<< 21-11-2015 >>
00:10:22*Matthias247 quit (Read error: Connection reset by peer)
00:12:20*cryzed quit (Ping timeout: 246 seconds)
00:15:38*makoLine joined #nim
00:24:29*cryzed joined #nim
00:37:09*desophos_ joined #nim
00:41:42*desophos_ quit (Ping timeout: 260 seconds)
00:46:15*yglukhov joined #nim
00:49:34*desophos joined #nim
00:50:29*yglukhov quit (Ping timeout: 246 seconds)
00:59:56*jaco60 quit (Ping timeout: 246 seconds)
01:09:54*Trustable quit (Remote host closed the connection)
01:31:54*AMorpork left #nim ("Leaving")
02:16:59*tomp joined #nim
02:17:10*strcmp4 quit (Ping timeout: 240 seconds)
02:18:56*delian66 joined #nim
02:18:57*ayeganov_ joined #nim
02:18:57*Xe` joined #nim
02:18:59*Xe quit (Quit: *.yolo *.swag)
02:19:02*ayeganov quit (Write error: Broken pipe)
02:19:04*tomes quit (Write error: Broken pipe)
02:19:05*delian66_ quit (Remote host closed the connection)
02:21:05*Xe` is now known as Xe
02:30:35*desophos_ joined #nim
02:35:17*desophos_ quit (Ping timeout: 272 seconds)
02:39:53*C10RD joined #nim
02:41:37*pregressive joined #nim
02:50:10*C10RD quit (Remote host closed the connection)
02:57:12*strcmp4 joined #nim
03:03:11*C10RD joined #nim
03:05:39*C10RD quit (Remote host closed the connection)
03:13:19*Guest48911 is now known as Guest48911isaway
03:24:59*C10RD joined #nim
03:30:16*C10RD quit (Quit: Bye)
03:30:40*C10RD joined #nim
03:30:55*C10RD quit (Client Quit)
03:48:53*brson quit (Quit: leaving)
04:10:01*xtagon joined #nim
04:18:10*thume joined #nim
04:23:58thumeSo I’m having a super weird issue with a nim program. It won’t let me pipe its stdout to anything.
04:24:25thumeUnlike every other program, doing ./server outputs to stdout but doing “./server | cat” gives no output whatsoever.
04:24:54thumeThis is an issue because it means that the systemd service for my server can’t get any logging data to put in journald
04:26:13thumeIt’s just a normal jester app that uses “echo”: https://github.com/trishume/ratewithscience/blob/master/server.nim
04:27:56thumeIt does work if I redirect to a file though like “./server > test.txt”
05:07:06*thume quit (Quit: thume)
05:09:53*thume joined #nim
05:10:23thumekk result is I gave up on piping the logs. My app now has no logging.
05:10:33*thume quit (Client Quit)
05:18:35*pregressive quit (Remote host closed the connection)
05:19:29*Varriount-Laptop joined #nim
05:36:30*darkf joined #nim
05:56:36*cyraxjoe_ joined #nim
05:57:34*cyraxjoe quit (Ping timeout: 276 seconds)
06:09:07*Varriount-Laptop quit (Ping timeout: 265 seconds)
06:20:52*xtagon quit ()
06:31:42*strcmp4 quit (Ping timeout: 272 seconds)
06:35:08*desophos quit (Read error: Connection reset by peer)
06:38:03*Demon_Fox joined #nim
06:52:02*strcmp4 joined #nim
07:01:28*BitPuffin|osx quit (Ping timeout: 272 seconds)
07:15:14*gour joined #nim
07:38:57*smodo joined #nim
07:55:28*Demon_Fox quit (Ping timeout: 250 seconds)
08:08:39*Demon_Fox joined #nim
08:30:52*darkf quit (Read error: Connection reset by peer)
08:31:17*darkf joined #nim
08:32:29*yglukhov joined #nim
08:51:51*yglukhov quit (Remote host closed the connection)
08:52:03*yglukhov joined #nim
09:01:30*makoLine quit (Ping timeout: 260 seconds)
09:09:19*Kingsquee quit (Quit: https://i.imgur.com/qicT3GK.gif)
09:23:12*Matthias247 joined #nim
09:37:55*yglukhov quit (Remote host closed the connection)
10:01:06*gokr joined #nim
10:22:00*gokr quit (Quit: Leaving.)
