|
Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm Set by Zoffix on 27 July 2018. |
|||
| MasterDuke | nqp: my $a := "Inf"; say(+$a) | 01:07 | |
| camelia | Inf | ||
| MasterDuke | where does that numification happen in NQP? that's broken in my branch (it gives 0 instead), but there aren't any tests for it | 01:08 | |
| hm. i changed prefix:<+> to mean intify instead of numify, i think that's the problem | 01:14 | ||
| nqp: my $a := Inf; say(nqp::isnanorinf($a)) | 01:17 | ||
| camelia | 0 | ||
| MasterDuke | huh | 01:18 | |
|
01:54
lizmat left
|
|||
| MasterDuke | well, now we're passing `make m-test` | 01:54 | |
| some spectest failures though | 01:58 | ||
| `Cannot find method 'Num': no method cache and no .^find_method` in Actions.bare_complex_number(), not sure how that happens | 02:11 | ||
| here: github.com/rakudo/rakudo/blob/mast...8279-L8289 | 02:16 | ||
|
02:38
Kaiepi left
03:49
MasterDuke left
|
|||
| Geth | rakudo: MasterDuke17++ created pull request #2630: Remove unneeded line in bare_complex_number |
04:31 | |
|
04:43
Kaiepi joined
04:44
p6bannerbot sets mode: +v Kaiepi
|
|||
| Geth | rakudo: b6030d3745 | (Daniel Green)++ | src/Perl6/Actions.nqp Remove unneeded line in bare_complex_number |
07:08 | |
| rakudo: 5555bbf0ba | niner++ (committed using GitHub Web editor) | src/Perl6/Actions.nqp Merge pull request #2630 from MasterDuke17/remove_unneeded_line_in_bare_complex_number Remove unneeded line in bare_complex_number |
|||
|
07:46
huguesmax joined,
p6bannerbot sets mode: +v huguesmax,
huguesmax left
08:55
AlexDaniel left
09:42
lizmat joined,
p6bannerbot sets mode: +v lizmat
|
|||
| nine | tell patrickb I wonder if we can salvage the RAKUDO_HOME/NQP_HOME idea. So openSUSE wants binary libs to be stored in lib/lib64. The same is true for virtually all other distros I guess. So why do we store stuff in /usr/share anyway? Perl 5 and Python libs are stored in /usr/lib already. | 10:24 | |
| .tell patrickb I wonder if we can salvage the RAKUDO_HOME/NQP_HOME idea. So openSUSE wants binary libs to be stored in lib/lib64. The same is true for virtually all other distros I guess. So why do we store stuff in /usr/share anyway? Perl 5 and Python libs are stored in /usr/lib already. | |||
| yoleaux | nine: I'll pass your message to patrickb. | ||
| nine | .tell patrickb The FHS states that /usr/lib is for "Libraries for programming and packages" while /usr/share is for "Architecture-independent *data*" (emphasis mine). Considering that I find plenty of *.gif, *.png, *.ogg, *.wav and whatever in my /usr/lib I'd guess "if in doubt, put it in /usr/lib" is the rule of thumb | 10:47 | |
| yoleaux | nine: I'll pass your message to patrickb. | ||
| nine | I wonder if we could be smarter about integer constants. Right now passing a 1 as literal to a function with an int32 argument, the int will be boxed and we end up accessing it via sp_p6oget_bi which is really expensive for a tiny literal | 10:56 | |
|
11:16
ufobat joined
11:17
p6bannerbot sets mode: +v ufobat
|
|||
| Geth | rakudo: MasterDuke17++ created pull request #2631: Move some Num constant rvs to their signature |
12:36 | |
|
12:37
MasterDuke joined,
p6bannerbot sets mode: +v MasterDuke,
MasterDuke left,
MasterDuke joined,
herbert.freenode.net sets mode: +v MasterDuke,
p6bannerbot sets mode: +v MasterDuke
12:49
gfldex left
13:18
AlexDaniel joined,
p6bannerbot sets mode: +v AlexDaniel
16:31
ExtraCrispy joined,
p6bannerbot sets mode: +v ExtraCrispy
16:43
gfldex joined,
gfldex left,
gfldex joined
16:44
p6bannerbot sets mode: +v gfldex
17:20
lucasb joined,
p6bannerbot sets mode: +v lucasb
|
|||
| Geth | rakudo: debe6b6ea4 | (Elizabeth Mattijsen)++ | 7 files Make sure we do "my constant %foo = nqp::hash" only on moar As a temporary fix for R#2628, to not longer block the release. |
17:24 | |
| synopsebot | R#2628 [open]: github.com/rakudo/rakudo/issues/2628 [JVM][⚠ blocker ⚠] [JVM] build is broken (error in stage parse during generation of CORE.d.setting.jar) | ||
| Geth | rakudo: f02bceb7c2 | (Daniel Green)++ | src/core/Num.pm6 Move some Num constant rvs to their signature |
17:26 | |
| rakudo: 35a81d5448 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/core/Num.pm6 Merge pull request #2631 from MasterDuke17/move_some_Num_constant_rvs_to_signature Move some Num constant rvs to their signature |
|||
| lucasb | hmm, better move R#2627 to user-experience, no? | 17:38 | |
