github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
02:42 Kaiepi left 06:20 AlexDaniel left 06:27 brrt joined 06:50 domidumont joined 07:17 patrickb joined
lizmat twitter.com/Aearnus/status/1150980890257793024 07:42
perhaps updating Unicode version / other encodings would be a relatively simple task ? 07:43
nwc10 IIRC from Perl 5, "updating Unicode version" is rarely a simple task, because some of the stuff that changes needs to be expressed as new code. 08:00
it was part of why the idea of "just drop in some newer Unicode data files into your old checkout and regen" was actually not going to fly as a "user upgradable part" 08:01
do "other encodings" change that much to need updating?
lizmat well, afaik, there's a bunch of encodings that some people would like to see 08:06
and I guess an EBCDIC one would be cool as well :-) 08:07
nwc10 wash your mouth out! 08:17
(and technically, "which EBCDIC"?)
there were/are at least 2, and they disagree on where '~' is IIRC 08:18
lizmat aaah...so many to choose from: en.wikipedia.org/wiki/EBCDIC_code_pages
nwc10 or at least, one of the code points that's happily in one place in ASCII, and is used as a "magic number" for one type of magic in Perl 5
09:29 sena_kun joined 10:07 zakharyas joined 10:22 brrt left 11:44 sena_kun left 11:45 sena_kun joined 13:29 pamplemousse joined 13:50 zakharyas left 14:33 lucasb joined 15:16 pamplemousse left, pamplemousse joined 15:20 domidumont left 15:31 pamplemousse left 15:35 AlexDaniel joined 15:58 brrt joined 16:11 patrickb left 16:54 Kaiepi joined 17:00 pamplemousse_ joined 17:03 domidumont joined 17:33 rfold joined 17:42 domidumont left 17:44 robertle joined
pamplemousse_ \o 18:23
18:27 pamplemousse_ left, pamplemousse_ joined 18:28 Kaiepi left 18:32 Kaiepi joined
pamplemousse_ I'm running into a bit of a weird error, and was wondering if anyone might have any ideas. When I have my script run the command "perl6 --bytecode <MBC file>, it throws an error here: github.com/MoarVM/MoarVM/blob/mast...unit.c#L56 18:38
But if I run "perl6 --bytecode <MBC file>" directly, it's totally fine and executes the MBC file. All the --bytecode flag does differently is open the file in evalfiles and then calls loadbytecode on the compunit.
I'm not really sure why there would be a difference between calling "perl6 --bytecode <file>" from within a script and outside a script 18:39
brrt ohai pamplemousse_ 18:46
pamplemousse_ hi brrt 18:47
brrt pamplemousse_: I'd suggest implementing the "FIXME" error
not sure if we have something for that yet, but we probably do
and also, duplicate the line in the script, and put 'echo' in front of it 18:50
see if your script is really pointing to the same file
pamplemousse_ Alright, I'll try those, thank you! 18:56
brrt I don't know if we have a MVM_os_strerror equivalent... we probably should 18:59
did it help? 19:18
pamplemousse_ I still don't have it working, no. The script is pointing to the same file, so that's not it. I'm trying to implement the FIXME error now 19:19
brrt Ok 19:21
weird
Geth App-MoarVM-ConfprogCompiler/master: 4 commits pushed by (Timo Paulssen)++ 19:31
pamplemousse_ Another weird thing, if in the script I drop the -b flag and use a source file, it doesn't throw an error, but it doesn't output anything either. And it's not an issue where using system() is just not doing anything, because if I chain some other command on the end (I used ls), it definitely outputs that 19:34
brrt okay, I'm wondering what's actually going on with your script 19:37
I think you can run bash in a verbose mode 19:38
pamplemousse_ I'm sorry, I used the wrong word. I'm not running a script, I'm running a c program. 19:40
brrt I see 19:44
hmmm
and, quoting and stuff, is all correct?
19:45 zakharyas joined
pamplemousse_ Yes, as far as I can tell. This is the program: github.com/pamplemoussecache/p6_li...mpt2.c#L33 19:47
brrt has a look 19:53
pamplemousse_ All I've changed between this working and this no longer working was switching to using a file. When I change the command back to using -e and then have it directly plug in the contents of the program (say "Hello world!";), it still works just fine.
brrt so, couple of points 19:54
- try moving fclose sooner 19:55
before system
- system is not a production interface for reasons, but let's ignore that for a second :-)
pamplemousse_ Moving it sooner got it to work, sorry, should have caught that. Thanks brrt 19:57
brrt np :-)
oh, the other thing is 19:58
the template string will be modified, you should copy it to an array (rather than use the pointer to the string constant) 19:59
(the argument to mkstemp, I mean)
20:09 brrt left
pamplemousse_ brrt: Isn't that the desired behavior, that it will be modified? I thought that was the reasoning behind using mkstemp, so that I know the name of the temp file I want to delete after I'm done with it 20:12
Oh, never mind, I see now. Thank you for your help 20:20
When I add in the modules, I'll need to make than just the one temp file, so that makes sense 20:21
20:33 lucasb left 20:49 sena_kun left, sena_kun joined
Geth App-MoarVM-ConfprogCompiler: 8e8a1bd2fc | (Timo Paulssen)++ | 3 files
actually enable xor; enable eq_i and ne_i (== and !=)
20:56
App-MoarVM-ConfprogCompiler: 89f64c4744 | (Timo Paulssen)++ | 3 files
0.0.6
20:56 zakharyas left 21:02 sena_kun left 21:05 sena_kun joined 21:07 sena_kun left 21:08 sena_kun joined 21:09 sena_kun left 21:10 rfold left 21:40 MasterDuke joined
MasterDuke anyone have any comments on github.com/MoarVM/MoarVM/pull/1139 or github.com/MoarVM/MoarVM/pull/1142 ? 21:58
22:17 MasterDuke left
Geth MoarVM: 9cf16bb7b1 | (Timo Paulssen)++ | src/profiler/configuration.c
confprog: add band_i, bor_i, bxor_i ops
22:43
MoarVM: e07c0f2525 | (Timo Paulssen)++ | src/profiler/configuration.c
confprog: don't return null from filename()
22:55 Kaiepi left 22:56 Kaiepi joined 23:24 pamplemousse_ left