Re: Please Tables in HTML+

"William M. Perry" <wmperry@strawberry.ucs.indiana.edu>
To: marca@ncsa.uiuc.edu (Marc Andreessen)
Cc: Jim Davis <davis@dri.cornell.edu>, www-talk@nxoc01.cern.ch
Subject: Re: Please Tables in HTML+ 
In-reply-to: Your message of "Tue, 02 Nov 1993 10:51:02 PST."
             <9311021851.AA04792@wintermute.ncsa.uiuc.edu> 
Date: Tue, 02 Nov 1993 13:55:57 -0500
Message-id: <15153.752266557@strawberry.ucs.indiana.edu>
From: "William M. Perry" <wmperry@strawberry.ucs.indiana.edu>

In response to your message dated: Tue, 02 Nov 1993 10:51:02 PST
>Terry Allen writes:
>> It's simple.  My docs have lots of tables, and tables are common
>> in other people's docs.  It's a common problem, not to be evaded
>> on the argument that WWW can't do everything.  
>
>Terry, there are *lots* of common problems.  The resources required to
>solve them all simply aren't present.  Neither Jim nor I are trying to
>torpedo your efforts by suggesting that tables aren't important, but
>at the same time I think that "tables are common" isn't a sufficient
>standalone argument to justify the complexity they introduce.

  Hear hear. :) I think tables would be difficult to parse correctly.
Especially when you have links in the <TD> elements.  When do you fill
the paragraphs within them?  I think using a <PRE> segment would work
just as well in 99 out of 100 cases.  And in that 100th case, HTML
isn't going to be the best way to present the document anyway - it
should use postscript or PDF or whatever the latest and greatest full
blown presentation language is at the time.  If we simply must have
tables, how about specifying them a little more at the start?
Something like <TABLE COLS="5" WIDTH="15">, with width being optional,
and defaulting to (Screen Width / COLS)?  And the vertical argument to
TH is going to be a bear.  (Where one headers takes up two vertical
slots)

  When I grabbed the new HTML+ specification and saw all the nice
things in there, I thought 'Hey, this will be great to write it', then
I thought 'God this is going to be a pain to implement'.  Especially
that bit about Math Presentation.  I don't think it would be a good
idea to have the executables bloated by the TeX math code.  People who
want to do math-related documents will be more comfortable with TeX,
and limited by HTML+.  Either an IMG tag to a gif image or a hyperlink
to the dvi file would be better.

   HTML+ is getting too far away from `a light weight delivery format
for browsing and querying [...] HTML+ embodies a pageless model making
it suitable for efficient rendering on a wide range of display types,
for example, VT100 terminals, X11 Workstations [...]`  It will be
almost impossible to do the math markup on a vt100, and seems like it
would be difficult even in Xwindows.  This is akin to the desire for
'sliders' and 'knobs' for forms support - yes, it would be nice, BUT
not everyone could use them.  There are DVI viewers for almost every
system out there, as well as postscript, and eventually PDF (if adobe
doesn't sue too much).

   I for one feel the Xmosaic binary is large enough as it is - I
don't think we need TeX-like math support in it to bloat it anymore.
(Marc - no offense - its Motif's fault :)

   Hope I didn't ramble too much...

   On to another subject from the HTML+ specification - what happened
to the <EM B> tag?  I thought this was better than the <RENDER> hints
that we can put in the documents.  But, if the author is going to have
to put the RENDER attribute in anyway, why wouldn't he just use the
attribute he is saying is equivalent to begin with?  Did something
just get cut & pasted away by accident, or was it excised on purpose?

   -Bill P.