Closed Bug 437372 Opened 18 years ago Closed 18 years ago

Fennec should implement psm security error ui

Categories

(Firefox for Android Graveyard :: General, defect, P2)

x86
Linux
defect

Tracking

(Not tracked)

VERIFIED FIXED
fennec1.0a1

People

(Reporter: romaxa, Assigned: johnath)

References

()

Details

Attachments

(1 file, 2 obsolete files)

Currently Fennec does not have proper PSM ui implementation for handling cert errors. We have to put preference for using error pages and add netError.dtd override with implementation...
browser.xul.error_pages.enabled - should be enabled to get rid of ugly alert dialog showing..
Assignee: nobody → johnath
Flags: wanted-fennec1.0+
Depends on: 446684
Attached patch Straight port from Firefox (obsolete) — — Splinter Review
This gives us the buttons, strings, and browser.js logic to mimic Firefox's handling. It's not clear to me if that's the best outcome on this platform, but it is substantially better than what we have now, which is nothing. :) The string duplication is an annoying side effect of the way netError.xhtml is built, and the browser.js code is a way for us to handle in-content events with chrome privilege, when appropriate.
Priority: -- → P2
Target Milestone: --- → Fennec M7
Attachment #331162 - Flags: review?(enndeakin)
Comment on attachment 331162 [details] [diff] [review] Straight port from Firefox Are all those strings used? Or did you just copy the dtd file? Do we not need to worry about the case where the homepage is changed?
Target Milestone: Fennec M7 → Fennec A1
(In reply to comment #3) > (From update of attachment 331162 [details] [diff] [review]) > Are all those strings used? Or did you just copy the dtd file? The way that override works on netError, if you decide to provide your own strings anywhere in that dtd, you get to provide them everywhere in that dtd - you can't do per-entity overrides. So yes, they are used, because yes, I did just copy the dtd file. :) > Do we not need to worry about the case where the homepage is changed? Hrmph - yeah, we should probably use the default branch there, not gPrefService, that way we end up at a known-good page.
Attachment #331162 - Flags: review?(enndeakin)
Comment on attachment 331162 [details] [diff] [review] Straight port from Firefox New patch coming I'm told
Attached patch Use default branch, un-bitrot (obsolete) — — Splinter Review
Uses the default branch to get the pref. A lot of the generic event handling code has moved to browser-ui.js in the interim, but this isn't really "UI" in the way that scroll events and onmousedown are, so I kept this code in browser.js and just removed the parts that hooked into the Browser-wide handleEvent(). I've confirmed that this does do the basics, get me out of here, exception adding, but it's not really optimal for mobile, and my testing shows (unsurprisingly) that Fennec is just as beset upon by bug 431712 as Firefox is. Annoying. Still, this patch is better than not having this patch.
Attachment #331162 - Attachment is obsolete: true
Attachment #337650 - Flags: review?(enndeakin)
Whoops - last version didn't have the added netError.dtd file
Attachment #337650 - Attachment is obsolete: true
Attachment #337651 - Flags: review?(enndeakin)
Attachment #337650 - Flags: review?(enndeakin)
Attachment #337651 - Flags: review?(enndeakin) → review+
Status: NEW → RESOLVED
Closed: 18 years ago
Resolution: --- → FIXED
Depends on: 456015
verified with 1.9.2 20091001 beta4 on my n810
Status: RESOLVED → VERIFIED
Flags: in-litmus?
can someone point me to what the psm security error UI should look like? im trying to put together a manual testcase with steps to verify this. thanks.
I'm reassigning this to myself for adding litmus testcases for this (aakashd discussed this on irc with blassey and ted on #mobile).
Assignee: johnath → martijn.martijn
Assignee: martijn.martijn → johnath
Flags: in-litmus? → in-litmus?(martijn.martijn)
Flags: in-litmus?(martijn.martijn)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: