mozilla :: #devtools

14 Mar 2017
09:28nchevobbedon't know if it's me, but the vidyo sound is terrible
09:35julienwnchevobbe, do you hear bryan correctly ?
09:35nchevobbejulienw: I can barely hear what he says
09:36nchevobbeit cuts and lags :/
09:37julienwnchevobbe, devtools vidyo ?
09:37nchevobbemight be my connexion, I don't know
09:37julienwnchevobbe, no I have it here too, let me look
09:38nchevobbejulienw: a bit better
09:38nchevobbediffernt i'd say
09:38julienwbetter ?
09:38julienwnchevobbe, ok, I'll raise it again
09:38nchevobbei thought, but no. Not a big deal though, I'll catchup later
09:40julienwnchevobbe, I hear it's terrible but I'll try to fix it after this one
09:46solenchevobbe: so weird to hear you talk French!
09:47soleHonza: https://github.com/devtools-html/extension-examples/pull/5
09:47nchevobbesole: ^^ my girlfriend use to say I don't have the same voice when I speak english XD
09:48solenchevobbe: maybe there's some truth on that - depends on the intonation of each language! French tends to sound way more lower notes
09:48julienwwe speak usually lower in english :)
09:48julienwand higher in french
09:48julienwand _yet_ higher in spanish
09:48soleyou mean louder?
09:48soleor higher?
09:48julienwhigher pitched :)
09:48nchevobbehigher-pitched ?
09:48nchevobbeyep
09:49soleI find french tends to use lower notes (or that's how it sounds to me!)
09:49Honzasole: nice, thanks!
09:49julienwah interesting !
09:49soleSpecially when you have that sound between words
09:49sole"oooooooooh"
09:49soleit's between an e and u and an o
09:49soleVery French sound
09:50julienwI find that american people speak very low and I think I imitate them unconsciously :)
09:50julienwsole, "euuuh" ?
09:50julienw)
09:53solejulienw: yes!
09:53solejulienw: the "sound of disapproval and not very silent judgment"
09:58nchevobbedon't bother for me, not a big deal
10:02julienwnchevobbe, we want it to work eventually
10:02julienwnchevobbe, I think we'll end up using the laptop's mic
10:02julienwthis is frustrating because microphones work here
10:02julienwbut not on vidyo
10:03julienwi'll look during lunch
10:03nchevobbeokay
10:04julienwbonus points if we can record :)
10:52nchevobbejulienw: the weird thing is that I hear you better than bryan
10:55pbrooh, I can also record meetings in the devtools vidyo room
10:56nchevobbeso i think the mic on bryan laptop was on, and vidyo was maybe trying to do some smart noise cancelling between the laptop mic and the other one
10:57nchevobbeor it was just a matter of turning vidyo off and on again
10:58ochameaunchevobbe: there shouldn't be anything until 2pm now.
10:59nchevobbeochameau: okay, I was wondering if you were still need help for testing the mics :)
10:59nchevobbehave a good lunch everyone
10:59ochameaunchevobbe: all looks good now, thanks !
11:31rickychienlol. right now is in lunch time
12:11julienwrickychien, it is :)
13:48rindolfHi all! jlast: here is the output - http://www.shlomifish.org/Files/files/text/debugger-html-run-test-fedora-25-x86-64.txt
14:04pbrojryans: we're taking a 5 minutes break
14:04jryansok!
14:04nchevobbei'm +1-ing the last part of the speech
14:05nchevobbeit felt really good when someone thanked me on a bug, talked about a fix in the weekly meeting or in IRC
14:08jryansnchevobbe: yes, same here, i had similar experiences as a contributor
14:10pbro+1
14:11nchevobbei think blake should mute his vidyo on his laptop
14:12nchevobbethis isn't any better in fact ^^
14:12jryansit was at first, maybe he's too far from the mic now?
14:12jryanspbro: ^^
14:12julienwis it ok now ?
14:13jryansjulienw: yes, that's better :)
14:13julienwnot _so_ good
14:13jryansbetter than it was at least...
14:16julienwcould you hear the room's questions properly for the previous talk ?
14:17jryansyes, the previously talk was fine
14:17julienwthe talk yes, but the questions ?
