On Sunday, March 17, 2002, at 04:23 PM, Ian Summerfield wrote:
> I didn't know any of what you said. I'm working on OS X, and this > evening > just finished a little project to monitor an APC UPS. I used "accept > datagram" but then found the "have some work to do" bit you refer > to was > only a case of calling "read from socket nnnn with message xxxxxx" > at the > end of my message handler. It works fine. Is this the proper > approach? I don't know if it is proper. I didn't even think of the idea. I didn't even know the "read from socket" callbackMessage had a second parameter until I saw it in the dictionary for Revolution 1.1.1 RC1. > on mouseup > accept datagram connections on port 3052 with message "gotGram" > end mouseup > > on gotGram fromwho,what > if what is not empty then > -- do whatever I want with the data > end if > read from socket 3052 with message "gotGram" > end gotGram This has a problem on my system (OS X with R 1.1.1 RC1). You might have it in yours. Try it with a beep or counter at the end of your "gotGram" handler. I get a call back with an empty datagram not only the first time but also repeatedly about 100 times a second. Then I get socket timeout callbacks, even when I'm receiving datagrams. What I had in mind for "have some work to do" was closing and and accepting again after each valid datagram. Dar Scott Dar Scott Consulting Albuquerque _______________________________________________ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
