Re: Status and whats the staus?

["Steve Hall" <digitect mindspring com>]

From: "Jeff Waugh" <jdub perkypants org>
> <quote who="digitect mindspring com">
>
> > *** Bottom line: are we doing server-side detection? ***
>
> *boggle* Yes of course we *can* do it.
>
> > 1. If so, then multiple HTML/CSS combos, graphics, JS is easy, why
are we
> > arguing for template simplicity?
>
> It's just been assumed because people thought we were going to be
doing an
> inane dynamic-everything system.

Who thought this? Since the first NS4 CSS template I thought we had
agreed to:

0. Get page request
1. Determine browser
2. Serve HTML, style CSS (colors) and mouseover JS. (same for everybody)
3. Serve browser's layout CSS (placement and graphics) based on #1.
(browser specific)
4. Client requests graphics based on the CSS, server does nothing but
serve them.

In this scheme, the *only* thing beyond a static site is #3, where the
server determines which layout CSS gets returned (2-3 letters). I can't
imagine anything easier short of being 100% static.


> I'd just like to see more thought put into simplifying the current
design,
> in terms of browser-applicability, graphics, etc. The foot is a good
example
> because it's a pretty-only thing, it doesn't actually enhance
usability at
> all.

I'd like to see proof that we can get something up first, frankly. This
list has been going since November 2000 and we still have only
templates, graphics and drfickle's software map. Nothing else that I've
seen, outside of list and IRC logs.

The delivery won't depend on the templates. We can serve three different
scrap HTML pages to prove we can determine the browser, and can plug in
*any* template/CSS/JS thingy we want, so let's not continue to make that
an issue. Regardless of what templates we use, we *still* need:

A. A server
B. Server applet to parse the browser from the request, and do a few
if/thans to select the CSS for it.
C. An authoring method which translates a submitted document
(HTML/text/XML/whatever) to our storage format (XML).
D. Code to auto-create the navigational HTML based on page location
within the structure. (NOT dynamically, but when authored in #C above.)

I think we're contending #3/#B here. Rather than say we *can* do it, I'd
rather us agree to whether we *are* or *are not* and proceed. Can we
please make this decision and move forward? There's still a lot to do.


Steve Hall  [ digitect mindspring com ]