<< 24-07-2016 >>
00:26:57*chrisheller joined #nim
00:35:15*zodiak_ quit (Ping timeout: 264 seconds)
00:41:35*chrisheller quit (Remote host closed the connection)
00:42:10*chrisheller joined #nim
00:43:01*desophos joined #nim
00:46:28*chrisheller quit (Ping timeout: 252 seconds)
00:54:14*desophos quit (Read error: Connection reset by peer)
00:57:35*desophos joined #nim
01:10:21*abruanese joined #nim
01:16:36*desophos quit (Ping timeout: 250 seconds)
01:17:11*kulelu88 quit (Quit: Leaving)
01:21:39*arnetheduck joined #nim
01:21:48*chemist69_ quit (Ping timeout: 250 seconds)
01:33:59*chemist69_ joined #nim
01:34:04*kssreeram joined #nim
01:35:57*brechtm quit (Read error: Connection reset by peer)
01:36:51*brechtm joined #nim
01:38:24*josh joined #nim
01:39:05*josh is now known as Guest33000
01:39:12*Guest33000 quit (Client Quit)
02:09:36*Demon_Fox joined #nim
02:16:57*kssreeram quit (Quit: My Mac has gone to sleep. ZZZzzz…)
03:02:56*jeffc quit ()
03:13:28*desophos joined #nim
03:28:56*zodiak joined #nim
03:43:39*zodiak quit (Ping timeout: 260 seconds)
03:43:56*kssreeram joined #nim
03:45:56*kssreeram quit (Client Quit)
03:47:42*zodiak joined #nim
04:15:03*xet7__ joined #nim
04:16:01*xet7 quit (Read error: Connection reset by peer)
04:39:22*yglukhov joined #nim
04:44:30*yglukhov quit (Ping timeout: 276 seconds)
04:45:39*nsf quit (Quit: WeeChat 1.5)
05:09:53*kssreeram joined #nim
05:58:59*brechtm quit (Read error: Connection reset by peer)
05:59:06*brechtm_ joined #nim
06:19:48*xet7__ quit (Quit: Leaving)
06:20:32*xet7 joined #nim
06:41:07*brechtm_ quit (Read error: Connection reset by peer)
06:41:13*brechtm joined #nim
06:49:48*yglukhov joined #nim
06:54:30*yglukhov quit (Ping timeout: 276 seconds)
07:13:46*der-landgraf quit (Quit: WeeChat 1.5)
07:23:12*Varriount quit (Read error: Connection reset by peer)
07:25:39*yglukhov joined #nim
07:30:15*yglukhov quit (Ping timeout: 276 seconds)
07:37:35*yglukhov joined #nim
07:38:45*pafmaf joined #nim
07:47:06*fastrom joined #nim
08:00:57*yglukhov quit (Remote host closed the connection)
08:01:34*kssreeram quit (Quit: My Mac has gone to sleep. ZZZzzz…)
08:04:50*Matthias247 joined #nim
08:12:49*yglukhov joined #nim
08:17:42*yglukhov quit (Ping timeout: 276 seconds)
08:48:40*yglukhov joined #nim
08:53:27*yglukhov quit (Ping timeout: 276 seconds)
08:54:41*elrood joined #nim
08:57:52*zaquest quit (Remote host closed the connection)
09:01:55*zaquest joined #nim
09:14:04*Demon_Fox quit (Quit: Leaving)
09:21:19*yglukhov joined #nim
09:30:45*kssreeram joined #nim
09:41:58*GangstaCat joined #nim
09:46:22*der-landgraf joined #nim
10:08:39dom96Anybody worked with protobuf in Nim before?
10:09:30federico3dom96: looked at some binary protocol libs, ended up using jangko/msgpack4nim instead
10:12:17dom96federico3: pity. A protobuf wrapper would be really nice.
10:12:51dom96Right now I'm doing something very hackish: calling the `protoc` binary to encode/decode packets and then parsing its output.
10:16:19federico3dom96: sounds like you just volunteered for a protobuf wrapper
10:20:06dom96hah maybe
10:20:16*xet7_ quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com ))
10:20:46dom96The interesting thing about protobuf is that you can transform proto files into .c, .py etc. source files which contain type definitions which match the protocol.
10:21:17dom96I wonder how easy that would be to do.
10:21:43dom96Currently my program is crashing on `writeFile` after some time and I have no clue why
10:22:28dom96But other than that my status page is working well so far https://pokemongostatus.org/
10:24:20ofelasi wrote a .thrift parser in nim once to generate C, .proto to nim would be along the same lines, anything protobuf these days seems to be related to pokemon go
10:24:35ofelas8)
10:35:51*kssreeram quit (Ping timeout: 264 seconds)
10:38:24dom96ofelas: cool, is it on github anywhere?
