mozilla :: #fx-team

13 Jul 2017
06:25adwmikedeboer: you might not be there yet but i just needinfo'ed you on bug 1374477 if you have a moment
06:25firebot ASSIGNED, Photon API for adding, removing, and moving page actions
11:33* Standard8 discovers htmllint
12:00* johannh runs away
13:51johannhugh, why is nothing about this bug easy
13:52johannhI don't really understand why I can't debug BookmarkingUI.onCommand when clicking on the bookmark menu button
13:52johannhmak maybe^? :)
13:53johannhis it running in the content process?
13:53johannhlike, I set a breakpoint and it doesn't hit it
13:53makno, it's all in the chrome process
13:53johannhalthough it's supposed to do that, right?
13:54makoff-hand it looks so
13:54johannhit has oncommand="BookmarkingUI.onCommand(event);"
13:54johannhmanually calling the function in the console makes me hit the breakpoint
13:54johannhI wonder what's up there
13:54makI wonder if the difference is between the menu and the menu-button
13:54makin which mode are you? photon or not?
13:55johannhand in my case it only matters for Photon
13:55makif it's a menu oncommand may not fire?
13:55maksince it's not a button
13:55johannhand it just shows the menu
13:55makyeah, I don't think oncommand applies to "menu"
13:55makI may be wrong, but my gut tells me that
13:56makthis sounds like an oncommand for the button part of a menu-button
13:56johannhyup you're right
13:56johannhthat makes me thoroughly unhappy
13:57makit should have been ifdef-ed
13:57mak(likely I missed that in review, not trivial to notice though)
13:57johannh(why is it even implemented that way)
13:58makI ask myself the same question every time I see code I wronte 1 year ago.
13:58johannhaka is there any way to make it a menubutton consistently again?
13:58* johannh gives up and will solve that in a follow up bug
13:58makI don't know what you are trying to do, but in photon this is not supposed to be a button
13:59johannhbut the other things in the nav toolbar are also buttons, no?
13:59johannhthis works entirely differently from all other items in the toolbar
14:00johannhactually everything in that damn toolbar works entirely differently
14:00* johannh calms down
14:00makxul === fun
14:58johannhI think there was a bug for being able to drag the customize mode items, right? Does anyone remember the bug #?
14:58johannhlike touch-drag
15:13nhnt11hmm, my home tab crashes at startup on my local build (pulled central an hour or two ago)
15:14nhnt11(homepage is about:home)
15:14* nhnt11 investigates
15:14nhnt11Heh, tried to view a crash report and the tab crashed again
15:16nhnt11so, about:home crashes 100% of the time
15:16nhnt11nothing in the error console
15:16nhnt11###!!! [Parent][MessageChannel] Error: (msgtype=0x28007E,name=PBrowser::Msg_Destroy) Channel error: cannot send/recv
15:18nhnt11hmm, all is fine on a fresh profile
15:51_6a68Is `Services.appinfo.defaultUpdateChannel` the right way to check which Firefox channel a user is on (nightly / dev / release)?
15:58johannhmikedeboer: ping
15:58aswanjaws: is there a way to get to chrome://browser/content/sanitize.xul in the new preferences pages?
15:58jawsaswan: there should be, where was it before?
15:59jawsi know that name isn't the name we use anymore
15:59aswanjaws: i don't remember... :/
15:59jawsaswan: oh, i don't think that is every from the preferences
16:00jawsaswan: i may have found it actually
16:01jawsaswan: it is in preferences
16:02aswansomewhere under privacy & security?
16:02jawsaswan: to get to it, you have to go to Preferences > Privacy & Security > Change "Remember history" to "Use custom settings for history" > Check the box for "Clear history when Nightly closes" > Then click on the "Settings..." button that becomes enabled
16:03aswanah ha! thanks, i would have had a hard time finding that myself
16:12johannhmaybe paolo: ping? :)
16:14* johannh realizes it's too late in Europe
18:02OsmoseSo I have an about: page running in a content process, and it needs to get some data from the parent process. I can send a message over Services.cpmm to ask for the data, but I can't figure out how to know which message manager to send the data back on from the parent process.
18:02OsmoseI could use Services.ppmm.broadcastAsnycMessage, but the docs say to avoid that
18:06MossopOsmose: Where you receive the message you should get a reference to the message manager it came from
18:08OsmoseMossop: Is that what target is here?
18:08OsmoseThe "element owning the message manager" bit threw me off
18:09MossopOsmose: It's either the message manager or the browser element in which case target.messageManager is what you want
18:09OsmoseAh, so target.messageManager || target would do
18:10OsmoseMossop: Rad, thanks!
18:10MossopOsmose: It wont change unless you change the message manager you send with
18:41adwshorlander: thanks for the quick response on that bug
19:45gfritzsche_6a68: this is what we use for telemetry (not sure about your constraints though):
19:46_6a68gfritzsche: ah, cool! I hadn't seen that API before, I'll check it out
19:46gfritzschealso, we have "esr"
19:46gfritzsche"nightly", "beta", "release", "esr"
19:48gfritzsche_6a68: plus there are the "-cck" suffixes for channels, not sure if those are still an active thing
19:49gfritzschebut you can skip those partner bits
20:24OsmoseMossop: It's totally the async handler
20:24OsmoseIf receiveMessage is not async it works fine
20:24OsmoseEven if it defers to another async function
20:24MossopWait. Is this code going through babel and the asyncness getting transpiled somehow?
20:26MossopThen that seems broken to me. Maybe an xpconnect problem
20:26OsmoseI'll see if I can get a reduced testcase
20:31Mossopgandalf: Are there more tests for L10nRegistry and MessageContext somewhere?
20:50daofirefox on android still can't download mp3s? :(
20:50* dao copies URL and opens chrome
20:51daoah, firefox on android has "save audio" in the context menu nowadays
20:51daochrome has a more obvious download button in the audio player UI
21:38Mossopianbicking: You seem to be missing the string change in bug 1380817
21:38firebot NEW, Update Screenshots to version 10.5.0
21:43dmosek88hudson: mconley: is there some reason we can't create the preloaded browser at the time we're loading the initial tab for a window, rather than the first time about:newtab opens in a window?
21:45mconleydmose: likely because that'll impact the time to paint that initial tab
21:45dmoseactually, i guess we do it whjen the second tab opens
21:45dmosebut that's a fair reason, i suppose
21:46mconleydmose: not that opening the second time is a great time to do it either
21:46mconleyopening after the second tab, I should say
21:46mconleySee bug 1353013
21:46firebot NEW, Kick off preloading about:newtab at a better time
21:46dmosemconley: i suppose we could use idleDispatch
21:46* mconley nods
21:48dmoseah, indeed, already suggested
21:48dmosethanks for the tip
14 Jul 2017
No messages
Last message: 14 days and 55 minutes ago