On 10/30/14 17:40, Jeff Morriss wrote:
Hmm, I don't do much with TCP (from a protocol perspective) but that
certainly makes sense. And, in fact, it appears that the TCP dissector
*wants* to act that way, based on the comments in the code:
/* This is called for SYN and SYN+ACK packets and the purpose is to
verify
* that we have seen window scaling in both directions.
* If we can't find window scaling being set in both directions
* that means it was present in the SYN but not in the SYN+ACK
* (or the SYN was missing) and then we disable the window scaling
* for this tcp session.
*/
It would be good if someone can open a bug report (attaching this
already-public capture file) so this isn't forgotten.
Oh, I just noticed there's already a bug open for it:
https://bugs.wireshark.org/bugzilla/show_bug.cgi?id=10514
