To all Lorenzo Fiorini wrote:
Today we have hbodbc, hbmysql, hbpgsql, hbfbird contrib libs and probably more in the future. Their pourpose is to connect Harbour to SQL databases but they use different ( class ) interfaces.
I would venture that many Harbour programmers use these SQL databases to route around the DOS share based problems of the DBF world (well certainly the free databases).
Perhaps the discussion could include some ideas for NETRDD mentioned in the past. I have used LetoDB and it seems quite good, but I find the coding a bit opaque (particularly the C) and it is a bit on the "fringe" of the community.
I guess the biggest obstacle has been removed by Przemek, which was multithreading which would be required for a server (particularly on Windows which lacks a reasonable fork call).
Sockets appear to me to be reliable in Harbour but I think Przemek has alluded to problems with them? I don't know what they are. I know ADS uses UDP, I'm not sure why.
Do we have some form of RPC? I thought so but I can't find anything. We need to consider "serialising" the data in the current record of a file, not strictly a return value in the RPC sense. I guess also we need actually a copy of the current workarea data in the server, as we would like to know about BOF(), EOF() etc. Clipper errors need also to be "serialised" and returned/avaialable when something fails.
We have the USER RDD to allow a "simple" interface on the client side. Any thoughts? Regards Alex _______________________________________________ Harbour mailing list Harbour@harbour-project.org http://lists.harbour-project.org/mailman/listinfo/harbour
