github.com/moarvm/moarvm | IRC logs at colabti.org/irclogger/irclogger_logs/moarvm
Set by AlexDaniel on 12 June 2018.
00:07 lucasb left 00:25 sena_kun left 01:21 Voldenet joined, Voldenet left, Voldenet joined 02:30 squashable6 joined 03:06 ZzZombo_ joined 03:08 ZzZombo_ is now known as ZzZombo 04:13 squashable6 left 04:15 squashable6 joined 04:38 angelds joined 05:17 angelds left 05:18 angelds joined 05:39 angelds left 05:40 angelds joined 05:47 angelds left, angelds joined 05:52 angelds left 05:56 robertle left
nine dogbert17: 329f669c and ea9b29e8 can be merged into 1 changelog entry 05:56
Personally I'd prefer a rather more condensed ChangeLog that focused more on the user visible impact. E.g. several entries could be summed up as "destroying threads no longer leads to obscure crashes". 05:59
openSUSE restricts changelog entries for RPM files to IIRC 30 lines which always forces me to condense entries. This is actually what takes most of my time when packaging rakudo and it's tedious work, but the result really feels much more useful. 06:00
For example as a user I'd not have the slightest idea what "Oops don't overwrite fields; add a new one instead" does for me 06:01
06:01 colomon joined
nine It probably fixes a bug that was introduced after the previous release, so not something I'd have encountered anyway 06:02
06:05 ZzZombo left 06:23 ZzZombo joined 06:38 ZzZombo_ joined 06:40 ZzZombo left, ZzZombo_ is now known as ZzZombo 08:49 sena_kun joined 09:01 robertle joined 09:08 Altai-man_ joined 09:11 sena_kun left 10:12 squashable6 left, squashable6 joined 11:10 sena_kun joined 11:11 Altai-man_ left 11:30 Kaeipi joined 11:31 Kaiepi left 11:41 angelds joined
dogbert17 nine: I've fixed some of the things you pointed out, however the changelog is still rather large 12:14
timotimo, jnthn, MasterDuke: any comments wrt the Changelog draft? gist.github.com/dogbert17/1c151a7a...d852a18d36 12:15
and samcv of course 12:18
sena_kun dogbert17, I am not an expert, but is it necessary to mention every single commit? if yes, I think small ones can be gathered under a single entry, e.g. "C99 fix", "Fix duplicate includes", they are mostly style, OTOH "Experiment to develop new heapsnapshot format" <- there are a bunch of them and it can be written as "Preparation work for new heapsnapshot format". 12:23
dogbert17 sena_kun: I'm no expert either but I'll try to implement your suggestions 12:24
sena_kun for example,"Add some safety checks for macOS due to homebrew binutils and Xcode incompatibilities" is easy to understand for the user, but "[8efd3825] Make scripts executable" <- what scripts?
"Spesh bisect should learn about PEA" <- it certainly should, maybe "Add better support of PEA to spesh bisect" 12:25
dogbert17 I definitely could remove some of the more vague commits, hopefully I won't piss anyone off
sena_kun most of them are looking nice, dogbert17++
"Add obviously missing break; to switch statement in instrument_graph" <- "Fix a subtle bug in instrument_graph" 12:26
dogbert17 sena_kun: please go on 12:30
sena_kun "Fix incorrect conditional" <- it might be hard to understand what bug or warned was caused by the commit alone, but at least mentioning a place can make things a bit more clear 12:31
"Give error message improvement to libffi as well" <- not sure about that, maybe "Improve error message for libffi"? Can it be merged with "Add more context to error messages" before that? 12:33
the same for the next line
"If getaddrinfo in debugserver fails, get better error" <- "Better error if ..."
dogbert17 cool, will fix
sena_kun "Make debugserver errors/output clearer with "Debugserver:"" <- not very clear what "Debugserver:" means in this context... Like a string is prepended for easy grepping? Or something related to envvars responsible for debugserver logging? 12:34
"Refer to Rakudo as just “Rakudo”" can be just "Improve naming in documentation", I think...
