Closed
Bug 786890
Opened 12 years ago
Closed 6 years ago
No plugin support error does not appear on Dinopanic application
Categories
(Firefox OS Graveyard :: General, defect, P2)
Tracking
(blocking-basecamp:-)
RESOLVED
WONTFIX
blocking-basecamp | - |
People
(Reporter: jsmith, Unassigned)
Details
## Build * Device: Otoro * Gaia: cc7fac1524a56bbd34cf99066bb3f9d213177d9f * Mozilla Central: b8c791a10915326098e940eaf61f76d47713580b Steps: 1. Go to marketplace.mozilla.org 2. Install Dinopanic 3. Launch Dinopanic Expected: The app should launch to it's default content with no errors. Actual: Startup crash when the app launches. Cannot load the content even on a retry.
Reporter | ||
Updated•12 years ago
|
Keywords: reproducible
Whiteboard: [startupcrash]
Reporter | ||
Comment 1•12 years ago
|
||
The only thing that looked interesting in the log was this: 08-30 00:36:33.622: E/GeckoConsole(105): Content JS INFO at app://system.thisdomaindoesnotexist.org/js/window_manager.js:397 in appendFrame: %%%%% Launching Dinopanic as remote (OOP) The app appears to be using flash in the content of the app, but the page didn't even load at all to get the point where it would be shown.
Comment 2•12 years ago
|
||
Doug, can you look into this one? A stack for this crash would be a good first step... If it's not in code you're familiar we can find a more appropriate owner.
Assignee: nobody → bugzilla
blocking-basecamp: ? → +
Comment 3•12 years ago
|
||
afaik, in general cjones is the best the person to debug OOP-related crashes. I have a lot on my plate still within the panning/zooming code.
Comment 4•12 years ago
|
||
Also it was mentioned to me that we may have crashes when we try to load unknown content types, which would explain why a flash app may be crashing it before it's even loaded. I don't know the bug # for this, or if there even is one.
Reporter | ||
Comment 5•12 years ago
|
||
(In reply to Doug Sherk (:drs) (:dRdR) from comment #4) > Also it was mentioned to me that we may have crashes when we try to load > unknown content types, which would explain why a flash app may be crashing > it before it's even loaded. I don't know the bug # for this, or if there > even is one. I know of the bug you are talking about, although this one is loading a webpage with flash content, not an unknown content type.
It's not unknown content per se, but content that doesn't have an in-gecko content handler registered. For that we'll sometimes try to kick out to an external helper application.
Updated•12 years ago
|
Whiteboard: [startupcrash] → [startupcrash] [WebAPI:P0]
Comment 7•12 years ago
|
||
Unassigning myself. I think this is out of the scope of code I know. If anyone debugs it and it actually is, feel free to reassign to me.
Assignee: bugzilla → nobody
Reporter | ||
Updated•12 years ago
|
Whiteboard: [startupcrash] [WebAPI:P0] → [startupcrash] [WebAPI:P0] [blocked-on-input Jason Smith]
Reporter | ||
Comment 8•12 years ago
|
||
Still 100% reproducible, although now I know what the error I saw means, the error I mean is a generic error, so not sure if it's a crash. Although whatever is happening prevents access to the content. I tried going to the URL in the browser for comparison - http://beta.swartag.com/?task=view&id=1159. The page tries to load the page, but many parts of the page are missing and scrolling content appears to be entirely busted.
Keywords: crash,
qawanted,
reproducible
Summary: Content crash when launching Dinopanic application → Cannot launch Dinopanic application - Error on startup
Whiteboard: [startupcrash] [WebAPI:P0] [blocked-on-input Jason Smith] → [WebAPI:P0]
Reporter | ||
Updated•12 years ago
|
Summary: Cannot launch Dinopanic application - Error on startup → Generate a more graceful fallback content UI when flash content is encountered, don't error out or generate a blank white content
Reporter | ||
Comment 9•12 years ago
|
||
Talked about this in triage - the root cause of the problem here is that we do a really bad job of handling the case of when we get lots of flash content in web content. The app errors out immediately on launch. Accessing it directly in the browser looks really bad (missing many layout pieces). We should handle in-content flash content gracefully by producing a "not supported" UI in FF OS.
Comment 11•12 years ago
|
||
(In reply to Jason Smith [:jsmith] from comment #9) > Talked about this in triage - the root cause of the problem here is that we > do a really bad job of handling the case of when we get lots of flash > content in web content. The app errors out immediately on launch. Accessing > it directly in the browser looks really bad (missing many layout pieces). We > should handle in-content flash content gracefully by producing a "not > supported" UI in FF OS. The plugin overlay code that shows an error message in desktop/mobile Firefox is in core/toolkit code, and it does show up in the browser app on my b2g desktop build: http://cl.ly/image/192A0E2G0y0s Are you sure the issue with this app is the fact that there is plugin content in it? I'm new to b2g development, so I'm not really sure how to start debugging this.
Reporter | ||
Comment 12•12 years ago
|
||
(In reply to Margaret Leibovic [:margaret] from comment #11) > (In reply to Jason Smith [:jsmith] from comment #9) > > Talked about this in triage - the root cause of the problem here is that we > > do a really bad job of handling the case of when we get lots of flash > > content in web content. The app errors out immediately on launch. Accessing > > it directly in the browser looks really bad (missing many layout pieces). We > > should handle in-content flash content gracefully by producing a "not > > supported" UI in FF OS. > > The plugin overlay code that shows an error message in desktop/mobile > Firefox is in core/toolkit code, and it does show up in the browser app on > my b2g desktop build: > http://cl.ly/image/192A0E2G0y0s > > Are you sure the issue with this app is the fact that there is plugin > content in it? I'm new to b2g development, so I'm not really sure how to > start debugging this. Hmm...I see your point. I tried this - http://mozqa.com/data/firefox/plugins/flash/test_swf_embed.html and saw the plugin error message appear. So it isn't a problem in the general case then. I compared the behavior of http://beta.swartag.com/?task=view&id=1159 on the following devices when the game (not the advertisement starts loading): - Galaxy Nexus, FF Android - plugin content shows up - FF Desktop - plugin content shows up - FF OS - the plugin warning doesn't show up at all (on device) I did confirm that this game is indeed a flash game though. If it would be helpful, we could outreach to the app developer here, as Lisa (app review lead) could probably get us in contact with them. Btw, if you need a device to directly test this in the office in either SF or Mt view, you could find me if you are in Mt View (I sit near Zombocom) or find Naoki in SF.
Summary: Generate a more graceful fallback content UI when flash content is encountered, don't error out or generate a blank white content → No plugin support error does not appear on Dinopanic application
Updated•12 years ago
|
Priority: -- → P2
Whiteboard: [WebAPI:P0]
Reporter | ||
Comment 13•12 years ago
|
||
Given that we are really cracking down on priorities here for blocking, I'm going to renom this bug for a minus. The behavior isn't terrible (no block of content vs. plugin not supported) and it's not happening in the general case.
blocking-basecamp: + → ?
Comment 14•12 years ago
|
||
Given that it's not happening in the general case, let's not block on this.
blocking-basecamp: ? → -
Comment 15•12 years ago
|
||
I haven't been working on this since it isn't a blocker. Unassigning myself in case someone else wants to pick it up.
Assignee: margaret.leibovic → nobody
Comment 16•6 years ago
|
||
Firefox OS is not being worked on
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
You need to log in
before you can comment on or make changes to this bug.
Description
•