What does this mean?
(In reply to Ben Francis [:benfrancis] from comment #1) > What does this mean? Solving this will fix bug 1052154 and bug 1053396. I assume for the future, we should just show rocketbar with a close button here, but maybe it shouldn't be editable. If there is a way to keep existing functionality, without regressing scrolling - we should do that as we're running out of time.
Do we have a UX spec for how Rocketbar should behave with popup windows? As I understand it, the current desired behaviour is: * window.open() or <a target="_blank"> called from a browser window should open a new browser window (currently broken, see bug 1054215) * window.open() from an app should open an overlay (popup) window with a close button and nothing else. There's currently a regression which causes the top of the page to be obscured (bug 1052154) which could just be fixed by adding padding above the iframe? * window.open(url, '_blank') and <a target="_blank"> called from an app should always open a browser window (since bug 1030433 changed this behaviour) Once "app scope" is implemented (see bug 972320): * Calls to window.open() or <a> to a URL which falls within the scope of an app should open in the app (we need a UX spec to define when it should navigate the main app frame and when it should create a popup window) * Calls to window.open() or <a> from an app to a URL which matches the "open-in-app" prefix in its manifest should open within the app (again, UX spec needed). * Calls to window.open() or <a> to a URL which doesn't fall within the scope of any app should open a browser window, unless being called from an app where the URL matches an "open-in-app" URL prefix in its manifest. Once full sheets is implemented: * window.open() should always create a new sheet * Perhaps the only time the popup window with the close button should appear is if the "dialog" feature of window.open() is used top opt in to it? In conclusion, I don't think we need to implement Rocketbar for PopupWindow, do we? We should fix the regression and gradually reduce the number of places PopupWindow is used? The one edge case I can think of is that if a popup window is created from an app with a collapsed Rocketbar, you'll be able to see the title of the app in the statusbar while the popup window (which could be a completely different website) is being displayed with its own title. What should the behaviour be then? Note that I have a test app for some of these things which you can load in the browser or install as an app http://people.mozilla.org/~bfrancis/hyperlink
> > The one edge case I can think of is that if a popup window is created from > an app with a collapsed Rocketbar, you'll be able to see the title of the > app in the statusbar while the popup window (which could be a completely > different website) is being displayed with its own title. What should the > behaviour be then? > I think ideally we should not show the rocket bar when pop-up windows are displayed, since the pop-up window has its own title displayed in the chrome below.
I have renamed the bug title accordingly, please let me know if this is wrong.
The two dependent bugs are only related to page height, so renaming the bugs. We could still file a polish bug to remove the rocketbar, but I don't think that should block FL - we should definitely get the page height fixed ASAP though as this is causing all kinds of problems.
Assignee: nobody → kgrandon
Status: NEW → ASSIGNED
Summary: Hide Rocketbar for popup windows → Fix page height for popup windows
Target Milestone: --- → 2.1 S3 (29aug)
Created attachment 8478737 [details] [review] Github pull request Tim or Etienne - could one of you give this a quick review? Thanks!
Comment on attachment 8478737 [details] [review] Github pull request This can land for the code perspective, if that's what you want (before FL and polish etc.)
Attachment #8478737 - Flags: review?(timdream) → review+
Status: ASSIGNED → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.