Fonts hinting

[Mario Torre]

Is there anything changed in the way OpenJDK handles fonts in the latest
fedora?

I update the system:

java-1.6.0-openjdk-1.6.0.0-43.1.8.2.fc13.x86_64

And now my nice antialiased fonts are gone. I do have freetype with the
patented hinting enabled (and the patent *is* expired), but it looks
like the setting is not picked up by OpenJDK.

I thought we used the system freetype.

Setting awt.useSystemAAFontSettings="lcd_hrgb" also doesn't change
anything, so I think something broke.

Cheers,
Mario
-- 
pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF

Proud GNU Classpath developer: http://www.classpath.org/
Read About us at: http://planet.classpath.org
OpenJDK: http://openjdk.java.net/projects/caciocavallo/

Please, support open standards:
http://endsoftpatents.org/

-- 
pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF

Proud GNU Classpath developer: http://www.classpath.org/
Read About us at: http://planet.classpath.org
OpenJDK: http://openjdk.java.net/projects/caciocavallo/

Please, support open standards:
http://endsoftpatents.org/

[Andrew Hughes]

On 11:40 Fri 26 Nov     , Mario Torre wrote:
>  Is there anything changed in the way OpenJDK handles fonts in the
> latest
>  fedora?
>  
>  I update the system:
>  
>  java-1.6.0-openjdk-1.6.0.0-43.1.8.2.fc13.x86_64
>  
>  And now my nice antialiased fonts are gone. I do have freetype with the
>  patented hinting enabled (and the patent *is* expired), but it looks
>  like the setting is not picked up by OpenJDK.
>  
>  I thought we used the system freetype.
>  
>  Setting awt.useSystemAAFontSettings="lcd_hrgb" also doesn't change
>  anything, so I think something broke.
>  
>  Cheers,
>  Mario
>  -- 
>  pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
>  Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF
>  
>  Proud GNU Classpath developer: http://www.classpath.org/
>  Read About us at: http://planet.classpath.org
>  OpenJDK: http://openjdk.java.net/projects/caciocavallo/
>  
>  Please, support open standards:
>  http://endsoftpatents.org/
>  
>  -- 
>  pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
>  Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF
>  
>  Proud GNU Classpath developer: http://www.classpath.org/
>  Read About us at: http://planet.classpath.org
>  OpenJDK: http://openjdk.java.net/projects/caciocavallo/
>  
>  Please, support open standards:
>  http://endsoftpatents.org/
>  
>  --
>  java-devel mailing list
>  java-devel(a)lists.fedoraproject.org
>  https://admin.fedoraproject.org/mailman/listinfo/java-devel 
1.8.2 is quite an old release now:

New in release 1.8.2 (2010-10-13):

* Security updates
  - S6914943, CVE-2009-3555: TLS: MITM attacks via session renegotiation 
  - S6559775, CVE-2010-3568: OpenJDK Deserialization Race condition
  - S6891766, CVE-2010-3554: OpenJDK corba reflection vulnerabilities
  - S6925710, CVE-2010-3562: OpenJDK IndexColorModel double-free
  - S6938813, CVE-2010-3557: OpenJDK Swing mutable static
  - S6957564, CVE-2010-3548: OpenJDK DNS server IP address information leak
  - S6958060, CVE-2010-3564: OpenJDK kerberos vulnerability
  - S6963023, CVE-2010-3565: OpenJDK JPEG writeImage remote code execution
  - S6963489, CVE-2010-3566: OpenJDK ICC Profile remote code execution
  - S6966692, CVE-2010-3569: OpenJDK Serialization inconsistencies
  - S6622002, CVE-2010-3553: UIDefault.ProxyLazyValue has unsafe reflection usage
  - S6925672, CVE-2010-3561: Privileged ServerSocket.accept allows receiving connections
from any host
  - S6952017, CVE-2010-3549: HttpURLConnection chunked encoding issue (Http request
splitting)
  - S6952603, CVE-2010-3551: NetworkInterface reveals local network address to untrusted
code
  - S6961084, CVE-2010-3541: limit setting of some request headers in HttpURLConnection
  - S6963285, CVE-2010-3567: Crash in ICU Opentype layout engine due to mismatch in
character counts
  - S6980004, CVE-2010-3573: limit HTTP request cookie headers in HttpURLConnection
  - S6981426, CVE-2010-3574: limit use of TRACE method in HttpURLConnection
  - S6990437: Update with correct copyright info for source and test files from SSR10_02
fixes
* Fixes:
  - G244901: Skip test_gamma on hardened (PaX-enabled) kernels
  - G266295: Provide font configuration for Gentoo.
  - Provide font configuration for RHEL 6.
  - RH633510: OpenJDK should use NUMA even if glibc doesn't provide it
* Backports:
  - S6951319: enable solaris builds using Sun Studio 12 update 1 (fixes PR398)
  - S6539464, RH500077: Ensure java.lang.Math functions provide consistent results.
  - S6638712: Inference with wildcard types causes selection of inapplicable method
  - S6650759: Inference of formal type parameter (unused in formal parameters) is not
performed
  - S6623943: javax.swing.TimerQueue's thread occasionally fails to start
* NetX:
  - Fix browser command in BasicService.showDocument(URL)
  - Run programs that inherit main(String[]) in their main-class
  - Run JNLP files that use 1.6 as the spec version
  - RH601281: Possible NullPointerException in splash screen code
  - New man page for javaws
* Plugin 
  - RH560193: Fix zip error when applet jar contained another 0-byte jar
  - PR519: 100% CPU usage when displaying applets in Webkit based browsers

