On Thu, 17 Jan 2008, nash wrote: > > The clients should be able to, without pipelining do the following: > > connect frame > > <- reply > > getFeatures > > <- reply > > setFilters > > <- reply > > <set filters active> > > then start pipelining. > > Most annoying part of the protocol to be frank ;-) Hence if there were > mandatory filters, I'd do something like: > connect mesg > -> reply > getFeatures (protocol requires I believe)
Nope, only the connect message is required. > setFilters > -> reply features > ->record anything > -> filters okay > [set internal parser replies to use this] You could do this, but you have to be careful about when you enable the filters. The server too will have to be careful (of which tpserver-cpp isn't yet) to make sure that synchronised. > > Although.... you probably don't want to pipeline around login.... (state > > change). > > Depends on how much you hate round trips... ;-) True, but it will save bandwidth if you wait and the the reply is fail. > > > > - client's implementation is known to be incorrect > > > > > > Personally I'd say for the last case: > > > - Let the client be incorrect, fix it in the client. Work arounds for > > > other bugs suck... this is one area open source can get advantage ;-) > > > > What do we do about the client in the mean time? In particular if it's > > like tpsai-py and automatically re-establishes the connection after it is > > drop for a good reason (like lost synchronisation, corrupted packets, > > etc). > > Nothing. > > If the amount of time required to fix the client is that drastically > different, then the change to the server would be to add that client to > a blacklist. Except if there is a client from outside our immediate project, or (heaven help us) a commercial client with a bug in it. Just look at the various provisions in Apache for working around broken HTTP/1.1 implementations (early versions of Java springs to mind). > Regards, > nash Later Lee
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ tp-devel mailing list [email protected] http://www.thousandparsec.net/tp/mailman.php/listinfo/tp-devel
