<    March 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
00:06 maxbeizer joined
00:07 refriedchicken joined
00:11 duane joined
00:35 jeregrine joined
00:35 gvaughn_ joined
00:46 sevenseacat joined
00:46 vicsy joined
00:49 chavito joined
00:52 gvaughn__ joined
00:53 <vicsy> hi
00:54 nighty- joined
00:56 <vicsy> hi
00:56 griffinb_ joined
00:57 nighty- joined
01:00 vicsy left
01:02 io_bora joined
01:07 icapurro joined
01:10 griffinbyatt joined
01:11 jschneck joined
01:13 webdev007 joined
01:15 jkreeftmeijer joined
01:17 nahtnam joined
01:33 harfangk joined
01:34 raycoll joined
01:39 tomterl joined
01:46 DTZUZO joined
01:52 raycoll joined
02:15 jkreeftmeijer joined
02:19 dmilith joined
02:27 gvaughn joined
02:28 gvaughn_ joined
02:29 <cmk_zzz> The ecto documentation says: An Ecto schema is used to map any data source into an Elixir struct"
02:29 <cmk_zzz> and the ecto book gives an example mapping a From to an ecto schema.
02:29 <cmk_zzz> Wouldn't it make sense to decouple Ecto schemas from ecto? I.e if they can map to any data source why ship them with the database layer?
02:31 raycoll joined
02:42 io_bora joined
03:08 <benwilson512> ecto only sort of has a database layer
03:08 <benwilson512> you still need postgrex or mariaex to actually talk to the database
03:13 <cmk_zzz> benwilson512: yes, but they are listed (although optional) as dependencies to ecto. Just wondering if schemas and changesets are more generic and the database pooling and drivers could've been separate or if there actually is such a strong coupling that it makes sense to ship them together as it is today
03:14 icapurro joined
03:15 <benwilson512> well it's about goals
03:15 <benwilson512> frankly "any data source" may be a slightly strong statement
03:16 <benwilson512> more generic doesn't always make for a good experience with specific platforms
03:17 <sevenseacat> I think the schema stuff is a bonus as to how ecto was designed to interact with databases
03:17 jkreeftmeijer joined
03:18 <cmk_zzz> yes. I think the recent changes are great though. For example ecto schemaless is generally what I need. In smaller applications where I already have a database I am only after the raw queries and don't care about schemas or changeset. In other applications I would love to use changesets to validate my data
03:18 <cmk_zzz> without using a database
03:19 <cmk_zzz> anyway I usually write erlang so I am looking for any oppurtunity to bash elixir, but you are making it harder and harder ;)
03:22 <sevenseacat> one day I'll set aside some time to look at erlang
03:23 <sevenseacat> it fits in with my 'learn things that start with E' philosophy (only half j/k)
03:25 rschmukler joined
03:25 <cmk_zzz> sevenseacat: how far have you come? elm?
03:41 BeerLover joined
03:43 racycle joined
03:47 shankardevy joined
03:49 <BeerLover> which files are in scope so i can import/use modules from them?
03:49 <BeerLover> and where can we modify this?
04:00 nomicflux joined
04:00 lessless joined
04:03 shankardevy joined
04:15 io_bora joined
04:19 <Nicd-> all the files in the project and the dependencies
04:20 aeliton joined
04:21 <Nicd-> the paths are defined in mix.exs in elixirc_paths
04:23 <aeliton> hi all, does anybody knows how can I make my phoenix application to print all logs to a file? I'm using exrm and systemd to run the application, but I can't see what is causing http 500.
04:24 <Nicd-> by default it should print error stack traces to the console
04:25 <Nicd-> if you are running it as a systemd service the output should be in journalctl
04:27 <aeliton> Thanks Nicd, I have never heard about this journalctl before, I take a look on it!
04:28 <Nicd-> it's systemd's logging system
04:28 <Nicd-> you can read logs from all the services
04:29 shankardevy joined
04:55 gvaughn joined
04:58 m_m joined
05:05 dec0n joined
05:08 bungoman joined
05:09 io_bora joined
05:15 dec0n joined
05:15 cryo28 joined
05:19 jkreeftmeijer joined
05:23 meandi_2 joined
05:25 m_m joined
05:42 rschmukler joined
05:46 josevalim joined
05:46 icanhazbroccoli joined
05:47 <BeerLover> Nicd- If i manually want to add path then?
05:48 <BeerLover> without making mix.exs
05:49 <BeerLover> elixirc_paths: elixirc_paths(Mix.env)
05:49 <BeerLover> Where is Mix.env?
