WebKit Bugzilla
New
Browse
Search+
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED WORKSFORME
20258
HTML5 canvas: fillRect() and friends do not support negative heights
https://bugs.webkit.org/show_bug.cgi?id=20258
Summary
HTML5 canvas: fillRect() and friends do not support negative heights
David Jones
Reported
2008-08-01 14:20:17 PDT
PROBLEM fillRect(x, y, w, h) will fail with "INDEX_SIZE_ERR: INDEX_SIZE_ERR: DOM Exception 1" (or something like that) when h is negative. There's nothing in the spec to prohibit fillRect() from taking a negative h. The specification requires that the path be formed from the four coordinates: "(x, y), (x+w, y), (x+w, y+h), (x, y+h)". These four coodinates still have a perfectly sensible interpretation when either h or w is negative, and there's nothing I can see in the spec from prohibiting h and w from being negative. It is sometimes even quite natural to use a negative h, for example if one has flipped the Y-axis by issuing a scale(1,-1) then one still might want (x,y) to be the top-left of a rectangle, in which case h will be negative.
Attachments
Negative width/height
(455 bytes, text/html)
2008-08-21 21:14 PDT
,
Dirk Schulze
no flags
Details
View All
Add attachment
proposed patch, testcase, etc.
Dirk Schulze
Comment 1
2008-08-21 21:14:13 PDT
Created
attachment 22929
[details]
Negative width/height This test has a negative with and height for fillRect() and strokeRect(). Can't confirm it on cairo but can not test Safari.
David Jones
Comment 2
2008-08-22 00:32:53 PDT
Ah thanks for the test (odd, I had one lying around when I reported the bug, I wonder why I didn't add it). Running the test on Safari results in no visible output and "INDEX_SIZE_ERR: DOM Exception 1" in the error console.
Dirk Schulze
Comment 3
2008-12-07 23:56:47 PST
Can't confirm it on Mac with the latest trunk. Possibly fixed in some earlier versions. Do you still have problems with it? Please try one of the
http://nightly.webkit.org/
builds to check.
Oliver Hunt
Comment 4
2008-12-29 04:09:53 PST
I agree with dirk, the code obviously supports this behaviour, and the testcase passes. I assume this has been fixed at some point.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug