The web in a box - a next generation web framework for the Perl programming language

IRC log for #mojo, 2016-03-01

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:11 punter joined #mojo
01:14 bpmedley joined #mojo
01:20 asarch joined #mojo
01:29 zivester joined #mojo
01:47 bpmedley sri : I would donate to kickstarter for your Mojolicious + Minion features that have a price on them.. I wager a few other people would, as well.. :)
01:49 sri bpmedley: you can put a bounty on issues on bountysource
01:54 bpmedley Cool, if I contribute do y'all get most of it?
01:56 sri yes
01:56 sri https://www.bountysource.com/teams/mojolicious/issues
01:58 bpmedley Woohoo!  Just contributed
01:59 sri bpmedley++
02:02 sri you can also post bounties directly on issues, if you want something specific to get done https://www.bountysource.com/issues/30414668-admin-ui
02:03 sri anyone can then work on them and claim them afterwards
02:06 sri ibm has a lot of bounties https://www.bountysource.com/teams/ibm
02:13 lluad joined #mojo
02:20 sri bpmedley: btw. i don't think there is actually enough people to get one of the big features fully funded
02:21 bpmedley Understood
02:21 sri a few bigger companies would have to chip in
02:26 sri honestly, i think the only way to make it work would be the sidekiq way
02:27 sri mojolicious enterprise :)
02:28 bpmedley That would be sweet..
02:31 sri bpmedley: what minion features are you interested in anyway?
02:33 bpmedley I moved physical locations and the perl shop I'm at now uses dancer.  So, before we use Minion I need a way to start using Mojo at work.. and I'm trying to get Mojo::UserAgent added to the test suite.
02:33 sri ouch
03:15 acamari joined #mojo
03:26 good_news_everyon joined #mojo
03:26 good_news_everyon [mojo] kraih created alpn (+1 new commit): https://git.io/v2K1H
03:26 good_news_everyon mojo/alpn a7602f7 Sebastian Riedel: add ALPN support
03:26 good_news_everyon left #mojo
03:27 sri that means #888 is done :)
03:29 sri https://github.com/kraih/mojo/compare/alpn
03:29 sri there is no need to merge it yet
03:31 sri but it's there when we need it
03:36 good_news_everyon joined #mojo
03:36 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2KDJ
03:36 good_news_everyon mojo/master 1ee33fa Sebastian Riedel: we support RFC 7725
03:36 good_news_everyon left #mojo
03:40 sri https://github.com/kraih/mojo/issues/888#issuecomment-190520892
03:50 inokenty joined #mojo
03:52 zivester joined #mojo
03:53 noganex joined #mojo
04:50 inokenty-w joined #mojo
05:54 chienjo joined #mojo
05:55 chienjo Hello guys. I have a problem with linux mint 17.3 Rosa. Everything was working fine until about a week ago when it started to freeze randomly from time to time. Especially when I am streaming a movie.
05:56 chienjo Plus, the normal alt+ctr+sysresq keys do not respond, only a hard reboot by the power button.
05:56 chienjo Sorry, wrong channel.
06:23 Vandal joined #mojo
06:42 hesh joined #mojo
07:05 dod joined #mojo
07:08 dod joined #mojo
07:25 pink_mist ... I was really starting to wonder what Mojo had to do with any of that until I saw the last line
07:26 Adura They're using a realtime kernel for Mojo's realtime web abilities...!
07:42 osfabibisi joined #mojo
07:44 batman The kernel has a name: Mojo::IOLoop :)
07:48 batman Hm.. Or is it Mojo::Reactor..?
08:16 AndrewIsh joined #mojo
08:38 trone joined #mojo
09:36 Jonis does mojo run fine on the raspberry pi?
09:36 Jonis I haven't actually tried
09:43 bpmedley Jonis : http://raspberrypi.stackexchange.com/questions/5513/does-raspbian-come-with-perl
09:44 Lee Jonis: installed mojo on my pi zero at the weekend no problem
09:50 anon joined #mojo
10:20 bpmedley Anyone use gitprep?
11:16 HtbaaPi_ joined #mojo
11:42 cpan_mojo MojoX-Renderer-JSON-XS-0.06 by YOWCOW https://metacpan.org/release/YOWCOW/MojoX-Renderer-JSON-XS-0.06
11:49 batman bpmedley: i was curious, but dropped it when i saw https://github.com/yuki-kimoto/gitprep/tree/master/mojo
11:54 pink_mist how long ago was it you dropped it? :P
11:55 bpmedley Egads batman, I wonder why it's such an old version
11:57 batman bpmedley: maybe same as convos...? lack of maintenance :P
11:58 batman (btw: convos is on the move these days)
11:58 batman or maybe the author is running the app on some old OS
11:58 bpmedley OS/2 Warp?
