mozilla :: #devtools

10 Oct 2017
08:11no_gravi1yHello! Is it possible to inspect the request that was sent for a link that opened in a new window?
08:29nchevobbeno_gravi1y: hi
08:29nchevobbenot sure what you mean
08:30nchevobbeno_gravi1y: if you open the netmonitor in the opened window you will see the request
08:32no_gravi1ynchevobbe: No, then it's too late.
08:32no_gravi1ynchevobbe: When you click on <a target=_blank href=&quot;somewhere.com&quot;> and then open devtools in the new window, they are empty.
08:33no_gravi1yA manual workaround would be to edit the link in the console to open in the same window. Or in a named window where you have devtools open already.
08:33no_gravi1yLooks like that&#39;s the only solution at the moment.
10:44ellipticaldoorHi, I&#39;m trying to understand what this code does
10:44ellipticaldoorin /devtools/client/dom/dom-panel.js the line 81
10:45ellipticaldoorthe destroy function
10:55Kwanellipticaldoor: looks like it removes a couple of event listeners
10:55phlixi_owell no_gravi1y already left... i had the same problem once (checking request in new window) i think i activated and used some kind of devtols for the browser, then it was possible
10:56phlixi_oalso, sometimes changing the target is not that trivial
10:58ellipticaldoor@Kwan I want to change the defer part to use a promise, how I can now that my code removes the event listeners correctly?
11:05ondrasso, from a devtools extension, is it possible to &quot;switch to Console&quot; ?
12:12pbroondras: I don&#39;t think so. But maybe you should confirm by asking elsewhere. The WebExtensions APIs are being developed by a different team than DevTools at mozilla. They would know better. I&#39;m sure they have a channel here on IRC. Let&#39;s ask rpl
12:13ondraspbro: well I suppose that such API would be ff-devtools-specific; that is why I asked here :)
12:13pbroyes, here is a good place too, :)
12:13pbrojust suggesting that a webext channel might be helpful too
12:14rplpbro: ondras: do you mean an API to open the console panel from the devtools panel created by the extension?
12:16rplpbro: ondras: there isn&#39;t an API that explicitly aims to do that, but you should be able to do something similar using the `inspect(...)` helper from javascript code executed using devtools.inspectedWindow.eval
12:16ondrasrpl: &quot;inspect&quot; would switch the devools panel?
12:17ondrasrpl: that sounds like a good way. let me try that.
12:17ondrasrpl: I already do devtools.inspectedWindow.eval(&quot;console.log(...)&quot;), so I just s/console.log/inspect/ ?
12:18rplpbro: ondras: on Firefox, when you call something like browser.devtools.inspectedWindow.eval(&quot;inspect(window.body)&quot;) then the devtools toolbox switches to the elements panel, but if the passed parameter is not a DOM node (e.g. something like browser.devtools.inspectedwindow.eval(&quot;inspect({})&quot;), it opens the split console and shows a preview of that object
12:18ondrasrpl: that is probably exactly what I need. thanks!
12:19rplondras: yw!
13:50CrazyLazyDazyFirefox 56.0 64bit triggers my rest-endpoint twice, but when I see the network request in developer tools, only one request is sent. This does not happen in Chrome. I have no add-ons installed.
14:10nchevobbeCrazyLazyDazy: Hello. do your page is public so we can check it ?
14:10CrazyLazyDazynchevobbe, it is not public.
14:10nchevobbeCrazyLazyDazy: or if it&#39;s not, could you tell us what your setup looks like (how the endpoint gets called in the first place, )
14:11CrazyLazyDazyThe user clicks a button to send a form
14:11tromeydevtools can re-fetch style sheets at least
14:12CrazyLazyDazyDoes firefox resend request if there is no reply within X seconds?
14:13tromeyand if devtools does fetch something - the style sheets or source maps - these fetches won&#39;t show up in the content network monitor
14:13tromeyyou have to open the browser toolbxo to see those
14:13nchevobbeCrazyLazyDazy: i don&#39;t think so. is it an xhr call or a regular form submit POST ?
14:14soleCrazyLazyDazy: I think in the past I saw something similar when there was some sort of img being loaded with the wrong path (accidentally empty), and it would assume the current path (so it would hit the same url twice)
14:14CrazyLazyDazynchevobbe, xhr call (angularjs post)
14:15nchevobbeCrazyLazyDazy: does it happen only if you have the devtools open ?
