Let me join this discussing. I've been lurking now for a while
and didn't have the time to give back my RT or opinion on this...

First of all: can someone please enlighten me - Sorry, I don't
see a conceptual difference between a FSM and a Turing machine.
AFAIK they are pretty much the same!?

I always wanted to avoid page-cenctric logic (as Ovidiu postulates)
in my webapps - but real life has shown that forcing this is not
really useful at all. It often leads to an over-abstraction of the
problem that doesn't help a thing.

Consider a form where you collect data over a couple of pages
(I call this a multi-page form) If you have only an abstract
view of your logic you will not be able to derive a validation/
redirection behaviour. You simply have no idea which page to go
back to if validation fails - the information is missing in the abstract
description of the application flow.

I also doubt that building a webapp as FSM or Turing machine is
really that hard. We already using this approach and we are quite
happy about the gain of possibilities. I wrote a Flowmap
action to integrate smoothly with Cocoon. Not having one
big Flowmap but a couple of small ones makes it quite easy
to maintain.

Also take into account: even if the flowmap gets bigger
and becomes a PITA to maintain via emacs (or any of your
favorite editors;) there is always an option to write
a visual tool!! Don't lets us limit ourself because it will
be hard to maintain from commandline. I cannot think of
a much easier, faster and natural development of having
a visual tool describing a Turing machine! This also reminds
me of a computer science technique - I guess it was "data
flow diagrams".

Although I have to admit I haven't had the time yet
to look into the resources Ovidiu sent I have quite
a bad feeling about continuations.
Isn't a flow with a lot different choices/transitions
also a PITA with continuations?!

Well, I get back to continuations when I did my homework;)

> So, if in your context FSM programming means "page-centric programming",
> I totally agree with both of you that we should "get control back" when
> building webapps.

What do you mean by "get control back", Stefano?
--
Torsten


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, email: [EMAIL PROTECTED]

Reply via email to