10:33:06Araqmeh, "not nil" as a default is such a pita
10:33:41Araqhow can I express "ptr Foo not nil if someCondition"
10:35:27*vendethiel joined #nim
10:40:39reactormonkAraq, what wou dbe someCondition?
10:41:02Araqsomething I cannot express via Nim's type system
10:41:11Araqlike
10:41:23Araqif t.kind == tyObject
10:42:08reactormonktime for a case object?
10:43:41*Trustable joined #nim
10:44:04Araqcan't, it's some really low level stuff
10:44:45reactormonkIntegrating with C or lowlevel hardware?
10:45:25*Matthias247 quit (Read error: Connection reset by peer)
10:48:21Araqwith my very own codegen which I'm not gonna change for this
10:51:34*gokr joined #nim
11:22:43dom96Araq: Sounds like you're setting yourself up for "No, this can't work"
11:23:14dom96If `not nil` is default then why would you need to express that?
11:35:10*gokr quit (Quit: Leaving.)
11:54:05*jaco60 joined #nim
11:57:17Araqdom96: because sometimes it can be nil, sometimes it can't
11:57:28Araqdepending on a condition that I cannot express in the type system
12:27:28dom96Araq: What is this condition?
12:27:48dom96if it can't be expressed can't you just revert back to the default behaviour?
12:29:43Araqwell I'm evaluating what the language looks like if 'not nil' were the default
12:29:57Araqso surely I can go back but that misses the point
12:42:19dom96Araq: Surely some types will still need to be marked `has nil`
12:42:29dom96The point is that most will not need to be
12:43:25*C10RD joined #nim
12:49:45*C10RD left #nim (#nim)
12:55:51dom96Found a really good talk about RNGs: https://www.youtube.com/watch?v=45Oet5qjlms
12:55:56strcmp4i think not nil will be a good idea
13:01:56*yglukhov joined #nim
13:30:46*Guest48911isaway is now known as Guest48911
13:32:02*Demon_Fox quit (Quit: Leaving)
13:52:01*smede joined #nim
13:52:51*smede quit (Client Quit)
13:53:00*zepolen joined #nim
13:59:02*vendethiel quit (Ping timeout: 250 seconds)
14:01:32*smede joined #nim
14:39:12*yglukhov quit (Remote host closed the connection)
14:53:44*vendethiel joined #nim
15:14:40*zepolen quit (Remote host closed the connection)
15:39:46*yglukhov joined #nim
15:46:19*Trustable quit (Remote host closed the connection)
16:13:38*rkj-b joined #nim
16:15:12*desophos joined #nim
16:19:43*pmbauer left #nim (#nim)
16:19:54*pmbauer joined #nim
16:41:54*strcmp5 joined #nim
16:44:15*strcmp4 quit (Ping timeout: 240 seconds)
16:49:50*zepolen joined #nim
16:54:16*zepolen_ joined #nim
16:55:21*zepolen quit (Ping timeout: 272 seconds)
17:02:29*zepolen joined #nim
17:02:49*rkj-b quit (Quit: ChatZilla 0.9.92 [Firefox 42.0/20151029151421])
17:03:45*zepolen_ quit (Ping timeout: 250 seconds)
17:04:05*aziz_ joined #nim
17:14:15*BitPuffin|osx joined #nim
17:15:40*pregressive joined #nim
17:18:13*mrkishi_ quit (Remote host closed the connection)
17:20:18*Trustable joined #nim
17:20:51*xet7 joined #nim
17:23:05BitPuffin|osxAraq: something like C++'s unique_ptr isn't implementable in nim, right?
17:36:28*zepolen quit (Remote host closed the connection)
17:36:48*Sornaensis quit (Excess Flood)
17:37:12*Sornaensis joined #nim
17:40:14*strcmp1 joined #nim
17:42:50*strcmp5 quit (Ping timeout: 276 seconds)
17:44:46*mikolalysenko quit (Ping timeout: 240 seconds)
17:44:47*Guest4360 quit (Ping timeout: 240 seconds)
17:45:06*Learath2 quit (Ping timeout: 240 seconds)
17:45:26*darkf quit (Ping timeout: 240 seconds)
17:46:52*Guest4360 joined #nim
17:47:28*mikolalysenko joined #nim
17:47:48*zepolen joined #nim
17:50:49*Learath2 joined #nim
17:54:11AraqBitPuffin|osx: it is once I finally fix destructors in Nim
17:54:54BitPuffin|osxAraq: ah, so you'll finally have deterministic destructors for value types going out of scope?
