<     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
00:03 potatosalad joined
00:04 PaReeOhNos joined
00:04 potatosalad joined
00:05 potatosalad joined
00:06 potatosalad joined
00:07 potatosalad joined
00:08 rodolfojcj joined
00:08 dimitarvp` joined
00:11 m00dy joined
00:12 __charly__ joined
00:12 mrEngineer joined
00:22 Cohedrin joined
00:24 jschneck joined
00:32 cp__ joined
00:32 wpcarro joined
00:37 codestorm joined
00:41 bhahn joined
00:45 sillyotter joined
00:58 elgenie joined
00:59 kronicdeth joined
01:00 rozap joined
01:01 wpcarro joined
01:03 robbie joined
01:03 <robbie> What is the runtime for appending strings look like?
01:03 <robbie> What does*
01:04 <robbie> appending to a string* gosh.
01:04 <robbie> ie: "aaaa" <> "b"? o(n)?
01:05 justelex joined
01:07 codestorm joined
01:12 m00dy joined
01:12 PaReeOhNos joined
01:13 sp4rrow joined
01:14 dj_goku joined
01:14 dj_goku joined
01:14 srxa joined
01:17 cdg joined
01:17 wpcarro joined
01:26 cdg joined
01:27 <wwwd> *** ssswitch (~ssswitch@unaffiliated/ssswitch) has joined channel #elixir-lang
01:27 <wwwd> [18:23]
01:27 <wwwd> *** dj_goku (~dj_goku@unaffiliated/dj-goku/x-000000001) has joined channel
01:27 <wwwd> #elixir-lang [18:26]
01:27 <wwwd> *** gvaughn (~gvaughn@2605:6000:8d44:e300:7003:68e3:9e2f:c977) has quit: Ping
01:27 <wwwd> timeout: 240 seconds [18:30]
01:27 <wwwd> *** arquebus (~ed@ has joined channel #elixir-lang
01:27 <wwwd> *** dj_goku (~dj_goku@unaffiliated/dj-goku/x-000000001) has quit: Ping
01:28 <wwwd> timeout: 260 seconds [18:31]
01:28 <wwwd> *** gvaughn_ (~gvaughn@cpe-70-121-11-255.tx.res.rr.com) has joined channel
01:28 <wwwd> #elixir-lang
01:28 <wwwd> *** cdg (~cdg@ has quit: Remote host closed the connection
01:28 <wwwd> [18:32]
01:28 <wwwd> *** rodolfojcj (~rodolfojc@190-206-136-247.dyn.dsl.cantv.net) has quit: Ping
01:28 <wwwd> timeout: 260 seconds [18:37]
01:28 <wwwd> *** isubasti_ (~isubasti@ has joined channel #elixir-lang
01:28 <wwwd> [18:40]
01:28 <wwwd> *** codestorm (~codestorm@2605:e000:9196:e300:b5f9:d987:c130:13e1) has joined
01:28 <wwwd> channel #elixir-lang [18:43]
01:28 <wwwd> *** wpcarro (~wpcarro@cpe-66-108-244-101.nyc.res.rr.com) has joined channel
01:28 <wwwd> #elixir-lang
01:28 <wwwd> *** kronicdeth (~textual@cpe-70-121-97-113.austin.res.rr.com) has quit: Quit:
01:28 <wwwd> My MacBook has gone to sleep. ZZZzzz… [18:47]
01:28 <wwwd> *** codestorm (~codestorm@2605:e000:9196:e300:b5f9:d987:c130:13e1) has quit:
01:28 <wwwd> Ping timeout: 240 seconds
01:28 <wwwd> *** gvaughn_ (~gvaughn@cpe-70-121-11-255.tx.res.rr.com) has quit: Ping
01:28 <wwwd> timeout: 240 seconds [18:50]
01:28 <wwwd> *** gvaughn_ (~gvaughn@cpe-70-121-11-255.tx.res.rr.com) has joined channel
01:29 <wwwd> #elixir-lang [18:52]
01:29 <wwwd> *** wpcarro (~wpcarro@cpe-66-108-244-101.nyc.res.rr.com) has quit: Ping
01:29 <wwwd> timeout: 246 seconds [18:53]
01:29 <wwwd> *** cemilowski (~Thunderbi@unaffiliated/cemilowski) has quit: Ping timeout:
01:29 <wwwd> 240 seconds [18:56]
01:29 <wwwd> *** wpcarro (~wpcarro@cpe-66-108-244-101.nyc.res.rr.com) has joined channel
01:29 <wwwd> #elixir-lang [18:58]
01:29 <wwwd> *** potatosalad (~potatosal@184-155-112-206.cpe.cableone.net) has quit: Quit:
01:29 <wwwd> My MacBook has gone to sleep. ZZZzzz… [19:02]
01:29 <wwwd> *** wpcarro (~wpcarro@cpe-66-108-244-101.nyc.res.rr.com) has quit: Ping
01:29 <wwwd> timeout: 260 seconds [19:03]
01:29 <wwwd> *** potatosalad (~potatosal@184-155-112-206.cpe.cableone.net) has joined
01:29 <wwwd> channel #elixir-lang
01:29 <wwwd> *** chenghiz (~Adium@96-89-96-217-static.hfc.comcastbusiness.net) has quit:
01:29 <wwwd> Quit: Leaving.
01:29 <wwwd> *** potatosalad (~potatosal@184-155-112-206.cpe.cableone.net) has quit: Client
01:29 <wwwd> Quit [19:04]
01:29 <wwwd> *** PaReeOhNos (~pareeohno@cpc81197-farn9-2-0-cust129.6-2.cable.virginm.net)
01:30 wwwd joined
01:32 gvaughn_ joined
01:38 tomterl joined
01:44 jschneck joined
01:47 <benwilson512> oook
01:47 <benwilson512> wwwd: what happened?
01:49 <benwilson512> robbie: http://erlang.org/doc/efficiency_guide/binaryhandling.html
01:49 <benwilson512> it sorta depends
01:50 <benwilson512> there are some special optimizations around binaries
01:51 <robbie> I see. So there's some extra space allocated at the tail of binaries?
