| |
GM7HUD > TECH 20.12.05 00:14l 29 Lines 1068 Bytes #999 (0) @ WW
BID : 773135GM7HUD
Read: DL1LCA GUEST OE7FMI
Subj: re: A puzzling fault
Path: DB0FHN<DB0MRW<DB0SON<DB0ERF<DB0EAM<DB0SMG<DB0RES<ON0BEL<ZL2BAU<GB7ESX
Sent: 051219/2135z @:GB7ESX.#31.GBR.EU #:71440 $:773135GM7HUD [Witham, Esx]NNA
K0CQ wrote:-
> I have found through long experiments that unless a frequency is
> totally private with only two users that I get more throughput
> (as does bbs forwarding) if maxframe is set to 1 than when set to
> 7. That's because when there are collisions they almost always
> are in the first frame and the protocol has no way of accepting
> the other 6 good frames and asking for the first only to be sent.
> The protocol can only reject that damaged frame and so all 7 get
> sent again.
And a long time ago in Europe we found the same thing.
*LAST CENTURY* ;-)
Thus DAMA was developed. Whereby the parameters of the connection between
you and the BBS/Node are varied dynamically.
Each user connected to the node is polled by the node and does not transmit
until commanded, minimising collisions between hidden stations.
Areas where DAMA is used have better throughputs for all users... even
those who have TNCs that dont support DAMA as long as many users do support
the protocol.
Have a Google for DAMA and give it a whirl.
73 de Andy GM7HUD
Read previous mail | Read next mail
| |
