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

IRC log for #mojo, 2017-06-14

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

All times shown according to UTC.

Time Nick Message
00:02 disputin joined #mojo
00:03 stryx` joined #mojo
00:45 stryx` joined #mojo
01:24 aborazmeh joined #mojo
02:07 karjala_ joined #mojo
02:21 zivester joined #mojo
02:33 noganex joined #mojo
02:34 tchaves joined #mojo
04:04 dboehmer joined #mojo
05:03 htaunay joined #mojo
06:04 dod joined #mojo
06:09 dod joined #mojo
06:35 karjala_ joined #mojo
07:00 AndrewIsh joined #mojo
07:01 prg joined #mojo
07:15 inokenty-w joined #mojo
07:31 trone joined #mojo
08:10 rshadow joined #mojo
10:29 stryx` joined #mojo
10:45 stryx` joined #mojo
11:26 tchaves joined #mojo
11:33 ashimema Mojo::File + Mojo::Collection = Awesome!
11:34 ashimema loving how easy they just made my job ;)
11:56 * nic agrees
11:58 nic I find Mojo::Collection quite funny, cos when you look at it it's very simple, but from time to time utilising that in one of your classes is like sprinkling a bit of magic on it (in terms of usability)
12:11 CandyAngel Hear hear
12:15 ashimema totally agree
12:17 pink_mist people keep saying they want a separated out Mojo::DOM or Mojo::UserAgent ... but Mojo::Collection does seem a lot more useful - of course, I expect Grinnz's Mojo::DOM58 includes that one
12:17 pink_mist s/a lot more useful/$& to separate out/
12:30 Pyritic joined #mojo
12:38 gizmomathboy joined #mojo
12:45 CandyAngel Hm
12:45 CandyAngel Perl.. y u do dis
12:48 CandyAngel Am I going crazy, or would you expect splitting "a1" on 1 to give you ('a', '')?
12:49 pink_mist depends on the semantics of the splitting operation
12:49 pink_mist very possibly, yes
12:50 CandyAngel It doesn't.. even 'a1111111' only gives you ('a') :|
12:51 CandyAngel Oh,  you have to give split a negative number.. what
12:51 CandyAngel o.o
12:51 CandyAngel https://stackoverflow.com/questions/4787217/perl-split-function-consecutive-delimiters
12:51 pink_mist split /1/, 'a1111', -1
12:52 pink_mist all explained in perldoc -f split
12:52 CandyAngel It is but.. I don't recall needing to do that before
12:53 pink_mist then you weren't faced with this issue before :P
12:53 pink_mist the default spliting semantics are quite useful after all
12:54 CandyAngel Yeah, sure.. just.. surprised
12:56 maschine joined #mojo
13:25 gryphon joined #mojo
13:29 kaare joined #mojo
13:46 tholen joined #mojo
13:54 Intern42 joined #mojo
14:06 kaare joined #mojo
14:11 kaare_ joined #mojo
14:22 zivester joined #mojo
14:30 kaare joined #mojo
14:31 sri i learned about that split feature very late too
14:32 CandyAngel That genuinely makes me feel 1000x better :)
14:42 dod joined #mojo
14:43 kaare_ joined #mojo
14:59 kaare joined #mojo
15:02 jabberwok Is this too 'tricksy' for production code? in a class that uses Mojo::Pg, with a debugging attribute 'rebuild', in the new method: » $self->db->migrations->tap(sub{if ($self->rebuild) { $_->migrate(0); } })->migrate; « ... or am I just trying too hard to use 'tap'?
15:03 kaare_ joined #mojo
15:04 kaare_ joined #mojo
15:05 sri trying too hard imo
15:06 sri ->tap(sub{ $self->rebuild and $_->migrate(0) })
15:06 sri that's not too bad though
15:08 sri once there is blocks or more than one statement i tend not to put it into a closure
15:23 dod joined #mojo
15:30 PopeF0 joined #mojo
15:46 PryMar56 joined #mojo
16:30 tholen joined #mojo
16:41 disputin joined #mojo
16:48 dod joined #mojo
16:53 chandwki joined #mojo
16:53 kgoess joined #mojo
17:05 disputin joined #mojo
17:22 haarg sri: the & with wide characters thing should be fixed in blead now