01:52 codestorm joined
01:52 <benwilson512> yeah they end up getting allocated a lot like vectors in other languages at times
01:53 <robbie> I was going to say it's reminiscent of how dynamic arrays are implemented.
01:54 sevenseacat joined
01:54 <robbie> My main reason for asking is because I'm debating whether or not to work directly with binaries or work with lists inside of a tokenizer. It very may well be faster to append characters the end of a binary opposed to prepending, reversing and converting a list to a string.
01:55 <benwilson512> ah
01:55 <benwilson512> well if the end result can be iodata
01:55 justelex joined
01:55 <benwilson512> you can do a few novel things as well
01:55 <benwilson512> instead of appending or prepending to a list, build an improper list
01:56 <benwilson512> new_list = [existing_list | new_item]
01:56 <benwilson512> then at the end you can `IO.iodata_to_binary`
01:56 <robbie> Yeah, I know of that trick.
01:56 jleon joined
01:56 <robbie> It's interesting. I need to do some profiling.
01:57 <benwilson512> at the end of the day you'll likely need to benchmark
01:57 <benwilson512> heh, on the same page I see
01:58 nhooyr joined
01:58 wpcarro joined
01:59 __charly__ joined
02:01 cschneid_ joined
02:05 Cohedrin joined
02:07 Sgeo joined
02:12 m00dy joined
02:13 travis-ci joined
02:13 <travis-ci> xiamx/elixir#15 (escript_test_windows - 154e05e : Mengxuan Xia): The build passed.
02:13 <travis-ci> Change view : https://github.com/xiamx/elixir/compare/8ffc2cdc6b5d...154e05e1689b
02:13 <travis-ci> Build details : https://travis-ci.org/xiamx/elixir/builds/236226970
02:13 travis-ci left
02:14 WhiskyRyan joined
02:17 dw2 joined
02:18 <dw2> My company is looking to hire a development firm for a medium-large project. We've been looking at Thoughtbot. Are there any other companies we should consider?
02:22 PaReeOhNos joined
02:23 travis-ci joined
02:23 <travis-ci> xiamx/elixir#16 (windows_compatibility - 04ada8e : José Valim): The build passed.
02:23 <travis-ci> Change view : https://github.com/xiamx/elixir/compare/52503d2bb1a7^...04ada8e1b86a
02:23 <travis-ci> Build details : https://travis-ci.org/xiamx/elixir/builds/236228493
02:23 travis-ci left
02:25 smt_ joined
02:28 gvaughn_ joined
02:35 nhooyr joined
02:36 ssswitch joined
02:39 Cohedrin joined
02:54 srxa joined
02:55 wpcarro joined
02:58 __charly__ joined
03:03 dj_goku joined
03:03 dj_goku joined
03:04 gvaughn_ joined
03:05 elgenie joined
03:12 smt_ joined
03:14 gvaughn_ joined
03:16 utkarsh joined
03:26 jeramyRR joined
03:29 Cohedrin joined
03:31 PaReeOhNos joined
03:31 srxa joined
03:34 snowowl joined
03:41 srxa joined
03:41 <snowowl> Hi theree, so I'm working through the phoenix book and ran into an issue with the database part and i'm not exactly sure what's going wrong
03:41 demio joined
03:41 <snowowl> i'll paste the error in just a sec
03:42 <demio> is anyone here super well versed in optimizing binary pattern matching on long strings
03:42 <snowowl> https://paste.pound-python.org/show/GdvKTdJn6Wfutqnz87vf/
03:43 <sevenseacat> so the error says ExpSite.UserController.create/2
03:43 <sevenseacat> is undefined or private
03:43 <sevenseacat> have you defined that function?
03:43 <snowowl> i believe so
03:44 <snowowl> UserController was defined in user_controller.ex
03:44 <sevenseacat> can you gist that controller?
03:44 <snowowl> yeah
03:47 jerel joined
03:48 wpcarro joined
03:50 <snowowl> https://gist.github.com/k-cross/ea6d51597a96bb007bdb4c60cbf966af
03:50 <snowowl> sorry, never actually created a gist before
03:50 <sevenseacat> right, there's no create/2 function there
03:50 <snowowl> hmmm, yeah i thought that was a little funny
03:51 <snowowl> it's not mentioned anywhere in the book and i kinda thought it was dynamically generated
03:51 <sevenseacat> which page are you looking at
03:52 <snowowl> 62
03:53 <snowowl> get "/", PageController, :index
03:53 <snowowl> resources "/users", UserController, only: [:index, :show, :new, :create]
03:53 <snowowl> particularly the resources part
03:54 <sevenseacat> ok I can see the create action being built at the top of page 65
03:54 Cohedrin joined
03:54 <sevenseacat> so you might just be getting a bit ahead of the book :)
03:55 <snowowl> ahh just adds routes
03:55 <sevenseacat> yep yep
03:55 <sevenseacat> then you build the form, then the action that processes the form
03:55 <snowowl> he tells us to run it after that and hmm
03:55 <snowowl> but it's good to know that that's what i need
03:55 <snowowl> thanks
03:55 <sevenseacat> no probs
03:56 <snowowl> i'm new to both the language and framework. it's been fun
03:56 <sevenseacat> glad you're enjoying it!
