00:16 Guest14 left
aruniecrisps the plot thickens 00:22
my distribution is showing up on 360.zef.pm/
but it's not showing up on raku.land
incredibly bizarre 00:23
00:56 lichtkind left
oof i really bungled the release of this package didnt i 00:58
01:31 ingy left 01:32 ingy joined 01:41 ACfromTX left, ACfromTX joined
@tbrowder_04810 do you think you could help? 02:06
.landyacht. Voldenet - I'm not changing the string, just passing a new one into the same label-setting routine which is supposed to take care of the memory management from there 02:33
interesting that you got a better result after assigning into a variable... I'm also not assigning it into anything, so I wonder if that's the problem 02:34
nope, changed it up and still getting the same issue 02:36
the C example code they give does use OPENSSL_malloc instead of standard malloc, so I wonder if I have to do that...
03:05 Aedil joined
Got it working using CRYPTO_malloc from libcrypto and then copying the label data in (from $label.encode('utf8')) using mempcy from libc, and treating everything as Pointer[uint8] on the Raku side - whether that fixed it by preventing Moar from GCing, or because libcrypto just demands stuff be alloc'd with its own malloc version, I don't know 03:18
03:21 stanrifkin joined
Voldenet you could check if regular malloc works, but it should 03:24
Yes, I'm guessing that specific mallocs do something differently, but in the end you get a pointer and that's it 03:25
it should only matter if you don't explicitly call free, but the lib does this for you
er the change - explicitly-manage returns CStr and free remains the original Str with mixed in ExplicitlyManagedString 03:28
erm, I just read what I wrote and it makes no sense
regarding the difference between the var and return – explicitly-manage returns CStr and its parameter gets ExplicitlyManagedString mixed in 03:29
.landyacht. yeah, I am leaning toward thinking this is a bug with explicitly-manage not doing what it says on the tin, but it's late so I'll check what results regular malloc gets me tomorrow to confirm 03:47
04:00 stanrifkin left 04:53 derpydoo joined 05:59 derpydoo left 06:59 broquain1 left, broquaint joined 08:15 stanrifkin joined 08:31 crnlskn joined 08:39 crnlskn left 09:14 MoC joined 09:43 apac joined 09:49 Sgeo left
ds7832 m: your code goes here #raku channel evalbot 10:57
camelia ===SORRY!=== Error while compiling <tmp>
Undeclared routines:
code used at line 1
goes used at line 1
here used at line 1. Did you mean 'WHERE'?
your used at line 1
11:03 lichtkind joined
melezhik. Made few mentions of Sparky on relevant fosstodont posts 11:38
fosstodon.org/@melezhik/115242052395806133
fosstodon.org/@melezhik/115242008926072373
12:03 apac left
tbrowder aruniecrisps: did you use "mi6 publish" to publish your repo? 12:10
lizmat you mean "mi6 release" ? 12:13
tbrowder yes, thnx
12:19 dg left, dg joined
tbrowder aruniecrisps: if you did *not* use mi6 to publish, i suggest you make some small change and republish using "mi6 release". note your README may get updated in the process unless you comment out the line in the "dist.ini" file about the "readme from". a leading semicolon should take care of that. 12:29
12:40 crnlskn joined 13:07 lizmat left, lizmat joined
.landyacht. Voldenet - so yes, it does also work fine with regular malloc. At some point I'll try to pare this down to a minimal example with both C and Raku sides and submit an issue 13:24
13:26 apac joined 13:29 crnlskn left
okay, so a very minimal example can reproduce the issue, which is good - I'm just not sure if this is a Rakudo or MoarVM bug now 13:34
github.com/rakudo/rakudo/issues/5955 13:55
14:20 apac left 14:24 apac joined 14:32 apac left 15:23 ACfromTX left
aruniecrisps @tbrowder_04810 for version 0.0.1 i used fez publish, and then for 0.0.2 I used mi6 release 15:32
