<< 30-06-2014 >>
00:00:23Jehan_flaviu: Yeah.
00:00:49Araqyou can try "consolidation" with language design, or you can try to push the state of art
00:01:04Araqyou can't do both at the same time
00:01:33AraqNimrod is about pushing the state of the art
00:01:56Jehan_My current design for Talis requires you to use custom operators within a named DSL section (which also allows you to redefine precedence and associativity). This way, the name of the section serves as a readable reference.
00:02:34Jehan_Araq: I'm fine with pushing the state of the art. In fact, I'm very much in favor of it.
00:02:39flaviuJehan_: There was a language on reddit that did that
00:02:51AraqRust and D do the same btw.
00:02:52Jehan_flaviu: I'm not surprised, it's a pretty logical idea.
00:03:14AraqGo however is all about what it leaves out because it's considered "dangerous".
00:03:32Jehan_Araq: I think the "dangerous" argument is pretty bad.
00:03:59Jehan_The features that Go leaves out aren't dangerous (at least the ones I can think of), they just add to the learning curve.
00:04:08Jehan_Worse, in some cases they probably would reduce the learning curve.
00:04:13gooblesGo is dumb nuff said
00:04:31flaviuJehan_: Its also good. I don't want to read code by a guy who thinks operators are the best and only uses them for method names.
00:04:41flaviuBut outright banning operators is insane
00:04:53Jehan_flaviu: Agreed.
00:05:14Jehan_Having to write a.mul(b) for matrices and vectors is annoying as hell.
00:09:05*reactormonk_ is now known as reactormonk
00:09:28*XAMPP-8 quit (Ping timeout: 248 seconds)
00:10:04Araqugh I feel stupid
00:11:06gooblesi think ur probably suffering from low blood sugar
00:11:27AraqI always thought the basic layout of the alloctor's data structures doesn't allow for pointer bumping
00:11:41*retsej left #nimrod (#nimrod)
00:11:58Araqbut it should work out just fine with minor tweakings
00:14:03*ehaliewicz joined #nimrod
00:15:35gooblesnimrod lets me not have header files woot
00:18:58*q66 quit (Quit: Leaving)
00:19:00Jehan_I think OCaml is about the only (semi-)recent language that still decided to have a concept of an interface file separate from the implementation.
00:26:39gooblesu proficient in nimrod jehan
00:27:09gooblescan you make sweet code fast?
00:30:30Jehan_No idea, and I'm about to go to bed, so no way to find out tonight. :)
00:32:24*Jehan_ quit (Quit: Leaving)
00:33:12gooblescan you do hot reloading of nimrod code
00:33:15gooblesthat would be sweeeeeeeeeeeeet
00:33:50flaviugoobles: Not easily, its pretty much impossible, just like in C
00:39:44Araqgoobles: you can do it with the VM ... perhaps
00:41:40*goobles quit (Ping timeout: 246 seconds)
00:59:08*brson quit (Ping timeout: 240 seconds)
01:01:15*brson joined #nimrod
01:13:00*ehaliewicz quit (Remote host closed the connection)
01:26:51*eigenlicht_ quit (Ping timeout: 240 seconds)
01:28:21*eigenlicht_ joined #nimrod
01:37:17*wan quit (Ping timeout: 245 seconds)
01:47:32*Jesin quit (Quit: Leaving)
01:47:54*Jesin joined #nimrod
02:12:13*Nimrod_ joined #nimrod
02:14:51*Nimrod quit (Ping timeout: 240 seconds)
02:15:09*goobles joined #nimrod
02:30:48*Varriount|Mobile quit (Ping timeout: 248 seconds)
03:05:25*brson quit (Quit: leaving)
04:22:30reactormonkgoobles, what do you need header files for?
04:28:06*xtagon joined #nimrod
04:36:39*kshlm joined #nimrod
04:44:35*flaviu quit (Ping timeout: 272 seconds)
04:56:54*crippa joined #nimrod
05:01:08gooblesC++
05:01:30*crippa quit (Ping timeout: 255 seconds)
05:01:42reactormonkgoobles, well, you can just give the generated code directly to the compiler, see the extern pragma
05:03:15gooblesare you good at nimrod?
05:05:20reactormonkyour call: https://github.com/Araq/Nimrod/graphs/contributors
05:06:17gooblesoh good u do know this language !
05:06:26goobleswow araq guy has 750k lines
05:06:44gooblesthats alot of lines how the fuck does nimrod even have that many
05:06:53reactormonkchanges.
05:07:28reactormonkassume that some of it is generated.
05:07:50goobleshave you used nimrods macros much
05:07:58gooblesare they as powerful as lisp macros?
05:08:32*Jesin quit (Ping timeout: 252 seconds)
05:09:22reactormonkNot too much, but yes, they can call almost all nimrod stuff. IIRC you can't call FFI stuff during macro resolution because the VM doesn't support that yet.
05:09:41reactormonkBut otherwise, yes, they are as powerful.
05:10:15gooblescool
05:10:28gooblesis it fast at incremental building? Like change 1 line rebuild?
05:10:39goobles(for larger project)
05:10:45gooblesnot 1 file things
05:11:12Demoswell partial compilation has been broken for a while. There is nothing technical preventing it from being fixed but nobody really needs it
05:11:32Demosthe compiler can compile itself in like 15s
05:11:39reactormonkAnd it's got LOT of lines
05:12:00gooblesic
05:12:04reactormonkmore like 20s. But yes, it's fast enough that no one cares.
05:12:33DemosI mean it is like a order of magnitude faster than c++
05:14:08reactormonkgoobles, no, no partial compliation. But you don't care.
05:14:53gooblescause it compiles teh fast?
05:14:58gooblesthats good ya
05:22:06*Demos quit (Read error: Connection reset by peer)
05:45:55*ARCADIVS quit (Quit: WeeChat 0.4.3)
05:50:45*xtagon quit (Read error: Connection reset by peer)
05:53:42*pafmaf joined #nimrod
05:55:12*Jesin joined #nimrod
06:15:50*gkoller joined #nimrod
07:03:40*pafmaf quit (Quit: This computer has gone to sleep)
07:41:10*kunev joined #nimrod
08:05:27*vendethiel quit (Read error: Connection reset by peer)
08:16:38*Trustable joined #nimrod
08:27:31*BitPuffin quit (Ping timeout: 272 seconds)
08:33:12*pafmaf joined #nimrod
08:58:38*kshlm quit (Ping timeout: 240 seconds)
09:15:44*BitPuffin joined #nimrod
09:25:56*kshlm joined #nimrod
10:00:03*kshlm quit (Remote host closed the connection)
10:00:30*kshlm joined #nimrod
10:09:34*q66 joined #nimrod
10:25:15*Raynes quit (Ping timeout: 252 seconds)
10:26:27*dom96_w joined #nimrod
10:26:39*dom96_w quit (Client Quit)
10:26:51*Raynes joined #nimrod
10:31:15*q66_ joined #nimrod
10:34:41*q66 quit (Ping timeout: 252 seconds)
11:44:21*Varriount|Mobile joined #nimrod
11:44:26*pafmaf quit (Quit: Verlassend)
11:47:39gooblescompiling to c++ has produces a fair amount of errors with pointer conversion
11:47:56gooblesis there a way to compile to a readable c or C++?
11:50:19*kshlm quit (Ping timeout: 240 seconds)
11:55:18*EXetoC joined #nimrod
12:16:54Varriount|Mobilegoobles: Compiling to C with linedir on should embed lines from the Nimrod source into the generated c code
12:19:33*kshlm joined #nimrod
12:19:46*kshlm quit (Client Quit)
12:25:35*untitaker_ quit (Ping timeout: 264 seconds)
12:26:56*q66_ is now known as q66
12:30:35*untitaker joined #nimrod
12:35:28Araqgoobles: you need to do something like --passC:"-d const" to make the preprocessor get rid of 'const' for C++
12:35:53Araqit's an open problem how to deal properly with C++'s type system. "portable assembler" my ass
12:48:23*Varriount|Mobile quit (Remote host closed the connection)
13:26:33*fran__ joined #nimrod
13:29:32*Fr4n quit (Ping timeout: 240 seconds)
13:33:25*asterite joined #nimrod
13:35:37*asterite quit (Client Quit)
13:35:54*asterite joined #nimrod
13:38:23*darkf quit (Quit: Leaving)
13:56:39*io2 joined #nimrod
14:47:52*gkoller quit (Ping timeout: 248 seconds)
14:50:36*kunev quit (Quit: leaving)
15:04:36*io2 quit (Ping timeout: 260 seconds)
15:33:42*blamestross_ joined #nimrod
15:39:00blamestross_Hey I am considering pull-requesting babel to add my vectors package. I figured I'd ask first.
15:41:30blamestross_well in https://github.com/nimrod-code/packages not babel
15:44:07*asterite quit (Quit: Leaving.)
15:53:05flyxblamestross_: did you test it?
15:54:45*zahary joined #nimrod
15:56:59*Demos joined #nimrod
15:58:19*blamestross_ quit (Ping timeout: 246 seconds)
15:59:04*Jesin quit (Quit: Leaving)
16:05:00*Jesin joined #nimrod
16:08:11*vendethiel joined #nimrod
16:08:25*blamestross-work joined #nimrod
16:08:51blamestross-workyeah, I wrote it because I need it for some research code I am working on anyway
16:09:10blamestross-workI might end up doing a graph library.
16:09:40blamestross-workI'm doing real paying work to avoid implementing BigInt division
16:09:44Demosblamestross-work, that would be sweet. Do you know if there exists a graph library that is better than BGL?
16:09:58Demosassuming you mean that sort of graph
16:10:58blamestross-worklikely no, what I would write would be closer to networkX. I need to do analysis of network topologies
16:12:40*goobles quit (Ping timeout: 246 seconds)
16:17:54*asterite joined #nimrod
16:19:35*gkoller joined #nimrod
16:21:12blamestross-workEvery time I suggest "lets make a library to do x" everybody gets excited and says language Blah has a library that does X really well that people have sunt years of effort into lets make it like/better then that. Then I go "no, I just want it to work!"
16:22:05blamestross-works/sunt/sunk
16:29:17Araqhey zahary. welcome back
16:29:30zaharyhi
16:31:41NimBotAraq/Nimrod devel 79cdb26 def [+0 ±1 -0]: Add missing operators for uint and uint64: `+=`, `-=`, `*=`
16:31:41NimBotAraq/Nimrod devel 4f24fac def [+0 ±1 -0]: Add inc and dec for uint and uint64
16:31:41NimBotAraq/Nimrod devel 5439699 def [+0 ±2 -0]: Add type constraints for uint and uint64 to the System module's procs instead
16:31:41NimBotAraq/Nimrod devel 5555d4a Andreas Rumpf [+0 ±1 -0]: Merge pull request #1312 from def-/missing-unsigned-operators... 2 more lines
16:32:13Araqcan you comment on #1184 ?
16:32:39zaharylet me take a look. I'm dying to get back to working on Nimrod, but still can't find the time
16:33:50NimBotAraq/Nimrod devel 92e958b def [+0 ±1 -0]: Add keepIf proc and keepIfIt template to sequtils
16:33:50NimBotAraq/Nimrod devel f6c4864 def [+0 ±1 -0]: Rename keepIfIt to keepItIf
16:33:50NimBotAraq/Nimrod devel dc15d37 Andreas Rumpf [+0 ±1 -0]: Merge pull request #1301 from def-/sequtils-filter-directly... 2 more lines
16:34:52zahary1184 doesn't look like something that should be merged, but it's still not clear in what situations the error messages are not quite good
16:38:36Demosif you have a proc taking a typeclass and a call fails because the provided type does not match it does not tell you why
16:38:54BitPuffinAraq: do you know if anyone has successfully ran nimrod on ios or android?
16:38:58BitPuffinand android rather
16:39:18DemosBitPuffin, I know Grada has used nimrod on iOS
16:39:49BitPuffinohreally?
16:39:52*gkoller quit (Ping timeout: 248 seconds)
16:40:27Demosyes
16:40:56BitPuffinI wonder if he used threads and stuff
16:44:30BitPuffinwell our httpclient is pretty fail
16:44:40BitPuffinbecause it's synchronous
16:45:02BitPuffinah wait there is an async version
16:46:01BitPuffinbut you can't cancel downloadFile
16:47:33*Jesin quit (Ping timeout: 255 seconds)
16:48:54Araqzahary: I thought about some makeFit(f, x) define that checks if f's type is "const" in C++ and if so casts x to "const" otherwise passes x to f
16:49:24Araqthe codegen would then generate that everywhere to make C++'s "const" behave
16:49:47Araqbut I'm not sure if that will work out
16:50:42Araqwell hrm f's return type is clearly not what we need here
16:55:17blamestross-workIs there a work around for no math.random in windows?
16:55:27blamestross-workshort of implementing my own prng?
16:56:42Araqhu? windows has its own math.random
16:57:10*BitPuffin quit (Read error: Connection reset by peer)
16:57:59blamestross-workonly for int
16:58:15*BitPuffin joined #nimrod
16:58:17*io2 joined #nimrod
16:58:27blamestross-worktoroid.nim(17, 11) Error: type mismatch: got (float) but expected one of: math.random(x: TSlice[T]): T math.random(max: int): int
16:58:37Araq when defined(windows):
16:58:38Araq proc random(max: float): float =
16:58:40Araq # we are hardcodeing this because
16:58:41Araq # importcing macros is extremely problematic
16:58:43Araq # and because the value is publicly documented
16:58:44Araq # on MSDN and very unlikely to change
16:58:46Araq const rand_max = 32767
16:58:48Araq result = (float(rand()) / float(rand_max)) * max
17:00:23*Jesin joined #nimrod
17:00:50blamestross-workok, I'm off to figure out what stupid mistake I am makign that prevents me from using that
17:03:33blamestross-workinteresting
17:03:40*brson joined #nimrod
17:03:45blamestross-workthe source for math.nim lacks that bit of code
17:03:51blamestross-workat least locally
17:03:58AraqI'm on devel
17:03:59Demosyou on devel or master?
17:04:03Demosyou should be on devel
17:04:09Araqcould be a recent addition
17:04:09blamestross-workaha, master
17:04:32Demosyup it is a recent addition
17:04:44Araqwell this master vs devel split worked out nicely ... :-/
17:05:22Araqoh well at least I can blame Varriount for not doing his job :P
17:06:53blamestross-workI used the windows installer
17:07:08Araqwell then you're not even on master
17:07:13Araqbut on 0.9.4
17:07:21blamestross-workso It looks like I shoudl just replace /programs/nimrod with a pull off the github
17:07:50blamestross-workor will that do horrible things to configurations?
17:08:22Demoswell it will not have the binaries in it
17:08:48DemosI just pulled the source to a folder, built the compiler, and added ./bin to my path
17:09:56blamestross-workwow
17:10:06AraqI copy stuff from lots of places around, mixing different git, gcc versions and installations
17:10:18blamestross-worktrying to delete nimrod using powershell results in a lot of no-permissions errors
17:10:26Araqit all works but then I know what I'm doing
17:10:55blamestross-workAraq: I'm working on it
17:11:31NimBotAraq/Nimrod devel 9ed9842 Grzegorz Adam Hankiewicz [+0 ±2 -0]: Hyperlinks together importc like pragmas.
17:11:31NimBotAraq/Nimrod devel 6a2bb00 Grzegorz Adam Hankiewicz [+1 ±3 -0]: Starts backends.txt with bits from nimrodc.txt.