03:57 akeating joined
03:57 jleon joined
03:57 <demio> im having trouble with this specific function
03:57 <demio> https://github.com/cblage/elixir-json/blob/optimizing/lib/json/parser/bitstring.ex#L150-L170
03:57 <demio> its very slow
03:58 <demio> for very long strings
04:01 m00dy joined
04:10 cschneid_ joined
04:12 griffinbyatt joined
04:29 elgenie joined
04:30 <bitwalker> demio: I would run fprof on it and see where it's spending all the time
04:31 wsieroci joined
04:33 Cohedrin joined
04:38 <demio> bitwalker already did that
04:38 Cohedrin joined
04:39 <demio> its spending all its time in the function i linked
04:40 PaReeOhNos joined
04:42 wpcarro joined
04:43 <demio> https://www.dropbox.com/s/vprj9fjokxyhe9n/Screenshot%202017-05-26%2005.43.42.png?dl=0
04:43 picasso joined
04:51 dj_goku joined
04:51 dj_goku joined
04:52 elgenie joined
04:53 pookleblinky joined
04:57 <bitwalker> demio: I think the unnecessary match in the terminate_* function is forcing copies
04:58 <bitwalker> I don't know that this is what's causing you problems, but it's probably one thing worth changing
04:59 Cohedrin joined
05:02 jeffweiss joined
05:04 <bitwalker> you could also compile with ERL_COMPILER_OPTIONS=bin_opt_info to see what optimizations you may be missing
05:07 cemilowski joined
05:09 <demio> im already using the ERL_COMPILER_OPTIONS
05:11 dec0n joined
05:14 srxa_ joined
05:14 griffinbyatt joined
05:14 meandi_2 joined
05:17 Cohedri__ joined
05:17 cjhowe joined
05:18 cjhowe joined
05:19 cjhowe joined
05:20 cjhowe joined
05:21 cjhowe joined
05:21 cjhowe joined
05:22 cjhowe joined
05:23 potatosalad joined
05:36 cjhowe joined
05:39 cjhowe joined
05:40 cjhowe joined
05:49 m00dy joined
05:51 PaReeOhNos joined
05:55 cemilowski joined
05:57 cemilowski1 joined
05:57 jleon joined
06:03 Cohedrin joined
06:04 l14n joined
06:08 gvaughn joined
06:17 codestorm joined
06:18 codestorm joined
06:18 griffinbyatt joined
06:21 <micmus> demio: probably, the problem is here: https://github.com/cblage/elixir-json/blob/optimizing/lib/json/parser/bitstring.ex#L174
06:21 <micmus> the function does not start by matching on the first argument, so any caller can't be optimised
06:21 <micmus> replacing `json` with `<<json::binary>>` should fix it
06:28 <demio> hmm
06:31 wpcarro joined
06:32 refriedchicken joined
06:38 <snowowl> wasted an hour on a old dependency problem :(
06:39 dj_goku joined
06:39 dj_goku joined
06:42 tuacker joined
06:43 tuacker joined
06:43 <demio> @micmus
06:43 <demio> better or worse
06:43 <demio> https://github.com/cblage/elixir-json/commit/e37ad563815c617fde0a47bf65d15f9cc5d84cb5
06:46 <demio> x_x
06:48 <micmus> demio: I'm not 100% sure to be honest, I would need to spend some more time looking at it, but I can't right now.
06:48 <micmus> Are you trying to implement a parser faster than poison?
06:48 <demio> im trying to at least match it
06:48 <demio> i feel ashamed about my parsers performance
06:49 <demio> and idk where i went wrong
06:49 <demio> :(
06:49 <demio> i still get this from fprof
06:50 <demio> micus: https://www.dropbox.com/s/9fi4gc1ktxhhb22/Screenshot%202017-05-26%2007.49.48.png?dl=0
06:50 <demio> micmus
06:50 <micmus> efficient binary handling is sort of an art. Profiling won't always be helpful since it changes the system under test :)
06:50 <micmus> ERL_COMPILER_OPTIONS=bin_opt_info is your best friend when working with binaries
06:50 mk[] joined
06:51 mk[] left
06:52 <demio> ive been using that
06:52 <demio> but i cant get every single function to optimize
06:54 <demio> i just feel like theres somethign fundamentally wrong with my current approach
06:55 <demio> for "parse_string_recursive" being so slow
06:55 <micmus> yes, doing every single one is probably impossible, but hot loops need to be in order to achieve good performance
06:55 <micmus> e.g. ^ that one
06:57 squallstter joined
06:57 <demio> but the only non-optimzed "parse_string_recursive" calls
06:57 <demio> are terminal ones
06:58 <micmus> all heads need to be optimized for the function to be optimized
06:58 ur5us joined
07:00 PaReeOhNos joined
07:00 dw2 joined
07:03 <demio> how do I fix these micmus https://gist.github.com/cblage/37a722aa301aef30204e2eff3e20f2e3
07:03 <demio> x_x
07:03 <demio> i mean
07:03 <demio> parse_number is not a problem
07:03 <demio> so dont need to optimize it
07:04 <demio> how do I optimize this one
07:04 <demio> https://github.com/cblage/elixir-json/blob/optimizing/lib/json/parser/bitstring.ex#L154
07:06 tuacker joined
07:12 codestorm joined
07:15 <fredsir> I've been looking for a way to manage multiple elixir versions on my system. I've found kiex... Any one here using it? Is it good? Is there anyone better you can recommend instead of it?
07:15 josevalim joined
07:22 PaReeOhNos joined
07:23 m00dy joined
07:23 mark_66 joined
07:24 <steffkes> fredsir: did you give docker a whirl?
07:25 <sevenseacat> asdf is also an option I think
07:25 cemilowski joined
07:33 cemilowski joined
07:34 <sasajuric> yp, I'm using asdf because it can manage both Elixir and Erlang versions
07:35 dw2 joined
07:46 jkreeftmeijer joined
07:48 PaReeOhNos joined
07:48 Guest95 joined
07:49 akeating joined
07:51 jkreeftmeijer joined
07:51 m00dy joined
07:54 m00dy joined
07:57 gvaughn joined
07:58 jleon joined
08:06 codestorm joined
08:06 travis-ci joined
08:06 <travis-ci> elixir-lang/elixir#14723 (jv-windows-xiamx - 4e06b4f : Mengxuan Xia): The build passed.
08:06 <travis-ci> Change view : https://github.com/elixir-lang/elixir/commit/4e06b4f31187
08:06 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236284959
08:06 travis-ci left
08:07 PaReeOhNos joined
08:09 PaReeOhNos joined
08:09 jleon joined
08:12 cschneid_ joined
08:15 cdg joined
08:19 PaReeOhNos joined
08:19 wpcarro joined
08:19 griffinbyatt joined
08:26 cemilowski joined
08:27 dj_goku joined
08:27 dj_goku joined
08:28 travis-ci joined
08:28 <travis-ci> elixir-lang/elixir#14725 (jv-windows-xiamx - 7d646f6 : José Valim): The build passed.