14:17jryansi could hear both the speaker and the questions pretty clearly
14:17julienwok
14:17julienwthis is _really_ weird
14:17zeSKBwhat's happening
14:17julienwit's the same mic we used for the questions :)
14:18jryansyeah, not sure...
14:18nchevobbei do really think vidyo is doing some kind of echo cancellation between the mic he's using and the one on his laptop, used to streal his screen
14:18nchevobbe*stream
14:19julienwyeah but i don't want to interrupt him again esp if you can hear something now
14:20jryansyeah, it's mostly understandable for now
14:40soleHonza: Landed!!!! https://hg.mozilla.org/mozilla-central/rev/e939d2c1a8c1
14:42Honzasole: Yep, I saw it, thanks for the patch!
14:43jryanssole: yay! :)
14:46solejryans: Despite the opposition haha
14:48jryanssole: maybe it's worth working out what can be improved...
14:49solejryans: I'll mention a couple of things later when I talk about what I discovered - maybe we can get the brain started on it
14:49soleThe process is not very well documented on our side to start with
14:50jryanssole: well, historically it's been a fairly rare thing to do. if we start measuring many more actions, then it would be quite different
14:51soleRight! Makes sense
14:51soleMaybe we're abusing the system
14:57solejryans: nchevobbe I've added a list at the end of the document to capture notes about the upcoming session, split by tools
14:57jryanssole: our desktop and android browsers have used telemetry for much more detailed UI action tracking in the past... trying to find a dashboard i saw a long time ago
14:57nchevobbesole:
14:58pbronchevobbe: jryans can you see Honza's slides? and hear him?
14:59soleand see his slide notes in Czech!!!?
14:59jryanspbro: yes, can see slides
14:59nchevobbeyep
14:59jryanssounds ok at the moment
14:59pbroawesome
14:59nchevobbespoke too fast XD
15:00solenchevobbe: which microphone is better?
15:00jryanssounds good with new mic :)
15:00pbrook thanks
15:00nchevobbesole: better, still some glitches
15:00jryans(or back to how it was with bwinton at least...)
15:01solejryans: same microphone than blake now!
15:02jryansyeah
15:05julienwpbro, wdyt of recording ? :)
15:06jryanssole: clarkbw: here are some docs on how mobile team tracks UI actions https://wiki.mozilla.org/Mobile/Metrics/Redash
15:06rplHonza: ping
15:07jryansrpl: might be some delay, there is devtools work week this week
15:08rpljryans: oh that's right, I forgot that the devtools work week was this week
15:08rplHonza: no rush, I was reading the email about the websocket monitor experiment and I wanted to briefly chat about it
15:23jryanspbro: changing rooms?
15:24jryansjulienw: what's going on?
15:29pbrojryans: we had to change rooms
15:29pbrosorry
15:29pbrothey're using the commons for an event later
15:29nchevobbethe shutdown was a bit brutal :)
15:30Honzarpl: I am here
15:30jryansyep, seems all is okay :)
15:31pbrojryans: nchevobbe sorry, I didn't say this on the microphone ...
15:31rplHonza: Vidyo or IRC?
15:32Honzarpl: IRC
15:32Honzarpl: (in a room full of people :-) )
15:33rplHonza: , so the issue with the listener is related to one of the current limitations of the experiments
15:34GorOmouleTokoyami: ok thank you :)
15:34Honzarpl: do I have any alternatives ? e.g. using events?
15:34rplHonza: the experiments are currently running only in the main process (well, the extensions are running in the main process currently, but the abstraction are always assuming that the experiments runs in a different process from the extension code)
15:34GorOmouleHello, is this repo (https://github.com/firebug/firebug.next) reference to the current firefox dev tools ? If not, where can I find the source code of the perfomance dev tool please ? :)
15:35rplHonza: sure, the events needs to be declared as an event in the schema.json file
15:36rplHonza: and the implementation should use our internal abstraction that are proxied as needed, so that the child and the parent process are able to work correctly
15:36rplHonza: let me take the related fragment from the webext-experiment-FirefoxRDP repo
15:36HonzaGorOmoule: No, Firebug.next has been integrated to DevTools
15:37rplHonza: this is where the events of the experiment are declared in the schema file: https://github.com/rpl/webext-experiment-firefoxRDP/blob/master/experiment/schema.json#L12
15:37Honzagregtatum: ^^ where can I find the source code of the perfomance dev tool please ?
