JDK-6817112 : JDK7: deterioration in rendering of TrueType fonts - notably the JDK's Lucida fonts.
  • Type: Bug
  • Component: client-libs
  • Sub-Component: 2d
  • Affected Version: 7
  • Priority: P2
  • Status: Resolved
  • Resolution: Fixed
  • OS:
    generic,linux_suse_sles_10,windows_xp generic,linux_suse_sles_10,windows_xp
  • CPU: generic,x86
  • Submitted: 2009-03-13
  • Updated: 2010-08-05
  • Resolved: 2009-06-30
The Version table provides details related to the release that this issue/RFE will be addressed.

Unresolved : Release in which this issue/RFE will be addressed.
Resolved: Release in which this issue/RFE has been resolved.
Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.

To download the current JDK release, click here.
JDK 7
7 b63Fixed
Related Reports
Duplicate :  
JDK-6798591 - test closed/java/awt/font/CompositeGlyph/SymmetryTest.java failed on 1.7.0-ea-b43 but passed on b42
Relates :  
JDK-4356282 - RFE: JDK should support OpenType/CFF fonts
Description
Bug Info:
==========

Text fonts are not rendering properly and it looks blur in motif look and feel with JDK7.But it works fine with JDK6.

JDK Info:
=========
java version "1.7.0-ea"
Java(TM) SE Runtime Environment (build 1.7.0-ea-b50)
Java HotSpot(TM) Client VM (build 15.0-b02, mixed mode)

Steps to reporduce:
===================

1. Launch the swingset2 demo using JDK7-b50 and JDK6 builds.
2. Choose the motif look and feel on both the demos.
3. Now compare the rendering font on both the screens.

Please check the attached images for reference.
Comments
EVALUATION We should use different approaches for TrueType and Type1/CFF fonts. Truetype fonts only allow quadratic Bezier curves and use "half-way" points as control points if 2 off-curve points are specified in the font.
21-05-2009
EVALUATION This is indeed regression caused by scan convertion improvements (sc.c) that were done as part of adding opentype support.
05-05-2009
EVALUATION b43 surprised me! Turns out its not what I had originally suspected, the only change of note in b43 is 4356282: RFE: JDK should support OpenType/CFF fonts It appears probable that the hinting changes there have had adverse affects on other fonts. The effect on the fonts shipped with the JDK appear particularly severe, and that's the fonts that are being picked up by the Motif L&F on RH 5. Its not obvious in Metal because that does anti-aliasing. You don't need SwingSet, or even RH 5 to see this. Font2DTest on Solaris will do.
29-04-2009
EVALUATION I have seen this issue on RHEL 5 x86 machine. It has been regressed from jdk1.7.0-b43.
17-03-2009
EVALUATION Looks to me as if this is picking up a font that doesn't render well in B&W (ie no hints) - Motif L&F only uses B&W since that's what X11 does. In JDK7 this would be a possible result of our new code to automatically synthesise a font properties on Linux, if the fonts on that distro are poor. That's my guess, but I can't be sure. - doesn't identify platform ! The bug report says its 'generic', which is supposed to mean all platforms. But what little I can see of the backdrop in the images. doesn't look like Solaris. ACTION: Please update the bug with the the exact version of all OSes on which this is seen. - doesn't say when this started (report is against the current b50 but the fontconfig generation is since b33) ACTION: Please update the bug report with the first build in which this is observed. Marking bug incomplete pending this info from the submitter. The previous evaluator noted : >Probably the antialising hint is switched on by default for some reason? I can see how you'd think that from the synopsis saying it "looks blur" but it doesn't. Its still B&W. I've changed the synopsis from "Text fonts are not rendering properly and it looks blur in motif look and feel." to something which I think is a whole lot more accurate.
13-03-2009
EVALUATION Motif Laf is obsolete and we don't fix many bugs in it at least for a couple of years, so I beleive that it is not our regression Probably the antialising hint is switched on by default for some reason? reassigned to 2D team
13-03-2009