08:28 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/4e06b4f31187...7d646f6b0f23
08:28 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236286402
08:28 travis-ci left
08:29 jkreeftm_ joined
08:33 cschneid_ joined
08:34 PaReeOhNos joined
08:38 Guest95 joined
08:38 jkreeftmeijer joined
08:40 dw2 joined
08:42 <fredsir> steffkes: Were using kubernetes in production, and im playing with minikube (which is as local one node kubernetes cluster) for local dev. What is still a problem is getting editors (in my case vim) to be able to know about the code without the tools it needs (for example elixir) being installed on the host instead of only inside a container running in the local cluster, so while it works, i have yet
08:42 Guest67_ joined
08:42 <fredsir> to have it functioning on par with local dev ...
08:42 <fredsir> ... where you install the tools directly on the host
08:44 inoas joined
08:47 jkreeftmeijer joined
08:48 travis-ci joined
08:48 <travis-ci> elixir-lang/elixir#14727 (master - 0d14e72 : José Valim): The build passed.
08:48 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/945c40ceec86...0d14e724d424
08:48 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236287139
08:48 travis-ci left
08:48 justelex joined
08:49 r00tl3ss joined
08:52 travis-ci joined
08:52 <travis-ci> elixir-lang/elixir#14728 (master - 70b6f97 : José Valim): The build passed.
08:52 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/0d14e724d424...70b6f973c063
08:52 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236292123
08:52 travis-ci left
08:52 notdaniel joined
08:57 jleon joined
08:59 lexmag joined
09:00 marr joined
09:00 codestorm joined
09:04 travis-ci joined
09:04 <travis-ci> elixir-lang/elixir#14729 (jv-windows-xiamx - 814287a : José Valim): The build passed.
09:04 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/7d646f6b0f23...814287a5fcc9
09:04 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236298747
09:04 travis-ci left
09:04 jkreeftmeijer joined
09:12 jkreeftmeijer joined
09:12 gregman_ joined
09:13 cdg_ joined
09:14 m00dy joined
09:15 m00dy joined
09:23 cschneid_ joined
09:24 srxa joined
09:26 jkreeftmeijer joined
09:29 akeating_ joined
09:32 meandi joined
09:39 proteusguy joined
09:40 travis-ci joined
09:40 <travis-ci> elixir-lang/elixir#14731 (jv-windows-xiamx - d5eedfa : José Valim): The build passed.
09:40 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/814287a5fcc9...d5eedfa60cef
09:40 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236308683
09:40 travis-ci left
09:45 gvaughn joined
09:54 codestorm joined
10:00 travis-ci joined
10:00 <travis-ci> elixir-lang/elixir#14733 (jv-windows-xiamx - 970c09a : José Valim): The build passed.
10:00 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/d5eedfa60cef...970c09a64fa2
10:00 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236313686
10:00 travis-ci left
10:02 jkreeftmeijer joined
10:04 ur5us joined
10:07 wpcarro joined
10:10 vili_ joined
10:15 dj_goku joined
10:16 bitmod joined
10:17 vili_ joined
10:19 ed joined
10:20 m00dy joined
10:20 griffinbyatt joined
10:21 travis-ci joined
10:21 <travis-ci> elixir-lang/elixir#14735 (jv-windows-xiamx - ec12984 : Mengxuan Xia): The build has errored.
10:21 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/970c09a64fa2...ec12984de708
10:21 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236316369
10:21 travis-ci left
10:25 Ilyes512 joined
10:30 travis-ci joined
10:30 <travis-ci> elixir-lang/elixir#14737 (jv-windows-xiamx - 4d5b423 : José Valim): The build has errored.
10:30 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/ec12984de708...4d5b42326794
10:30 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236318962
10:30 travis-ci left
10:31 m00dy joined
10:33 <micmus> Somebody's busy today with commits :P
10:38 bitmod joined
10:40 gvaughn joined
10:43 ur5us joined
10:45 m00dy joined
10:45 cemilowski joined
10:48 codestorm joined
10:48 <OliverMT> fredsir: why not both?
10:48 inoas joined
10:54 travis-ci joined
10:54 <travis-ci> elixir-lang/elixir#14741 (jv-windows-xiamx - 49fbb02 : José Valim): The build has errored.
10:54 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/825414e0a20c...49fbb028803f
10:54 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236323115
10:54 travis-ci left
11:05 m00dy joined
11:10 akeating joined
11:11 cschneid_ joined
11:18 <hmans> Do we know any rough kind of ETA on Phoenix 1.3?
11:21 griffinbyatt joined
11:22 travis-ci joined
11:22 <travis-ci> elixir-lang/elixir#14743 (jv-windows-xiamx - f4f9281 : José Valim): The build passed.
11:22 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/49fbb028803f...f4f9281366d0
11:22 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236331545
11:22 travis-ci left
11:26 akeating joined
11:28 akeating joined
11:29 griffinbyatt joined
11:41 greengriminal joined
11:43 codestorm joined
11:46 m00dy joined
11:47 m00dy joined
11:54 __charly__ joined
11:55 travis-ci joined
11:55 <travis-ci> elixir-lang/elixir#14746 (jv-windows-xiamx - ced5ec5 : José Valim): The build passed.
11:55 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/f4f9281366d0...ced5ec5d7bd7
11:55 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236337214
11:55 travis-ci left
11:56 <micmus> hmans: it's blocking on updating guides
11:56 PaReeOhNos joined
11:57 <hmans> micmus: thanks
11:57 wpcarro joined
11:58 bitmod joined
12:00 greengriminal joined
12:01 karreiro joined
12:02 bitmod joined
12:04 dj_goku joined
12:04 dj_goku joined
12:08 Douman joined
12:08 <Douman> Anyone been using sqlite3 with ecto?
