You are much better informed about WebObjects (WO) than I am. Thanks for enlightening me.
I didn't work with the group long, but I did move to WebObjects from ATG in 1997, although I did move on to try "that startup thing" in 1998. (Yes, I'm still trying...)
Which products (or combination of products) compete with WO? I'm still not clear about the things
one can do with WO: create a Web site? Build the front end that users interact with? Manage the back
end (ie substitute for traditional database programs, such as SQL Server or Oracle)? All of the above?
Technically, WebObjects is a funny space. On some level, they compete with Kiva (Netscape bought) and NetDynamics (Sun bought). More than anything, they compete with home-grown solutions. I believe IBM and Oracle also have offerings that are in the space.
WebObjects is based on the the Enterprise Objects Framework, which is made up of three layers:
Access Layer: Allows you to access a wide variety of databases abstractly, but with customized speed and features for each. This means you can deploy the same solution over Oracle, SQL Server, Sybase, Informix, ODBC, etc. (Great for development!)
Control Layer: Allows you to write object-oriented business logic over the access layer.
Interface Layer: Allows you to create flexible interfaces in 100% pure java or in YellowBox that you can deploy locally or remotely.
WebObjects adds classes to help manage an HTML "interface layer", including advanced session management features.
WebObjects feature a GUI WWW builder tool, but it is more often used for development features than for design and deployment of a website.
Think of it this way:
WebObjects is a software platform, that allows you to develop applications that can be deployed over a variety of databases and with a variety of front ends, all over the same application logic. To WebObjects, HTML is just another way for a user to interact with a distributed application.
In some ways, it is similar to PowerPlant, the Mac OS framework by Metrowerks that allows you to easily develop Mac applications that run over 68k and PPC Macs, Mac OS 7.1 to Mac OS 8.6, etc.
It is kind of weird to think of a website as an application, but the more programmatic websites get, the more need there is to integrate traditional software techniques for object oriented development, platform flexibility, and versioning. App servers are definitely still a nascent market, although Enterprise Java Beans are building momentum.
IMO, the crux of the matter is that a product's success these days depends on its accessibility -- price,
ease of use, and availability. Even the great (pricey) Oracle and SAP are starting to aim at smaller
businesses (still expensive, but from where they're coming, it's a big mark down). Supercomputers play
to an extremely small niche market and the companies that make them (are there any left?) are not in
the best of health. As an Apple fan, I hope WO doesn't go the way of the supercomputer.
Everything boils down to perception. Classic Mac vs Windows problem. WO can be the best of the
best, but if it's seen as expensive, it's not going to be very visible in the market. Apple should take a
lesson from MS here: Make it affordable enough to get the foot through the door. Let people try it out.
Those who want more (because the product is so insanely great) can add on, one bell or whistle at a
time; those who don't think the product is for them won't feel too bad about throwing away $299,
especially if it's a company. At least they'll have tried it and know about it. However, if it's going to cost
$5K just to take a test drive, the majority of folks are going to look somewhere else. Ironically, they'll
end up spending $5K anyway, but they'll have done it one piece at time, assembling all the components
they need. But unfortunately for Apple, they won't have spent the money on WO.
A development license for WebObjects is less than $5K, I think it is around a $1K. A deployment license for a server is $5K, while clients are $299 a seat. This is not high by any means for the market.
The fact that Apple is shipping a development version of WebObjects with every copy of Mac OS X Server ($499) makes it pretty accessible at this point.
I think the bigger problem Apple faces is that the app server market is still fledging, and enterprise is not looking to Apple for ground-breaking solutions in its market. Competing with Sun, IBM, and Oracle to just get recognized is going to be hard enough.
That being said, some pretty big sites, both internal and external are using WebObjects.
I do think, however, that WebObjects would do well on the IPO market, given its revenue, growth, and market potential. |
