On Sunday 16 July 2006 03:23, Bradley Arsenault wrote:
> I discussed it with several people (not just glob2'ers), they all
> agree TCP is the way to go. I don't understand why we chose UDP
> sockets, there so much less safe.

<advocate client="devil">UDP has much lower overhead and you have the 
possibility to drop non-critical packets to the floor, also dropped packets 
(by network congestion) don't stall the whole pipe but can be resent and 
integrated later.</advocate>

> At any rate all my sense seems to point to our choice of UDP sockets
> indirectly causing this bug, and that a change to TCP would do only
> good.

Presuming glob2 is neither latency- nor bandwidth-bound, I'd think that be a 
good idea too.


Regards, David

-- 
- hallo... wie gehts heute?
- *hust* gut *rotz* *keuch*
- gott sei dank kommunizieren wir über ein septisches medium ;)
 -- Matthias Leeb, Uni f. angewandte Kunst, 2005-02-15


_______________________________________________
glob2-devel mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/glob2-devel

Reply via email to