12:08 travis-ci joined
12:08 <travis-ci> elixir-lang/elixir#14748 (jv-windows-xiamx - 07ebcbc : José Valim): The build passed.
12:08 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/ced5ec5d7bd7...07ebcbc4d29e
12:08 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236348496
12:08 travis-ci left
12:14 juan_ joined
12:15 <juan_> anyone in here who can help me with an ecto problem?
12:15 <asonge> i'm not always the best with ecto problems, but go ahead.
12:15 <hmans> juan_: ask and someone shall answer... at some point... maybe.
12:16 <juan_> thanks, well I keep getting this error when trying to do a double join
12:16 <Douman> ah... it seems sqlite3 adapter for ecto isn't compatible with v2
12:16 <juan_> value in `where` cannot be cast to type :id in query:
12:17 <asonge> juan_: you want to gist the full error and the code?
12:17 <juan_> yes will do
12:19 travis-ci joined
12:19 <travis-ci> elixir-lang/elixir#14750 (jv-windows-xiamx - 4b9238e : José Valim): The build has errored.
12:19 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/07ebcbc4d29e...4b9238ee3741
12:19 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236354120
12:19 travis-ci left
12:21 <juan_> ok : https://gist.github.com/jwaterfaucett/b9c0eae9afed9d7e731b8fed342a3f37
12:21 <juan_> basically, I have your typical User, UserRoles, Role system
12:22 <juan_> and want to see if a user_role entry having a particular role exists
12:22 <juan_> I don't need the user so I don't have to do a three way join
12:23 <benwilson512> juan_: the second argument to get should be an id
12:23 <benwilson512> not a query
12:23 <benwilson512> pretty sure you don't want Repo.get
12:23 <benwilson512> maybe Repo.one
12:23 <asonge> juan_: you can just add more joins.
12:23 <asonge> as well
12:24 <benwilson512> your error about casting to type `:id` is almost certainly because of the use of Repo.get
12:25 <juan_> thanks guys, great help, sorry I kept thinking :id had to with the query because of queryable error
12:28 gvaughn joined
12:28 akeating joined
12:30 dimitarvp joined
12:32 <juan_> I got another question that's also slightly confusing me while we're at it :)
12:32 <juan_> what is the preferred way to require associations in a schema file?
12:32 <asonge> require associations?
12:33 <juan_> yes, for instance in rails say u have a Comment model, then you have validate_presence_of :post
12:34 <juan_> to ensure that the post is always set when putting a model in the db
12:34 <juan_> I figured out how to do database unique indexes and not null validations in ecto in the migrations
12:34 <asonge> changesets can be used to require associations to be present in an update or insert, and you can control whether an association can be null or not in the schema and in the migration to make it database-level.
12:34 <juan_> yes so with changesets is it cast_assoc or put_assoc?
12:35 <juan_> the difference isn't clear to me
12:35 cemilowski joined
12:35 <asonge> yes, though they're subtly different ( i have to read the docs every time)
12:35 <sevenseacat> i have to check every time too lol
12:35 <juan_> ok, so I'm not alone there :)
12:37 <asonge> ah, the docs for cast_assoc discuss the differences. cast_assoc does a lot more heavy lifting.
12:37 codestorm joined
12:37 <benwilson512> I've been meaning to contribute examples to put_assoc forever
12:37 <benwilson512> I think that's part of why it's so confusing
12:37 srxa joined
12:37 travis-ci joined
12:37 <travis-ci> elixir-lang/elixir#14752 (jv-windows-xiamx - 141f4b4 : José Valim): The build passed.
12:37 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/4b9238ee3741...141f4b472b9a
12:37 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236356314
12:37 travis-ci left
12:38 mirela666 joined
12:38 jkreeftmeijer joined
12:38 <juan_> benwilson512: yes, that would be very nice
12:39 <juan_> Overall, I'm really liking ecto coming from active_record, its much more explicit which I think is a good thing
12:39 wpcarro joined
12:39 <benwilson512> yeah there's still a rough edge or two but I feel so much more familiar with SQL after doing ecto
12:39 <ciawal> I have a phoenix app on heroku with 2 dynos
12:39 <ciawal> I wanted to create a genserver which would do some periodic processing, but if I do this (I think) it would be run on both dynos, is there any common hack I can use to avoid this?
12:39 <juan_> but there's still lots cases that I immediately encountered where the docs don't help that much
12:40 <benwilson512> ciawal: you've got a few options. an easy one is that if you already have a database you run the process on both nodes, and each tries to get a lock on a database row
12:41 <benwilson512> select for update that kinda thing
12:45 jleon joined
12:46 m00dy joined
12:46 akeating joined
12:46 <ciawal> seems there is a $DYNO env var available, so I could in theory check this and only start the server on web.1
12:46 <ciawal> seems messy though
12:48 akeating joined
12:48 <juan_> benwilson512: Do you care to look at how I'm validating associations now and tell me what you think?
12:48 wpcarro joined
12:48 jkreeftmeijer joined
12:48 <juan_> Here's the link : https://gist.github.com/jwaterfaucett/01709888680e076e1e769cd0d99292fb
12:49 <ciawal> my situation is that I have a cms with users editing data, but edited records need some processing which takes some time and can be better done in bulk
12:49 <ciawal> so the plan was to have a genserver which would store a last processed timestamp, and every 5 minutes would look for records with updated_at > this date to run the processing
12:49 <ciawal> is there perhaps a better option for what I want to do?
12:49 <juan_> I'm somewhat confused by fields of type id and the associations
12:50 <juan_> coming from rails, passing the ids directly (instead of using the models) feels dirty
12:50 <juan_> when doing the insert
12:50 m00dy joined
12:50 <ciawal> you should be able to use the struct too
12:50 <ciawal> instead of struct.id
12:54 rodolfojcj joined
12:58 travis-ci joined
12:58 <travis-ci> elixir-lang/elixir#14754 (jv-windows-xiamx - 1ef4baa : José Valim): The build passed.
