Imre,
Funny you should respond since I have a copy of EJIPT (among others) that I
use
for testing. I saw your use of RMIClassLoader in an example provided with
EJIPT -
nice.
There's a problem though with universal support among EJB servers for this.
For example, WebLogic requires a username/password pair for HTTP access even
if the objects served up by its "T3" JNDI service are unsecured. Anyway, do
all EJB
servers respond to HTTP?
I'm toying with the idea of just storing the code in the directory. I'm
assuming that
there will be in practice only a handful of factories that need to be
stored. I foresee
some problems with this approach but I don't see a better alternative (right
now
at least). As the stranger said, "I am an egg." Any other suggestions?
Thanks,
Bob
Bob Pecor
Vanguard Cellular Systems, Inc.
336.286.1742 (Voice) 336.286.1881 (Fax)
[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>
-----Original Message-----
From: Imre Kifor [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, March 17, 1999 9:46 AM
To: [EMAIL PROTECTED]
Subject: Re: Bean access to server info
Bob,
> I'm finding though that there is an issue with the JNDI
>InitialContextFactory code: how to get it? The factory
code
>is needed by the client to create the initial context but
it may
>not be resident on the client.
>
>Any ideas?
>
We are using RMIClassLoaders to explicitly download the
factory class from
the server if it is not already available on the client.
This also allows us
to require only one piece of contact info for the server
(which is the
string http URL of our class server). You have to use JNDI
1.1.2 (which is
1.1.1 compiled for Java 2) to work around the security
policy issue while
instantiating the factory class.
Imre Kifor
Valto Systems
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