05:50 Duckily joined
05:50 <Nicd-> Mix.env() returns the current Mix environment
05:50 <Nicd-> if you want to compile manually, call elixirc with all the files you want to compile
05:51 <BeerLover> Nicd- can you explain with a simple example
05:51 <BeerLover> lets say I have A.ex with a module
05:51 <BeerLover> and B.ex with a struct definition
05:52 <BeerLover> how to use the struct in A.ex
05:52 <BeerLover> to patternmatch and what not
05:52 <Nicd-> I haven't used elixirc directly myself but looking at the help it would work like: $ elixirc -o _build path/to/files/*.ex another/path/*.ex some_file.ex etc etc
05:53 <Nicd-> if you have a module B with a defstruct, you use that like %B{} in other modules
05:53 smferris joined
05:54 <BeerLover> but i get error that %Identicon.Image is not defined
05:54 <BeerLover> somehting like that
05:54 <BeerLover> i run iex main.ex
05:55 <Nicd-> is this a project with a mix.exs?
05:55 <sevenseacat> and if not, why not? <_<
05:55 <BeerLover> no
05:55 <BeerLover> 2 files
05:55 <Nicd-> you are only giving iex the file main.ex, it won't compile any other files
05:56 <BeerLover> ohh wait
05:56 <BeerLover> herein lies the issue
05:56 <BeerLover> Nicd- one more thing, where is elixirc_paths/1 defined?
05:57 <Nicd-> it's in mix.exs if you create a project with Mix
05:57 <BeerLover> i see a function call to elixirc_paths(Mix.env) in my mix.exs
05:57 <Nicd-> sooo you have one?
05:57 <Nicd-> I thought you didn't
05:57 <BeerLover> this is a separate proj
05:57 <BeerLover> that was something different
05:58 <Nicd-> out of interest, why are you not using a Mix project?
05:59 <Nicd-> I don't know how to do it outside of a Mix project, I guess you need to compile them all with elixirc and run that output somehow, or jump into iex and compile all the files there before calling stuff
05:59 <BeerLover> Nicd- yes. I was just wondering that
05:59 <BeerLover> that wasn't working for me
06:00 <BeerLover> hence asked here
06:00 <Nicd-> what specifically?
06:03 <BeerLover> is it possible to get just some variables from .exs to other .exs?
06:04 <Nicd-> what kind of variables? what are you doing?
06:06 robinsjdotcom joined
06:11 imack joined
06:16 luke_nukem joined
06:20 nd__ joined
06:21 io_bora joined
06:24 shankardevy joined
06:25 nd___ joined
06:25 shankardevy joined
06:28 jkreeftmeijer joined
06:31 rschmukler joined
06:34 Guest67_ joined
06:35 <evax> Hi, we're using an umbrella project as a mono-repository, cutting releases for different apps within in. The way config files are handled - using the umbrella's config including in turn all the apps configs - makes it hard to have dep specific configurations (e.g. different logger config for different apps) because I guess due to the wildcard include the apps config are included in alphabetical order. Is
06:35 <evax> there a way we could get the target app's config to be included last so that we have a chance to finetune?
06:36 jkreeftm_ joined
06:38 refriedchicken joined
06:39 dignifiedquire joined
06:40 <evax> The issue is not for releases (using conform makes eveything tunable) but for development mode
06:42 jkreeftmeijer joined
06:43 gvaughn joined
06:46 jkreeftmeijer joined
06:46 milad joined
06:59 tuacker joined
07:01 <evax> I could hack something:
07:01 <evax> app = Mix.ProjectStack.peek()[:config][:app]
07:01 <evax> import_config "../apps/#{app}/config/config.exs"
07:02 <evax> after the wildcard include config
07:02 <josevalim> evax: get rid of the wildcard and then list them all explicitly
07:02 <josevalim> you don't want to do that because 1. Mix.PorjectStack is not public API
07:02 <josevalim> 2. it means each project is going to have a different configuration which may lead to weird testing behaviour
07:03 <josevalim> if you really want to have different configs, then hvae different configuration directories
07:03 <josevalim> like config1/config.exs, config2/config.exs, config3/config.exs
07:03 <josevalim> and pick accordingly
07:06 squallstter joined
07:10 cemilowski joined
07:12 shankardevy joined
07:13 ensyde joined
07:13 nd__ joined
07:15 milad joined
07:16 <evax> josevalim: I see, this makes sense, thanks
07:19 <the_voice_> I am trying to find a good example of using belongs_to and put_assoc in phoenix 1.3
07:20 ejpcmac joined
07:20 <the_voice_> basically inside my Accounts context I have 2 schemas, user and auth_token schemas in which auth_token references accounts_users
07:24 <the_voice_> Is it still identical to how it works in 1.2 with only the addition of the accounts_ prefix to users when using the generator (as the generator does not create any references in the schema at the moment)
07:24 m_m joined
07:24 Guest67_ joined
07:30 imack joined
07:31 flaviodesousa joined
07:33 ZippoWeb joined
07:37 mattyw joined
07:39 <ivan> something in elixir regressed mix test --trace + async test cases recently; output is all mixed up
07:40 m_m joined
07:45 chuff joined
07:45 <OliverMT> shouldnt async be all mixed up by definiotion?
