Found on The Serverside, BEA have published a Websphere Reality Check document, alongside Ten Questions You Should Ask IBM About Websphere (Disclaimer: I work for an IBM business partner. Websphere pays my bills)

Some of BEA's claims squarely hit the mark. Many of them are cringe-worthy marketing FUD. I'm particularly amused by the line: “IBM's WebSphere solution is built on an outdated middleware technology called CORBA”, that could only have been written by Marketing. None of this actually matters, though. From a marketing perspective, you only ever publish this sort of document in two circumstances. Either (a) you're losing, or (b) you're Larry Ellison.

I've been saying for a while that there's no room in the market for BEA. They're an historical artifact, fading away as we speak. They're being eaten from below by servers that are cheaper, more able or both (Orion, JBoss). Meanwhile in the big corporate playing-field, IBM, Sun and Oracle all have pre-existing relationships with the big customers, and adding an application server to that bundle seems far more natural than dealing with Yet Another Vendor.

BEA may indeed have a better product than Websphere. I don't know, I've never used it. But the reason this latest marketing effort completely fails to hit the mark is because it doesn't matter whether they do or not. It's not going to change anybody's minds, because once a company is in the realms of shelling out six-, seven- or eight-figure sums for hardware, software, programming, training and support, the differences between the Websphere and Weblogic application servers become a pretty small part of the deal.

If it weren't for the collateral issues, nobody would buy Websphere. But by the same token, nobody would buy Weblogic either. Orion and JBoss would rule the world.