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

IRC log for #mojo, 2015-11-30

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

All times shown according to UTC.

Time Nick Message
00:16 Adura joined #mojo
01:24 woz joined #mojo
02:10 sawtooth joined #mojo
02:39 jberger not only are the prioritizations hard to follow, its a really silly scheme
02:58 bpmedley_ joined #mojo
03:12 noganex_ joined #mojo
03:41 kaare joined #mojo
04:08 voldemortensen joined #mojo
04:15 Zoffix joined #mojo
05:47 bpmedley joined #mojo
06:21 cpan_mojo Mojolicious-Plugin-HandlebarsJSRenderer-0.04 by RMZG https://metacpan.org/release/RMZG/Mojolicious-Plugin-HandlebarsJSRenderer-0.04
07:00 melo joined #mojo
07:15 dod joined #mojo
07:20 dod joined #mojo
07:27 McA joined #mojo
07:30 bpmedley joined #mojo
07:43 n16gel joined #mojo
07:55 jontaylor joined #mojo
07:57 woz joined #mojo
08:06 stephen joined #mojo
08:09 sri jberger++
08:13 Vandal joined #mojo
08:15 eseyman joined #mojo
08:22 AndrewIsh joined #mojo
08:23 trone joined #mojo
08:25 berov joined #mojo
08:26 bpmedley joined #mojo
08:27 meshl joined #mojo
08:29 sri btw. the header parser/builder from Mojo::Headers can't be moved to the channel layer because it gets reused for multipart bodies
08:29 sri and the start-line builder actually gets used for nph mode in Mojo::Server::CGI too
08:30 sri so keeping them in the container classes makes sense there
08:30 sri otherwise i'm all for moving protocol stuff to the channel layer
08:31 sri of course there's also the debug aspect, it's kinda nice to be able to stringify an http request or response for debugging, and i think the http/1.1 representation will be used for that for a long time
08:38 woz joined #mojo
08:49 woz joined #mojo
08:51 odc joined #mojo
09:08 punter joined #mojo
09:11 woz joined #mojo
09:17 woz joined #mojo
09:17 osfabibisi joined #mojo
09:22 vytas joined #mojo
09:23 Jonis -win 4
09:26 cpan_mojo Toadfarm-0.71 by JHTHORSEN https://metacpan.org/release/JHTHORSEN/Toadfarm-0.71
09:30 n16gel joined #mojo
09:42 woz joined #mojo
09:43 eseyman left #mojo
09:44 eseyman joined #mojo
09:48 woz joined #mojo
10:03 marty joined #mojo
10:21 vanHoesel joined #mojo
10:26 woz joined #mojo
10:38 woz joined #mojo
10:48 marty joined #mojo
10:50 woz joined #mojo
10:54 woz joined #mojo
11:02 melo joined #mojo
11:04 vanHoesel joined #mojo
11:06 vanHoesel1 joined #mojo
11:09 marty joined #mojo
11:12 cpan_mojo AnsibleModule-0.01 by MRAMBERG https://metacpan.org/release/MRAMBERG/AnsibleModule-0.01
11:16 cpan_mojo AnsibleModule-0.1 by MRAMBERG https://metacpan.org/release/MRAMBERG/AnsibleModule-0.1
11:26 marcusr ¯\_(ツ)_/¯
11:26 dysfun left #mojo
11:31 woz joined #mojo
11:34 Zoffix heh
11:38 marcusr accidentally the wrong version
11:41 preaction joined #mojo
11:49 berov joined #mojo
11:56 marty joined #mojo
12:04 neilhwatson joined #mojo
12:19 meshl joined #mojo
12:33 marty joined #mojo
12:41 woz joined #mojo
12:44 kes joined #mojo
12:50 marty joined #mojo
12:52 sugar joined #mojo
12:57 woz joined #mojo
13:07 cpan_mojo Mojo-SNMP-0.11 by JHTHORSEN https://metacpan.org/release/JHTHORSEN/Mojo-SNMP-0.11
13:11 ajr_ joined #mojo
13:11 moritz everybody wants to speak SNMP!
13:15 batman moritz?
13:15 batman jberger: ^ 0.11: "Fix infinite bulk_walk and walk"
13:16 moritz batman: just a comment regarding cpan_mojo's announce of Mojo-SNMP-0.11
13:16 batman i didn't understand your comment :/
13:23 bpmedley joined #mojo
13:25 marty joined #mojo
13:29 cpan_mojo Test-Clustericious-Cluster-0.24 by PLICEASE https://metacpan.org/release/PLICEASE/Test-Clustericious-Cluster-0.24
13:36 kaare joined #mojo
13:38 n16gel joined #mojo
13:46 woz joined #mojo
13:51 bwf joined #mojo
14:17 woz joined #mojo
14:23 dod joined #mojo
14:45 tencendur joined #mojo
14:50 woz joined #mojo
14:51 stephen joined #mojo
14:53 jberger batman: cool
14:54 jb360 joined #mojo
14:56 marty joined #mojo
14:56 nic I'm really liking that cached templates show their proper name in the log
14:57 nic sri++
14:57 woz joined #mojo
15:00 bpmedley /msg Phil21 Hey, can you login to irc.freenode.org?
