Hi everyone.
After fixing the memory leak tests (made a couple of object references
WeakReferences in the test classes—although it seems now that running
those tests from the command line is fine, but in the whole regard test
run I still get JFrame objects not cleared), I ran regard on my Linux
(1.6.0-b105), Windows (1.5.0_10-b03) and Mac (1.5.0_06-64) machines.
There are a number of problems on the Mac, which I would prefer to be
fixed before the release if possible. The problems fall in to a few
different categories:
▪ Text layout
http://mcc.id.au/temp/t/show?mac,fontChoice
http://mcc.id.au/temp/t/show?mac,smallFonts
http://mcc.id.au/temp/t/show?mac,textFeatures
Incorrect horizontal text placement of italic text.
http://mcc.id.au/temp/t/show?mac,longTextOnPath
http://mcc.id.au/temp/t/show?mac,textAnchor2
http://mcc.id.au/temp/t/show?mac,textAnchor3
http://mcc.id.au/temp/t/show?mac,textLayout2
http://mcc.id.au/temp/t/show?mac,text
http://mcc.id.au/temp/t/show?mac,text_children2
http://mcc.id.au/temp/t/show?mac,text_dylist1
http://mcc.id.au/temp/t/show?mac,text_dylist2
http://mcc.id.au/temp/t/show?mac,text-alignment-BE-11
Incorrect vertical text placement.
http://mcc.id.au/temp/t/show?mac,textLayout
http://mcc.id.au/temp/t/show?mac,textLength
lengthAdjust="spacingAndGlyphs" not computed properly.
http://mcc.id.au/temp/t/show?mac,verticalText
http://mcc.id.au/temp/t/show?mac,verticalTextOnPath
Vertical text aligned horizontally instead.
http://mcc.id.au/temp/t/show?mac,textGlyphOrientationHorizontal
glyph-orientation-horizontal rotated text goes in the wrong
direction.
▪ Flow text layout
http://mcc.id.au/temp/t/show?mac,flowBidi
http://mcc.id.au/temp/t/show?mac,flowText
http://mcc.id.au/temp/t/show?mac,flowText2
http://mcc.id.au/temp/t/show?mac,flowText3
http://mcc.id.au/temp/t/show?mac,flowText4
http://mcc.id.au/temp/t/show?mac,flowRegionBreak
http://mcc.id.au/temp/t/show?mac,lineHeightFontShorthand
Text is all over the place.
▪ Flow text extension layout
http://mcc.id.au/temp/t/show?mac,flowText,1,3
▪ Gradients
http://mcc.id.au/temp/t/show?mac,feTile
http://mcc.id.au/temp/t/show?mac,feTileTarget
http://mcc.id.au/temp/t/show?mac,markersShapes
http://mcc.id.au/temp/t/show?mac,linearGradientRepeat
Colours slightly off.
http://mcc.id.au/temp/t/show?mac,gradientLimit
Top right gradient wrong.
http://mcc.id.au/temp/t/show?mac,gradientPoint
Gradients wrong. Got this ouptut on the console, so it may be a VM
bug:
fox:~/work/svn/batik/trunk $ java -jar batik-1.6/batik.jar
samples/tests/spec/paints/gradientPoint.svg
java(10415,0x1867800) malloc: *** vm_allocate(size=3200004096) failed (error
code=3)
java(10415,0x1867800) malloc: *** error: can't allocate region
java(10415,0x1867800) malloc: *** set a breakpoint in szone_error to debug
java(10415,0x1867800) malloc: *** vm_allocate(size=3200004096) failed (error
code=3)
java(10415,0x1867800) malloc: *** error: can't allocate region
java(10415,0x1867800) malloc: *** set a breakpoint in szone_error to debug
java(10415,0x1867800) malloc: *** vm_allocate(size=3200004096) failed (error
code=3)
java(10415,0x1867800) malloc: *** error: can't allocate region
java(10415,0x1867800) malloc: *** set a breakpoint in szone_error to debug
java(10415,0x1867800) malloc: *** vm_allocate(size=3200004096) failed (error
code=3)
java(10415,0x1867800) malloc: *** error: can't allocate region
java(10415,0x1867800) malloc: *** set a breakpoint in szone_error to debug
java.lang.OutOfMemoryError: Requested array size exceeds VM limit
http://mcc.id.au/temp/t/show?mac,masking-mask-BE-05
http://mcc.id.au/temp/t/show?mac,masking-mask-BE-06
Mask gradient wrong.
▪ Non pixel aligning
http://mcc.id.au/temp/t/show?mac,markersMisc
The squares on the bottom right test are not pixel aligned.
http://mcc.id.au/temp/t/show?mac,patternRegions
Squares around each pattern are not pixel aligned. (Also slightly
off gradient colours on the last pattern.)
http://mcc.id.au/temp/t/show?mac,enclosureList2
Coloured squares are not pixel aligned.
▪ Opacity
http://mcc.id.au/temp/t/show?mac,polygon_points2
http://mcc.id.au/temp/t/show?mac,paintOpacity
http://mcc.id.au/temp/t/show?mac,enableBackground
http://mcc.id.au/temp/t/show?mac,svgEnableBackground
Semi-opaque paints are slightly different from the reference image.
▪ Miscellaneous
http://mcc.id.au/temp/t/show?mac,batik70
The vertical stripe background renders incorrectly. The aqua stripe
seems to be pink. Also, the path forming the bottom part of the ‘i’
glyph isn’t closed.
http://mcc.id.au/temp/t/show?mac,moonPhases
The lat/long lines on the earths don’t show unless zoomed right in.
http://mcc.id.au/temp/t/show?mac,histogramNormalization
Image normalisation not done.
http://mcc.id.au/temp/t/show?mac,feConvolveMatrix
Bottom right case doesn’t have black in the right places.
http://mcc.id.au/temp/t/show?mac,maskRegions
Masking is not anti-aliased. (Maybe that’s just the Mac Java2D
implementation that defaults to non-anti-aliasing.)
http://mcc.id.au/temp/t/show?mac,fontOnPath
Text selection highlight looks a bit strange (white border is not
right on the edge of the highlight).
These on Windows:
▪ Opacity
http://mcc.id.au/temp/t/show?win,enableBackground,1,3
http://mcc.id.au/temp/t/show?win,svgEnableBackground,1,3
http://mcc.id.au/temp/t/show?win,paintOpacity,1,3
http://mcc.id.au/temp/t/show?win,polygon_points2,1,3
Semi-opaque paints are slightly different from the reference image.
▪ Gradients
http://mcc.id.au/temp/t/show?win,feTile,1,3
Colours slightly off.
These on Linux:
▪ Opacity
http://mcc.id.au/temp/t/show?linux,enableBackground,1,3
http://mcc.id.au/temp/t/show?linux,svgEnableBackground,1,3
http://mcc.id.au/temp/t/show?linux,paintOpacity,1,3
http://mcc.id.au/temp/t/show?linux,polygon_points2,1,3
Semi-opaque paints are slightly different from the reference image.
▪ Gradients
http://mcc.id.au/temp/t/show?linux,feTile,1,3
Colours slightly off.
▪ ICC colour
http://mcc.id.au/temp/t/show?linux,paintType,1,3
http://mcc.id.au/temp/t/show?linux,color-colorProf-BE-03,1,3
ICC colours render slightly differently from the reference image.
Some of these may be just due to different JRE versions doing
compositing a bit differently; whether they’re incorrect I’m not sure.
--
Cameron McCormack, http://mcc.id.au/
xmpp:[EMAIL PROTECTED] ▪ ICQ 26955922 ▪ MSN [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
