Hi Samuel, apologies for the late reply, there were some changes in the mail system that lead to being offline last week. Thank you for the patch. What I am asking myself is whether the problem is * specific to links with a long RTT or * can be used in general for all links? I believe the patch is useful, but am not sure using it for the general case, need to work through the implications. Also the patch did not apply due to line-wrap, I tried to fix this manually but there were several places. The 'increase ISS on retransmission' issue was first discovered 2006 by Vladimir, http://www.mail-archive.com/dccp@xxxxxxxxxxxxxxx/msg00710.html Also, there are other things in DCCP that currently will not work well over long-delay links: * DCCP_TIMEOUT_INIT of 3 seconds (you mentioned this already in your email) * DCCP_SANE_RTT_MAX = 3 seconds (only used by ccid3/4) I wonder thus if it makes sense to introduce a compilation flag "DCCP_LONG_DELAY_LINKS" which changes these parameters and which would conditionally enable your patch? Gerrit -- To unsubscribe from this list: send the line "unsubscribe dccp" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
