Note: This file has moved to notablog.

> Anyone here had the opportunity to compare/contrast an ASP
> (application service) package solution such as [name deleted]
> [...]
> to building off of a component-base such as WebLogic with the
> commerce package?

> If so, what is the breakpoint at which it would make sense to go
> with something such as one of these packages, over a more custom
> solution built upon a WebLogic core? Is it all about speed to
> market, or is there more to it than that? I frankly don't see the
> advantage yet...but it seems that a lot of people do, lately. What
> am I missing?

I have deleted specific "e-business platform" names from the above because a) the ones I *am* familiar are owned by companies that are prone to suing and/or pressuring anybody who says anything bad about their technology and b) I'm not familiar with the others and it would be unfair to directly name and condemn them.

In my general experience, there is no value in going with any of these "e-business platform" solutions, except in very rare cases. Those cases are restricted to situations in which the vendor has EXACTLY the system you want, ALREADY up and running, and you'll ALWAYS be happy to use what they have. For the most part, the rest of it is pure snake oil.

A large chunk of the companies selling in that space are really selling extensive (and expensive) consulting services, built around selling you their half-assed "product" on the promise that it will do everything. Once the MBA making the decision has committed a huge chunk of money to it, he/she can't back out without looking colossally stupid. So they commit even more money for the consulting services.

An even larger chunk are selling various limited pieces of the puzzle, but you won't be able to tell just what the heck it is they're selling until you get them into a conference room and bring out the hot lights and rubber hoses.

The rest, from what I've seen, are selling a hodge-podge collection of applications that have been stuck together with bailing wire and chewing gum.

A lot of software developers have an overdeveloped skepticism about such products. That's why they/we tend to lean towards buying development tools and platforms instead. And when I say platforms, I mean real platforms, like operating systems, databases and, nowadays, application servers. Not the currently-fashionable trend of calling vertical niche applications "platforms".

What I've noticed over the years is that, as the industry itself gradually matures, the general trend is to abstract different pieces of the puzzle out into different layers. It's only when there's been enough time and market-driven erosion/evolution that those layers are truly separated, open and interoperable.

Mind you, I don't dispute that it is _theoretically_ possible for such products to actually be useful. But I have seen, time and again, that the overall process of adopting an "e-business platform" ends up being far slower and more costly than building your own. By overall process, I mean that far more time and energy is wasted trying to find candidates, get information about them, select one, negotiate a purchase (because they are *never* going to just give you a list price and let you buy it), figure out how it works, and then figure out how to work around its limitations, not to mention troubleshooting bugs that you don't have full documentation and source code to diagnose.

I suspect why you hear so much about it is because a lot of big businesses are getting into/onto the 'net these days. As a result, a lot more businessmen are getting into the act. Businessmen quite often do not understand (and hence fear) technology. However, they _do_ understand cutting deals. By taking this approach, they move the risk from an unknown area (technology) to a known area (doing business). This is not an altogether unsound strategy.

However, any strategy can be improperly implemented, and I think most often this strategy is poorly implemented at best. Usually this is because of underestimating the amount of discovery necessary; there's a lot of snake oil and a lot of bullshit artists out there, and this is a very, very popular field. Add to that unwillingness to actually listen to the technical staff (not to mention the simple fact that you can't remain completely ignorant, no matter what you do) and you start to see where the problems come from.