11:59 berov joined #mojo
12:01 batman hehe
13:10 asarch joined #mojo
13:13 kes joined #mojo
13:14 orev joined #mojo
13:28 kaare joined #mojo
13:52 diegok Jonis: I use mojo on several rpi's using a perlbrew perl :)
13:52 diegok ^ and raspbian
13:54 neilhwatson joined #mojo
13:55 dotan joined #mojo
13:56 ramortegui joined #mojo
14:00 jberger the new pi 3 looks nice
14:24 neilhwatson joined #mojo
14:26 Jonis jberger: yeah, that's what prompted it
14:26 Jonis I've already got a 2 somewhere though, guess I should dig that out
14:42 mcsnolte joined #mojo
14:42 * sri yawns
14:43 * jberger tosses a gummy bear in sri's open yap
14:43 * sri coughs
14:44 jberger you have to chew it!
14:45 pink_mist hope it isn't one of those that gives you the runs
14:45 sri sugar free *shudders*
15:44 geheimnis` joined #mojo
15:44 orev joined #mojo
15:47 * sri wonders if there are other uses for alpn than http/2
15:48 sri this is the old problem of adding Mojo::IOLoop features we don't actually need in core :S
15:49 sri (yet)
15:54 disputin joined #mojo
16:03 sue joined #mojo
16:11 PryMar56 joined #mojo
16:17 disputin joined #mojo
16:25 acamari joined #mojo
16:26 jberger again, if you think toolkit I think adding features is a nice thing
16:27 jberger I don't know too much about ALPN but it seems like a lot of the lifting comes from lower level modules (if I read your patch correctly)
16:33 jabberwok LPT: "toolkits" is not a good word to Spoonerize with your boss lady.
16:38 sri jberger: correct, but it's still 8 lines
16:39 sri and i'm a little unsure about $stream->protocol, name and implementation
16:42 sri and existence i guess
16:43 sri we already use $stream->handle->sockport and friends in Mojo::UserAgent/Mojo::Server::Daemon
16:43 sri $stream->handle->alpn_selected would fit too
16:45 sri hmm, no, we just don't need alpn support
16:51 jberger is alpn required for http/2?
16:51 sri yes
16:51 jberger so far I've really gotten by with doing ssl termination in nginx
16:51 jberger and I think (if I'm reading this correctly) that alpn comes into the same sphere as that
16:52 jberger but if we want native http/2 (which I think would be a cool selling point for mojo) then if we are going to need it, why not have it now?
16:52 genio https://github.com/http2/http2-spec/wiki/Implementations   ALPN, NPN, I'm not sure what Upgrade and Direct mean
16:53 sri you can theoreically do http/2 in other ways, but browsers require alpn
16:54 sri Upgrade means the Upgrade HTTP/1.1 header
16:54 sri like WebSockets use for the handshake
16:55 a6502 joined #mojo
16:55 sri direct means no handshake at all, your client and server only speak http/2
16:55 sri nobody does that though, so it might as well not exist
16:56 sri npn was the first prototype of alpn, and is on the way out, chrome already removed it i believe
16:57 genio https://www.imperialviolet.org/2013/03/20/alpn.html describes the difference pretty clearly for me
16:57 genio between npn and alpn, that is
16:59 bradjm joined #mojo
17:04 lluad joined #mojo
17:04 disputin joined #mojo
17:11 sri jberger: i'm not entirely sure if we'll ever actually get http/2 support
17:13 sri we'd prolly be fine without it for a very very very long time
17:14 sri sure it would be cool, but what are the odds that one of us will ever have the 2-3 months of free time it would take to implement
17:20 sri using a reverse proxy to handle http/2 is just too easy
17:25 sri and i think i might have been the only one who wanted to do it, but now even with funding i prolly wouldn't have the time anymore
17:27 sri the basic infrastructure is there though, so who knows, maybe someone will step up and surprise us with an amazing pull request :)
17:29 taichi joined #mojo
17:53 sri guess alpn would only be interesting for us if websockets could be negotiated with it
17:53 sri but that's not gonna happen, too much investment in the Upgrade handshake
17:54 sri cookies, subprotocol negotiation....
17:54 sri even extensions are a header field
17:59 jberger it does feel like that doesn't it
18:14 * sri closes 2 of the 3 remaining http/2 issues
18:15 punter joined #mojo
18:26 dotan left #mojo
18:29 dod joined #mojo
18:30 taichi joined #mojo
18:34 sri well, we got a lot of value out of the http/2 effort, countless bugfixes and performance improvements all over the codebase
18:34 sri even if it ends with this ;p https://github.com/kraih/mojo/issues/423#issuecomment-190842909
18:34 sri still worth it
18:45 good_news_everyon joined #mojo
18:45 good_news_everyon [mojo] kraih deleted polyfill_signatures at 6e2b0b8: https://git.io/v2P8k
18:45 good_news_everyon left #mojo
18:46 good_news_everyon joined #mojo
18:46 good_news_everyon [mojo] kraih deleted without_formats at fd93399: https://git.io/v2P8Z
18:46 good_news_everyon left #mojo
18:47 good_news_everyon joined #mojo
18:47 good_news_everyon [mojo] kraih deleted channels/ws at 61b3fdc: https://git.io/v2P8R
18:47 good_news_everyon left #mojo
18:49 bjoernfan joined #mojo
18:59 orev what is the use of steady_time and why would I need to use it?  is it better than the builtin 'time' function for some things?
