Re: V7.2 and threads (non-threads safe?)

In article <edward-0512951704420001_at_r-618macip18.hac.com>, edward_at_igate1.hac.com (Ed Bruce) writes:
|> In article <49vcda$fmb_at_inet-nntp-gw-1.us.oracle.com>, surman_at_oracle.com
|> (Scott Urman) wrote:
|>
|> > In article <49q92u$mcl_at_spruce.citicorp.com>, renato.ghica_at_citicorp.com
|> (Renato Ghica) writes:
|>
|> > Oracle 7.2 is NOT thread safe. The client libraries for Oracle 7.3 will be.
|> > Having said that, you can still write multi-threaded programs with Oracle 7.2.
|> > Make sure that each thread has a separate connection. In Pro*C, if you have
|> > more than one thread in the same source module, use the AT clause for each one
|> > with a different name to insure that each will have a different connection.
|> > You also should mutex all of your SQL statements to insure that a thread
 doesn't
|> > get swapped out during execution of SQL.
|> >
|> > In 7.3, you will be able to pass context information among threads (and
|> > therefore share a connection), and the mutexing restriction is less.
|>
|> Will the Pro*Ada for 7.3 support this for Ada tasks?

I would imagine so. I don't know for sure, however. The ada runtime library is different from c/cobol/fortran runtime library, so it is more code which would have to be modified. There are plans afoot to combine adalib and sqllib into one library, which would be thread safe by the time it is actually done. I don't know when (or if) that is actually going to happen, however.

|>
|> --
|> Ed Bruce
|> edward_at_igate1.hac.com
|> Key fingerprint = 62 8D FC 8F 27 2F 89 D0 8B 38 7E 34 33 74 C2 36
  Received on Wed Dec 06 1995 - 00:00:00 CET