Much ado about P [Was: Revised Beginner's Guide to HTML...]

"Daniel W. Connolly" <>
Date: Wed, 20 Apr 1994 01:44:40 --100
Message-id: <>
Precedence: bulk
From: "Daniel W. Connolly" <>
To: Multiple recipients of list <>
Subject: Much ado about P [Was: Revised Beginner's Guide to HTML...]
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Content-Length: 2502
In message <199404192318.QAA01132@rock>, John Labovitz writes:
>Daniel W. Connolly said:
>> I agree that nothing in this primer should conflict with HTML
>> specifications. I suggested to Mr. Chang that he encourage the use
>> of <P> at the beginning rather than then end of paragraphs 
>That doesn't work so well in lynx (2.2).  An example:
>  <H1>header</H1>
>  <P>an elephant
>  <P>a zebra
>Lynx puts extra space between the header and the first paragraph.

Lynx is broken. ;-)

Actually... I didn't know this. I wish we could get all the implementors
in a room and do rock-scissors-paper to decide how this is going to work.
There are lots of arguments out there, but none of them is very convincing.

I've been in the "P is a container" camp for a while, but I'll admit
it's a little awkward to write docs that way.

It comes down to two basic issues, to me:

	(1) HTML is or is not defined strictly in terms of SGML?

I'm pretty sold on the idea that it is a real application of SGML; i.e.
no HTML+-style inferring <P> tags when the SGML standard says you can't
do that.

Then, the question is: what's the structure of an HTML doc? If
everybody's mostly happy with tag soup, then there's no problem with
having P as a separator. But if you want to be able to reason
effectively about things like "the 3rd paragraph in section 2," then
there's motivation for seeing P as a container.

I don't care... I just hope that I get to go to WWW94 and join
in the shouting match that settles it once and for all. :-)


p.s. I've resumed hacking on html-mode... I started with a little
ditty that attempts to put in all these missing <P> tags to bring
a doc up to date... it also sticks "<!DOCTYPE ..." at the top.

I also changed the "html-add-code" etc. to work on the region, like a
good little word processor should: you select some text with the mouse
and choose Emphasize from the menu, and it wraps the selection with
<em>...</em>. How many times to you go "hmmm.. I think I'm going to
write some emphasized text..." versus "ah... that span of text should
be marked as CODE." Headers are the same way... type in the header,
select it, then make it a header.

This way is _much_ easer for hand-converting text documents into HTML.
Anybody interested?

Daniel W. Connolly        "We believe in the interconnectedness of all things"
Software Engineer, Hal Software Systems, OLIAS project   (512) 834-9962 x5010