both versions are showing up on 360.zef.pm but neither are showing up on raku.land 15:33
15:36 ACfromTX joined
tbrowder weird 16:10
time to have the big guns go to work to save the day 16:12
16:42 apac joined 17:04 apac left 17:10 yewscion_ joined 17:13 Aedil_ joined 17:17 yewscion_ left 17:18 Aedil left, gabiruh_ left, yewscion_ joined 17:24 gabiruh_ joined
Voldenet If I really wanted truly unmanaged string, I'd never pass any GC-managed object (Str) to the native sub - it might work in some cases but it'd be very error-prone IMO, I'd rather allocate bytes and strncpy or memcpy manually 17:34
aruniecrisps @tbrowder_04810 what do you mean? 17:54
if there's a way to take those distributions off the list so i can revert back to version 0.0.1 that would be amazing 17:58
.landyacht. Voldenet - yeah, I think that’s my takeaway going forward 18:09
18:32 abraxxa-home joined 18:40 abraxxa-home left 18:43 abraxxa-home joined 18:46 Aedil_ left 19:04 Sgeo joined
tbrowder no, i don’t think you can do that. when i said “big guns” i meant lizmat, ugexe, tonyo, [Coke], etc. 19:19
19:20 abraxxa-home left 19:21 abraxxa-home joined
aruniecrisps ah understood 19:23
thanks for the help
librasteve weekly: fosstodon.org/@melezhik/115242052395806133 19:42
notable6__ librasteve, Noted! (weekly)
20:00 elcaro left, sivoais left, kst left, elcaro joined 20:01 kst`` joined 20:02 sivoais joined
ugexe aruniecrisps 20:03
raku.land/build
/cache/zef/93c2c3331961d83405b44b4f9a8be7fa051036dc: invalid input value for enum license: "Hippocratic 20:04
that seems related
presumably raku.land only allows distributions with certain licenses (i have no idea) 20:06
aruniecrisps oof 20:13
gotta change the license then
at least in the build file 20:15
20:15 librasteve_ joined 20:22 stanrifkin left
okay i changed the license in META6.json 20:24
well it seems to be showing up in 360.zef.pm, not sure why it's still not indexing 20:26
cdn.discordapp.com/attachments/633...de3d4&
apologies folks 😅 didn't mean to gum up the system 20:27
20:29 apac joined 20:35 oodani left 20:36 oodani joined, greppable6__ left, notable6__ left, nativecallable6_ left, tellable6 left, bisectable6 left, unicodable6 left, linkable6 left, evalable6 left, committable6 left, coverable6 left, releasable6 left 20:39 nativecallable6 joined, sourceable6 joined, evalable6 joined, linkable6 joined, quotable6 joined 20:40 unicodable6 joined, coverable6 joined, committable6 joined, bloatable6 joined, shareable6 joined, benchable6 joined
FINALLY: raku.land/zef:arunvickram/DataStar 20:40
20:40 greppable6 joined
looks like it's fixed 20:40
20:41 bisectable6 joined 20:42 tellable6 joined, releasable6 joined, notable6 joined
librasteve \o/ 20:44
btw - I see that raku.land is open to many licences raku.land/stats so not sure why this is a problem ... maybe open an issue over there if that's important to you... 20:45
aruniecrisps do you have the repo where i can place the issue? i just googled raku land repo and it doesn't seem to be popping up for me for some reason 20:50
librasteve the link is on raku.land 20:52
gitlab.com/raku-land/raku-land
hmmm not sure how to file issues on GitLab
tbrowder my problem too. gitlab be strange duck. 20:55
librasteve oh - its in the left sidebar 20:56
aruniecrisps just added it 20:58
added an issue to the backlog i mean
librasteve =b 20:59
21:14 MoC left 21:24 zuphinx joined 21:35 oodani left 21:36 oodani joined 21:52 abraxxa-home left
SmokeMachine weekly: dev.to/fco/typed-named-endpoints-f...elpers-g9j 22:15
notable6 SmokeMachine, Noted! (weekly)
22:42 guifa_ joined 22:43 guifa left 23:10 apac left 23:37 librasteve_ left 23:45 guifa_ left 23:46 guifa joined 23:59 lichtkind left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!