17:11:31NimBotAraq/Nimrod devel 2353b56 Grzegorz Adam Hankiewicz [+0 ±4 -0]: Hyperlinks back JavaScript target.
17:11:31NimBotAraq/Nimrod devel 9c8ce45 Grzegorz Adam Hankiewicz [+0 ±1 -0]: Starts nimrod calling backend section.
17:11:31NimBot7 more commits.
17:22:40blamestross-workC:\Nimrod\csources [master]> .\build.bat gcc -w -Ic_code -c c_code\1_1\nimrod.c -o c_code\1_1\nimrod.o In file included from c_code\1_1\nimrod.c:5:0: c_code/nimbase.h:382:13: error: size of array 'assert_numbits' is negative C:\Nimrod\csources [master]>
17:22:58Araquse build64.bat
17:23:37blamestross-workthat makes sense
17:23:48Demosalternitively download a binary from the buildfarm and use that to bootstrap koch and the compiler
17:29:56*Jesin quit (Ping timeout: 252 seconds)
17:31:26blamestross-workDemos: that looks like the plan, build64.bat keeps crashing
17:32:00dom96I suggest still using 32bit, finding 64bit DLLs is difficult.
17:32:58DemosI have had no problems with that dom96
17:33:25Demosmsys2 has them all avalible in a package management system
17:33:37Demosand it is not that hard to build most libraries with cmake and visual studio
17:34:52*io2 quit (Ping timeout: 260 seconds)
17:35:31*io2 joined #nimrod
17:35:44blamestross-workI just re-installed nimrod so I could compile nimrod. bootstrapping allways makes me feel dirty
17:38:43*silven_ is now known as silven
17:40:58NimBotAraq/Nimrod devel 678f3d7 Araq [+0 ±3 -1]: fixes constant array indexing bug
17:40:58NimBotAraq/Nimrod devel 1d5938a Araq [+1 ±10 -0]: Merge branch 'devel' of https://github.com/Araq/Nimrod into devel
17:40:58NimBotAraq/Nimrod devel b89495e Araq [+0 ±4 -0]: corrected backends.txt
17:42:06Amrykidyay for the mention of powershell
17:42:44Araqoh hi Amrykid. glad you're still around
17:43:08Amrykidhi Araq. i lurk.
17:43:28AraqKerrigan would be proud of you.
17:43:41AmrykidI learn from the best.
17:44:53*Jesin joined #nimrod
17:46:34dom96Demos: Can you build me GTK, GtkSourceView and its dependencies for 64bit?
17:46:46dom96Demos: I think it's harder than you realise.
17:47:04*Jehan_ joined #nimrod
17:47:21Demoserm.... yeah probably not
17:47:35Demosalthough they do have 64 bit versions on linux...
17:48:45Jehan_Dependency hell is one of the primary reasons why I don't use Linux. :(
17:49:34Demoswell at least they don't need 8 versions of every library
17:50:02Jehan_Demos: Yeah, but … have you ever looked at the Debian Mono packages, say?
17:50:17Demosno... thank god
17:50:43Jehan_There are like a couple dozen. Why? Apparently because there was an argument about which language could have the smallest distribution footprint in its minimal packaging.
17:51:13Demoswhy is c# and mono even in that race
17:51:19Araqin practice duplication works, sharing does not.
17:51:28DemosI agree with araq
17:51:49Jehan_Demos: As it turns out, you can chop Mono in pretty small pieces.
17:52:12Demosyeah but I am guessing that you need all of them 99.99% of the time
17:52:29*retsej joined #nimrod
17:52:40Demosbesides we should just like forth win that battle and walk away :D
17:52:48Demoss/like/let/
17:52:53Jehan_Araq: More importantly, this kind of optimization made sense when hard drives were maybe 500 MB in size and RAM size was around 64MB.
17:53:29DemosJehan_, but it must run on my PDP-11
17:53:42AraqJehan_: yeah that too, but Unix/Linux is stuck in the 70ies anyway
17:53:47Araqlol, Demos
17:54:24Jehan_Araq: Userland … well, let's say, the 90s. I have little complaints about the kernel itself.
