Re: DevCo: A look from outside the window.

Like most models for application development this has strengths and
weaknesses depending on the kind of system you're building.

Indeed, from the developer's perspective. But the SaaS story is much larger
than just what us developers think about it. It really unlocks the next
level of computing potential; during the late 80's, users realized how novel
it was to have a computer sitting in front of them at their desks. Via word
processors, spreadsheets, accounting software, they could simplify and
automate a dozen tasks that used to be tedious. Computers and software made
their jobs easier.

But now everybody takes legacy computing power for granted. Even using the
latest desktop software and the latest OS, if we use computers the same way,
this is still an example of legacy computing power. Companies have
integrated the benefits and costs associated with legacy computing power
into the cost of doing business now, and believe it or not they no longer
can show the cost benefit of using computers and traditional software.

Companies want more and that's what SaaS provides them:

1) More cost effective use of computing resources, software and hardware
included
2) Instant access to more functionality via services
3) Maybe juciest of all, true ability for end-users to provide resources to
each other

Salesforce.com is a great SaaS example, but we shouldn't get stuck on this
example as the end of the story. SaaS is a platform in and of itself
independent of the OS; the web browser is the engine for the presentation
layer. So now theres the potential for end-users to extend their desktops
onto the internet. There's no boundary between the way we use local and
internet computing resources, as there has been in the past.

This is a whole new level of collaboration that we haven't seen before:
end-users can offer services to other end-users, companies can share
services. And this has a huge effect with regard to devaluing the OS. If
Linux and Vista can both provide the end-user with the same service at
generally the same efficiency and with the same security, and if some
platform agnostic development tool can help make this happen with equal
ease, then the OS is just a commodity choice. It's like deciding whether or
not to use a Gateway box or a Dell box as a server. Believe me MSFT is very
aware of this danger to their market.

This is the opportunity for Delphi that I'm talking about.

Where software-as-a-service hasn't been so effective is for local client
applications such as word processing or anything else that demands local
processing or a very responsive UI.

I think Google will probably provide some good counter examples to this.
Besides, we don't care about the word processor market, except in those
instances where this functionality needs to be embedded into our
applications. And these applications are still never designed to compete
against MS Word. All this is moot anyway because a good word processor is
perfectly free these days.

Some of this can be done using Terminal
Services or similar, but local processing power is so cheap compared to
the
cost of network bandwith that it's generally more economic to install and
maintain a full PC.

Hard for me to believe that this is the case now, and I know this won't be
true in the future. Massive computing power is cheap, and even cheaper if
you can deploy less of it. Couple that with with fact that for minimal cost
we will eventually be "awash in bandwidth" as Metcalf put it in one of his
"From the Ether" articles back in 1999, and there's simply no point in every
end-user being his own island.

The bottom line is, SaaS as its own platform makes too much sense cost-wise
and collaboration wise to be ignored. Google knows it, MSFT knows it and is
trying to adapt and control it before it's too late. The challenge for Devco
and Delphi is, from Windows to Vista to Windows Live, will it continue
sucking whatever hind-teat MSFT provides, or will it venture out on its own,
at least a little bit?

James


.