17:23 haarg if you see more failure reports let me know
17:24 sri \o/
17:24 sri haarg++
17:36 trone joined #mojo
17:49 sri after watching The Expanse i actually started reading the books... and now i can't stop anymore, they are really really good :S
18:11 haarg I may have to read those at some point
18:43 stryx` joined #mojo
18:46 cheako joined #mojo
18:49 cheako joined #mojo
19:46 cheako joined #mojo
19:46 Pyritic joined #mojo
19:50 gizmomathboy joined #mojo
19:53 disputin joined #mojo
20:04 genioweb joined #mojo
20:12 gizmomathboy joined #mojo
20:48 disputin joined #mojo
21:04 cheako joined #mojo
21:04 jstores0711 joined #mojo
21:06 stryx` joined #mojo
21:17 preaction ... apparently i have a server with a misconfigured IPv6 DNS. is there a way i can force IPv4 for Mojo::UserAgent?
21:22 sri ipv4 address
21:23 Grinnz i ran into this problem before... i don't think i found a solution before i just got rid of the broken ipv6
21:23 sri $ua->local_address($local_ipv4_address) should work too
21:24 Grinnz ideally there would be a way to pass Family => AF_INET to the underlying IO::Socket::IP
21:27 Grinnz hardcoding IP addresses for destinations or the local IP wasn't exactly something I wanted to do...
21:27 preaction sorry, it's not the local server, it's the remote server i'm trying to fetch a webpage from
21:28 pink_mist preaction: doesn't change anything
21:28 pink_mist preaction: if you bind the local address to an ipv4 address it'll use ipv4
21:28 preaction okay, i'll try it
21:30 stryx` joined #mojo
21:30 preaction okay, so i need to use Net::Address::IP::Local to find the right IPv4 address to bind to and then set that
21:34 disputin joined #mojo
21:35 Grinnz something like adding an address_family option to Mojo::IOLoop::Client and an address_family attribute to Mojo::UserAgent could work, but getting the AF_INET constants would be annoying and testing it would be difficult
21:37 trippeh if mojo respects /etc/gai.conf you may get away with changing the priorities there
21:38 Grinnz IO::Socket::IP just uses getaddrinfo, so probably
21:39 trippeh ...and then fix the actual problem :)
21:40 preaction i can't fix someone else's server
21:41 Grinnz or someone else's DNS records, or nameserver, or whatever
21:41 Grinnz anyway it's 2017, why isn't ipv6 mainstream yet :P
21:42 trippeh it is fairly common client side. server side not so much
21:42 preaction the server worked up to a few days ago, so i'm guessing they messed something up and might fix it later
21:45 trippeh may want to have happy eyeballs-ish behaviour in Mojo::UserAgent. at least fall to ipv4 fairly quickly if it is an available option.
21:45 Grinnz happy ... eyeballs?
21:46 trippeh the thing browsers do to paper over broken ipv6 networks and services
21:46 Grinnz a fallback mechanism would probably be more in the realm of a separate CPAN extension
21:47 trippeh (maybe not be as aggressive as them, but)
21:47 Grinnz but it would be hard to do unless Mojo::IOLoop::Client itself supports the option
21:49 trippeh like, some browsers will connect to both at the same time, then killing the slowest one. ugh
21:50 trippeh kind of abusive really ;-)
21:50 Grinnz perfectly allowable as long as it's a GET
21:51 Grinnz but yeah dont make GET endpoints that have a problem with being called extra times :P
21:55 trippeh I dont think they actually issue the HTTP part multiple times - they time connection setup (with TCP Fast Open it could happen I suppose)
21:55 trippeh it just burns some extra sockets and state in firewalls and such
22:04 Grinnz that makes sense

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