17:56:39BitPuffin|osxwhat about move semantics, I thought that's also a key recipe for unique_ptr
18:04:24*zepolen quit (Remote host closed the connection)
18:09:12*ephja joined #nim
18:11:30*smede quit (Ping timeout: 240 seconds)
18:12:31*smodo quit (Remote host closed the connection)
18:13:50*pregressive quit (Remote host closed the connection)
18:14:11*pregressive joined #nim
18:16:15AraqBitPuffin|osx: we have overloading of '=' too, override it to give it move semantics
18:16:59BitPuffin|osxwell move doesn't only happen when you assign with =
18:17:08BitPuffin|osxit can also happen when you pass an argument, or return a value
18:17:17BitPuffin|osxor would = be called then too?
18:23:55*zepolen joined #nim
18:26:12dom96Really wish somebody bothered to document why there is now a parseopt and parseopt2
18:26:55strcmp1more accomodating worldview
18:27:26strcmp1python philosophy
18:30:35dom96will we rename parseopt2 back to parseopt in the future?
18:34:04*zepolen quit (Remote host closed the connection)
18:35:54*cyraxjoe_ quit (Read error: Connection reset by peer)
18:36:06*brson joined #nim
18:37:15*cyraxjoe joined #nim
18:37:55AraqBitPuffin|osx: return and parameter passing are already move operations
18:38:41BitPuffin|osxAraq: can you overload that move?
18:39:07BitPuffin|osxor how do you mean, for the GC?
18:39:16BitPuffin|osxbecause we wanna implement unique_ptr here, which would not use the gc
18:41:11*smede joined #nim
18:41:59*zepolen joined #nim
18:45:46*zepolen quit (Remote host closed the connection)
18:47:10*yglukhov quit (Ping timeout: 260 seconds)
18:48:56*yglukhov joined #nim
18:52:15*yglukhov quit (Remote host closed the connection)
19:00:10*smede quit (Ping timeout: 240 seconds)
19:28:18*pregressive quit (Remote host closed the connection)
19:31:24*brson quit (Ping timeout: 250 seconds)
19:35:26*strcmp1 quit (Quit: return 0;)
19:35:37*strcmp1 joined #nim
19:41:12*Demon_Fox joined #nim
19:43:46*brson joined #nim
19:45:30*vendethiel- joined #nim
19:47:08*vendethiel quit (Ping timeout: 244 seconds)
19:56:37*gokr_ joined #nim
19:57:41*elrood joined #nim
20:01:01*terry_ joined #nim
20:05:38terry_hi Araq, can I follow up on a comment you made on the forum?
20:05:42terry_see http://forum.nim-lang.org/t/1804/2
20:06:00Araqhi :-)
20:06:02terry_Can you explain a little more to me why libmill's light weight threading is a deadend when interacting with Nim's GC? I am not sure I understand it
20:06:37Araqwell tbh I don't know how libmill works but there are only a couple of solutions possible
20:06:48Araq1. setjmp hacks
20:07:08Araq2. stack switching via assembler instructions
20:07:43terry_it is indeed based on setjmp hacks. and it is very similar to musl system inside Nim's coroutine system
20:07:53Araq3. posix's setcontext
20:09:16Araqok, well setjmp hacks are strictly speaking not compatible with Ansi C iirc
20:09:39Araqand you need to be aware of the volatile requirements etc
20:10:08Araqplus I found the code much harder to understand than just inline assembler
20:10:22Araqso what's the benefit?
20:11:06terry_mostly just b/c I didn't know about -d:nimCoroutines (i didn't see it in the documentation)
20:12:00terry_but i am more interested in whether there are some special things you needed to do in your -d:nimCoroutines that resolved some bad interaction b/w setjmp and Nim's GC
20:12:26Araq-d:nimCoroutines just supports Nim's GC
20:12:45Araqyou don't need to avoid it because it tells the Gc properly about the available stacks
20:13:02terry_but does a regular C library not handle GC if I only pass alloc'd memory to it?
20:13:18terry_or if I use GC_ref/GC_unref with it?
20:13:42Araqr-ku knows the details, but he started with something with libmill and now it uses inline assembler and works much better
20:14:20Araqyou can indeed use GC_ref/GC_unref as well as a couple of other solutions
20:15:00Araq(*he started with something like libmill)
20:15:13terry_okay, that makes sense to me
20:15:17terry_I definite have no issues with using nimCoroutines. I am just more worried about GC interactions in general when interfacing with C library.