15:00 bpmedley haha
15:03 jberger bpmedley++
15:03 jberger and btw yes I'm on
15:03 bpmedley I'm getting a lot of timeouts.  Hrmm.
15:09 asarch joined #mojo
15:09 bpmedley joined #mojo
15:16 voldemortensen joined #mojo
15:20 asarch Is FreeNode down?
15:21 asarch Maybe an scheduled service
15:22 * jberger is using freenode currently
15:23 jberger maybe it matters that I've been connected for a while
15:23 jberger possible that they aren't accepting new connections?
15:23 stephan48 i think it is flapping a bit at the moment
15:23 stephan48 it was down for me yesterday i think
15:23 stephan48 and somewhen over night it resolved
15:24 woz joined #mojo
15:32 sri jberger: thinking some more about http/2, your design might work if the transaction got a stream attribute where the id is stored
15:32 sri the channel would be shared by multiple transactions
15:33 jberger so the connection == channel and stream == transaction (http/2 vs mojo parlance)?
15:34 sri not sure what you mean with ==
15:34 sri the channel has to keep track of all active streams"
15:35 jberger in that article you linked to me last night
15:35 sri so there would be a data structure where the compression context for each stream is stored
15:35 sri in the channel
15:36 jberger are we saying that what you/we are calling the channel is what the article calls the stream?
15:36 sri no
15:36 sri channel is an arbitrary concept for protocol abstraction
15:37 sri a channel is tied to the connection
15:37 sri not to a stream
15:37 sri the channel manages all streams on the connection
15:37 sri (tcp connection)
15:39 woz joined #mojo
15:39 sri so, the channel object couldn't have a tx attribute
15:39 sri since there's a bunch of transactions tied to the channel
15:40 sri so, a transaction having a channel attribute makes sense this way
15:40 sri the opposite not so much
15:42 sri anyway, bottom line is that we need a design where each tcp connection has exactly one channel object, which may have one or more transactions
15:43 sri on the server side at least
15:43 sri client side we prolly will never do concurrent transactions
15:44 sri you know what would be cool too? if one of us actually did some research on the topic ;p
15:45 sri like looking into how curl, nginx or firefox abstract protocols
15:46 sri or closer to our use cases, netty and node.js
15:52 jberger "you know what would be cool too? if one of us actually did some research on the topic ;p" why let something like that get in the way?!
15:52 jberger :D
15:53 sri i fear the correct way of handling more protocols is a total rewrite of Mojo::UserAgentand Mojo::Server::Daemon
15:54 sri with a protocol layer n top of Mojo::IOLoop::Stream
15:54 sri *on
15:54 stephan48 research? that is nonsense. it will just lead to edjucated decisions and that is just not good
15:55 jberger sri: eeek
15:59 * sri shrugs
16:00 marty joined #mojo
16:00 sri i guess as long as a channel can have multiple transactions we'll be fine though
16:04 woz joined #mojo
16:14 woz joined #mojo
16:18 lluad joined #mojo
16:21 gryphon joined #mojo
16:26 sri jberger: looks like your branch still has few big problems https://travis-ci.org/kraih/mojo/jobs/93835485
16:29 sri crab: have you thought about making your branch public too?
16:29 jberger sri: it blows up on websockets
16:29 jberger that's known
16:29 jberger its no where near done
16:30 meshl joined #mojo
16:31 marcusr it is known
16:33 marcusr http://cdn.meme.am/instances/53532811.jpg
16:33 jberger it is known
16:33 cpan_mojo AnsibleModule-0.2 by MRAMBERG https://metacpan.org/release/MRAMBERG/AnsibleModule-0.2
16:35 sri the more i look at the branch, the more i want to try another approach :)
16:36 sri a module that gets attached to the stream here https://github.com/kraih/mojo/blob/master/lib/Mojo/Server/Daemon.pm#L172-L181
16:36 sri and takes over the connection
16:38 jberger sri: be my guest
16:39 sri wish i had the time :(
16:40 sri so frustrating not being able to participate in such an important refactoring task
16:40 sri this will decide the future of mojolicious i believe
16:41 sri no pressure :)
16:41 marcusr lol
16:41 marcusr don't fuck it up, jberger!
16:42 jberger marcusr: we'll see when the next time I have the time to work on it comes around
16:42 jberger I gave it most of a weekend day and got that far
16:42 jberger (not very far)
16:44 marcusr jberger: after this pep talk, I'm sure it will be never ;)
16:45 jberger there are some really compelling things in this refactor
16:46 jberger but it isn't a cakewalk
16:46 marcusr https://www.stthomas.edu/media/star/thumb_image.jpeg
16:59 woz joined #mojo
17:07 jnbek joined #mojo
17:14 woz joined #mojo
17:18 good_news_everyon joined #mojo
17:18 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vB7VI
17:18 good_news_everyon mojo/master d463cfc Sebastian Riedel: link to implementations
17:18 good_news_everyon left #mojo
17:45 ZoffixW joined #mojo
17:46 ZoffixW is perl MyApp.pl daemon the standard way to run Mojolicious on Windows? I'm having some issues with random freezes :/
17:51 woz joined #mojo
17:52 ZoffixW Random 4-5-second latency before I see "Rendering template foobar" and the "200 OK" in the log output :S
17:52 preaction i think also prefork
17:52 sri marcusr: dude, it's time you commit something again! your last commit was may 2014 :o
17:52 marcusr sri: wow mom wow
17:53 ZoffixW k, gonna try prefork
17:53 ZoffixW "Preforking is not available for Windows"... well, at least that's conclusive :P
17:53 marcusr because forking
17:54 ZoffixW yeah, I figured
17:54 Grinnz_ windows doesn't fork, it spoons
17:54 sri also pipes
17:54 Grinnz_ or threads-ish
17:54 marcusr oh danny boy / the pipes the pipes are calling
17:55 marcusr From glen to glen /and down the mountain side
17:55 ZoffixW I'm starting to think convincing the IT department to let me install Perl on this Windows server as opposed to paying for an SSH cert to do CORS to a linux box was a bad idea.
17:56 marcusr You should have convinced them to install linux on it instead
17:56 Grinnz_ wipe all; install linux
17:56 marcusr no rugrats
18:23 good_news_everyon joined #mojo
18:23 good_news_everyon [mojo] kraih tagged v6.34 at 2b9f999: http://git.io/vB7hF
18:23 good_news_everyon left #mojo
18:25 good_news_everyon joined #mojo
18:25 good_news_everyon [mojo] kraih pushed 2 new commits to master: http://git.io/vB7jy
18:25 good_news_everyon mojo/master cf103c8 Sebastian Riedel: bump version
18:25 good_news_everyon mojo/master 0ac5726 Sebastian Riedel: the return value is not important
18:25 good_news_everyon left #mojo
18:29 sh4 joined #mojo
18:33 trone joined #mojo
18:40 jberger CORS isn't so bad and SSH certs are about to be free
18:41 sawtooth joined #mojo
18:44 woz joined #mojo
18:47 dod joined #mojo
18:47 woz joined #mojo
19:08 disputin joined #mojo
19:15 neilhwatson joined #mojo
19:35 woz joined #mojo
19:48 woz joined #mojo
19:49 PryMar56 joined #mojo
19:50 pink_mist yes! not long now until letsencrypt goes into public beta!
20:09 jberger its already in semi-public beta
20:10 disputin joined #mojo
20:10 pink_mist it's in limited beta
20:15 lluad I'm in the beta. It's ... pretty damn good.
20:16 * lluad blogged about it today, in fact :)
20:16 pink_mist link? =)
20:16 lluad Uh .... https://blog.wordtothewise.com/ should get there.
20:17 lluad https://wordtothewise.com/blog/ I mean. Oops. I should fix that...
20:18 Grinnz_ google domains has easy subdomain forwarding, i might use it for that someday if i get lazy enough
20:18 Grinnz_ actually you know what, i should do that
20:19 lluad Yeah. I'm redirecting blog. on http, but not on SSL, apparently.
20:19 * lluad goes to have words with nginx
20:20 Grinnz_ hmmm can't tell it to redirect on the same protocol though :P
20:21 lluad Will it take //example.com/foo as a URL?