12:58 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/141f4b472b9a...1ef4baa5cbb3
12:58 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236359221
12:58 travis-ci left
13:02 kronicdeth joined
13:03 ramortegui joined
13:06 srxa_ joined
13:06 srxa_ joined
13:10 ikcerog joined
13:10 ssswitch joined
13:12 travis-ci joined
13:12 <travis-ci> elixir-lang/elixir#14756 (jv-windows-xiamx - 5e2f8c9 : José Valim): The build passed.
13:12 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/1ef4baa5cbb3...5e2f8c992c85
13:12 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236364749
13:12 travis-ci left
13:13 <OliverMT> ciawal: are you on aws?
13:14 <OliverMT> no, heroku
13:17 cdg joined
13:17 <OliverMT> ciawal: I normally reccomend scheduled aws lambda jobs for people who need global exactly once semantics on recurring, seems like heroku has https://devcenter.heroku.com/articles/scheduler
13:17 <OliverMT> set it up to call your api every 5-10 mins or whatever
13:17 <OliverMT> and the load balancer will take care of making sure only one node runs :)
13:18 <OliverMT> I spent a day trying to make a non-complicated "run every 5 mins but on only one node in cluster" before thinking to use aws lambda
13:18 <OliverMT> it seems trivial, but there is a lot of pitfalls about electing masters, handling splits and all sorts of crap
13:18 <OliverMT> its even harder if you are on heroku, as you arent even running disterls, so the nodes are stateless
13:19 travis-ci joined
13:19 <travis-ci> elixir-lang/elixir#14758 (jv-windows-xiamx - 8f236b0 : José Valim): The build has errored.
13:19 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/5e2f8c992c85...8f236b0d5dde
13:19 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236368792
13:19 travis-ci left
13:20 <sorentwo> If you're using a modern version of Postgres (9.5+) you can use SELECT .. FOR UPDATE SKIP LOCKED to build a very simple queue with exactly once semantics.
13:21 <OliverMT> it's still manual handwaving
13:22 <OliverMT> you'd need to pull often on the non master to see if you should be amster, and master needs to keep the lock active to tell the slaves its not split off blabla
13:22 jschneck joined
13:22 <benwilson512> well, you can select for update and immediately change a state column to "processing" or something
13:22 <smeevil> o/
13:22 <benwilson512> and then other select for updates only look for state pending
13:22 <sorentwo> Yes, that's precisely what you do.
13:23 <benwilson512> for cron like stuff I think lambda is fine
13:23 <benwilson512> but the select for update stuff is very handy if you want to treat a table as a queue
13:23 <OliverMT> yes, but thats a queue, not a "run on exactly one node every 5 minutes"
13:24 <benwilson512> yup
13:24 <OliverMT> updating your cluster could make it end up as each node skewed 2.5 minutes from eachother
13:24 <OliverMT> if node1 goes down the job will be skewed 2.5 minutes until node2 tries to run again
13:24 <smeevil> I'm trying to persist a record in the database with a date field. my changeset has start_date: ~D[2005-03-16]}, the changeset is also valid. when i then try to insert it i get (ArgumentError) could not encode date/time: {2005, 3, 16} but i have no idea what i'm doing wrong. In the migration and schema i use the :date attribute for the field.
13:24 <OliverMT> unless node1 keeps saying its up and node2 keeps asking if node1 is up
13:24 <sorentwo> You can have as many workers as you like that poll every N minutes from the queue, the select for update is atomic, only one worker can grab a record at a time.
13:24 <OliverMT> unless you lock it to exactly 05, 10, 15 etc
13:24 <OliverMT> thats not the issue sorentwo
13:25 <OliverMT> what if the end result is a mail? you'd end up suddenly splitting records between two nodes
13:25 <OliverMT> producing two mails for a single aggregate
13:25 <benwilson512> yeah time skew can produce weird issues
13:25 <OliverMT> since your jobs can run on whatever node
13:25 kronicdeth joined
13:26 <OliverMT> the nodes needs to either be told externally which one should run an exactly-on-one-node job *or* they need to chatter behind the scenes continously to figure out who should be running next window
13:26 <OliverMT> either way its certainly waaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaay off trivial
13:26 <OliverMT> heroku scheduler that simply calls your api with someo sort of secret key or a heroku clock dyno that does the same sounds like a pretty cheap and easy solution
13:28 jerel joined
13:30 <smeevil> anyone might have a pointer on the encoding problem ?
13:31 codestorm joined
13:31 <OliverMT> smeevil: are you going via cast and is the start_date mapped as Ecto.Date in model?
13:32 <OliverMT> oh you said :date in schema
13:32 <smeevil> aye, but also if i set it to Ecto.Date in the schema it gives the same error
13:32 <smeevil> the changeset then has start_date: #Ecto.Date<2005-03-16>},
13:33 <smeevil> wait... I'm being stupid
13:33 <* OliverMT> can roll with that
13:34 <smeevil> oh i thought i messed around the month / day but that was correct , i rectify what i said :P
13:34 gvaughn joined
13:34 <smeevil> at least the changeset gives an error when giving 2015-16-03 so that is working as expected
13:35 <ciawal> OliverMT: the reason I wanted to use a genserver rather than a scheduler in this case was to keep some state in memory
13:35 <OliverMT> well you could write to db_
13:35 <OliverMT> ?
13:35 <ciawal> if needed, but I'd rather avoid it
13:35 <OliverMT> its either external and easy or internal and hard :p
13:36 wpcarro joined
13:36 lexmag joined
13:37 <smeevil> might you have an other suggestion that i could check ?
13:38 travis-ci joined
13:38 <travis-ci> elixir-lang/elixir#14760 (jv-windows-xiamx - 91693fa : José Valim): The build has errored.
13:38 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/8f236b0d5dde...91693fa0adbb
13:38 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236372956
13:38 travis-ci left
13:39 Tallys__ joined
13:39 ianclark joined
13:39 <OliverMT> nah, heroku is kinda worst of both worlds
13:39 hbraun joined
13:39 <OliverMT> why not write to db?