07:45 <ivan> OliverMT: no, when using --trace it's supposed to restrict it one test case at a time
07:45 <OliverMT> oh
07:46 <ivan> I've found a good commit so I'll file a bug once I finish bisecting it
07:49 tuacker1 joined
07:52 <Nicd-> hey OliverMT my near favorite Norwegian, have you been to Trolltunga/Kjeragbolten?
07:52 <OliverMT> No and No, but I know them well
07:53 <OliverMT> I grew up right next to geiranger though
07:53 <OliverMT> so I do have some norwegian nature cred
07:53 <OliverMT> :D
07:53 <Nicd-> I'm going in the summer
07:53 <OliverMT> ohh nice
07:53 <OliverMT> trolltunga is a long ass walk
07:53 <Nicd-> was wondering if I'll die during the hike
07:53 <OliverMT> news is filled with tourists getting stuck in the dark with too little clothes
07:53 <OliverMT> so bring soft shell jackets, proper shoes and pants
07:53 <OliverMT> kjeragbolten is like 45min walk from the car and up
07:53 <Nicd-> and a tent is the plan
07:54 <OliverMT> I think trolltunga is camping restricted, but not sure
07:54 DeadTrickster joined
07:54 <OliverMT> the walk isnt THAT long
07:54 <OliverMT> the newsworthy people are often chinese or other asians
07:54 <OliverMT> going up in converse shoes
07:54 <OliverMT> and maybe a light tshirt
07:54 <OliverMT> one woman had to be carried down after she went up in pumps (heels)
07:54 <Nicd-> I read to prepare for 11 hours (6 hours up, 1 hour there, 4 down)
07:54 <Nicd-> whaat
07:55 <OliverMT> that sounds about right
07:55 <OliverMT> bitwalker: ping
07:55 <Nicd-> here's the rough plan: https://trips.furkot.com/board/cDcquT
07:56 <Nicd-> the stops are autogenerated by the planner though
07:57 stephen_m joined
07:58 <OliverMT> https://usercontent.irccloud-cdn.com/file/8kGkjGva/where%20I%20live
07:59 <Nicd-> what's there? you?
07:59 <OliverMT> how many are you?
07:59 <OliverMT> thats me yes
07:59 <Nicd-> 2 or 3
07:59 <OliverMT> Måløy
07:59 <Nicd-> I think I've expressed my jealousness before
08:01 <Nicd-> anyway I'm so looking forward to it
08:01 <Nicd-> and kvikk lunsj, that's good shit <_<
08:02 <OliverMT> :D
08:03 <smeevil_> Morning :)
08:07 <OliverMT> why do you spend time going south?
08:08 <OliverMT> btw Nicd-
08:08 <OliverMT> the spectacular stuff in norway is all west and northwest
08:08 <OliverMT> the coast from bergen and south is very flat, except for the one spot in kjerag
08:10 Ioyrie joined
08:13 io_bora joined
08:20 robinsjdotcom joined
08:22 mattyw joined
08:24 akeating joined
08:25 vmoravec joined
08:26 <OliverMT> I wonder how Contexts in phoenix will be viewed in a year from now
08:26 <OliverMT> my first impression is that a lot of users get very confused
08:26 <OliverMT> what do they put in there, when?
08:26 <OliverMT> how often does a schema need its own context blabla
08:26 <sevenseacat> +1
08:27 <OliverMT> I don't really like neither the naming nor the way the generators make a default context
08:27 <OliverMT> is Context a prior existing naming? I havent heard it before phoenix introduced it
08:27 <OliverMT> it is sorta like a service, but not quite
08:28 <OliverMT> I think it was maybe a step too far to stop people putting code in their controllers
08:29 <micmus> The problem is that you have to think about it - people are lazy and don't like thinking
08:29 <OliverMT> I disagree
08:29 marr joined
08:30 <OliverMT> I have been enforcing a service layer based model since day one
08:30 <OliverMT> adopting more or less the spring IoC
08:30 <Nicd-> OliverMT: after Kjerag it's just the fastest way back to Finland
08:30 <OliverMT> from java
08:30 <OliverMT> oh... I read your trip the other way
08:30 <sevenseacat> people do like being told exactly what to do
08:30 <OliverMT> clockwise
08:30 <micmus> Oh, of course. My apps have a sevice/function wrapper layer too
08:30 <Nicd-> heh :)
08:30 <Nicd-> it's a matter of how far we can drive in the time
08:30 <OliverMT> anyways, I use a service layer for all things
08:30 <OliverMT> and I am still a bit confused about contexts
08:31 <OliverMT> it is just foreign naming and a bit unclear on what chris really thinks is the best practice in general
08:31 <micmus> But figuring out the name of the context requires significantly more brain cycles than running the odl old generator.
