Re: Positioning HTML Elements with Cascading Style Sheets

On Sat, 1 Feb 1997, Todd Fahrner wrote:

> This is of course crucial. But would such a tool be capable of generating
> HTML whose presentation adapts gracefully to varying browser window
> resolutions and aspect ratios? If not, then it's WYSIWIM
> ("what-you-see-is-what-I-might") and therefore not truly portable. Absolute
> positioning is possible today with GIFs and tables (cf. NetObjects Fusion),
> but this printlike capability is a mixed blessing at best. PDF is at least
> resolution-independent.

I think it's possible to at least design a WYSIWIM editor that will
show you exactly what you will see in one particular browser.

Whether it would show any difference between <em> and <i> is an
interesting question though.  There would probably be no noticeable
difference, and to the clueless (ie most of the world), <i> would
be just as good as <em>.

Most people don't use more than one or two browsers.  I usually
use lynx; and occasionally Netscape.  Even after E-scape is
working (the browser I'm writing), I'll probably still use lynx
at least as much as E-scape.

Eric Naggum was stating on gnu.misc.disc that he things we should
teach people how _HARD_ programming is, instead of how _EASY_ it
is.  (Eric is one of the key contributors to GNU emacs as I understand
it.)  I think maybe the same applies to good web page design--if
you show people what it physically looks like, they might not
understand the differences for other browsers.

How do you solve this problem?  I don't know.  If you show the tags
in the text, then you don't have WYSIWYG.

You might be able to color the text in editing mode; so <em> could
be one color and <i> another; but that is only helpful for the people
who understand that they should care.  And some of them might
not bother with that feature.

For the people who don't choose to notice that there's a difference
between <em> and <i>, I'm not convinced that there is any hope
at all.  There are those who think we have all day to spend downloading
callow plugins in order to browser their pages.

I personally haven't been convinced that validation is a Good Thing.
I can see that it might be useful...how about another daemon
spamming me when I write bad html...

(Some GNU hackers love getting all sorts of automated mail--diffs
of changed files, reminders to release RCS locks, reminders to
start backups, which I get in Hawaii even though there's now way
I could load tapes in Cambridge MA, etc)

But at least I'm convinced that writing correct HTML is a Good Thing.
I'd generally use <em> instead of <i>.

So the real question is not how to make nice tools but how to change
the thinking of the people who write the pages.

I happen to think that the ``best viewed in any browser'' campaign
is a Good Thing.

(That does not imply that I will implement only accepted features in E-scape.
E-scape will support some extensions that Netscape
and Internet Explorer probably never will, like support for Guile,
but I still like lynx.)

So why is it that people don't bother to care about supporting all
browsers?  Is it that they don't bother learning what they're doing?
(Possibly.  My mother had to be taught how to click on a window
to select it in Win95 within the last few months.  And we consider
that user friendly.  I have a friend who insists that UN*X is
user friendly; it's just selective about its friends.)
Or do people think that allowing part of their audience to lose
is OK, as long as the people with the high end browsers are
attracted to their site?

I think that different web authors have different values.  The
people who created www.gnu.ai.mit.edu and www.lpf.org are more
interested in presenting information, while other sites perfer
to have the most attractive eye candy in the hope that it will
attract people.

(Notice the alignment of the right edge of the previous paragraph ;-)

Personally, I hate the advertising that seems to have conquered large
web sites.  Does anyone know of a good search engine that doesn't
have any advertisements?  I'd love to use it.

To sumarize: we can create tools that will acurately show what something
will physically look like, but raw HTML is probably the only way
to reasonably show semantic structure.  If we want good HTML written,
we're going to have to convince people that it's a Good Thing, and
that will be more work that creating the right software.

<nemo@koa.iolani.honolulu.hi.us>                    <devnull@gnu.ai.mit.edu>
"...For I have not come to call the righteous, but sinners." -- Matthew 9:13