On Aug 17, 2006, at 13:09 UTC, Richard wrote: > > The DataAvailable event handler is probably the problem. It can, > and will, > > allow other DataAvailable events to trigger from inside the > DataAvailable > > Handler. > > That, is disastrous.
Not necessarily. It depends on the circumstances -- does this happen only if you do something to service the main event loop (as I would expect)? What about if you disable background tasks? Finally, even if it really does fire reentrantly, whenever data comes in, that's still not a disaster as long as you're aware of it. But it seems like we need more detail on this. > So this bug has been present as far back as 1999 and it's still with > us to this day. Has RS decided to let go of serial port support? I'm quite sure they have not. Let's not exaggerate. > In my DataAvailable event, if all I do is read the buffer and then > send a fixed string back out the serial port, it is delayed from the > received packet anywhere from around 2 ms to 80 ms (yes, 80 > milliseconds). > > What must I do to insure that this Quad G5 can send a reply within 10 > ms or so? Probably, you must install an operating system that lacks preemptive multitasking. Milliseconds are very short -- much smaller than the time-slice on a typical preemptive OS. Such an OS is not suitable for uses where such real-time responsiveness is required. Best, - Joe -- Joe Strout -- [EMAIL PROTECTED] Verified Express, LLC "Making the Internet a Better Place" http://www.verex.com/ _______________________________________________ Unsubscribe or switch delivery mode: <http://www.realsoftware.com/support/listmanager/> Search the archives of this list here: <http://support.realsoftware.com/listarchives/lists.html>