08:31 <whodidth1s> contexts are for minimizing the available fields, virtual or not, when using schemas
08:31 <OliverMT> we have our own generators now, that generates an anemic controller + a service with all the repo code
08:31 <OliverMT> repo is never used in a controller
08:32 <OliverMT> whodidth1s: so sort of like a two stage changeset
08:32 <OliverMT> one "outer" and one inner for the repo?
08:32 <micmus> likewise, except I don't have generators. Ctrl-c + Ctrl-p FTW!
08:32 <Nicd-> ctrl-p?
08:32 <whodidth1s> just have separate schemas
08:32 <micmus> v
08:32 <OliverMT> we have such an abundance of Acl stuff
08:32 <micmus> whatever :P
08:32 <OliverMT> that varies based on what the controller should handle
08:32 <OliverMT> so its nice to have a genny
08:35 io_bora joined
08:35 <OliverMT> quotes like this from the forum kinda ironically underlines what I am trying to say
08:35 <OliverMT> "Phoenix is not the foundation to the application. It's the foundation of the web layer, which is part of your application. E.g. the website I talked about above. All it's functionality is accessable through the actual website powered by phoenix, but also via iex bypassing everything phoenix is doing. It's an elixir app, which includes phoenix to make
08:35 <OliverMT> serving web requests easy. And it's not like phoenix would offer anything more I couldn't use because of that."
08:36 <OliverMT> The *vast* majority of people starting an app doing mix phoenix.new
08:36 <OliverMT> has web as their overshadowing main concern
08:36 <OliverMT> so trying to argue you're not building a phoenix application is so weird to me
08:36 <OliverMT> the web parts of it will in almost all cases drive most of your design decisions further down in the app, like schemas, integrations and whatnot
08:37 <OliverMT> if thats not the case, then you are already modular and might as well slap phoenix in an umbrella alongside what other stuff you have
08:37 <OliverMT> further invalidating the need for this context stuff
08:37 <sevenseacat> I do like the idea of the web app (controllers, views, templates) being one app in an umbrella.
08:38 <OliverMT> you need a pretty damn huge codebase before that becomes a pressing concern imo
08:38 <OliverMT> we have four major phoenix apps in prod for about a year now
08:38 <OliverMT> and not a single one of them is close to requiring umbrellas
08:39 <sevenseacat> I have none in prod :) so I'll believe you
08:41 <micmus> Speaking about umbrellas and separate applications in context of isolating functionality, is a bit of a misconception, IMO. You can do exactly the same with modules/parts in a regular app. Separate applications allow for separate supervision trees - that's most important.
08:41 <Nicd-> I have two but they are a mess. I was thinking of refactoring them to the umbrella style but now I don't know what to do :(
08:41 <Nicd-> it would be much easier if I didn't have to think myslef
08:41 <OliverMT> micmus: that as well
08:44 cemilowski left
08:46 shankardevy joined
08:48 vendethiel joined
08:50 gvaughn joined
08:51 hipertracker joined
08:54 io_bora joined
08:57 lexmag joined
09:01 rschmukler joined
09:07 kgge joined
09:07 pHcF joined
09:09 kgge left
09:09 <OliverMT> tristan__: do you guys do one kubernetes cluster per app, or do you mix all your apps into one big cluster? I am leaning towards the latter
09:11 io_bora joined
09:13 cemilowski joined
09:17 zv joined
09:26 <milad> Have anyone experience that Ecto preload doesn’t work when it’s called from another app in umbrella project?!
09:26 <milad> This the `Ecto.Query` but when I load the data relations are not preloaded: https://gist.github.com/slashmili/a1b46ff5a0a6ad594b3da8b95096e450
09:26 <milad> I call it from another app which handles the web requests
09:29 cemilowski left
09:32 dani0_ joined
09:35 gmcabrita joined
09:37 lessless joined
09:40 the_voice- joined
09:45 cemilowski joined
09:53 hipertracker joined
10:11 imack joined
10:13 elasticdog joined
10:34 DeadTrickster joined
10:34 hylh joined
10:38 gvaughn joined
10:46 shankardevy joined
10:49 rschmukler joined
10:58 Or1on joined
11:03 robinsjdotcom joined
11:06 cemilowski joined
11:08 io_bora joined
11:09 Guest67_ joined
11:13 shankardevy joined
11:19 mattyw joined
11:20 drewolson joined
11:23 Gasher joined
11:29 shankardevy joined
11:44 nomicflux joined
11:46 shankardevy joined
11:48 renl joined
11:49 <renl> hi for ecto schema I want to have a string field that is longer than 255 chars how do i do that?