"New GC debug level" <- maybe write a bit more? Like how to invoke it, --debug=2, --debug=10, --debug-level=super-tiny? 12:36
gist.github.com/dogbert17/1c151a7a...xt-L83-L84 <- these two can be merged
"Actually make async sockets work again" <- and some below that can be merged into... we need to look at Kaeipi++ PRs to think of some short and clear description of what was done 12:37
"Fix for M#1172, possible uid/gid confusion" <- confusion of what?
synopsebot M#1172 [open]: github.com/MoarVM/MoarVM/issues/1172 possible uid/gid confusion
sena_kun "Jit some num ops" <- can we specify which ones? 12:38
"Make jit-dump.pl useful again" <- no idea if it can be made more clear... 12:39
"Say hello to MVM_CONFPROG_DEBUG env var" <- hm? If it is about introducing it, then "Introduce new env variable ... which enables ...", if something else, then no idea 12:41
as I wrote already, gist.github.com/dogbert17/1c151a7a...-L194-L208 can be merged into a single item, probably 12:42
gist.github.com/dogbert17/1c151a7a...1-txt-L251 <- "Fix clang warnings in GB2313 implementation"? 12:43
also, possibly gist.github.com/dogbert17/1c151a7a...-L251-L254 can be merged
dogbert17 Excellent feedback, truly appreciated 12:44
sena_kun + [986f3403] Add GB18030 encode, decode and decodestream support and + [73f28d38] Add GB2312 encode, decode and decodestream support are duplicates
ouch, they are not, sorry
"Configure.pl: I assume I forgot this line in has_native_library" <- "I want to eat pizza right now...", no, stop, not this one, "Fix ??? in Configure.pl" 12:45
phew, it is huge!
dogbert17++ for working on it, great job.
robertle sena_kun: I can answer the "possible uid/gid confusion"... 12:47
sena_kun robertle, if you can rephrase it to be clearer for the end user what was fixed for dogbert17, that'd be awesome, I think
dogbert17 sena_kun: if you refresh the gist I have fixed some of the things you pointed out, will continue after a short lunch 12:48
robertle perhaps "fixed a problem around group permissions that could lead to files/directories not being regarded readable or writable when in fact they are" 12:49
perhaps a tad too long
"file permission interpretation problems around groups"?
Kaeipi my pr mainly focused on issues related to dns resolution and gc for async sockets iirc 12:53
12:56 angelds left 12:57 angelds joined 13:09 Altai-man_ joined 13:11 sena_kun left 13:23 angelds left 13:34 robertle left 13:36 robertle joined
dogbert17 sena_kun: we're now down to 'only' 245 lines 14:25
tellable6 dogbert17, I'll pass your message to sena_kun
14:31 lucasb joined 14:44 Altai-man_ is now known as sena_kun
sena_kun 14:44
tellable6 2019-11-19T14:25:54Z #moarvm <dogbert17> sena_kun: we're now down to 'only' 245 lines
sena_kun dogbert17, I don't see what else can be reduced myself. last release was 4 month ago and having a long changelog is expected, I think 14:45
dogbert17 Cool, I'll PR it against the 'release-2019.11' branch then 14:49
sena_kun (even 5 month ago, I think) 14:54
nine dogbert17: the headlines will already help me a lot when condensing it further :) 14:56
Geth MoarVM: dogbert17++ created pull request #1213:
Update ChangeLog for release
dogbert17 phew 14:57
14:58 squashable6 left 15:00 squashable6 joined
nine dogbert17, sena_kun: thank you for working on this! 15:08
15:08 Altai-man_ joined 15:11 sena_kun left 15:36 vrurg_ joined, vrurg_ left, vrurg_ joined 16:08 robertle left, ZzZombo left 16:26 ZzZombo joined 16:53 domidumont joined 17:04 robertle joined 17:11 Altai-man_ left 17:58 brrt joined 18:39 brrt left 19:06 robertle left 19:47 domidumont left 20:48 brrt joined
brrt btw, nine++ 20:49
(for the fix in the unshift of ConcBlockingQueue) 20:50
I still confess I don't entirely get what could've gone wrong
21:23 Kaeipi left 21:25 Kaiepi joined 21:38 AlexDaniel joined, AlexDaniel left, AlexDaniel joined 21:39 domidumont joined 21:41 domidumont left 22:13 brrt left
samcv . 22:22
I think somebody sent me a message on here on #raku-dev. maybe guest##### 22:33
23:50 lucasb left