17:55:47*mmatalka quit (Ping timeout: 252 seconds)
17:56:12Jehan_Ah, here we go, pretty much what I remembered: http://apebox.org/wordpress/rants/416/
17:57:36blamestross-workhttps://gist.github.com/blamestross/9f18d56a0bc37f495f4b
17:59:28Araqbbl
18:00:11*Jesin quit (Ping timeout: 252 seconds)
18:02:30*kunev joined #nimrod
18:04:58blamestross-workbbl too
18:05:02*blamestross-work quit (Quit: Page closed)
18:10:22*Jesin joined #nimrod
18:16:55*mmatalka joined #nimrod
18:18:58*Jesin quit (Ping timeout: 240 seconds)
18:24:22*Matthias247 joined #nimrod
18:34:39*BitPuffin quit (Ping timeout: 255 seconds)
18:36:43*Jesin joined #nimrod
18:43:20*kunev quit (Quit: Lost terminal)
18:58:28*gsingh93_ joined #nimrod
19:27:38*vendethiel quit (Remote host closed the connection)
19:27:48*vendethiel joined #nimrod
19:31:31*kunev joined #nimrod
19:38:32Varriountdom96: Has a model been worked out yet for multi-thread async io?
19:39:11AraqVarriount: I don't think so, but what I'm implementing is passing a closure between threads per copy
19:39:18VarriountAraq: Blame me for what exactly?
19:39:30Araqthat master doesn't get any love
19:41:40Demoswe are so far from the critical mass required to do the stable/beta/canary thing though
19:43:47VarriountHm.. maybe Araq was right...
19:44:13VarriountMaybe Multi-threaded async-io is uneeded, at least for the usual jobs.
19:44:38AraqDemos: it's not about critical mass. it's about professionalism.
19:45:16DemosI suppose. I guess I figure that since the language as a whole is not stable yet the three branch split seems strange.
19:45:33*mwpher joined #nimrod
19:46:01Araqthree branches?
19:46:05Araqwe have 2.
19:46:07VarriountAraq: To make a release candidate process, we would need a third branch dedicated to fixes, so that
19:46:10Demoswell two branches and the main releases
19:46:24Varriountso that the new branch could be merged into both master and devel.
19:46:37Demosmaybe it is just that we are pretty far from 0.9.6 right now
19:46:52Araqwe're not.
19:47:17Araqmany things on the roadmap to 0.9.6 have been implemented
19:49:13VarriountAraq: If you want me to port fixes to master, I need to know which commits are fixes. :/
19:50:48*mwpher left #nimrod (#nimrod)
19:51:36Araqusually the commit message tells you
19:54:30VarriountAre documentation changes 'fixes' or 'updates'?
19:55:06Araqup to you
20:03:17*hoverbear joined #nimrod
20:03:38*Nimrod_ is now known as Nimrod
20:03:50*Nimrod quit (Changing host)
20:03:50*Nimrod joined #nimrod
20:09:20*asterite quit (Quit: Leaving.)
20:13:22dom96IMO the master/devel split is pointless
20:14:40fowldoesnt master just get updated from devel periodically
20:14:47dom96Varriount: We can worry about multi-threaded async when single-threaded is stable.
20:14:54Varriountdom96: Yeah.
20:15:03dom96fowl: nope, it's literally the same as the v0.9.4 tag
20:15:24Araqfowl: however that *was* the plan
20:15:39VarriountAraq: Trying to cherry pick fixes from the devel branch isn't working.
20:15:47fowlwell lets just update it once in a while a call it a micro non-release
20:15:57fowlalso tag it 0.9.4.1
20:15:57dom96Yeah. I'm pretty sure cherry picking fixes is pretty much impossible.
20:16:07dom96Because the branches have diverged so much.
20:16:08VarriountI'm getting too many conflict errors.
20:16:15Araqwe had this discussion before
20:16:23VarriountYes.
20:16:29Araqwe decided to try the 2 branches.
20:16:38dom96I don't remember that conclusion.