You could try using strace on a small AWT app to see if the right fontconfig.properties is
being
picked up.  Maybe the Gentoo or RHEL additions broke something?  Strange it hasn't
been reported
before if so.
-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

Support Free Java!
Contribute to GNU Classpath and the OpenJDK
http://www.gnu.org/software/classpath
http://openjdk.java.net
PGP Key: 94EFD9D8 (http://subkeys.pgp.net)
Fingerprint = F8EF F1EA 401E 2E60 15FA  7927 142C 2591 94EF D9D8

[Omair Majid]

On 11/26/2010 05:40 AM, Mario Torre wrote:
>  Is there anything changed in the way OpenJDK handles fonts in the
> latest
>  fedora?
>
>  I update the system:
>
>  java-1.6.0-openjdk-1.6.0.0-43.1.8.2.fc13.x86_64
>
>  And now my nice antialiased fonts are gone. I do have freetype with the
>  patented hinting enabled (and the patent *is* expired), but it looks
>  like the setting is not picked up by OpenJDK.
>  
AFAIK the default freetype packages have subpixel rendering disabled. 
This is different from the bytecode interpreter (which does hinting). 
Can you please ensure that supixel rendering is enabled in freetyp? 
(Note that most other applications work around it, so they wont be 
affected if subpixel rendering is disabled in freetype). When I build 
freetype with subpixel rendering enabled, it shows up fine with OpenJDK.

Cheers,
Omair

[Mario Torre]

Il giorno ven, 26/11/2010 alle 09.50 -0500, Omair Majid ha scritto:
>  On 11/26/2010 05:40 AM, Mario Torre wrote:
>  > Is there anything changed in the way OpenJDK handles fonts in the latest
>  > fedora?
>  >
>  > I update the system:
>  >
>  > java-1.6.0-openjdk-1.6.0.0-43.1.8.2.fc13.x86_64
>  >
>  > And now my nice antialiased fonts are gone. I do have freetype with the
>  > patented hinting enabled (and the patent *is* expired), but it looks
>  > like the setting is not picked up by OpenJDK.
>  >
>  
>  AFAIK the default freetype packages have subpixel rendering disabled. 
>  This is different from the bytecode interpreter (which does hinting). 
>  Can you please ensure that supixel rendering is enabled in freetyp? 
>  (Note that most other applications work around it, so they wont be 
>  affected if subpixel rendering is disabled in freetype). When I build 
>  freetype with subpixel rendering enabled, it shows up fine with OpenJDK.
>  
>  Cheers,
>  Omair 
Yep, indeed,

It was my fault, the update also updated freetype, but I resumed from
suspend, that means the freetype library used by the running the other
applications was not reloaded, but the first time I launched OpenJDK...
well...

I fixed this by recompiling freetype... Funny that Fedora still has the
subpixel rendering off by default.

Thanks for the hints :)

Cheers,
Mario
-- 
pgp key: http://subkeys.pgp.net/ PGP Key ID: 80F240CF
Fingerprint: BA39 9666 94EC 8B73 27FA  FC7C 4086 63E3 80F2 40CF

Proud GNU Classpath developer: http://www.classpath.org/
Read About us at: http://planet.classpath.org
OpenJDK: http://openjdk.java.net/projects/caciocavallo/

Please, support open standards:
http://endsoftpatents.org/

[Omair Majid]

On 11/26/2010 10:20 AM, Mario Torre wrote:
>  Il giorno ven, 26/11/2010 alle 09.50 -0500, Omair Majid ha scritto:
> > On 11/26/2010 05:40 AM, Mario Torre wrote:
> >> Is there anything changed in the way OpenJDK handles fonts in the latest
> >> fedora?
> >>
> >> I update the system:
> >>
> >> java-1.6.0-openjdk-1.6.0.0-43.1.8.2.fc13.x86_64
> >>
> >> And now my nice antialiased fonts are gone. I do have freetype with the
> >> patented hinting enabled (and the patent *is* expired), but it looks
> >> like the setting is not picked up by OpenJDK.
> >>
> >
> > AFAIK the default freetype packages have subpixel rendering disabled.
> > This is different from the bytecode interpreter (which does hinting).
> > Can you please ensure that supixel rendering is enabled in freetyp?
> > (Note that most other applications work around it, so they wont be
> > affected if subpixel rendering is disabled in freetype). When I build
> > freetype with subpixel rendering enabled, it shows up fine with OpenJDK.
> >
> > Cheers,
> > Omair
>
>  Yep, indeed,
>
>  It was my fault, the update also updated freetype, but I resumed from
>  suspend, that means the freetype library used by the running the other
>  applications was not reloaded, but the first time I launched OpenJDK...
>  well...
>
>  I fixed this by recompiling freetype... Funny that Fedora still has the
>  subpixel rendering off by default.
>  
Almost nothing (aside from OpenJDK) uses freetype for subpixel 
rendering. Almost every toolkit including Xft, Cairo (gnome) and Qt 
(KDE) performs subpixel anti aliasing itself.

>  Thanks for the hints :)
>  
You're welcome.

Cheers,
Omair