[
Lists Home |
Date Index |
Thread Index
]
bry@itnisk.com (bryan) writes:
>>What is more questionable in the assertion made elsewhere that
>>hypertext (eg browser based web pages) is the
>>only realistic client for the web. It condemms complex
>>operations to novice mode. Because any compute
>>process is linearizable doesn't mean it is a good idea
>>for any given case.
>
>Certainly doesn't mean it's a bad idea for every case either. Anyway
>the linearity of a solution may be bad or good depending on how many
>other solutions the architecture supports and on how that architecture
>allows the solution to interact with other solutions in its space.
>From my perspective, the browser is a classic 80/20 (or better). It
offers 20% of the capabilities of richer hypertext systems and GUIs
generally, maybe even less. It turns out that there's an incredible
amount of stuff (probably more than 80%) you can do within that tiny
capability subset.
The Web did force a lot of work onto the back end, where interaction
with other solutions was a lot easier. I have a hard time seeing that
as a problem for most applications.
On Len's question, I don't see hypertext as "the only realistic client
for the Web". For certain, it's the cheapest client out there, and I
can't say I see much benefit to building other systems except in a
relatively small number of cases. From my perspective, XForms takes
care of about 90% of my remaining Web-interactivity problems - though
I'm still concerned about the odds on its widespread implementation.
If you need something else, though - build it. For all the effort the
Curl and Flash folks have put into Rich Internet Apps, though, I don't
see much need for a generic toolkit with powers beyond the existing
Web+XForms. Custom clients, though - sure. Sounds good.
--
Simon St.Laurent
Ring around the content, a pocket full of brackets
Errors, errors, all fall down!
http://simonstl.com -- http://monasticxml.org
|