10:40:39ofelasdom96: nope, unfortunately not, i may have it laying around somewhere, was just trying/learning Nim at the time
10:49:05dom96aww too bad
10:49:24federico3dom96: the code autogeneration works well, especially for java, but protobuf has the downside of unpacking entire messages and create object structures even for fields that you are never going to access
11:04:17kierDom96: check out cap'n proto - imo it's much nicer than protobuf and allows more idiomatic data structures
11:04:42dom96kier: I don't have a choice unfortunately
11:06:45kierAh too bad - cap'n proto is designed so that you only need to parse what you use (rather than parsing everything and copying it into another data structure)
11:07:45kierIt's been a while since I used protobuf, so not sure I can be much with that
11:22:41*yglukhov quit (Remote host closed the connection)
11:33:13*yglukhov joined #nim
11:44:29*arnetheduck quit (Ping timeout: 258 seconds)
11:44:52*arnetheduck joined #nim
11:50:44*brechtm quit (Read error: No route to host)
11:50:56*brechtm joined #nim
12:29:20federico3the unittest lib needs some more features
12:33:59*yglukhov quit (Remote host closed the connection)
12:39:16*kssreeram joined #nim
12:45:53*yglukhov joined #nim
12:50:23*yglukhov quit (Ping timeout: 244 seconds)
13:04:37dom96ah, of course, I'm leaking FDs
13:07:51dom96RAII sure would be nice in Nim.
13:10:24*yglukhov joined #nim
13:14:40*yglukhov quit (Ping timeout: 244 seconds)
13:15:24*kssreeram quit (Ping timeout: 276 seconds)
13:24:25elrooddepending on how literal your definition of raii is, that won't necessarily help with leaking, though
13:50:50*pafmaf quit (Ping timeout: 244 seconds)
14:15:14*arnetheduck quit (Remote host closed the connection)
14:17:46*Ven joined #nim
14:18:49*yglukhov joined #nim
14:49:01*rolha joined #nim
14:50:26rolhahi everyone. is it possible to have a proc reference in a Table? i.e. let table: Table[string, proc(x: string):string] ?
14:50:59rolhathen initialised as proc a(x:string):string = x ; { "x" : a }
15:00:28*rolha quit (Quit: Textual IRC Client: www.textualapp.com)
15:09:53*Jesin quit (Quit: Leaving)
15:12:03*Jesin joined #nim
15:30:05*Matthias247 quit (Read error: Connection reset by peer)
15:33:05kierrolha: that looks like it should work. Have you tried it?
16:08:17*rolha joined #nim
16:22:02*nsf joined #nim
16:38:51*Trustable joined #nim
16:50:19*kulelu88 joined #nim
17:00:53*reyDaJavaNub joined #nim
17:24:32*yglukhov quit (Remote host closed the connection)
17:24:54*reyDaJavaNub quit (Ping timeout: 250 seconds)
17:34:16kieris there a straightforward way to dump the AST of a source code file after templates have been expanded during the compilation process?
17:36:24*yglukhov joined #nim
17:36:52*brechtm_ joined #nim
17:38:15*brechtm quit (Ping timeout: 264 seconds)
17:40:45*yglukhov quit (Ping timeout: 244 seconds)
17:53:53*yglukhov joined #nim
17:58:48*yglukhov quit (Ping timeout: 276 seconds)
18:16:00*kulelu88 quit (Quit: Leaving)
18:22:23*shadwick joined #nim
18:23:18shadwickHi all. I was trying to use some OO combined with generics but I hit a problem.
18:24:07shadwickI have some type, "Entity", which has several subtypes. I then also have a "Repository[T: Entity]" to be able to load/store these generically.
18:25:22Araqkier: no.
18:25:30Araqit's often requested though.
18:25:52Araqshadwick: and then?
18:25:56shadwickHowever, I wanted to use dynamic-dispatching with the repository to allow different implementations at runtime, meaning something like "SqliteRepository[T: Entity] = ref object of Repository[T]", but that I get "invalid type SqliteRepository[SqliteRepository.T]" from a var declaration for it
18:26:31shadwickAnd if I drop the ": Entity" from the Sqlite generic def. it appears to hang the compiler, but I am still looking into thaty
18:29:53avsejdom96, what do you think about this fix? https://github.com/nim-lang/nimforum/pull/89/files
18:30:11avsejit works for me, but I'm not sure about implementation
18:30:24avsejmaybe just replace <document> with <div> at all?
18:37:16Araqshadwick: "invalid type" usually means you try to use a generic type where the compiler cannot resolve it to a concrete type
18:37:40Araqand when you manage to hang the compiler, please create proper bug reports
18:38:08shadwickAraq: certainly, I'm trying to find the minimal case where that happened, and if I can get any more verbosity from Nim
18:39:09shadwickSo I might be wrong here, but I assume that if I have "SomeType[T: Blah]", then the "Blah" constraint also includes all types that are "= object of Blah"?
