🚀 go-pugleaf

Yeah, I know, but I'm willing to take the hit on performance to meet a
deadline. The real fix would be to have the vendor properly handle modem's
with hardware flowcontrol, but the product must be released at the end of
the month and the vender is not what I consider to be reliable. (OK,
actually the real fix would be to replace the vendor.)  None of their code
works the first time. They consider us to be their test bed. The contoller
that we are talking to is maxed out as far as resources and can't be made to
go any faster anyway, so the hit on performance is acceptable in this
iteration.  Our new hardware is about 60 times faster and then I'll need to
worry about performance of everything, but now I can make a poor technical
decision in order to better serve the business imperatives.

"Andrue Cope [TeamB]" <no.spam@spammy.com> wrote in message
news:40c5c721$1@newsgroups.borland.com...
> Ron Sawyer wrote:
>
> > (Yes, if it were up to me, we would tell the vender to get stuffed,
> > that we are the customer, that he should fix his own da#% code, etc,
> > etc. My boss says we have been doing business with him too long to
> > easily make changes now...)
>
> Lol. I'd be more concerned about the fact that 80% of your test modems
> didn't support hardware flow control. I would have thought it better to
> get it working on the 80% of modems that aren't operating correctly
> than to turn it off for the 20% of modems that are.
>
> Resorting to XON/XOFF is going to hit you with a 10 to 20% throughput
> reduction as well being less reliable.
>
> -- 
> Andrue Cope [Team]
> [Bicester, Uk]
> http://info.borland.com/newsgroups/guide.html