|
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. |
|||||||||||||||||||||||||||||||||||||||
|
01:47
Tux__ joined,
|Tux| left,
p6bannerbot sets mode: +v Tux__
06:12
ufobat_ joined
06:13
p6bannerbot sets mode: +v ufobat_
06:16
lizmat left
08:41
patrickb joined,
p6bannerbot sets mode: +v patrickb
08:43
robertle left
|
|||||||||||||||||||||||||||||||||||||||
| patrickb | o/ | 08:45 | |||||||||||||||||||||||||||||||||||||
| So, the gsoc mentor organisation applications are open for real now. (summerofcode.withgoogle.com/) | 09:11 | ||||||||||||||||||||||||||||||||||||||
| I have taken a look at how this works in general. First thing needed is a so called Organization Administrator. That's the one person that has the last call on most gsoc matters and also the one google contacts should the need arise. | 09:12 | ||||||||||||||||||||||||||||||||||||||
| I think it's important the org admin has some esteem in the community. | |||||||||||||||||||||||||||||||||||||||
| It seems the last gsoc perl participated in was 2014: www.google-melange.com/archive/gso...4/orgs/tpf | 09:13 | ||||||||||||||||||||||||||||||||||||||
|
09:13
lizmat joined,
p6bannerbot sets mode: +v lizmat
|
|||||||||||||||||||||||||||||||||||||||
| patrickb | mst: I think you have been involved in past gsocs. In your opinion, does it make more sense to keep perl 5 and 6 separate wrt to gsoc or try to cooperate and apply as a single organization? | 09:14 | |||||||||||||||||||||||||||||||||||||
|
09:14
Tux__ is now known as |Tux|
09:40
ExtraCrispy left
|
|||||||||||||||||||||||||||||||||||||||
| lizmat | Files=1265, Tests=88040, 383 wallclock secs (19.15 usr 6.12 sys + 2659.28 cusr 233.75 csys = 2918.30 CPU) | 09:55 | |||||||||||||||||||||||||||||||||||||
| Geth | rakudo: 836c18ae26 | (Elizabeth Mattijsen)++ | src/core/MixHash.pm6 Fix copy pasto Fixes test failures in S32-basics/xxKEY.t since 87b71c2758df5a7f5 |
10:22 | |||||||||||||||||||||||||||||||||||||
| nine | patrickb: just wanted to mention, that I feel your pain WRT library paths in rakudo: github.com/niner/Inline-Perl6/comm...02f5e1bc15 | 10:55 | |||||||||||||||||||||||||||||||||||||
| releasable6 | Next release in ≈3 days and ≈7 hours. 5 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft | 11:00 | |||||||||||||||||||||||||||||||||||||
| patrickb | nine: The approach I'm taking in my relocatable refactor is to introduce the concept of an nqp_home and perl6_home. Those point to share/nqp and share/perl6 respectively. Do you think this is the best approach? The code you just linked seems to indicate one is sometimes interested in setting only the lib dir and not the entire nqp/perl6 dir. | 11:02 | |||||||||||||||||||||||||||||||||||||
|
11:11
lizmat left
|
|||||||||||||||||||||||||||||||||||||||
| nine | patrickb: Linux distros usually don't like binaries in /usr/share. That's why the openSUSE packages for nqp set --libdir=/usr/lib64. We actually should do something similar with other precomp files | 11:55 | |||||||||||||||||||||||||||||||||||||
| patrickb | nine: They put the actual .moarvm files directly in /usr/lib64? That changes things. Ugh. | 12:41 | |||||||||||||||||||||||||||||||||||||
| The rakudo package on the other hand was left in tact. It even puts its nqp files in /usr/share/nqp/lib/perl6 | 12:49 | ||||||||||||||||||||||||||||||||||||||
| nine: That means the idea of an nqp_home is dead. | 12:59 | ||||||||||||||||||||||||||||||||||||||
| AFAIK the share/perl6 folder only contains three compunit repos (/, /site, and /vendor) and an empty /lib folder. Does that mean the share/perl6 folder in its entirety would be moved to /usr/lib64? Or would it be taken apart? | 13:03 | ||||||||||||||||||||||||||||||||||||||
| I fear the idea of a perl6_home with a fixed directory layout might also be unrealistic. :-( | 13:04 | ||||||||||||||||||||||||||||||||||||||
|
13:42
pmurias joined,
p6bannerbot sets mode: +v pmurias
|
|||||||||||||||||||||||||||||||||||||||
| nine | patrickb: yeah, these issues are _hard_ :/ | 13:48 | |||||||||||||||||||||||||||||||||||||
|
13:49
pmurias left
|
|||||||||||||||||||||||||||||||||||||||
| nine | patrickb: I actually think that even /usr/lib64 is the wrong directory as the .moarvm and precomp files are not architecture dependent. So /usr/lib seems to be the most appropriate | 13:49 | |||||||||||||||||||||||||||||||||||||
|
13:49
pmurias joined,
p6bannerbot sets mode: +v pmurias
13:53
pmurias left,
pmurias joined,
p6bannerbot sets mode: +v pmurias
|
|||||||||||||||||||||||||||||||||||||||
