On Mon, 21 Jan 2002, Martin Wilck wrote:
On Thu, 17 Jan 2002 lawson_whitney@juno.com wrote:
Perhaps I screwed up the blocking semantics of non-overlapped IO.
That's what it looks like to me.
I can't make sense out of these traces. The numbers of bytes transferred are different, but you were transferring different data. The general pattern of function calls and messages is very similar with and without the patch. Especially the error patterns are similar.
Martin
Btw I don't think I can help with the serial app. I see no indication that overlapped IO plays a role there.
Maybe I was hallucinating, or maybe the juno server was just having a bad half-hour. Sometimes, not often, these sessions can fail for no good reason, or at least not one that has anything to do with program code on this end. I thought maybe trace was changing the timing of things, so I tried the app without a trace, but with tcpdump active. AFAICT, it worked fine. Twice. One other possibility, - the script I used to run the app using wine as last built (instead of the installed wine) invokes it --winver nt40. I don't think I did that earlier, and the app might behave differently. I've installed the patches for now and I'll beat on them a bit more and see if anything shakes loose.
Sorry, maybe I put you and me both to a lot of trouble because of an unrelated glitch. I would feel better if somebody else besides me would take a swat at testing these winsock patches, too.
Lawson