18:41:55Araqyes
18:47:51*yglukhov joined #nim
18:51:24*krux02 joined #nim
18:52:03*yglukhov quit (Ping timeout: 244 seconds)
18:52:13krux02can it be that the defer statement is broken? ``Error: internal error: expr(nkDefer); unknown node kind``
18:52:31*Sembei quit (Read error: No route to host)
18:58:06Araqkrux02: more likely is that your macro found a way to produce an AST that's not transformed properly
19:00:13krux02Araq: actually I get that error when I try to compile the example from the tutorial here: http://nim-lang.org/docs/manual.html#exception-handling-defer-statement
19:00:42krux02so no macro magic here
19:00:49AraqTop level defer statements are not supported since it's unclear what such a statement should refer to
19:01:01Araq^ insufficient error checking for this?
19:01:14krux02Ah, ok thank you
19:05:00shadwickI've produced a several line example of what I was attempting to do with the generics and subtypes here: http://ideone.com/8jlxYB
19:06:28shadwickI've come to find Nim a week ago so I'm sure I'm doing simple wrong
19:12:23*yglukhov joined #nim
19:16:51*yglukhov quit (Ping timeout: 244 seconds)
19:22:43krux02shadwick: so what are you trying to express with your code?
19:25:00krux02my knowledge so far of generics is, that you should not use constraints on generic types, because concepts are not yet ready
19:27:05*kssreeram joined #nim
19:27:35shadwickkrux02: Coming from Ada2012 or Java, I would consider an interface of "Repository" to read or write entity objects of T. I then want to provide several implementations of Repository. This is due to it changing via runtime configuration, etc.
19:28:12shadwickThe thing here is that I have a repository for domain type X and one for domain type Y, both of why are just "Repository[X]" and "Repository[Y]", where X and Y are both of type T
19:28:19krux02I don't know ada, but I do know java
19:28:29krux02generics do not work like they work in java
19:28:49*Matthias247 joined #nim
19:29:16krux02I think the nim compiler doesn't care weather X and Y are both of type T
19:29:33krux02you can just leave out the `` : Entity `` part
19:29:40shadwickkrux02: But then what is the point of specifying constraints like "Something[T: Entity]"?
19:30:29shadwickkrux02: The reason for that constraint/bound was to guarantee the T objects have a specific field that is common to all entities
19:30:37shadwickfor this DB layer
19:30:48krux02In my experience they have a very good purpose, but in my experience and what I've read in the forums, they are not a stable enough language feature to use them in real code
19:30:57krux02so for now, just leave them out
19:31:35krux02you should know that unlike java, in generics the nim compiler generatas code for every type you put in the generic type
19:32:04krux02java does not do that, java relies on the virtual dispatch of interfaces
19:32:48krux02yes, concepts are meant to give you that certainty, but sadly concept are not stable yet.
19:33:53*kulelu88 joined #nim
19:34:33shadwickkrux02: So that's the same as Ada generics. Because I'm defining an "interface" that actually requires a generic parameters, this might be the issue
19:35:14*kulelu88 quit (Max SendQ exceeded)
19:35:19shadwickIn Java, I could always pass a "Repository<? extends Entity>" around, but there is no such "type" in Ada if done this way since the Repository belongs inside a generic unit
19:35:40*kulelu88 joined #nim
19:35:58*cnu- joined #nim
19:36:02*kulelu88 quit (Max SendQ exceeded)
19:36:23*kulelu88 joined #nim
19:37:46krux02http://ideone.com/gu1405
19:37:52*yglukhov joined #nim
19:39:03shadwickkrux02: However, with that, you cannot define some proc elsewhere that is "proc whatever[R: Repository](repo: R): \ r.someMethod()" since SqliteRepository is not a Repository
19:39:22krux02shadwick: I removed all inheritance (in nim you don't need inheritance to use generics)
19:39:47shadwickunless the missing piece is that the generic checking here is like templates in C+
19:39:58krux02it is
19:40:17shadwickas-in, a method taking T and using "T.field" will not fail-typechecking unless a particular instantiation provides a T without ".field"
19:40:28krux02generic checking is like templates in c++, except that it does not support template specialization
19:40:40shadwicki.e. checking only the concrete expansions as opposed to the abstract
19:42:32krux02yes that's exactly the case, and concepts are there, to give you better error messages in case of generic parameters are not suitable
19:42:48*yglukhov quit (Ping timeout: 276 seconds)
19:43:01shadwickkrux02: Ok, thanks for the help. I will revisit how I'm designing this now with that info
19:51:59*yglukhov joined #nim
20:01:06*desophos quit (Ping timeout: 250 seconds)
20:05:41*gokr joined #nim
20:06:07*brechtm_ quit (Remote host closed the connection)
20:06:58*shadwick quit (Ping timeout: 250 seconds)
20:11:59*gokr quit (Ping timeout: 244 seconds)
20:12:44*brechtm joined #nim
20:27:26*gokr joined #nim
20:28:37*brechtm quit (Remote host closed the connection)
20:35:39*xet7_ joined #nim
20:40:24*brechtm joined #nim
20:42:31*gokr quit (Read error: No route to host)
20:45:47*brechtm quit (Remote host closed the connection)
20:46:03*brechtm joined #nim
20:50:31*Trustable quit (Remote host closed the connection)
20:58:34*krux02 quit (Quit: Verlassend)
21:06:15*brechtm quit (Remote host closed the connection)
21:07:54*roose joined #nim
21:08:23*brechtm joined #nim
21:19:10*krux02 joined #nim
21:20:05krux02I can't find the nim equivalent of std::numeric_limits, where is the maximum of an int32 etc defined?