14:15soleAnyone here have experience with uplifting a patch? want to uplift the fix for that screenshot bug we mentioned yesterday
14:15CrazyLazyDazynchevobbe, no
14:15soleMaybe Honza ? jdescottes ?
14:16jdescottessole: can help, what is the bug#?
14:17CrazyLazyDazysole, I will see if this is the case
14:17solejdescottes: https://bugzilla.mozilla.org/show_bug.cgi?id=1391738
14:17firebotBug 1391738 UNCONFIRMED, spenades@mozilla.com Shift-F2 > screenshot fails with unknownError
14:17solejdescottes: according to this it has to land on m-c first anyway https://wiki.mozilla.org/Release_Management/Uplift_rules
14:17jdescottesyes
14:18soleCrazyLazyDazy: thanks! Also sometimes missing favicon paths can do it too (weird)
14:19jdescottessole: first land it on central as usual, it will need to be verified before it can be uplifted. I usually wait for the patch to land on m-c before setting beta approval flag on the patch
14:19solejdescottes: what do you mean with need verified? Do I need to set a special flag? :)
14:20jdescottessole: status needs to be VERIFIED FIXED
14:20soleoh riiight
14:20soleOK :D
14:20solejdescottes: I know how to do that!
14:22jdescottessole: as I said you will need to set the approvalmozillabeta flag on your attachment, that&#39;s what will trigger the whole process
14:22solejdescottes: perfect
14:22jdescottesthe flag comes with a &quot;form&quot;, ping me if you need help filling it
14:23solejdescottes: haha love the &quot;form&quot; 8-)
14:23soleI will - first to wait for the land
14:23jdescottesyep
16:16bgrinsmikeratcliffe: have you been able to work on the react update in https://bugzilla.mozilla.org/show_bug.cgi?id=1399493#c1 ?
16:16bgrinsreally want this for the new console
16:16mikeratcliffebgrins: Sorry, yes... I have a fix
16:16mikeratcliffeBut I had laptop issues and had to rebuild
16:17mikeratcliffeBut I still have the patch.
16:17mikeratcliffeLet me see my last try run...
16:17jryansochameau: did you mean to mark chrome frame visibility patch for review?
16:19solejulienw_: you have to ask paris IT to fix the sound in that room, it&#39;s so cutting out
16:19ochameaujryans: waiting for try results first
16:20jryansochameau: ok :)
16:23mikeratcliffebgrins: Still some failures but jsonview tests... I will need to look into that but I will make it my first priority after I have caught up on emails etc.
16:23bgrinsmikeratcliffe: ok thanks :)
16:23mikeratcliffenp
16:26ochameausole: is there also echo from our room? isn&#39;t it better now?
16:27soleochameau: it&#39;s really bad, echo or cutting out, or beeps
16:27soleas in high pitched beep
16:27soleochameau: I find it hard to understand when you two speak :(
19:56felipetromey: ping
19:56tromeyfelipe: hi
19:57felipetromey: hey! do you know anything about the code in performance-controller.js? (i&#39;m ping&#39;ing semi-randomly to find someone)
19:57tromeyI do not. maybe nobody on the team does any more, though you could ask gregtatum
19:58felipetromey: do you at least know if this is a feature that never rolled out? staying only on nightly
19:58felipeor gregtatum ^
19:59tromeylet me look
19:59tromeythe performance panel is definitely in release
19:59felipeI&#39;m asking because I&#39;m removing the E10S_TESTING_ONLY define
19:59felipeand it&#39;s used here: http://searchfox.org/mozilla-central/source/devtools/client/performance/performance-controller.js#527
20:00felipeso on not-nightly, this supported var is defined as false
20:00felipebut i&#39;m guessing it&#39;s just being ignored later on
20:00tromeyhmm
20:00gregtatumfelipe pto today, hit me up tomorrow
20:01tromeythere are two callers of this function
20:01tromeyone in views/overview.js and one in performance-controller.js
20:01tromeyonly the latter checks &quot;supported&quot;
20:01tromeyanyway I think the idea is that this feature (or some features) only work well on e10s
20:02tromeyso if we have e10s everywhere this could probably just be removed
20:03felipeyeah.. it looks like if e10s is enabled, this &quot;supported&quot; is ignored
20:03felipeit was only used to check for the case where e10s was not enabled
20:04felipeI&#39;ll make a simple change on my end to remove the ifdef, and ask greg for review.. If anyone wants to do some cleanup here I&#39;ll let that be a followup
11 Oct 2017
No messages
   
Last message: 6 days and 17 hours ago