Re: HTML 3.2 and TABLE ALIGN attribute

Peter Flynn (pflynn@curia.ucc.ie)
04 Feb 1997 00:08:51 +0000 (GMT)


Date: 04 Feb 1997 00:08:51 +0000 (GMT)
From: Peter Flynn <pflynn@curia.ucc.ie>
Subject: Re: HTML 3.2 and TABLE ALIGN attribute
In-reply-to: <199701231043.MAA26484@msa.tte.vtt.fi> (message from Markku Savela
To: www-html@w3.org
Message-id: <199702040008.AAA19116@curia.ucc.ie>

   Anyways, how did the definition of the TABLE ALIGN attribute slip by
   everyone into 3.2 description (Jan 97 version).

   1) As defined (LEFT, RIGHT, CENTER) the defition is very much useless,
      because those effects can easily achieved by use of DIV (and even

Because you might want a right-aligned table in a left-aligned DIV.
Omitting ALIGN from TABLE doesn't make sense.

      with <P ALIGN ..> too, because, IMHO table should never have been
      defined as block element, you should be able to have it inline).

I'm having some problems understanding what an inline table is.
A table by its nature is a rectangular (usually) object which
interrupts the flow of discourse (or is located separately in a
figure). I don't see any way you can have a table inline: it doesn't
fit logic: it must be a block element.

I think what you mean is you want TABLE as part of the content model
for P, like TeX does with display math. Can you explain a little more
and give some examples of why this is good?

   2) table ALIGN attribute (and the whole table in general) is much more
      meaningfull, if it is interpreted the same way as the IMG.

   The NS treats ALIGN already this way in respect to LEFT/RIGHT (text
   flows around table).

Ah, you want text flow around objects other than images? Go ask your
browser manufacturer about this. If some of them had bothered to read
and implement some of the earlier HTMLs, you'd have this by now.

///Peter