icedtea-sound and java9

[Omair Majid]

* Fridrich Strba <fridrich.strba at suse.com> [2017-11-07 10:42]:
> On 07/11/17 16:32, Andrew Hughes wrote:
> > I think it may be intended to be appended to sound.properties (if that
> > still exists
> > in 9), though that would have the effect of enabling the provider as default.
> 
> I found in the code just one place where the sound.properties is parsed,
> in JDK13Services.java
> 
> The experiment shows that icedtea.sound is used as default with empty
> sound.properties. It is maybe because of the order the providers are
> sorted out in the modules:
> 
> jlink --module-path /usr/lib64/jvm/java-9-openjdk-9/jmods
> --suggest-providers javax.sound.sampled.spi.MixerProvider
> 
> Suggested providers:
>   icedtea.sound provides javax.sound.sampled.spi.MixerProvider used by
> java.desktop
>   java.desktop provides javax.sound.sampled.spi.MixerProvider used by
> java.desktop
> 
> 
> Maybe it goes in alphabetical order, since when the module was called
> org.icedtea.sound, it was still using the internal one.

That sounds right to me: the order is not explicitly determined and may
change for many reasons. In the initial versions, we explicitly added a
provider to sound.properties to make the icedtea.sound provider *not*
the default: 

http://icedtea.classpath.org/hg/icedtea6/rev/8b85ec866923?revcount=10240#l7.1

> Concerning the pulseaudio.properties, it looks like it might have used
> some icedtea modification/patch. But I didn't find any place it would be
> parsed in jdk9 or in icedtea-3.6.0. Maybe just my eyes, maybe it is not
> used.

That sounds right to me too. I don't recall ever implementing a
pulseaudio.properties. Maybe it came out from the splitting of icedtea6
(sound.properties) and icedtea-sound?

Thanks,
Omair

-- 
PGP Key: 66484681 (http://pgp.mit.edu/)
Fingerprint = F072 555B 0A17 3957 4E95  0056 F286 F14F 6648 4681