15:37rplHonza: this is where the onDisconnected event is defined in the API implementation: https://github.com/rpl/webext-experiment-firefoxRDP/blob/master/experiment/api.js#L121
15:38gregtatumHonza: https://github.com/devtools-html/perf.html
15:38GorOmouleHonza: Ok, and do you know if devtools provide an api to use it through an extension ?
15:39HonzaGorOmoule: I dont think there are WebExtensions API for the perf tool
15:39Honzarpl: ^^ correct?
15:39rplHonza: GorOmoule: correct, not yet
15:40rplHonza: every event defined using this strategy is going to have the addListener and removeListener method as the other events defined in the existent WebExtensions APIs (so onDisconnected will have `browser.mynamespace.onDisconnected.addListener` and `browser.mynamespace.onDisconnected.removeListener`)
15:41Honzarpl: So, using events is better/recommended than what I have now?
15:41rplHonza: GorOmoule: this is the related bugzilla issue: https://bugzilla.mozilla.org/show_bug.cgi?id=1326572
15:41firebotBug 1326572 NEW, nobody@mozilla.org Provide an API for nsIProfiler
15:42rplHonza: yes, it is definitely the recommended way to define events in a new API proposal (especially because it will help to keep the API consistent by sharing the same conventions)
15:43Honzarpl: ok, thanks for the links!
15:44GorOmouleHonza: Okey thanks
15:47rplHonza: let me know how it proceed (I'm going to open more pull requests on the webextension experiments docs repo based on the doubts and issues that are coming out)
15:48Honzarpl: yep, will do
15:50GorOmouleIf I want to enhance a dev tool, is the devtools fully integrated (so, must to recompile firefox) or add-on like ?
15:51GorOmoulecf. https://wiki.mozilla.org/DevTools/Hacking
15:56GorOmouleIn fact, have you a link where I can find the process to contribute to the devTools, maybe all information that I need will be there :)
15:57HonzaGorOmoule: https://wiki.mozilla.org/DevTools/GetInvolved
15:57HonzaThis might help
16:00zeSKBhttps://hacks.mozilla.org/2016/02/introducing-devtools-reload/ does this still work?
16:00Honzasole: ^^^ ?
16:01solezeSKB: yes
16:01soleochameau can confirm
16:01zeSKByas. i wont need to fail compiling firefox
16:02solezeSKB: I mean I tried this last week
16:02soleit worked
16:04jryansbradwerth: devtools is having a work week this week, so it's not the standard meeting... currently we're listing off the state of every tool
16:05zeSKBwhat is this work week thing?
16:06bradwerthjryans: got it. I hope the work week goes well, everyone. good luck
16:06jryansbradwerth: thanks, sorry for confusion!
16:21jryansthe loudest bag of chips :)
16:22Honzazer0: ^^^ :-)
16:23zer0jryans: it's popcorn! sorry! :D
16:23jryanshaha, no worries :)
16:23jryansthe mics are working really well, at least! :)
16:40julienwyeah this room has a great av system
16:44nchevobbegot to go, see you tomorrow folks
17:09soleclarkbw: https://github.com/mozilla/gecko-dev/blob/master/toolkit/components/telemetry/Histograms.json#L7917
17:09solethere is?
17:09clarkbwsole: thx!
19:26rindolfHi all! jlast: here is the output - http://www.shlomifish.org/Files/files/text/debugger-html-run-test-fedora-25-x86-64.txt
19:27rindolfyarn run-test all fails here
19:28rindolfI mean yarn run test-all
19:48rindolfLet me try with nvm
19:56rindolfhappens with v6.10.0 LTS from nvm too
20:18rindolfseems to fail on mageia v6 x64 too
22:04smaugis it possible to keep console logs even cross page loads?
22:09jdescottessmaug: there is the option "Enable persistent logs" in the devtools option which should do this?
22:09smaugthanks
15 Mar 2017
No messages
   
Last message: 75 days and 6 hours ago