Patch for issue 1367 (was: Packet loss with JXTASockets)

Hi Mohamed,

coincidentally, I started testing that patch today. I have had some  
mixed results, so far, but I haven't been able to get something  
reproducible yet to share with you. I'll keep working on it until I  
do, but briefly what's happening is that (using the JxtaSocketServer  
and JxtaSocket example programs from the programmer's guide) the  
client *sometimes* attempts to make the second connection before the  
server has received EOT for the first batch of data, and that 2nd  
connection fails (connection attempt times out). I'm no longer seeing  
the error where the client closes the connection and the server  
misses the last of the data, however. As to the new problem, it's not  
happening consistently, and I'm not even sure of the cause yet. As  
soon as I can get a better description of it, I'll post it to the  
issue (1367).

Regards,

Vanessa

P.S. I'm also posting this to [email protected] and routing the replies  
there.

On 2-Aug-05, at 12:17 PM, Mohamed Abdelaziz wrote:

> Mohamed Abdelaziz wrote:
>
>
>> Matthias Horst wrote:
>>
>>
>>> Hi,
>>>
>>> me too :)  You understand it right. I implemented your idea and  
>>> it gives me the result. But still this is a "hack". I thought  
>>> that the principles of sending an acknowlege information is  
>>> already placed somewhere deeper in the communication stack (io  
>>> stream, TCP/IP etc), and must not be placed in the application  
>>> itself. At least it is said that JXTASockets are "reliable",  
>>> aren't they?
>>>
>>> Moreover I run a small example from P2PSocket project which is  
>>> based on JXTASockets. As far as I found out especially the close 
>>> () command is not overwritten, such that one calls  the same  
>>> method as in the original JXTASocket API. But to my surprise the  
>>> problem does not exist!
>>>
>>
>>
>> The P2PSocket project uses JxtaSockets, and therefore inherits the  
>> capabilities of JxtaSocket. It also appears that issue1367 <http:// 
>> platform.jxta.org/nonav/issues/showattachment.cgi/1356/ 
>> issue1367.patch> has already been applied to p2psocket (thanks for  
>> the validation point, I'll commit the patch shortly).
>>
>>
>
> After taking a closer look at p2psockets.  it does not appear  
> issue1367 has not been applied, never the less, you should apply  
> the patch as it addresses proper socket closure.
> Please report back your findings.
>
> Thanks,
> Mohamed
>
>
>
>
>>
>> Mohamed
>>
>>
>
>
> -- 
> http://blogs.sun.com/roller/page/hamada
> http://weblogs.java.net/blog/hamada
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>