Closed
Bug 600395
Opened 15 years ago
Closed 4 years ago
different d2d/gdi behaviour
Categories
(Core :: Graphics, defect)
Tracking
()
RESOLVED
INVALID
| Tracking | Status | |
|---|---|---|
| blocking2.0 | --- | - |
People
(Reporter: vlad, Assigned: bas.schouten)
Details
(Keywords: regression, Whiteboard: [estimated-time: 2 days])
Take a look at the testcase in bug 600390 -- the behaviour is different in GDI vs. D2D. Unfortunately, I think the GDI behaviour is correct. Basically, the first three large squares in each set of 4 should look identical (all 6 should be the same), and the two fourth squares should look identical. This is what I see with GDI. With D2D, I see the first three and the third in the second set with sharp edges, and the first two in the second set with fuzzy edges.
Given that we're using a canvas as a source in the first set, that means it's getting a D2D surface as a source; the second set is getting an image surface as a source. I think we're taking some incorrect fast paths for all the ones that don't have fuzzy edges, and correctly hitting the slow EXTEND_NONE emulation path for the two fuzzy ones.
| Reporter | ||
Updated•15 years ago
|
blocking2.0: --- → ?
Comment 1•15 years ago
|
||
Can you explain why you think this should block?
| Reporter | ||
Comment 2•15 years ago
|
||
Because if it doesn't, then we're going to be chasing that inconsistency in various ways for the next 6 months... we went through similar things in the last major graphics rework, and not fixing inconsistencies like this always led to lots of pain. Given the inconsistency is between using an image source and a d2d source, I can see some of that pain happening as soon as we start optimizing images to d2d surfaces.
Updated•15 years ago
|
Assignee: nobody → bas.schouten
| Assignee | ||
Updated•15 years ago
|
Whiteboard: [estimated-time: 2 days]
Updated•15 years ago
|
Keywords: regression
| Assignee | ||
Comment 4•15 years ago
|
||
I seriously doubt we should block on this. Although I do not like the behavior difference, it does not give any trouble to users, changing it would have performance implications(disabling some optimizations) and I'm not sure I want to risk that for firefox 4.
Updated•15 years ago
|
blocking2.0: final+ → -
User Agent: Mozilla/5.0 (Windows NT 6.1; rv:45.0) Gecko/20100101 Firefox/45.0
Build ID: 20160315153207
I have tested your issue on latest FF release (45.0.1), latest Nightly build (Build ID: 20160331030231) and managed to reproduce it. I have observed that the top row of images(provided in the attachment of bug 600390) have a darker outline around the green rectangles than the bottom row of images.
Updated•9 years ago
|
Version: unspecified → 2.0 Branch
| Assignee | ||
Updated•4 years ago
|
Status: NEW → RESOLVED
Closed: 4 years ago
Resolution: --- → INVALID
You need to log in
before you can comment on or make changes to this bug.
Description
•