20:15:34terry_and it sounds like it wasn't that there was a problem but that nimCoroutines is a better starting point
20:19:25Araqyup, but it doesn't work yet with the default GC iirc
20:19:45Araqoh wait, it does
20:22:27terry_okay, thanks for confirming. I will take a deeper look.
20:22:35*makoLine joined #nim
20:24:45Araqbut it kinda sucks because he was too lazy to use a bloom filter for the isOnStack primitive
20:26:25*mat4 joined #nim
20:26:27mat4hello
20:26:58*allan0 quit (Quit: no)
20:27:56*allan0 joined #nim
20:31:58terry_i see, he loops through and checks each element in stack
20:32:14Araqhi mat4
20:32:57mat4hi Araq
20:33:00*pregressive joined #nim
20:40:31mat4have you read the logs ?
20:42:43Araqno.
20:42:58AraqI haven't even read all my emails.
20:45:01*pmbauer left #nim (#nim)
20:47:26*pregressive quit (Remote host closed the connection)
20:47:59mat4I was wondering why ranges of type [1..n] will get initializes to zero
20:48:01*pregressive joined #nim
20:48:07*pregressive quit (Read error: Connection reset by peer)
20:48:15mat4^initialized
20:48:36dom96mat4: pretty sure that's a bug
20:48:48mat4I think so
20:48:51*smede joined #nim
20:50:51mat4(however someone here found this a rational initialisation choice)
20:54:54Araqit's not a bug, it's documented behaviour
20:55:21Araqdefault initialization is always binary zero because it's the most efficient thing that can be done
20:55:47Araqthe compiler does warn you though
20:56:13Araqand at some point we can make it a compile time error. I see nothing wrong with it.
20:56:48mat4ok, so I need to ensure manual initialisation for that cases
20:57:44*zepolen joined #nim
20:57:54dom96Araq: Doesn't the compiler know at compile-time the range of the array?
20:58:13dom96er, the range of the type
20:58:23Araqyes, it does
20:59:02dom96so why can't it just emit a '= 1' or '= 2' or whatever low(rangeT) is?
20:59:10BitPuffin|osxAraq: you never answered if you can overload the code that runs on a move on return value, also parameter passing doesn't always have to be a move
21:00:11Araqdom96: it could do that but it's some work to implement and the resulting machine code ain't pretty
21:00:54federico3are the buildbots running performance regression jobs?
21:01:11dom96Araq: hrm, I'm not sure I can see why it would be a lot of work.
21:01:18dom96But I'll take your word for it.
21:01:23dom96federico3: Don't think so.
21:01:54dom96Wish somebody would take over the maintenance of NimBuild off my hands ;)
21:02:07dom96Benchmarks were on my todo list for a long time
21:02:19federico3it would be nice to have some benchmarks and show their history
21:02:59*zepolen quit (Ping timeout: 272 seconds)
21:03:09dom96I agree
21:03:45federico3I could look for some simple tools to collect test run outputs and generate charts
21:04:07federico3I also have a cubieboard that could join the fleet for building and testing
21:05:45dom96The biggest problem we have is that none of us have the time for our infrastructure.
21:05:54dom96Varriount is even way too busy.
21:05:59dom96As far as I can see.
21:06:02*theduke joined #nim
21:07:56thedukeHey guys, dom just published my playground package: https://github.com/theduke/nim-playground
21:08:21AraqBitPuffin|osx: I don't know. By your logic 'const &T' would break C++ move semantics and I have never heard of that being the case.
21:08:26def-theduke: nice!
21:08:32thedukeweb-based light weight editor for playing with nimble code. I find it very convenient for testing things out
21:08:46thedukesuggestions welcome
21:08:54dom96theduke: that reminds me to test it!
21:10:52dom96theduke: works really well, nice!
21:11:09dom96One suggestion: make the "Press Ctrl+Enter to execute" a bit more visible.
21:11:39BitPuffin|osxAraq: I dunno what const &t does in nim necessarily
21:11:55BitPuffin|osxor why it would break
21:12:16Araqnim's parameter passing semantics are C++'s 'const &T'
21:12:35BitPuffin|osxyeah and C++ doesn't move every time you passa parameter
21:13:16BitPuffin|osxbut yeah you wouldn't move to a const &T in C++ I think
21:13:49BitPuffin|osxbut when you transfer ownership of a unique_ptr by calling some function you can move it
21:14:20BitPuffin|osxthen I guess it's just like taking a T in C++ but since you use std::move it gets casted to an rvalue and it is move constructed rather than copy constructed
21:14:55dom96bbl
21:16:50*Demos joined #nim
21:18:22federico3theduke: you might want to add some comments on its security
21:25:30*smede quit (Ping timeout: 260 seconds)
21:25:55federico3dom96: what kind of help do you need on the infrastructure?
21:26:17*theduke quit (Ping timeout: 252 seconds)
21:30:57*yglukhov joined #nim
21:31:58*vendethiel- quit (Read error: Connection reset by peer)
21:34:05*vendethiel joined #nim
21:38:12*theduke joined #nim
21:40:05*Trustable quit (Remote host closed the connection)
21:48:06BitPuffin|osxAraq: oh yeah I think I see what you mean now that I'm remembering more about nim, you don't really have pass by value you do proc foo(a: int) = let copied = a
21:53:24BitPuffin|osxso solution is take the parameter as a var
21:53:28BitPuffin|osxand overload = and move it
21:53:46BitPuffin|osxcan you define your own effects so that you can statically prevent use after move?
21:58:54*gour quit (Quit: WeeChat 1.3)
22:01:23BitPuffin|osxAraq: so a basic implementation of the basics would look like this https://gist.github.com/BitPuffin/a3e145bf43f934098faf ?
22:02:00*Trustable joined #nim
22:02:12BitPuffin|osxyou'd have to overload [] of course
22:02:19*Trustable quit (Remote host closed the connection)
22:02:29BitPuffin|osxbut for the moving and destruction that would be the implementation roughly
22:02:36*Ven joined #nim
22:02:40BitPuffin|osxjust that ideally you also wanna track the fact that it was moved with effects
22:03:46BitPuffin|osxyou could add a moved boolean but then you're introducing runtime overhead that the whole point was to avoid with a unique_ptr
22:18:18*lazypenguin quit (Ping timeout: 240 seconds)
22:22:22*Kingsquee joined #nim
22:28:34*mat4 quit (Quit: Leaving)
22:29:25AraqBitPuffin|osx: yeah, that's how I imagine a basic implementation to look like
22:29:50*brson quit (Ping timeout: 240 seconds)
22:30:38BitPuffin|osxAraq: that's pretty decent
22:31:56*Ven_ joined #nim
22:32:19BitPuffin|osxAraq: if you have the ability to get to the root of the AST I suppose you could implement the checking for use after move with a macro
22:32:30BitPuffin|osxjust that it would be tricky, and not very scalable
22:32:34BitPuffin|osxsince you want local checking
22:34:02*brson joined #nim
22:35:20*Ven quit (Ping timeout: 276 seconds)
22:35:25*yglukhov quit (Remote host closed the connection)
22:35:58AraqBitPuffin|osx: that's better done in the language directly though
22:36:39Araquntil a theory of programmable type systems comes along and Nim supports it
22:37:14Araqwhich is ten years away I guess
22:38:53BitPuffin|osxAraq: until then you could do pretty well with a static analysis tool that knows about uptr
22:39:31BitPuffin|osxyou could also make it a runtime error I guess in debug mode and do the bool thing then
22:41:56*lazypenguin joined #nim
22:44:50AraqBitPuffin|osx: I agree. the static analysis is the way to go
22:48:31BitPuffin|osxwhere did proc free go?
22:53:44Araqit was removed because it wasn't a wrapper for C's free, but an alias for dealloc
22:53:51Araqso use dealloc instead
22:54:11BitPuffin|osxyeah alright
22:57:37BitPuffin|osxAraq: wait so if it's not C free, what is it?
22:57:51AraqNim's free?
22:57:59BitPuffin|osxhow does that differ from C free?
