Skip to content
This repository has been archived by the owner. It is now read-only.
Permalink
Browse files

Try to fix the program not starting bug (thanks, Goble!)

  • Loading branch information...
SubJunk committed Feb 2, 2012
1 parent 80e7c43 commit 3c33d727cb93bf1758714e2684ad86d514192436
Showing with 1 addition and 1 deletion.
  1. +1 −1 PMS.launch
@@ -13,5 +13,5 @@
<stringAttribute key="org.eclipse.jdt.launching.MAIN_TYPE" value="net.pms.PMS"/>
<stringAttribute key="org.eclipse.jdt.launching.PROJECT_ATTR" value="ps3mediaserver"/>
<stringAttribute key="org.eclipse.jdt.launching.SOURCE_PATH_PROVIDER" value="org.eclipse.m2e.launchconfig.sourcepathProvider"/>
<stringAttribute key="org.eclipse.jdt.launching.VM_ARGUMENTS" value="-Xmx768M -Djava.net.preferIPv4Stack=true -Djava.encoding=UTF-8"/>
<stringAttribute key="org.eclipse.jdt.launching.VM_ARGUMENTS" value="-Xmx768M -Djava.encoding=UTF-8"/>
</launchConfiguration>

13 comments on commit 3c33d72

@taconaut

This comment has been minimized.

Copy link
Member

taconaut replied Feb 2, 2012

Any idea why this vm parameter had been added in the first place?

@gzsombor

This comment has been minimized.

Copy link
Contributor

gzsombor replied Feb 2, 2012

I think, this could be an explanation, but I don't know, it's correct/up to date/beneficial :
http://activemq.apache.org/multicast-watch-out-for-ipv6-vs-ipv4-support-on-your-operating-system-or-distribution-or-network.html

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 2, 2012

That has been in the code since january 1st 2009 and appears to be an Eclipse starting configuration?
Are you sure this is the thing?

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 2, 2012

Ah, hadn't read the discussion yet.
There are a couple more occurences in the code, I'll remove those as well.

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 2, 2012

Continuation committed in 73086af.

@ExSport

This comment has been minimized.

Copy link
Contributor

ExSport replied Feb 2, 2012

@gzsombor

This comment has been minimized.

Copy link
Contributor

gzsombor replied Feb 2, 2012

Are you sure, that pms works correctly on IPv6 ? Because for example, on the network interface detection, we explicitly skip IPv6 addresses.

@SubJunk

This comment has been minimized.

Copy link
Contributor Author

SubJunk replied Feb 12, 2012

We have a report of this revision and the next one breaking Linux compatibility:
http://www.ps3mediaserver.org/forum/viewtopic.php?f=14&t=8883&p=65177#p65177

It doesn't seem to have solved the bug it was supposed to solve - considering the debug.logs we've had about the issue, I suspect this only worked around the bug for some users instead of fixing it - so I will revert it since I was already thinking of doing that before that forum post.

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 13, 2012

The proper solution is probably to first fix the network code that requires the -Djava.net.preferIPv4Stack=true, and only then lose the option.

The question is: which code requires that option and why?

@taconaut

This comment has been minimized.

Copy link
Member

taconaut replied Feb 13, 2012

I'd suggest leaving this parameter enabled, except someone has a deeper knowledge of multicasting behaviour in ipv6 networks (having studied telecommunications, I can't help out without digging through the books)

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 13, 2012

At least Sun documented the whole IPv4 / IPv6 well.

If I understand the document correctly, Java shouldn't have a problem switching between the two unless we use IPv4 specific code (which we apparently do, or reverting wouldn't be necessary ;-)).

@gzsombor

This comment has been minimized.

Copy link
Contributor

gzsombor replied Feb 13, 2012

It would be an interesting test, to check, what happens, if we are announcing only on IPv6, and listening on that interface.

@Raptor399

This comment has been minimized.

Copy link
Member

Raptor399 replied Feb 13, 2012

Interesting perhaps. But it would be safer to make our code neutral and let the VM handle the IPv4 / IPv6 choice, as the documentation suggests.

Please sign in to comment.
You can’t perform that action at this time.
You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.