20:16:44fowlfowl's suggestion: kill all the lawyers
20:16:49EXetoChow do other projects do it then? manually try to apply the same patch?
20:17:06Araqdom96: *I* decided to try the 2 branches then.
20:17:16VarriountEXetoC: Other projects use a patch based system.
20:17:28dom96I have a vague feeling that there was another problem with the two branches.
20:17:32Araqbut ok, let's switch to something that's workable for Varriount
20:18:27VarriountIt would be best to make the fixes branch at the next release, when all the branches will be the same.
20:18:57VarriountWhich means working towards a release. Is there a todolist somewhere?
20:19:29Araqthere is a roadmap and there is my todo.txt
20:19:51Araqthe new comment handling breaks code so it's in a branch
20:20:04VarriountO_o
20:20:16VarriountHow does it break code?
20:20:27Araqfor instance:
20:20:29Araqelse:
20:20:34Araq # only a comment here
20:20:38Araqis not valid anymore
20:20:47VarriountThat was valid before?
20:20:51Araqyes
20:22:01VarriountWow.
20:23:34fowlyou animal
20:23:37fowl@ Araq
20:25:30Araqha, I knew fowl knows it and uses it
20:26:10Araqfowl: we could special case it, but
20:26:17Araqdiscard "only a comment here"
20:26:37Araqworks as well and people want me to break stuff
20:26:44Araqit's considered progress here
20:29:44*NimBot joined #nimrod
20:30:12*asterite joined #nimrod
20:31:22*asterite quit (Client Quit)
20:33:32*asterite joined #nimrod
20:36:21VarriountHello asterite
20:36:57asteriteHi Varriount
20:38:13*BitPuffin joined #nimrod
20:38:38Araqseriously guys
20:39:04Araqdo I have to do everything now? can't somebody at least tag new bug reports
20:39:33fowlhow do you do that?
20:40:52*BitPuffin quit (Read error: Connection reset by peer)
20:41:29*BitPuffin joined #nimrod
20:43:14NimBotAraq/Nimrod devel e21999c Grzegorz Adam Hankiewicz [+0 ±1 -0]: Forces babel package descriptions to end with a dot.
20:43:14NimBotAraq/Nimrod devel 1b78bdd Andreas Rumpf [+0 ±1 -0]: Merge pull request #1292 from gradha/pr_terminates_sentences_with_dot... 2 more lines
20:44:03*asterite quit (Quit: Leaving.)
20:44:07Demosdo nimrod enums start at 1 by default?
20:45:05*asterite joined #nimrod
20:45:06Araqno, 0
20:47:55*asterite quit (Client Quit)
20:50:17NimBotnimrod-code/packages master cf88aa4 BrendanBenshoof [+0 ±1 -0]: added vectors package
20:50:17NimBotnimrod-code/packages master 7f07c94 Billingsly Wetherfordshire [+0 ±1 -0]: Merge pull request #68 from blamestross/master... 2 more lines
20:52:44VarriountAraq: Tag bug reports as what, fixes?
20:52:54*asterite joined #nimrod
20:53:01AraqI could swear we have a test case for #1275 ... strange
20:53:02*asterite quit (Read error: Connection reset by peer)
20:53:29AraqVarriount: cannot follow
20:53:44Araqfixes #233 is the usual commit message
21:05:24VarriountAraq: What do you want me to work on?
21:09:27Araqgive bug reports about the broken Lambda lifting a new tag "lambda_lifting" please
21:09:40VarriountOn it.
21:09:55VarriountAny other tags you want me to give things?
21:09:59Araqclose c2nim related bugs as "c2nim moved into its own repositrory"
21:10:10Araqdunno if bug reports can be migrated
21:10:26AraqI created new tags: VM, semcheck and codegen
21:10:49Araqoh and please fix 'high' for the JS target
21:10:53Araqembarrassing bug
21:10:59VarriountI'll manually migrate bug reports.
21:11:14Araqok ty
21:20:50VarriountAraq: Is c2nim not going to be shipped with Nimrod anymore?