23:00:10AraqNim has its own allocator
23:01:10Araqwhich is thread local, doesn't return 'nil' but dies with OOM instead and supports Nim's GC
23:01:19BitPuffin|osxI see
23:01:35BitPuffin|osxI also wanted to add a way to push and pop a sort of allocator "stack"
23:03:06*Ven_ quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
23:03:18BitPuffin|osxif you've kept up with jonnifer blowpez' language he has a thing where you can push a new context to get a different allocator
23:03:23BitPuffin|osxand I thought that could be added fairly easy with macrros
23:09:09BitPuffin|osxError: implementation of 'uptr' expected o_O
23:09:15BitPuffin|osxtype
23:09:15BitPuffin|osx uptr[T]* = object of RootObj
23:09:15BitPuffin|osx p: ptr T
23:10:24*Demos quit (Ping timeout: 265 seconds)
23:11:15Araqthe star follows the identifier, not the brackets
23:11:28Araqbecause you export the identifier, not the brackets
23:11:40BitPuffin|osxah okay
23:12:36Araqyeah but pushing and poping allocators is just horrible for memory safety
23:12:48BitPuffin|osxyeah this isn't really for safety though
23:13:07Araqand almost impossible to model with a type system
23:13:19BitPuffin|osxit's for manual memory management
23:13:52BitPuffin|osxso that I can go, create arbitrary memory stack, push this one and every allocation should now come from there, and then free it all in one go later
23:14:12BitPuffin|osxanyhow, move semantics aren't possible because you can't modify the rhs
23:14:20BitPuffin|osxmem.nim(27, 1) Error: signature for '=' must be proc[T: object](x: var T; y: T)
23:14:30BitPuffin|osxif y can't be var then you can't move
23:15:34Araqif it has to be var it cannot be constructed though
23:16:33*aziz_ quit (Remote host closed the connection)
23:17:10BitPuffin|osxwait it compiled even though I'm mutating rhs
23:18:08Araqhu?
23:18:16BitPuffin|osxproc `=`*[T: object](lhs: var uptr[T], rhs: uptr[T]) =
23:18:17BitPuffin|osx lhs.p = rhs.p
23:18:17BitPuffin|osx rhs.p = nil
23:18:19BitPuffin|osxrhs.p = nil
23:18:26BitPuffin|osxwhy does it let me do that when rhs is not var?
23:18:37BitPuffin|osxI thought the default was const T&
23:22:47Araqbecause you don't instantiate this generic?
23:23:45BitPuffin|osxah, maybe that's why I'm getting can't instantiate deepCopy when I actually instantiate it :P
23:25:00BitPuffin|osxeither way since you can't make rhs var I guess you can't implement move on assignment
23:25:28BitPuffin|osxsince moving is supposed to clean up the object you're moving from
23:28:54*zepolen joined #nim
23:39:45BitPuffin|osxI guess you could make the operator <- isntead
23:40:25*vendethiel quit (Ping timeout: 240 seconds)
23:42:32*Sembei joined #nim
23:44:04Araqif '=' takes a rhs 'var' you cannot construct it. at all.
23:44:34BitPuffin|osxyeah but for move rhs needs to be a var
23:44:37BitPuffin|osxso move can't be with =
23:44:58BitPuffin|osxalso why not? you just can't construct it with another thing of the same type
23:49:34Araqso how do you construct something with something of a different type?
23:49:44Araqthat makes no sense.
23:49:59Araqvar s: string = 8 # i don't think so ...
23:50:30BitPuffin|osxI meant more like var s: something = something(3, 2)
23:50:35BitPuffin|osxthe construction isn't the assignment
23:51:12BitPuffin|osxit's the call to the constructor
23:51:44Araqyeah well
23:51:46*brson quit (Quit: leaving)
23:52:09BitPuffin|osxvar a: uptr[int] = uptr[int](p: sp)
23:52:09BitPuffin|osxvar b: uptr[int] = uptr[int](p: nil)
23:52:09BitPuffin|osxb <- a
23:52:13BitPuffin|osxdidn't work either
23:52:19BitPuffin|osxbecause b and a are apparently not var
23:52:44AraqNim rewrites 'var x = foo()' to 'var x; x = foo()' :P
23:52:55Araqif the type overloads '='
23:54:00BitPuffin|osxtype mismatch: got (uptr[system.int], uptr[system.int]) but expected one of: mem.<-(lhs: var uptr[<-.T], rhs: var uptr[<-.T])
23:54:17BitPuffin|osxbut yeah you really want move to be = here
23:54:26BitPuffin|osxso I dunno
23:57:45lazypenguinAraq: I enjoyed your talk at the Nim workshop (uploaded on Reddit)...too bad not a lot of people laughed at your jokes ;).
23:59:49BitPuffin|osxAraq: anyway I don't really see an obvious solution to it and I'm not so familiar with the language anymore, but it would be nice if there was one