On Apr 18, 2007, at 4:49 PM, Norman Palardy wrote: > > On 18-Apr-07, at 2:29 PM, Charles Yeomans wrote: > >> >> On Apr 18, 2007, at 3:33 PM, Guyren Howe wrote: >> >>> On Apr 18, 2007, at 2:09 PM, Charles Yeomans wrote: >>> >>>> Of course, the best solution is not to use use sockets >>>> synchronously. >>> >>> I don't agree with that as a blanket statement. If this is a quick >>> and dirty application, or the logic is such that splitting it up >>> between the events on the socket and other places is a pita to get >>> correct, I say do whatever's easiest. Make the computer work harder, >>> if it means you work easier. >> >> >> My experience is that not using sockets synchronously is easiest in >> the long run. > > Some tasks (like database interactions) are predicated on normally > synchronous activity. > > Calling SQLSelect and at some point in the future getting results > makes writing a client server app much harder than it needs to be. > > I'd be willing to bet that anyone using a database depends on > SQLSelect behaving synchronously.
Certainly my experience is not all-encompassing. But if you're implementing client-server communication for a database API, then I'd say that's not quick-n-dirty development. Charles Yeomans _______________________________________________ Unsubscribe or switch delivery mode: <http://www.realsoftware.com/support/listmanager/> Search the archives: <http://support.realsoftware.com/listarchives/lists.html>