21:21:05Araqit's a babel package now, Varriount
21:21:29Araqwe can still ship the c2nim.exe for a release
21:21:41Araqdunno if that's useful
21:24:06VarriountIt is..
21:24:35VarriountAraq: We can't rely on people being proactive enough to learn about babel...
21:25:12Araqwell the installer really needs to get better then
21:25:35VarriountAraq: What was the reason for moving c2nim into it's own package anyway?
21:25:49AraqI wanted it to have its own bug tracker
21:26:38Araqand we need more babel packages. this thing starts to get its own self fullfilling prophecy once we have one thousand packages
21:26:50Araq*to become its own
21:28:48VarriountHm.. It would be interesting to have a 'build-your-own-nimrod' web interface, where you could either choose some pre-built packages, or select modules to add on to the main bundle...
21:28:50*Nimrod_ joined #nimrod
21:30:18dom96Bundling babel is enough.
21:30:29dom96Using babel is a piece of cake.
21:30:43dom96And if you're translating C headers into nimrod you should be able to use a damn package manager.
21:35:58*Nimrod quit (Ping timeout: 240 seconds)
21:37:59NimBotAraq/Nimrod devel 0339b9d Araq [+1 ±1 -0]: fixes #1319
21:37:59NimBotAraq/Nimrod devel c41e099 Araq [+0 ±1 -0]: fixes #1144
21:37:59NimBotAraq/Nimrod devel 3c818c8 Araq [+0 ±10 -0]: fixes #1285
21:37:59NimBotAraq/Nimrod devel 4d5b7b7 Araq [+0 ±1 -0]: Merge branch 'devel' of https://github.com/Araq/Nimrod into devel
21:38:22VarriountOoo, shiny fixes.
21:41:36*flaviu joined #nimrod
21:42:00Araqflaviu: in transf.nim we do constant folding as an optimization pass
21:42:41AraqbindSym"myconst" is transformed into bindSym(myconst) and then it's transformed into
21:42:52AraqbindSym(14) if 14 is the value of myconst
21:43:50Araqthis is exactly what happens when you have lots of passes over the AST, there is no free lunch.
21:44:17Araqpasses also can interact in lots of other strange ways
21:46:25flaviuI'm not sure what you're referring to
21:46:45Araqyou once suggested to break up the compiler into more passes
21:50:56flaviuI'm not sure how that would be an issue
21:51:46Araqwell IF constant folding would be conflated with semantic checking completely this bug wouldn't have existed
21:52:12Araqbecause when I write the semantic checking of bindSym I know not to perform any further constant folding on the symbol
21:52:36flaviuI'm not sure I see the bug then
21:53:25flaviuWhat does bindSym do?
21:53:42Araqit's an example of where we have 2 passes instead of 1 and this producing a bug
22:13:31*retsej quit (Quit: Leaving.)
22:18:00*ARCADIVS joined #nimrod
22:30:23*Matthias247 quit (Read error: Connection reset by peer)
22:48:38*Trustable quit (Quit: Leaving)
22:48:57*Jehan_ quit (Quit: Leaving)
22:57:41*retsej joined #nimrod
23:01:31*Jesin quit (Quit: Leaving)
23:01:54*Jesin joined #nimrod
23:02:49*io2 quit ()
23:04:25*darkf joined #nimrod
23:10:55*Varriount|Mobile joined #nimrod
23:18:59NimBotAraq/Nimrod devel e5f2a8a Araq [+0 ±1 -0]: fixes #1268
23:18:59NimBotAraq/Nimrod devel 4860b46 Araq [+0 ±2 -0]: fixes #1264
23:18:59NimBotAraq/Nimrod devel 7c5f87e Araq [+0 ±1 -0]: fixes #1265
23:42:41*zahary quit (Ping timeout: 244 seconds)
23:43:27*retsej quit (Quit: Leaving.)
23:56:36*saml_ joined #nimrod
23:59:07*q66 quit (Quit: Leaving)