[jitsi-dev] ice4j patches


#1

I've been tracking my local changes in git-svn

I've exported them as patches that you can apply (attached)

Please let me know if you'd prefer me to just apply them directly into SVN

0001-Remove-extra-spaces-from-Candidate.toString-so-it-ca.patch (1.02 KB)

0002-Ensure-log-messages-contain-the-STUN-error-codes.patch (1.23 KB)

0003-Generate-log-messages-when-a-Harvester-is-disabled-d.patch (2.67 KB)

0004-Eliminate-race-condition-when-all-harvests-complete-.patch (1.59 KB)

0005-Add-log-messages-for-unusual-behavior-during-STUN-ha.patch (1.9 KB)

0006-Copy-attributes-into-a-vector-to-avoid-concurrent-mo.patch (1.1 KB)

0007-Use-notify-instead-of-notifyAll-because-only-one-oth.patch (1.61 KB)

0008-Check-for-null-when-closing-to-avoid-NullPointerExce.patch (875 Bytes)

0009-Enhance-DelegatingDatagramSocket-to-obtain-delegates.patch (6.38 KB)

0010-Don-t-overwrite-existing-system-properties.patch (1.26 KB)

0011-Additional-logging-about-STUN-candidate-harvest-prog.patch (4.19 KB)

0012-Add-more-log-messages-about-progress-and-retries.patch (1.94 KB)


#2

Hi Daniel,

Thank you for these patches!

I see there is a lot of logging output, I think I will and keep only the more useful. Anwyay I will try to review more deeply the patches at the end of the week or next week.

Best regards,

···

--
Seb

Le 18/01/12 15:57, Daniel Pocock a �crit :

I've been tracking my local changes in git-svn

I've exported them as patches that you can apply (attached)

Please let me know if you'd prefer me to just apply them directly into SVN


#3

I see there is a lot of logging output, I think I will and keep only the
more useful. Anwyay I will try to review more deeply the patches at the
end of the week or next week.

I've chosen the level (e.g. info or fine) appropriately so they should
only appear for people who want them

ICE negotiation only takes place once for each call, so I don't think
they have any significant performance impact. However, I found them
very useful for troubleshooting and I'm sure other people will too when
they try to add ice4j to a project.

The ICE support in my own project now works successfully (at least, for
the testing I've done) - I've tried it from various wifi NATs and the 3G
internet service on my phone. Thanks for all the help in getting to
this stage.

Regards,

Daniel


#4

Hi,

Le 18/01/12 16:24, Daniel Pocock a �crit :

I see there is a lot of logging output, I think I will and keep only the
more useful. Anwyay I will try to review more deeply the patches at the
end of the week or next week.

I've chosen the level (e.g. info or fine) appropriately so they should
only appear for people who want them

ICE negotiation only takes place once for each call, so I don't think
they have any significant performance impact. However, I found them
very useful for troubleshooting and I'm sure other people will too when
they try to add ice4j to a project.

OK.

The ICE support in my own project now works successfully (at least, for
the testing I've done) - I've tried it from various wifi NATs and the 3G
internet service on my phone. Thanks for all the help in getting to
this stage.

Glad to hear that your project works :).

Regards,

···

--
Seb

Regards,

Daniel


#5

I notice that some of my patches were added to ice4j in commit r301

Could anyone provide feedback about the remaining patches that were not
implemented, is it just waiting for someone to review, or should I
improve them in some way and submit them again?

Here is the link to the original message with the patches:
http://java.net/nonav/projects/jitsi/lists/dev/archive/2012-01/message/98

Without 0009-Enhance-DelegatingDatagramSocket-to-obtain-delegates.patch
people are unable to build Lumicall

···

On 18/01/12 15:24, Daniel Pocock wrote:

I see there is a lot of logging output, I think I will and keep only the
more useful. Anwyay I will try to review more deeply the patches at the
end of the week or next week.

I've chosen the level (e.g. info or fine) appropriately so they should
only appear for people who want them

ICE negotiation only takes place once for each call, so I don't think
they have any significant performance impact. However, I found them
very useful for troubleshooting and I'm sure other people will too when
they try to add ice4j to a project.

The ICE support in my own project now works successfully (at least, for
the testing I've done) - I've tried it from various wifi NATs and the 3G
internet service on my phone. Thanks for all the help in getting to
this stage.