#raku-dev on 9 January 2021 - Raku Programming Language Log

00:16 lucasb left

JRaspass

@Xliff let me know how you get on with cro and https, raku.land ended up putting cloudflare infront because both ff and chrome (and never curl) would sometimes hang on it, i turned off h2 and it still did it so i took the easy way out, its still tls between cf and cro though.

01:06

Xliff

JRaspass: Sure thing. I'm only using it to serve JSON (now JSONP) though.

01:43

This has been a bit of a chore.

JRaspass

which bit was tricky?

01:44

Xliff

Oh...where to begin?

JRaspass

hah

01:45

Xliff

Getting SSL working (takes latest IO::Socket::Async::SSL), getting a persistent DB connection up (still times out after an unknown amount of idle activity, so I need to address that)

Now it's CORS, which is not specitically Cro, but then getting everything working together to deal with it.

01:46

JRaspass

is jsonp a stop-gap? haven't seen that used much anymore

Xliff

Yes. And it's still relevant due to CORS.

The other thing I've noticed is that the deeper you get into Cro, you realized there is more that has to be done.

01:47

Take logging, for example.

That just send things out to STDOUT and STDERR. If you want to do something common like.. oh... both, you will need to roll your own.

I'll have to see if I can do something with Cro::HTTP::Log::File and then submit a PR.

01:48

None of these are necessarily show stoppers, but more of an additional hurdle to satisfying your precise needs.

JRaspass

i ended up changing the logging middleware to give me timings

01:49

Xliff

Nice! How did you do that?

JRaspass

hacky, it abuses request.annotations - gitlab.com/raku-land/raku-land/-/b...og.rakumod

does use the nice role to simplify the code though, was thinking of sending a pr to cro for using that role in the real log::file

01:50

Xliff

Ah, I see!

01:51

Why not just add that as its own headeR?

JRaspass

its not what annotations is for afaik, oh i didn't think of that

Xliff

We might want to move this to #raku.

JRaspass

yep, whoops

02:10 Xliff left 02:21 Kaeipi left, Kaiepi joined 02:24 Kaiepi left 02:25 Kaiepi joined 02:42 kjp left 03:05 Kaiepi left 03:15 frost-lab joined 03:34 Kaiepi joined 04:07 Kaiepi left 04:53 JRaspass left 06:14 linkable6 left, evalable6 left 06:16 linkable6 joined 06:17 evalable6 joined 06:38 Altai-man joined 07:25 squashable6 left 07:27 squashable6 joined 08:16 Kaiepi joined 09:16 linkable6 left, evalable6 left 09:17 linkable6 joined, evalable6 joined

[Tux]

Rakudo v2020.12-37-gc039b28fc (v6.d) on MoarVM 2020.12-34-g14708e989

csv-ip5xs	0.826	-	0.838
csv-ip5xs-20	8.038	-	8.341
csv-parser	25.864	-	26.436
csv-test-xs-20	0.382	-	0.391
test	7.641	-	7.720
test-t	1.847	-	1.875
test-t --race	0.865	-	0.876
test-t-20	31.635	-	32.707
test-t-20 --race	8.737	-	9.854

09:18

10:20 sena_kun joined 10:21 Altai-man left 11:29 JRaspass joined 13:00 JRaspass left

sena_kun

releasable6, status

13:20

releasable6

sena_kun, Next release in ≈14 days and ≈5 hours. 1 blocker. Changelog for this release was not started yet

sena_kun, Details: gist.github.com/0dfc1f17c7f96501a3...bd7eec659c

13:36 frost-lab left 14:19 Altai-man joined 14:21 sena_kun left 15:09 JRaspass joined 15:12 b2gills left

Geth

rakudo: 65a5cce11c | (Stefan Seifert)++ | src/Perl6/Metamodel/ParametricRoleHOW.nqp
Fix potential deadlock on specialization of roles

NQPLock's protect method doesn't cope well with return statements. Doing a return from a protected block will leave the lock in place. The only reason why this hasn't surfaced as deadlock before is that the lock is using a reeantrant mutex and most of the time the same thread will do the specializations. Fix by simply not using return and instead move the remainder into an if statement. Fixes a hang of role R { }; (^100).hyper.map({R.HOW.specialize(R, Any)})