[Biojava-l] Re: Biojava-l digest, Vol 1 #39 - 5 msgs

[Thomas Down]

On Fri, Mar 10, 2000 at 10:34:06AM -0500, Peter wrote:
> 
> There are no books on the use of Java or CORBA that I am aware of. Not
> only that, in the long term, I would advise against the use of CORBA. I
> expect that Jini will replace CORBA.

That surprises me.  Jini is based on the RMI proceedure call
system (I don't know if Sun have any plans for moving Jini
to anything other than RMI -- certainly the specs I've seen
haven't really been going in this direction).  RMI is great
for small/simple things, but is widely considered to be
unsuitable for `enterprise' deployment, primarily because it
requires the creation of a server socket for every remote
object.

Anything which is true for `enterprise' deployment is likely
to be doubly so when serving big biological databases.

A lot of Java developers are now looking favourably at IBM's
RMI implementation (bundled in JDK1.3) which allows the RMI
programming models to be used over the CORBA IIOP protocol.

Thomas.
-- 
When God gave out chins, I thought he said gins;
I asked for a double.
                      -- Anonymous