13:40 <OliverMT> lean on postgres until postgres starts to bend, *then* optimize, thats my mantra :D
13:41 <ciawal> it's not a scaling thing, it's just that I'd have to make a table with a single to store this timestamp, which is ugly
13:46 travis-ci joined
13:46 <travis-ci> elixir-lang/elixir#14762 (master - bb274ee : José Valim): The build passed.
13:46 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/ca1ec8a86f26...bb274ee00540
13:46 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236378958
13:46 travis-ci left
13:47 <smeevil> hmm it boils down to the fact that Ecto.Adapters.Postgres.Timestamp.encode!/1 is being used in stead of Ecto.Adapters.Postgres.Date/1
13:48 m00dy joined
13:48 jordan0day joined
13:48 potatosalad joined
13:51 m00dy joined
13:52 dj_goku joined
13:52 dj_goku joined
13:53 m00dy joined
13:56 m00dy joined
13:56 jleon joined
13:58 wpcarro joined
14:03 JEG2 joined
14:04 m00dy joined
14:15 kronicdeth joined
14:16 m00dy joined
14:16 maxbeizer joined
14:17 m00dy joined
14:18 <Douman> hm... is there a way for Plug.Static to on fly gzip files instead of looking up <file>.gz ?
14:19 sourpus5 joined
14:20 jeffweiss joined
14:25 codestorm joined
14:27 MaybeDragon joined
14:30 greengriminal joined
14:41 wsieroci joined
14:42 mwbrown joined
14:45 Lingo joined
14:56 akeating joined
15:03 smt joined
15:05 adgtl joined
15:05 adgtl joined
15:05 cemilowski joined
15:07 travis-ci joined
15:07 <travis-ci> elixir-lang/elixir#14763 (jv-unicode - 6eb05b9 : José Valim): The build passed.
15:07 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/004556957491...6eb05b9dd0ce
15:07 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236404471
15:07 travis-ci left
15:10 eddd joined
15:11 rozap joined
15:11 cschneid_ joined
15:18 Lingo joined
15:19 codestorm joined
15:24 tuacker joined
15:25 cjhowe joined
15:28 rschmukler joined
15:30 codestorm joined
15:40 dj_goku joined
15:40 dj_goku joined
15:47 srxa joined
15:47 meh` joined
15:52 chenghiz joined
15:55 cemilowski joined
15:56 __charly__ joined
15:57 davidw joined
16:06 squallstter joined
16:10 Lingo joined
16:16 PaReeOhNos joined
16:22 leons joined
16:24 dimitarvp joined
16:26 travis-ci joined
16:26 <travis-ci> elixir-lang/elixir#14764 (jv-unicode - 646821a : José Valim): The build passed.
16:26 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/6eb05b9dd0ce...646821a78f91
16:26 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236434781
16:26 travis-ci left
16:27 refriedchicken joined
16:31 jeffweiss joined
16:40 dw2 joined
16:40 refriedchicken66 joined
16:45 codestorm joined
16:46 m00dy joined
16:47 rschmukler joined
16:47 leons joined
16:47 wpcarro joined
16:50 rodolfojcj joined
16:51 wsieroci joined
16:52 LastWhisper____ joined
17:02 Cohedrin joined
17:05 jschneck joined
17:08 WhiskyRyan joined
17:17 OtherAllan joined
17:17 wpcarro joined
17:19 Cohedrin joined
17:22 PaReeOhNos joined
17:24 sourpus5 joined
17:27 WhiskyRyan joined
17:28 dj_goku joined
17:28 dj_goku joined
17:33 ani joined
17:36 maxbeizer joined
17:37 wpcarro joined
17:37 <ani> trying to use embedded schemas in ecto. have hit a road block.
17:37 <ani> Unable to update an embeds_many field similar to example on "http://blog.plataformatec.com.br/2015/08/working-with-ecto-associations-and-embeds/". But getting an error "If you are attempting to update an existing entry, please make sure you include the entry primary key (ID) alongside the data."
17:38 <ani> Anyone here done something similar ? any more blogs or examples out there that you can refer
17:41 inoas joined
17:42 jleon joined
17:45 Cohedrin joined
17:46 Guest67 joined
17:49 karreiro joined
17:49 jkreeftmeijer joined
17:49 PaReeOhNos joined
17:55 gvaughn_ joined
18:01 Lingo_ joined
18:05 JEG2 joined
18:06 Cohedrin joined
18:09 ikcerog joined
18:09 <meh`> with Ecto I'm getting a syntax error from Postgrex in a migration, is there some way to see the query it was trying to make?
18:11 <asonge> meh`: i think there's a log setting on ecto for catching all the queries?
18:13 <meh`> asonge, the error is coming from Ecto.Adapters.SQL.query! so I presume it's failing before even generating it
18:13 sourpus5 joined
18:17 JuanMiguel joined
18:20 PaReeOhNos joined
18:21 PaReeOhNos joined
18:25 Cohedri__ joined
18:26 ianclark joined
18:33 dorei joined
18:41 squallstter joined
18:44 jschneck joined
18:53 embik joined
18:56 ikcerog joined
19:01 ikcerog joined
19:12 eddd joined
19:12 Cohedrin joined
19:17 dj_goku joined
19:17 dj_goku joined
19:18 DisruptiveNL joined
19:25 attente joined
19:27 Lingo_ joined
19:40 cdg joined
19:43 srxa_ joined
19:45 fhoffmann joined
19:47 ur5us joined
19:51 sourpus5 joined
19:51 travis-ci joined
19:51 <travis-ci> elixir-lang/elixir#14767 (jv-unicode - 6164e50 : José Valim): The build passed.