19:00 jberger orev: it is a different purpose
19:01 jberger The main problem is when the clock is changed
19:01 jberger Like hitting daylight savings or the user changes the clock
19:01 jberger If you rely on that clock for checking timestamps as in
19:01 jberger "Do this thing in 10s"
19:02 jberger Then maybe it never happens
19:02 orev ok.  my servers use utc so maybe not a big concern, but I can see what it's for.
19:03 sri it's about time jumps in general
19:04 sri even accidental time jumps
19:04 sri and specifically those moving backwards in time
19:04 orev like ntp updates and vm drifts, etc..
19:04 sri which matters a lot for event loops
19:04 sri yea
19:05 sri say you have a Mojo::IOLoop->timer(15 => sub {...})
19:05 sri internaly it stores time + 15
19:05 sri now if your system clock jumps back a year, you're screwed
19:06 sri steady_time can prevent that, because it uses a monotonic clock if available
19:07 taichi joined #mojo
19:11 disputin1 joined #mojo
19:11 taichi joined #mojo
19:11 sri oh, does wireshark have a real os x ui now?
19:11 * sri just saw a screenshot on https://blog.newrelic.com/2016/02/17/http2-production/
19:13 sri oh my, apparently for about a year now :O
19:14 disputin joined #mojo
19:16 taichi joined #mojo
19:17 sue joined #mojo
19:18 jberger this article is really cool
19:18 jberger https://www.nginx.com/blog/nginx-1-9-5/
19:18 jberger thanks for sharing that
19:18 jberger so easy
19:25 sri indeed
19:26 sri wireshark for os x is really good
19:26 sri although, it's not hard to beat the old x11 ui... ;p
19:27 genio That's a long list of blacklisted ciphers.  *glares evilly at MS for not updating their stuff and forcing me to continue to allow blacklisted some blacklisted ciphers*
19:28 bjoernfan joined #mojo
20:16 jberger yay got my convos installation (with its generated letsencrypt cert) now has an A grade from ssllabs
20:16 jberger s/got//
20:17 taichi joined #mojo
20:19 mdom Woah, i was already writing tests to confirm that Mojo::Date can't handle fractional seconds in rfc3339 but then i discovered that my version of Time::Local is the culprit ...
20:41 asarch joined #mojo
21:17 sri debugging http/2 with wireshark is still pretty clunky though
21:17 sri due to tls
21:18 jberger oh hmm, I guess that that plus the framing makes it quite tough
21:18 sri not so much the framing, but the compression
21:18 sri miss one frame, and you can't debug anything in the connection
22:15 sue joined #mojo
22:57 sri haha, guess i'll vote -1 on this because of the semicolon ;p https://github.com/kraih/mojo/pull/924
23:05 Grinnz_ lol
23:05 Grinnz_ for bikeshedding I was just going to say the last "which is a..." phrase could be omitted
23:06 sri agree actually
23:06 sri it doesn't really add value
23:32 hernan604 joined #mojo
23:34 taichi joined #mojo
23:43 jberger Note that lindleyw == jabberwok
23:44 jberger jabberwok: o/
23:52 jabberwok jberger blows my cover
23:52 jabberwok 924 revised
23:53 jabberwok having both "K combinator" and "Kestrel" should give enough info for folks to search, and find the articles in the comments, without including actual references.
23:58 good_news_everyon joined #mojo
23:58 good_news_everyon [mojo] kraih pushed 3 new commits to master: https://git.io/v2X64
23:58 good_news_everyon mojo/master b60c6ac William Lindley: Clarify K combinator
23:58 good_news_everyon mojo/master 4bfbf4a William Lindley: Clarify K-combinator a.k.a. Kestrel
23:58 good_news_everyon mojo/master d4eac33 Sebastian Riedel: Merge pull request #924 from lindleyw/master...
23:58 good_news_everyon left #mojo
23:59 good_news_everyon joined #mojo
23:59 good_news_everyon [mojo] kraih pushed 1 new commit to master: https://git.io/v2X61
23:59 good_news_everyon mojo/master e64bcc2 Sebastian Riedel: fix formatting
23:59 good_news_everyon left #mojo

| Channels | #mojo index | Today | | Search | Google Search | Plain-Text | summary