so it's not 3) i guess?
23:52 <PeregrinePDX> Ok so that says it's not 1 or 2.
23:52 rkazak joined
23:52 <PeregrinePDX> Is this a phoenix app or is it just using postgrex directly?
23:53 <bitmod> PeregrinePDX: a phoenix app
23:53 <bitmod> PeregrinePDX: following this guide: http://teamon.eu/2017/deploying-phoenix-to-production-using-docker/
23:54 <bitmod> also i created a "test_prod" database manually, not sure if that's significant or not
23:55 <patientplatypus> hi
23:55 <bitmod> PeregrinePDX: and this is my config/prod.exs file: http://sprunge.us/bPYb
23:56 <patientplatypus> is there a good way to connect phoenix and ember that doesnt use the ja-serializer package? im looking for a good way to take advantage of phoenix's stateful streaming/socketing
23:56 <PeregrinePDX> What's in prod.secret.exs minus the password of course.
23:56 <patientplatypus> but pipe that to ember.js
23:56 <patientplatypus> and im having a fair amount of trouble finding resources
23:58 <patientplatypus> bueller
23:58 wpcarro joined
23:58 <bitmod> PeregrinePDX: config/prod.secret.exs -> http://sprunge.us/OiKg
23:59 <PeregrinePDX> I'm not positive about this but it looks like you're configuring for test in your production files and I think that might be the problem.