20:21 * lluad has no idea whether that'd even be valid at a protocol level
20:22 Grinnz_ nope
20:23 Grinnz_ has to have http:// or https:// and it sticks http on there if you don't
20:23 pink_mist lluad: nice, thanks for the link =)
20:24 tim joined #mojo
20:25 Grinnz_ " You do need a webserver to be live while you’re generating or renewing a certificate – but that’s actually easier to do for a host that doesn’t usually serve web pages than one that does." - good point, as long as port 80 is still open :P
20:26 Grinnz_ it's nice to hear that letsencrypt is still convenient as a more manual process
20:27 Grinnz_ i was somewhat put off by all of the "auto configure apache etc" that i heard it was doing
20:27 tim hi, can someone have a look on this: http://fpaste.org/295484/14487482/ (fix some warnings with malformed response cookies)
20:29 woz joined #mojo
20:29 Grinnz_ hm. it does look like that should be valid http://tools.ietf.org/html/rfc6265#page-19
20:32 lluad The certs expire after 90 days, so they're going to need some sort of automation - but the automation could be done on a management host that just copies files onto the webserver rather than having to be intimately coupled with apache.
20:32 Grinnz_ ouch, 90 days
20:34 pink_mist Grinnz_: that just means you'll definitely automate it, which means it'll be smoother in the long run
20:35 Grinnz_ perhaps
20:35 marcusr or perhaps he has nothing better to do.
20:38 tim I have some of those uninitialized warnings in my logs; would be nice if it can be fixed. I also get this warning "Mojo::Reactor::Poll: I/O watcher failed: attempt to get not ready result at Net/DNS/Native.pm line 68.", but unfortunately I can't reproduce it.
20:40 batman should i be able to do $pg->db->query("insert into t (x) values (?)", $c->param("x")); or must i decode "UTF-8", "x" first..?
20:40 batman i'm getting "Wide character" if i don't decode, but it seems like the data is stored incorrectly if i do decode :/
20:41 Grinnz_ parameters are UTF-8 decoded already
20:41 Grinnz_ wide character in what?
20:42 batman Wide character in subroutine entry at Mojo/Pg/Database.pm line 78
20:42 batman which is $sth->execute(map { _json($_) ? to_json $_->{json} : $_ } @_);
20:42 Grinnz_ hmm
20:43 Grinnz_ and it's not a json column right?
20:43 batman nope
20:43 batman it's TEXT
20:44 batman Devel::Peek says the parameter has FLAGS = (POK,IsCOW,pPOK,UTF8)
20:44 Grinnz_ no idea why that would complain about a wide character the
20:44 genio so, to_json doesn't encode.  your values before that have been decoded properly?
20:44 Grinnz_ is your DBD::Pg up to date?
20:44 Grinnz_ to_json should not be involved here
20:44 pink_mist _json would be
20:45 pink_mist as well as $sth->execute()
20:49 batman DBD::Pg 3.5.1 and Mojo::Pg 2.16
20:50 batman sorry. i meant 3.5.3
20:51 batman seems to be execute() that's complaining (just commented out the rest)
20:54 disputin joined #mojo
20:54 woz joined #mojo
20:56 disputin joined #mojo
20:58 good_news_everyon joined #mojo
20:58 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vBdOC
20:58 good_news_everyon mojo/master c6d00ca Sebastian Riedel: fix warnings in Mojo::Cookie::Response
20:58 good_news_everyon left #mojo
21:00 sri funny how this is blowing up right now https://github.com/http2/http2-spec/commit/ac468f3fab9f7092a430eedfd69ee1fb2e23c944#diff-8894168382f6487e5e38c4306e613a88
21:01 sri tim: should be fixed
21:03 jberger sri: ^ what is that
21:03 jberger ?
21:03 pink_mist I have no idea what what that diff means 0_o
21:03 sri there's PRISM in http/2
21:03 sri you know, the nsa thingy
21:03 sri it's a joke
21:03 batman Grinnz_: https://github.com/jhthorsen/app-mojopaste/commit/1922b00fc4739900532c31b87e028e5f3b74f2d9 <-- that's the commit that blow up, unless i decode/encode
21:04 pink_mist oh lawd
21:04 sri it used to be START, and they changed it to PRISM after the nsa scandal
21:06 sri now we've got a permanent reminder whenever we open a http/2 connection
21:07 tim thanks sri, but this won't fix the warning about the empty domain, right?
21:07 stephan48 sadly there will be a future time in which a student looks at a http/2 protocol dump and asks wtf is PRISM?
21:12 good_news_everyon joined #mojo
21:12 good_news_everyon [mojo] kraih pushed 1 new commit to master: http://git.io/vBdnU
21:12 good_news_everyon mojo/master 029683c Sebastian Riedel: fix another warning in Mojo::Cookie::Response
21:12 good_news_everyon left #mojo
21:23 woz joined #mojo
21:34 woz joined #mojo
21:53 woz joined #mojo
22:10 meshl joined #mojo
22:25 n16gel joined #mojo
22:54 disputin joined #mojo
22:54 padd173 joined #mojo
22:55 ipunter joined #mojo
23:58 disputin joined #mojo

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