> - one is that Apple is forcing me into the path of using open
> source when the existing tools are more than adequate to give me a
> serious productivity advantage over my competitors who use
> different technologies;
>
> - the other is that I believe that there is a philosophical bias in
> the whole WOLips approach to WebObjects development that reflects
> the interests of those who developed it: the approach to almost
> every development task seems to be to write Java code or edit text
> files (although Entity Modeler is a brilliant counter-example to
> this). This might be appropriate for programmers, who love to write
> code, but one of the main competitive advantages of WO for me is
> that it allows me, as the project manager, to rapidly whip up a
> demonstration system for the customer without writing any code,
> using visually-oriented tools (OK, now I am showing my bias-- like
> half the planet I am visually oriented rather than textual). Then
> when the customer buys in to the project, I (or my staff) can then
> use the very same development tools and the very same prototype to
> develop a world-class enterprise application. I used to joke with
> my staff that we were trying to achieve zero lines of code per
> day-- meaning that if they were writing code instead of using the
> high-level tools, they were probably going in the wrong direction.
> It seems to me that the path of least resistance with WOLips is to
> dive into code and the lower level details. Although Entity Modeler
> is a great tool and appears to do everything I used to to in
> EOModeler, there are no high-level visually-oriented tools in
> WOLips. And judging by the responses to recent discussions about a
> WO Builder replacement, there are many people who feel the same need.
The WO team didn't choose to drop the tools because they necessarily
wanted to, they chose to drop the tools to save the framework. The
WO apps were incredibly old code bases that have been on life support
for several years. If Leopard gets up to release and there's a
crashing bug in EOModeler, it's not Leopard that stops shipping, it's
WO. This was a pragmatic decision to attempt to keep a deserving
technology shipping that doesn't generate any direct revenue for the
company. The alternatives could have been much worse, so you should
focus on the bright side of this.
I also don't believe there's a "philosophical bias" in the WOLips
development team towards text-based tools. People write these
responses like I'm not aware of the pros and cons of visual vs non-
visual development, and apparently assume that because I didn't build
a graphical WOB right off the bat that I'm somehow anti-visual-
development and missing the boat on the concept. I assure you I
think about the concept quite a bit, and I'm not ignorant of the
benefits. Specifically, it's not that I don't believe in the concept
(I *love* IB), rather it is my opinion that that WOB does a lousy job
at solving the problem. It was a great tool for building modern apps
in 1999, but app demands have changed and WOB has most certainly not
kept up. There's also the reality of the situation which is that
developing text-based tools is a heck of a lot easier and faster-to-
do than developing visual tools (particularly cross-platform ones --
say goodbye to the productivity gain of Cocoa on that tool). So at
the moment, you're getting text-based dev tools. They're, IMNSHO,
some pretty kick-ass text-based development tools for WO. Don't
dismiss them because they don't let you drag a line between a square
and a word at this stage of their life.
ms
This archive was generated by hypermail 2.0.0 : Fri Jul 20 2007 - 18:56:03 EDT