11:49 <OliverMT> add :fieldname, :text
11:49 <OliverMT> assuming postgres
11:49 <OliverMT> whats the mysql one? :longtext?
11:50 akeating joined
11:50 <renl> oh im using postgres, k thanks :D
11:51 travis-ci joined
11:51 <travis-ci> elixir-lang/elixir#14129 (master - ab4805c : Adrian Mugnolo): The build passed.
11:51 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/62452ee7085b...ab4805c766bd
11:51 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/217139597
11:51 travis-ci left
11:52 nighty- joined
11:53 hipertracker joined
11:54 io_bora joined
11:54 dimitarvp joined
11:55 cristianR_ joined
11:58 io_bora joined
12:00 jerel joined
12:01 ctp joined
12:02 shankardevy joined
12:05 <Nicd-> it's TEXT in mysql too
12:06 <Nicd-> difference is that in postgres many people advocate for using TEXT in place of varchar too :P
12:07 nomicflux joined
12:15 hellofunk joined
12:15 <renl> close
12:15 renl left
12:16 io_bora joined
12:17 shankardevy joined
12:17 <OliverMT> not many, everyone who has a clue
12:17 <OliverMT> the reason for that is that its the same underlying type
12:17 <OliverMT> varchar just imposes a limit
12:17 <OliverMT> in mysql its completely different data types
12:18 <OliverMT> or it was before, havent used mysql in 5y aers
12:18 cemilowski left
12:19 <hmans> Ecto question: if I have a list of IDs, what's the canonical way of getting multiple records by these IDs? (ie., the equivalent to Repo.get(id), but with more than 1 ID)
12:19 <ciawal> does anyone know how much would be involved in renaming an app? I assume just renaming the app module and some mix file/config/supervisor changes?
12:19 <hmans> Ah, from(i in Item, where: i.id in ^ids) seems to do it
12:20 <hmans> but is there any other more immediate way?
12:20 <ciawal> there's no built-in helper for this AFAIK hmans
12:21 <hmans> Roger that, thanks
12:26 akeating joined
12:27 gvaughn joined
12:31 josevalim joined
12:31 <asonge> ciawal: are you just renaming the app name, or also the app module name as well?
12:32 <ciawal> I'd want to keep them in sync
12:34 <asonge> well, in elixir they're almost never the same thing. the app is :foo, where the app module is Foo
12:34 <asonge> but aside from all the stuff in lib/, just making sure to adjust mix.exs (include the mix file's module name) should be enough.
12:36 meh` joined
12:38 rschmukler joined
12:42 dani0_ joined
12:48 arquebus joined
12:53 jerel joined
12:54 Or1on joined
12:56 mitchellhenke joined
12:58 akeating joined
13:03 bambanx joined
13:09 ikcerog joined
13:11 mitchellhenke joined
13:12 tuacker joined
13:16 sfbw joined
13:16 meh`_ joined
13:19 mattyw joined
13:20 chef joined
13:21 dani0_ joined
13:26 chuff joined
13:27 Ven joined
13:29 robinsjdotcom joined
13:31 <hmans> OMG, I successfully contributed to Ecto: https://github.com/elixir-ecto/ecto/pull/2015#event-1024082489
13:31 <hmans> I am not a complete failure etc.
13:31 <hmans> Thanks ericmj
13:33 Ven_ joined
13:33 jeregrine joined
13:41 harfangk joined
13:42 maxbeizer joined
13:44 duane joined
13:45 aeliton joined
13:50 chrismccord joined
13:51 <cristianR_> Hi
13:54 akeating joined
13:55 <cristianR_> I'm trying to connect to an activemq stomp broker from elixir. I've found that it uses ssl for securing the connection.
13:56 <cristianR_> I googled a bit and I found that in order to use ssl I need to use the :ssl erlang app
13:57 <cristianR_> however, using the ssl app requires using its custom API, meaning that I have to change all the stomp client library gen_tcp calls
13:58 dani0_ joined
13:58 <cristianR_> Is there a simpler way of using SSL from elixir?
13:59 <asonge> cristianR_: use a library that extracts the differences?
14:00 <asonge> which means you need to change all the gen_tcp calls...