| synopsebot | R#2627 [open]: github.com/rakudo/rakudo/issues/2627 [6.e][RFC][decluttering] Perl 6, 6.d, Diwali, Raku, Rakudo, Moar, NQP, Star, * | ||
| lucasb | I know AlexDaniel++ explict acknowledged this in the issue, but... | 17:39 | |
| the own act of opening this in rakudo's repo already shows how confusing the current state is (maybe :) | |||
| and that twitter stream was fun to read :-) | 17:40 | ||
|
17:42
dogbert11 left
|
|||
| lucasb | iiuc, the issue is about how The Perl 6 Project wants to present itself to the outside world, right? | 17:43 | |
| lizmat | imi, yes | 18:13 | |
| *imo | 18:14 | ||
| MasterDuke | .ask samcv t/09-moar/Line_Break__LineBreak.t and t/09-moar/General_Category__extracted-DerivedGeneralCategory.t both have passing todos, think you could take a look? | 18:48 | |
| yoleaux | MasterDuke: I'll pass your message to samcv. | ||
| MasterDuke | anyone have an idea on about what would cause `Cannot find method 'Num': no method cache and no .^find_method` in Actions.bare_complex_number() ( github.com/rakudo/rakudo/blob/mast...8279-L8289 ) ? | 18:56 | |
| lizmat | MasterDuke: print a $foo.HOW.name($foo) and find out what it is you're calling .Num on | 19:05 | |
| MasterDuke: perhaps you need to hllize it first ? | 19:06 | ||
| MasterDuke | lizmat: this is just from running a spectest on my make_nqp_default_to_int branches, but yeah, i haven't tried printing that out yet | 19:11 | |
| lizmat | whenever I see that kind of error, I've usually had an "aha erlebnis" when I saw what it said :-) | 19:12 | |
| MasterDuke isn't sure if that's random typing or Dutch | 19:13 | ||
| lizmat | actually, it's german | 19:23 | |
| I guess a modern translation would be a "ah, duh! experience" | 19:24 | ||
| MasterDuke | sounds like an "ah ha! moment" or a "lightbulb going off moment" | 19:27 | |
| i would say my german is better than my dutch, but that's only because i don't know any dutch (and i do know more than 0 words in german) | |||
| huh, it's an Int (with a value of 5) | 19:29 | ||
| lizmat | are you sure it's not an int? | 19:32 | |
| MasterDuke | could be. this is when i'm running t/spec/S02-literals/allomorphic.t | 19:33 | |
| i don't get a line number | 19:34 | ||
| not an int, nqp::objprimspec is 0 | 19:40 | ||
| lizmat | weird | ||
| eh... perhaps you need to decont() it ? | 19:41 | ||
| afk for a bit& | |||
| MasterDuke | same message with decont() | 19:47 | |
| ok, i'm getting a BOOTNum. maybe i just don't need to call .Num on it at all? | 20:12 | ||
| yup, solves that problem | 20:17 | ||
| nice, down to just one spectest fail | 20:26 | ||
|
21:22
dogbert17 joined
21:23
p6bannerbot sets mode: +v dogbert17
|
|||
| MasterDuke | woohoo. now NQP builds ok and passes `make m-test` and Rakudo builds ok and passes `make m-test m-spectest` on my trilogy of make-NQP-default-to-int branches | 21:36 | |
| i assume everyone will be fine with it if i merge them all right before the release | 21:37 | ||
| Geth | rakudo: 94995d60ec | MasterDuke17++ (committed using GitHub Web editor) | t/02-rakudo/05-range-in-range.t Fix typo in test description |
21:40 | |
| AlexDaniel | lucasb: feel free to move it | 22:28 | |
| some of the points were rakudo-related, and a bunch of labels don't exist in other repos | 22:29 | ||
| (like `6.e`) | |||
| haha “the own act of opening this in rakudo's repo already shows how confusing the current state is” | 22:30 | ||
| AlexDaniel craughs | 22:31 | ||
| Geth | nqp: ugexe++ created pull request #517: Add uname opcode for MoarVM |
22:40 | |
|
23:08
lucasb left
|
|||
| japhb | MasterDuke: Sure! What could possibly go wrong? ;-) | 23:23 | |
| MasterDuke | japhb: i like the way you think | 23:28 | |
| greppable6: use nqp | 23:29 | ||
| greppable6 | MasterDuke, 96 lines, 65 modules: gist.github.com/8a6ffc031441866dff...876113b816 | 23:30 | |
| MasterDuke | at least two of those are false positives, hardly any need to worry | 23:31 | |
| japhb | MasterDuke: Any performance difference visible with your "trilogy of make-NQP-default-to-int branches"? | 23:38 | |
| MasterDuke | japhb: i'm somewhat surprised with myself, but i haven't actually tried any benchmarks | 23:49 | |
| japhb | Well, now you've got it working, so you can test whether you also made it fast. :-) | 23:50 | |
| MasterDuke | i did jit the new nqp::intify op, but there is some spesh code for the numify and strify ops that i haven't adapted yet | 23:56 | |