19:51 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/2a233aca603a...6164e5024266
19:51 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236499081
19:51 travis-ci left
19:54 m00dy joined
19:54 <LastWhisper____> could anyone point me in the right direction to set up a small working example? I want to open two iex consoles; have a supervisor in terminal A, and a child in terminal B; and send an IO.puts "hello" from terminal A, and have it appear in the console of terminal B
19:55 m00dy joined
19:59 Lingo_ joined
19:59 rschmukler joined
19:59 Lingo_ joined
20:00 Lingo_ joined
20:01 elgenie joined
20:05 rozap joined
20:05 <JEG2> LastWhisper____: There's an example here: https://elixir-lang.org/getting-started/mix-otp/distributed-tasks-and-configuration.html#our-first-distributed-code
20:05 DTZUZU joined
20:07 jdqx joined
20:09 Lingo_ joined
20:09 sourpus5 joined
20:11 Lingo_ joined
20:12 r00tl3ss joined
20:13 r00tl3ss_ joined
20:17 jschneck joined
20:17 <LastWhisper____> ty JEG2
20:19 <LastWhisper____> I have a follow up question: so I'm creating a supervisor in one iex session "foo", I want to open 2 more iex sessions "bar" and "baz", i would like them to be linked to "foo", and once I'm in "foo", i want to tell all children linked to IO.puts "hello world"
20:19 <LastWhisper____> How would I reference the children from the main supervisor ?
20:21 cemilowski joined
20:22 karmajunkie joined
20:23 cdg joined
20:24 refriedchicken joined
20:25 travis-ci joined
20:25 <travis-ci> elixir-lang/elixir#14769 (jv-unicode - 2e8eb15 : José Valim): The build passed.
20:25 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/6164e5024266...2e8eb159ca2d
20:25 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236508155
20:25 travis-ci left
20:29 Cohedrin joined
20:29 <JEG2> LastWhisper____ Can you show the code you're using to create a supervisor in iex? I don't believe I understand what you are trying to accomplish yet.
20:33 Lingo_ joined
20:34 pizzalord joined
20:34 rschmukler joined
20:37 notdaniel joined
20:37 fhoffmann joined
20:39 sp4rrow joined
20:42 refriedchicken joined
20:42 elgenie joined
20:44 <benwilson512> LastWhisper____: in general you don't want to supervise across nodes
20:45 <benwilson512> normally you would have a supervisor on each node, and they would start their children there
20:48 wpcarro joined
20:48 kronicdeth joined
20:48 rozap joined
20:52 cemilowski joined
20:55 wpcarro joined
20:57 WhiskyRyan joined
20:59 <OliverMT> am I blind? or is there no example in ecto for how to update_all with actual values? https://hexdocs.pm/ecto/Ecto.Repo.html#c:update_all/3
20:59 <OliverMT> oh wait, you use the actual column names in the keyword list of 2nd argument
21:02 WhiskyRyan joined
21:05 dj_goku joined
21:05 dj_goku joined
21:05 wpcarro joined
21:08 srxa joined
21:15 chrisi_ joined
21:17 chrisi_ left
21:19 WhiskyRyan joined
21:21 wpcarro joined
21:21 Lingo_ joined
21:21 travis-ci joined
21:21 <travis-ci> elixir-lang/elixir#14773 (master - 9825280 : José Valim): The build passed.
21:21 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/bb274ee00540...98252805eac1
21:21 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/236526108
21:21 travis-ci left
21:24 WhiskyRyan joined
21:26 rschmukler joined
21:32 rodolfojcj joined
21:42 <LastWhisper____> so I think i put the cart ahead of the horse; I was looking into https://hexdocs.pm/phoenix_pubsub/Phoenix.PubSub.html and saw they have two options, PG2 and point to your endpoint, or use redis... problem is I don't even have an endpoint, so that got me looking at https://hexdocs.pm/phoenix/Phoenix.Endpoint.html ... Before I even think of the whole publisher/subscriber model should I be looking at fleshing out an
21:42 <LastWhisper____> Endpoint and thinking of what it will support on that end ?
21:46 elgenie joined
21:56 sourpus5 joined
22:01 WhiskyRyan joined
22:02 <OliverMT> LastWhisper____: Endpoint is generated with phoenix.new
22:03 <OliverMT> when you set up a new phoenix project
22:03 <OliverMT> its nothing special, just plumbing
22:03 <LastWhisper____> I guess I'm trying to use phoenix pubsub without phoenix
22:04 <LastWhisper____> is my only "adapter" option redis in that case?
22:05 gvaughn_ joined
22:09 <adamkittelson> LastWhisper____: the pg2 adapter doesn't come from phoenix, pg2 is a built-in erlang module, you should be able to use it with phoenix_pubsub without phoenix
22:15 jleon joined
22:26 inoas joined
22:28 Cohedrin joined
22:30 jleon joined
22:33 cschneid_ joined
22:36 cemilowski joined
22:40 rschmukler joined
22:41 elixir-lang690 joined
22:42 <elixir-lang690> hi
22:42 rschmukler joined
22:42 <elixir-lang690> quit
22:42 elixir-lang690 left
22:45 rschmukler joined
22:45 droid joined
22:53 dj_goku joined
22:53 dj_goku joined
22:53 <benwilson512> LastWhisper____: in the docs you linked it merely said "often configured" with respect to the endpoint
22:53 <benwilson512> Direct usage section covers how to use it directly
22:54 <benwilson512> If you want to use PG2 you just use the PG2 module there instead of Redis
22:54 <LastWhisper____> you're referring to this: `supervisor(Phoenix.PubSub.PG2, [:my_pubsub, host: ""])` ?
23:02 <benwilson512> yes
23:03 <benwilson512> although you wouldn't need host there, that's likely a redis only config
23:03 <benwilson512> https://hexdocs.pm/phoenix_pubsub/Phoenix.PubSub.PG2.html covers the options for PG2 backend
23:06 bokayio joined
23:10 Cohedrin joined
23:10 jmcintosh joined
23:17 srxa joined
23:29 eddd joined
23:30 Cohedrin joined
23:37 rozap joined
23:41 __charly__ joined
23:42 jleon joined
23:42 jmiven joined
23:44 ur5us joined
23:44 dw2 left
23:49 griffinbyatt joined
23:50 m00dy joined
23:54 inoas joined
23:54 Tallys__ joined