14:00 gvaughn joined
14:00 <cristianR_> what do you mean by "extracts the differences"? I'm not quite sure if I am following you
14:02 <asonge> cristianR_: iirc, most of the ssl functions are really similar to the gen_tcp ones, and you can use an alias to swap between them
14:02 jeregrine joined
14:03 <cristianR_> asonge: can I alias an erlang module from elixir?
14:04 <asonge> yeah, alias :foo, as: Bar
14:04 duane_ joined
14:06 mitchellhenke joined
14:10 griffinbyatt joined
14:14 <hmans> Ecto.Schema question: is it possible to create a datetime column (that's not created_at or updated_at, es provided by timestamps() that is automatically set to the current time, without doing this in a changeset?
14:14 <hmans> I'm fine with doing it within a changeset, I was just wondering. autogenerate: true looked like a condidate, but the DateTime type doesn't support it
14:15 chavito joined
14:17 chriswarren joined
14:20 <ciawal> does anyone know when ecto.model was removed?
14:21 dani0_ joined
14:22 <cristianR_> asonge: can the alias between :ssl and :gen_tcp be made customizable? in our testing environment we do not have SSL, just in production
14:27 <asonge> cristianR_: yeah, you can grab the module from a config variable during compile time
14:27 robinsjdotcom joined
14:28 milad joined
14:28 milad joined
14:29 <cristianR_> asonge: wonderful!
14:30 icapurro joined
14:30 s_kilk joined
14:35 chriswar_ joined
14:35 racycle joined
14:35 travis-ci joined
14:35 <travis-ci> elixir-lang/elixir#14130 (master - b06be1f : José Valim): The build passed.
14:35 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/ab4805c766bd...b06be1f82141
14:35 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/217193909
14:35 travis-ci left
14:38 iam joined
14:38 robinsjdotcom joined
14:40 <Guest40344> test
14:48 wtetzner joined
14:54 dpehrson joined
14:54 <dpehrson> When using bitwalker's distillery+conform, should custom distillery commands be run with the proper conform config?
14:55 <dpehrson> I wrote a custom distillery command that does a database migration for ecto, but when I run it, the conform config is not applied and it fails due to a lack of a database config. Running the module/func the command is wrapping manually from a remote_console works fine.
14:55 <dpehrson> I'm not sure whether I did something wrong, or if this is not supported.
14:56 chuff joined
14:58 <bcardarella> in an app I'm working on there are migrations with `add :slug, :string, unique: true` but I don't see any evidence in the Ecto docs that `add` supports create unique indexes. Everything I'm reading requires this to be a separate function call. Did this API ever exist? Maybe < 2.x?
15:00 fantomik joined
15:01 <bcardarella> doesn't seem like it: https://hexdocs.pm/ecto/1.1.9/Ecto.Migration.html#add/3
15:02 MaSven joined
15:02 <ciawal> I don't believe so
15:06 tomku joined
15:09 <meh`_> does anyone have any advice for a library to deal with paypal payments (and possibly others)?
15:09 <meh`_> all I can find seems to be WIP one way or the other
15:09 <OliverMT> Stripe meg
15:09 <OliverMT> Meh
15:09 <OliverMT> Has a good lib for elixir
15:10 <OliverMT> meh`_: ^
15:10 <meh`_> OliverMT, I need PayPal sadly
15:13 dani0_ joined
15:16 akeating joined
15:19 rschmukler joined
15:28 <micmus> Nice, finally replaced exrm with distillery and the CI build time went a minute down :)
15:31 tuacker joined
15:32 josevalim joined
15:33 raycoll joined
15:34 <dpehrson> Woo, I like distillery.
15:34 <dpehrson> micmus, is there any chance you've used distillery custom commands with conform?
15:34 miwa_ joined
15:34 robinsjdotcom joined
15:34 <ciawal> I’m trying to test insert a record with a specific insertion date like so:
15:34 <ciawal> one_day_ago = Timex.now |> Timex.shift(days: -1) |> Timex.to_erl |> NaiveDateTime.from_erl!
15:35 <ciawal> one_day_ago_record = Repo.insert! %Record{inserted_at: one_day_ago}
15:35 <ciawal> when I fetch the same record to use in an assertion though it fails due to a microsecond difference, is there a correct way to handle this?
15:35 <micmus> dpehrson: no, not really. But I think there were some issues on the distillery repo for commands & confirm, maybe you'll find something useful there
15:35 <dpehrson> Yeah I dug through, didn't find any answers :(
15:40 <dpehrson> I mainly just want to know if it SHOULD use the conform config so I know whether or not I'm wasting my time.
