Re: March 1998 article comparing NT/Unix/Linux as web servers

Stephen D. Williams (sdw@lig.net)
Wed, 11 Feb 1998 09:57:03 -0500 (EST)


> On Tue, 10 Feb 1998, Stephen D. Williams wrote:
> > Note that until Java made it necessary (IIOB), Corba also had a fatal
> > flaw: Corba also didn't propose a wire format for data. That meant
> > you had to both buy into Corba AND a particular Corba companies
> > proprietary implementation!
>
> IIOP has been around since before Java. Even so, some vendors still tend
> to use their proprietary wire formats.

In several (earlier) books I have on Corba and lots of product
literature, I had seen no reference to IIOP. I'm glad to hear it was
there, but the vendors involved must have buried it enough that I
couldn't find it.

CORBA has NOT become as popular as it should be for one very simple
reason: there were no usable public implementations for us to play
with.

That, and the former dearth systems using a standard wire format, have
held it back. Now DCOM has a wide following...

Sybase is an example where they are holding onto the wire format to
have control over clients to their database. I can tell you that
their clients are not as good as they could be (CT_Lib took 80% of the
CPU in one large server that I wrote). I happen to have the spec for
the TDS protocol, but normally it's not public. Of course it's
interesting that MS's SQL Server is still using TDS, and was and/or is
based on an earlier Sybase engine. MS of course doesn't usually
benefit from publishing protocol specs.

Oracle on the other hand has had a published wire format for quite a while.

sdw

> -mike

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu