I wonder if it's even practical to continue to use the existing portability guidelines without some update. For example the limit on string length forces modules using TCP/IP to incur a lot of overhead and run much more slowly than they would otherwise. More generally, if OpenVistA is to interoperate with other applications, it will become increasingly difficult to ignore the fact that other applications will work with data in different sized chunks and different formats than we are accustomed to in Standard MUMPS. It's great that there is an XML parser in Kernel, but if we really want to *use* it, how long can we ignore the fact that we can't support UTF-8, or even ISO-8859? Saving incoming data into a global is great, but do we have to wait until the connection is closed before we can even start parsing it? I know DNS an other protocols will run over TCP, but can we really go on forever ignoring UDP? There isn't even a standard way to write a server socket. Running netstat while "listening" is illuminating because most MUMPS implementations do nothing of the sort, they poll the device! Is it really acceptable to make the network API completely implemntation dependent?
A practical man is a man who practices the errors of his forefathers. --Benjamin Disraeli ==== Greg Woodhouse [EMAIL PROTECTED] [EMAIL PROTECTED] ------------------------------------------------------- SF.Net email is sponsored by: Tell us your software development plans! Take this survey and enter to win a one-year sub to SourceForge.net Plus IDC's 2005 look-ahead and a copy of this survey Click here to start! http://www.idcswdc.com/cgi-bin/survey?id=105hix _______________________________________________ Hardhats-members mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/hardhats-members