15:40 squallstter joined
15:43 milad joined
15:43 milad joined
15:45 hipertracker joined
15:47 dimitarvp` joined
15:53 jeregrine joined
15:54 ikcerog joined
16:00 tomaz_b joined
16:00 amclain joined
16:01 racycle joined
16:04 ctp joined
16:07 ejpcmac left
16:14 marcdel joined
16:16 jeffweiss joined
16:19 miwa joined
16:19 josevalim joined
16:30 gvaughn joined
16:32 gvaughn_ joined
16:34 chrismccord joined
16:37 Or1on joined
16:38 akeating joined
16:39 maxbeizer joined
16:41 thistle joined
16:43 hipertracker joined
16:43 davidmn joined
16:46 <Nicd-> is Calecto deprecated?
16:47 refriedchicken joined
16:47 <Nicd-> apparently at least Calecto.DateTimeUTC is
16:50 <dpehrson> I was able to remove it from my app.
16:51 <dpehrson> Once they rolled the new date/time stuff into the standard lib, I no longer needed it.
16:53 montanonic joined
17:07 <Ankhers> I'm trying to deploy my Phoenix application. Things worked properly when it was 1.2. Now that I upgraded to 1.3, my assets do not seem to be compiled properly. Things work in development, but in production mode a decent amount of my javascript doesn't seem to work. Unfortunately, I am not getting any console errors either.
17:07 <Ankhers> Would anyone have any ideas?
17:13 gvaughn_ joined
17:14 <dpehrson> Shot in the dark: I know the name of the manifest file changed in 1.3 if you are using digesting.
17:14 <dpehrson> I haven't yet upgraded to 1.3, but I remember seeing that changed.
17:15 InternetFriend joined
17:17 imack joined
17:18 <gazler> Ankhers: Have you updated to the latest version of the buildpack? Is that a thing? https://github.com/gjaldon/heroku-buildpack-phoenix-static/commit/d4c042833c0e706ca3235b55b9e1986108e196a7
17:18 <Ankhers> Not using Heroku, so no.
17:18 rschmukler joined
17:18 <gazler> Oh - that's weird. I was sure I read Heroku.
17:19 <gazler> Are you running `mix phx.digest`?
17:19 <Ankhers> gazler: yes.
17:19 <Ankhers> dpehrson: manifest file?
17:19 <gazler> Are you using releases?
17:19 <Ankhers> no
17:19 <dpehrson> Ankhers: If you use the phoenix feature to digest your assets, it has to write a json file out that maps original filenames to hashed filenames.
17:20 <dpehrson> The filename of that manifest file changed for 1.3
17:20 <dpehrson> Just an off-the-top-of-my-head potential problem.
17:20 <gazler> The assets location has changed too.
17:21 <gazler> So you will need to `cd assets` before running brunch
17:21 <Ankhers> I did that.
17:21 <gazler> Are the files in your priv directory?
17:22 chriswar_ joined
17:22 chriswar_ joined
17:23 ZippoWeb joined
17:24 <Ankhers> There are files there that have been modified today.
17:28 duane joined
17:29 rschmukler joined
17:30 mary33 joined
17:30 <mary33> http://bit.do/djVWe
17:30 mary33 left
17:44 <Ankhers> So my CSS seems to be fine. It appears that my JavaScript isn't working though.
17:47 hashpuppy joined
17:53 akeating joined
17:54 lexmag joined
17:54 chrismccord joined
17:55 <Ankhers> dpeherson was right. It was the manifest file name change. I do not recall the being in the migration guide.
17:55 <Ankhers> gazler: Thanks for the help!
17:56 <Ankhers> Well, it appears it is in the migration guide now.
17:57 imack joined
17:57 hellofunk joined
17:58 akeating joined
18:05 cristianR_ joined
18:07 thistle joined
18:07 Duckily joined
18:10 rschmukler joined
18:14 chriswarren joined
18:20 rschmukler joined
18:35 rpip joined
18:36 chavito joined
18:42 chuff joined
18:43 Duckily joined
18:49 marr joined
18:50 ejpcmac joined
18:51 <cristianR_> asonge: wrapping tcp_gen in a module and delegating the functions to :ssl (or optionally :tcp_gen itself) did the trick!
18:54 <cristianR_> asonge: just a follow up question, the defdelegate targets a module that is set in a configuration value
18:55 <cristianR_> because this is a dependency of the project, when I change the configuration on the project it won't recompile it and the delegates might end up pointing to the wrong implementation
18:56 <cristianR_> What would be the proper approach to avoid such issues?