| AlexDaniel | patrickb: ooh gsoc, it would be lovely to make it happen | 13:57 | |||||||||||||||||||||||||||||||||||||
| patrickb: maybe [Coke] ? | 13:59 | ||||||||||||||||||||||||||||||||||||||
| patrickb | nine: The worst that can happen is that we have no guarantees about directory layout at all. Then, I think, the best we can do is to add three configuration options: 1. flat list of all libdirs, 2. flat list of comp repos, 3. path to main file (either perl6.moarvm or perl6-debug.moarvm). | 14:08 | |||||||||||||||||||||||||||||||||||||
| nine: I'll write some thoughts up on what approaches there might be. It'd be great if you could then have a look and give an opinion. | 14:16 | ||||||||||||||||||||||||||||||||||||||
| nine | patrickb: sure. Just keep in mind that I'm on vacation next week starting on Sunday and don't know how much time I'll have till then. So if you don't get any feedback in this time, that's why | 14:25 | |||||||||||||||||||||||||||||||||||||
| patrickb | nine: ok. OOC do you plan to go to FOSDEM this year? | 14:27 | |||||||||||||||||||||||||||||||||||||
| pmurias | 'use lib' is making tests fail in precompilation mode | 15:21 | |||||||||||||||||||||||||||||||||||||
| what do we want to do about that? | 15:22 | ||||||||||||||||||||||||||||||||||||||
| nine: is there a fundamental reason why 'use lib' doesn't work in precompilation mode or is it just unimplemented yet? | 15:23 | ||||||||||||||||||||||||||||||||||||||
| Geth | rakudo: 89fee86ace | (Paweł Murias)++ | 3 files Add a m-spectest5_precompile to run tests in precompilation mode Currently this fails for both tests that use 'use lib' or have precompilation only bugs. Fixing the 'use lib' part need further work. |
15:29 | |||||||||||||||||||||||||||||||||||||
| nine | patrickb: FOSDEM, yes :) | 15:36 | |||||||||||||||||||||||||||||||||||||
| pmurias: use lib changes the repository chain and therefore invalidates the dependency resolutions. This makes things quite complicated. We were extremely strict about it initially and IIRC we relaxed that somewhat | 15:39 | ||||||||||||||||||||||||||||||||||||||
|
15:46
ufobat_ left
|
|||||||||||||||||||||||||||||||||||||||
| Geth | rakudo: ec289a6c57 | (Paweł Murias)++ | t/harness5 Make m-spectest5_precompile skip runnings files that contain a 'use lib' |
15:49 | |||||||||||||||||||||||||||||||||||||
| [Coke] | I'm not a good gsoc resource. | 15:51 | |||||||||||||||||||||||||||||||||||||
| I am barely handling my current volunteer responsibilities with the TPF. | |||||||||||||||||||||||||||||||||||||||
|
15:52
ufobat_ joined
15:53
p6bannerbot sets mode: +v ufobat_
16:33
patrickb left
16:37
pmurias left
16:38
pmurias joined,
p6bannerbot sets mode: +v pmurias
16:48
ufobat_ left
|
|||||||||||||||||||||||||||||||||||||||
| |Tux| |
|
16:50 | |||||||||||||||||||||||||||||||||||||
|
17:05
pmurias left
17:06
pmurias joined,
p6bannerbot sets mode: +v pmurias
|
|||||||||||||||||||||||||||||||||||||||
| Geth | rakudo: bb36f04a68 | (Paweł Murias)++ | t/harness5 Fix bug in 'use lib' filtering in harness5 |
17:09 | |||||||||||||||||||||||||||||||||||||
|
18:07
patrickb joined,
p6bannerbot sets mode: +v patrickb
|
|||||||||||||||||||||||||||||||||||||||
| patrickb | nine: gist.github.com/patzim/710baab43da...3b5b81c923 | 18:33 | |||||||||||||||||||||||||||||||||||||
| please ignore the first file (gistfile1.txt) and just look at second_try.md below | |||||||||||||||||||||||||||||||||||||||
| Thank you! | |||||||||||||||||||||||||||||||||||||||
| ^ That's a small write up of where I'm currently struck with making rakudo relocatable. Whoever is interested in having a look, please do! I'm gratefull for feedback. | 18:35 | ||||||||||||||||||||||||||||||||||||||
|
19:01
patrickz joined,
p6bannerbot sets mode: +v patrickz,
patrickb left
19:27
robertle joined
19:28
p6bannerbot sets mode: +v robertle
19:36
patrickz left
20:32
pmurias left
20:34
pmurias joined,
p6bannerbot sets mode: +v pmurias
20:35
pmurias left
20:39
pmurias joined,
p6bannerbot sets mode: +v pmurias
21:32
lizmat joined,
p6bannerbot sets mode: +v lizmat
|
|||||||||||||||||||||||||||||||||||||||
| japhb | .ask nine Will Inline::Python support python3? Or will you EOL it next year? | 21:42 | |||||||||||||||||||||||||||||||||||||
| yoleaux | japhb: I'll pass your message to nine. | ||||||||||||||||||||||||||||||||||||||
|
21:45
pmurias left
22:02
pmurias joined,
p6bannerbot sets mode: +v pmurias
22:11
robertle left
22:18
pmurias left
22:19
pmurias joined,
p6bannerbot sets mode: +v pmurias
22:30
pmurias left
23:41
dct joined,
p6bannerbot sets mode: +v dct
|
|||||||||||||||||||||||||||||||||||||||