|
Cro: libraries and tools for building reactive services in Perl 6 | cro.services/ | Logs: irclog.perlgeek.de/cro/ Set by moderator on 1 February 2018. |
|||
|
00:06
Geth joined
00:08
Geth joined
00:10
Geth joined
01:00
lizmat joined
01:57
lizmat joined
02:59
ilbot3 joined
|
|||
| moderator | Cro: libraries and tools for building reactive services in Perl 6 | cro.services/ | Logs: irclog.perlgeek.de/cro/ | ||
|
10:46
jc_denton_ joined
|
|||
| jc_denton_ | Hey jnthn sorry to bother you again. I tried your solution but I think I missed something out. It isn't reacting as desired. We're emitting through the desired supplier but the client at the end of the websocket is not receiving it. | 10:52 | |
| I was hoping I could put a new line there. Here's a pastebin of what we had tried.pastebin.com/PvSnjvph | |||
| We also attempted putting it in the routes supply block but that did not appear to as planned. | |||
| jnthn | Have you stuck a say statement or something beofre `emit to-json $call-signal` to see if that line is ever reached? | 11:14 | |
| Also, not The Problem, but that line ends with a , and probably should be a ; | 11:15 | ||
| jc_denton_ | Just gave that a try, it does not seem to reach the emit to-json. | 11:24 | |
| jnthn | jc_denton_: Hm, something ain't right in here...will be less busy with work soon and can look a little more | 11:50 | |
| jc_denton: Ah, missed you, but if you read the log: it seems that you create `my $supplier = Supplier.new;` but never actually tap that. I think what you want is, after the call to new_socket_friend, to do whenever $supplier { emit to-json $_ } | 12:18 | ||
| The send-call-signal Supply never itself emits anything (using emit), it sends them into the $supplier, but that was never tapped | |||
| lunch & | |||
|
13:16
jc_denton1 joined
|
|||
| jnthn back | 13:18 | ||
| jc_denton1: Did you see my comments while you were away? | |||
| (logs link in topic if not) | |||
| jc_denton1 | Yep | ||
| just looked in the log now | |||
| gonna give it a try in a moment | |||
| Geth | cro-websocket: 7ef9a1cc86 | (Jonathan Worthington)++ | lib/Cro/WebSocket/FrameParser.pm6 Give frame parser a fresh frame after `emit` So that it will never modify a frame in-place that a downstream handler is working on. This somehow lay hidden until now. |
14:12 | |
| cro: 853ca70d30 | (Jonathan Worthington)++ | docs/reference/cro-websocket-message.md Update Cro::WebSocket::Message documentation |
14:22 | ||
| jnthn crosses his fingers for happy Travis CI | 14:23 | ||
|
14:28
jc_denton2 joined
|
|||
| jc_denton2 | I hope Travis likes it! | 14:28 | |
| hmmm getting some environment problems when I try and run it remotely. I'll look at it properly on Monday. I thought that may have been one of the problems. Thanks again, It looks a bit more like this now. pastebin.com/0BqYZ04y | |||
| jnthn | Might be that :host<localhost> or something in the server? | 14:29 | |
| (try :host<0.0.0.0> to bind to all interfaces) | |||
| Geth | cro: 025b2ee9fe | (Jonathan Worthington)++ | resources/web/index.html Remove commented out HTML |
14:37 | |
| cro-websocket: 4450a683d2 | (Jonathan Worthington)++ | lib/Cro/WebSocket/Client/Connection.pm6 Make :0timeout more likely to really time out |
14:46 | ||
| cro-websocket: 65c78229f8 | (Jonathan Worthington)++ | lib/Cro/WebSocket/Client/Connection.pm6 A supply block will be `done` for us So no need to call it explicitly when all we do otherwise is `emit` one thing. |
|||
| cro: 8bfb85a002 | (Jonathan Worthington)++ | t/tools-services.t Add diagnostic to test output Hopefully this will provide a clue as to why this test sometimes fails on Travis CI. |
14:56 | ||
| jnthn | Well, at least travis-ci.org/croservices/cro-websocket is happy again | 15:08 | |
| Geth | cro: 362dcc5e8b | Altai-man++ | resources/web/css/cro.css Font is already set in bootstrap file |
15:17 | |
| cro: feafa32a35 | Altai-man++ | lib/Cro/Tools/CLI.pm6 Remove debug leftover |
15:22 | ||
| cro: a4ae769691 | (Jonathan Worthington)++ | docs/index.md Add link from index to sessions/auth doc |
15:47 | ||
| jnthn | github.com/Leont/yamlish/issues/19 seems to be the reason the Cro tests (in the top-level distribution) sometimes fail | 16:13 | |
| And that in turn may hint at a Rakudo bug | 16:21 | ||
| Geth | cro: 3494dde21d | (Jonathan Worthington)++ | lib/Cro/Tools/CroFile.pm6 Workaround load-yaml concurrent usage crash This seems to have been at the heart of the test failures we've been seeing in this distribution. |
16:25 | |
| jnthn | This should mean we can at last stop needing to suggest --/test :) | 16:26 | |
| japhb | That would be nice! It's a very weird thing to see as one's first interaction with Cro. | 18:10 | |
| jnthn | Yeah, now you'll see 3 times as long waiting for an install :P | 19:45 | |
| (Not measured, but I'm guessing running the tests at least doubles installation time) | 19:46 | ||
|
20:24
jc_denton1 joined
|
|||
| japhb | True, but you release new ones rarely enough that I suspect that how sturdy it appears overrides how long it takes to install. Plus it just gives us incentive to improve zef and/or testing performance. | 20:59 | |
| jnthn | I wasn't being particularly serious. :-) And yeah, I'll be glad to be rid of the --/test advice. | 21:01 | |
| japhb | :-) | ||
| jnthn | Maybe it's time to drop it in the next website update, even. | 21:06 | |
| Geth | cro: a05f228c84 | Altai-man++ | 7 files Improve links handling |
21:15 | |
| cro: 70fb816c74 | Altai-man++ | resources/web/js/app.js Update generated code |
|||
|
21:59
colomon joined
|
|||