18:57 nd___ joined
18:59 hipertracker joined
18:59 kyrix joined
19:08 thistle joined
19:09 fernandomm joined
19:13 tomaz_b joined
19:13 hipertracker joined
19:15 ejpcmac joined
19:16 <micmus> I got strange errors in my app from logger - "no device registered with the name :user"
19:16 m_m joined
19:16 <micmus> (they went away on their own after some time, but still, I would expect the :user process to always be there)
19:19 sfbw joined
19:19 milad joined
19:28 hipertracker joined
19:29 nd__ joined
19:35 WhiskyRyan joined
19:36 griffinbyatt joined
19:40 chuff joined
19:47 luke_nukem joined
19:49 DTZUZO joined
19:50 dani0_ joined
19:54 hipertracker joined
19:54 proteus-guy joined
19:58 hellofunk joined
20:01 Duckily joined
20:05 milad joined
20:05 DLSteve joined
20:09 thistle joined
20:10 aeliton joined
20:15 hahuang61 joined
20:19 rschmukler joined
20:24 tolerablyjake joined
20:29 wsieroci joined
20:29 chavito joined
20:34 cemilowski joined
20:35 thistle joined
20:35 chriswar_ joined
20:35 rschmukler joined
20:42 bitshift joined
20:44 tomaz_b joined
20:52 klltkr joined
20:53 rschmukler joined
21:03 rschmukler joined
21:08 DTZUZO joined
21:17 rschmukler joined
21:18 jkreeftmeijer joined
21:20 bitshift_ joined
21:30 wsieroci joined
21:34 bitshift joined
21:39 cdg joined
21:48 rschmukler joined
21:48 thistle joined
21:49 thistle joined
21:51 chrismccord joined
21:57 MotherFlojo joined
21:57 InternetFriend joined
21:58 hellofunk joined
21:59 rschmukler joined
21:59 <Nicd-> soo, if I need a datetime as both UTC and the original local time, it would probably be best to store them both in separate columns (in Postgres)?
21:59 <Nicd-> instead of storing the UTC version and an offset/interval
22:04 stephen_m joined
22:06 io_bora joined
22:07 Duckily joined
22:08 chriswarren joined
22:09 chriswarren joined
22:10 chriswarren joined
22:11 chriswarren joined
22:11 chriswarren joined
22:12 chriswarren joined
22:13 chriswarren joined
22:14 chriswarren joined
22:15 chriswarren joined
22:24 <griffinbyatt> Hey, anyone know if rrrene of https://github.com/rrrene/elixirstatus-web is ever on irc or slack?
22:24 icapurro_ joined
22:28 DTZUZO joined
22:29 sfaxon joined
22:31 sfaxon joined
22:43 gmcabrita joined
22:57 <zv> Does anyone know if it's possible to use `proc_lib:start_links` within Elixir behaviors?
22:58 <zv> I'm trying to port one now and for whatever reason it isn't being added to the supervision tree and :erlang.register() is throwing an error
23:01 laut joined
23:05 SvenMa joined
23:06 mitchellhenke joined
23:06 hahuang61 joined
23:16 chriswarren joined
23:16 chriswar_ joined
23:19 BenMorganIO joined
23:19 <BenMorganIO> Hey guys, is there a way to do a dynamic `embeds_one`
23:19 <BenMorganIO> ?
23:20 <BenMorganIO> I'm wondering because I want to have an embeds_one whose module is dependent upon a `type` column.
23:20 Duckily joined
23:22 chriswarren joined
23:22 <BenMorganIO> https://gist.github.com/BenMorganIO/d541d086dd8043d7d26c8db05e585a23
23:24 nahtnam joined
23:25 meh` joined
23:28 <BenMorganIO> ** (ArgumentError) field/association :data is already set on schema
23:28 <BenMorganIO> lib/ecto/schema.ex:1697: Ecto.Schema.put_struct_field/3
23:28 leafs87 joined
23:28 <BenMorganIO> Can't do the latter...
23:28 <BenMorganIO> ** (CompileError) lib/ryal/web/models/payment_method.ex:16: undefined function embeds_one/1
23:28 <BenMorganIO> expanding macro: Ecto.Schema.schema/2
23:28 <BenMorganIO> And that's if I try to not specify a module.
23:28 leafs87 left
23:30 travis-ci joined
23:30 <travis-ci> elixir-lang/elixir#14133 (master - 7b009dd : Jon Parise): The build passed.
23:30 <travis-ci> Change view : https://github.com/elixir-lang/elixir/compare/b06be1f82141...7b009ddeed20
23:30 <travis-ci> Build details : https://travis-ci.org/elixir-lang/elixir/builds/217389127
23:30 travis-ci left
23:32 bitshift_ joined
23:45 jmiven joined
23:55 gvaughn joined
23:56 gvaughn_ joined
23:58 hellofunk joined
23:59 mitchellhenke joined