On Tue, 20 Nov 2001, Jean Tourrilhes wrote: > I've integrated a bit better the patch of Martin. I've also > took opportunity to fix a bit some of the sysctl stuff.
Hi Jean, thanks for the improvement - yeah, soon after sending this quick hack I've realized there was a thinko in the bits-handling :-( Because this wouldn't hurt people testing with their latency problems I decided to wait and see, whether it would help at all. > Try the patch. If it doesn't work, you know were to complain ;-) No complains - applies (2.4.14, with few lines offset for most hunks) and compiles fine, nothing broken for me and irlap/qos seems to do the right thing with the min_tx_turn_time. Only tested with mtt=1000 peer device - I have no such mtt=0 guy. > Also, I've been thinking of this FAST_RR fiasco. It's been > bitting regularly. I guess the fix is to change the config options to > default to the right thing (so you would have to say "Yes" to > *disable* FAST_RR. I would suggest to simply stay with the semantincs of the existing FAST_RR config option but just make it enabled by default. Which would still mean you have to say "No" in order to *disable* stuff, IMHO? "no need to say Yes to enable" != "Yes to disable" > But that will confuse all people that were starting > to get used to the current config options. Well, people saying "Yes" or "No" shouldn't get confused - people blindly hitting the spacebar to toggle the default might get surprized... > Well, just tell me what you think... I think just changing the default is fine - if it would revert the meaning like "Yes to disable FAST_RR", I'd consider renaming the thing to prevent confusion. I believe the Configure.help entry should change from "if unsure, say N" to "if unsure, say Y" as well. Martin _______________________________________________ Linux-IrDA mailing list - [EMAIL PROTECTED] http://www.pasta.cs.UiT.No/mailman/listinfo/linux-irda