21:23:54krux02and is there something I can use for a bitarray (I don't work on enums, something like c++ std::vector<bool> would be nice in this context)
21:28:08Araqintsets module comes close, I guess
21:28:17Araqhigh(int32)
21:29:29*brechtm_ joined #nim
21:31:54roosehi all
21:32:51*brechtm quit (Ping timeout: 264 seconds)
21:33:04rooseAraq you wrote "Python doesn't use OrderedDicts here, so no." about ordered json, but python have "object_hook" in json.load
21:33:39*brechtm_ quit (Ping timeout: 250 seconds)
21:34:28Araqroose: ok :-)
21:34:47rooseoh, *object_pairs_hook
21:36:18rooseand I can get ordered data with json.loads(file_name, object_pairs_hook=collections.OrderedDict)
21:37:02roosehow about make something like this in nim? )
21:37:08*fastrom quit (Quit: Leaving.)
21:40:54krux02Araq: thanks for high(int32) that works, but intset is not what I am looking for. I was looking for something that is guaranteed to not allocate, because I want to use it for an object pool of a particle system.
21:43:30AraqI fail to see the connection between "doesn't allocate" and "used for pooling"
21:46:32*kingofoz quit (Ping timeout: 260 seconds)
21:46:54krux02Araq: don't worry it's not that important for the moment. But `sparse` was the keyword is what i am explicitly not looking for. I was looking for something that is defined to be just one continious chunk of memory
21:47:06*kingofoz joined #nim
21:47:26Araqhey, but it's interesting
21:47:38Araqso what is the connection?
22:00:01krux02in memory pooling, or object pooling there is the requirement to create some sort of index to give out free objects. This index is usually a bit array every bit corresponds weather one object or one chunk of memory is currently in use. But when the memory pool itself might allocate when one object get's reserved, it kind of defeats the purpose of the whole pooling. And sparse datastructures do allocate on certain changes.
22:27:34Araqvector<bool> does reallocate ...
22:34:42rooseok, if ordered JSON is impossible =), how to compare two Ordered Table?
22:39:44*fvs joined #nim
22:41:03*elrood quit (Quit: Leaving)
22:42:47fvshi, it is possbile to populate an array using a proc before compilation? I want my code to use a fixed array instead of a seq and I'm too lazy filling it myself.
22:44:06*Ven quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
22:51:45Araqfvs: sure, use a compile-time proc or a macro
22:51:56Araqroose: we'll get ordered json
22:58:19rooseAraq: sorry for my english and i newbie to nim, i wanted copy json.nim and change Table to OrderedTable
23:00:45*Matthias247 quit (Read error: Connection reset by peer)
23:02:32roosetables.nim have comparsion procedure for Table but none for OrderedTable
23:04:43*kssreeram quit (Quit: My Mac has gone to sleep. ZZZzzz…)
23:11:17*der-landgraf quit (Quit: WeeChat 1.5)
23:12:11Araqroose: PRs welcome
23:18:41rooseAraq: ok, I'll try
23:30:27*der-landgraf joined #nim
23:30:27*der-landgraf quit (Changing host)
23:30:27*der-landgraf joined #nim
23:30:42*fvs left #nim ("ERC (IRC client for Emacs 24.5.1)")
23:41:06*kulelu88 quit (Quit: Leaving)
23:52:21*der-landgraf quit (Quit: WeeChat 1.5)
23:53:58*der-landgraf joined #nim
23:53:59*der-landgraf quit (Changing host)
23:53:59*der-landgraf joined #nim
23:56:05*rolha quit (Read error: Connection reset by peer)
23:57:28*rolha joined #nim