Images and formatting of lists, keywords in <a>

Kjetil Kjernsmo (kjetil.kjernsmo@astro.uio.no)
Mon, 28 Jul 1997 16:08:32 +0200 (MET DST)


Date: Mon, 28 Jul 1997 16:08:32 +0200 (MET DST)
From: Kjetil Kjernsmo <kjetil.kjernsmo@astro.uio.no>
To: www-html@w3.org
Subject: Images and formatting of lists, keywords in <a>
Message-ID: <Pine.OSF.3.96.970725230311.25341A-100000@procyon>

Dear all!

I am pretty new to this list, but have tried to search archives and
other documents for the stuff I am proposing here as instructed. It
isn't easy to search though, so I hope what I bring hasn't been
debated. Nor am I very experienced with programming. I've been writing
HTML for three years, and that is about all.

From the drafts, it seems the W3C are concerned with Big Things, I
just propose a few small things I think would make a big difference. 
Keywords: images with lists, formatting of lists, enumeration, 
indentation, keywords with links.

It has become quite common to use some graphical dots in unumbered
lists. My university has begun using it in the graphical profile
(which is a profile I dislike) at http://www.uio.no/
I haven't been using it, but there are instances where it would be
allright I think to have. Therefore, I would like to propose a new tag
for <ul>, src (or img, imgsrc or something), so for example a list with
green dots could be  <ul src="greendot.gif">.
Normal dots should be displayed before the dot is loaded, or by
non-graphical browsers. If the type-attribute is used, it would be the
alternate "text". 

We could, of course, take this further, and make <li> have a src
attribute too. Some might think this is taking it too far, but I have
a typical use for it. I am a norwegian, and host quite a few articles
in both norwegian and english. I usually indicate what language the
document that a link points to by a tiny flag somewhere. It would have
been very convienient if I could use <ul> and <li src="enflag.gif">
for this purpose. 

It should be possible to make this work with <ol> as well. Would it be
possible load images after a pattern, using wildcards? If you had a
set if files, say no1.gif, no2.gif, no3.gif etc. and you gave the 
<ol src="no*.gif">. The first <li> would display no1.gif, the second
no2.gif an so on. 

And now, to the format. If you have a long list of say, names (they are
normally not so long), it tends to take up a lot of space. It would
often be desireable to brake them up in coloumns. Today, it is
possible using tables to get the format
	1.	2.	3.
	4.	5.	6.
	7.	8.	9.
I feel it is a little awkward to do it this way, and I would like to
have	1.	4.	7.	
	2.	5.	8.	
	3.	6.	9.
as well. Actually, I think a numbered list could be a part of the
text, like 1. apples 2. oranges 3. other fruit. I think it would be
good to be able to use <li> in this situation as well. It is always
nice when you doesn't have to count. I propose yet a few attributes
to <ol>, format or something like that, the number of coloumns or
rows.
For example, the first list would be <ol format=horizontal cols=3>, 
the second <ol format=vertical rows=3> and the third <ol format=text>.


I started publishing transcripts of our meetings in the physics students
union 2.5 years ago, and tried to get the administration to do the same.
Now they are instructed to do so from "above", and I am standing here to
welcome them. Anyway, they argued two years ago that HTML wasn't good
because it couldn't be used to numerate cases like "EC-5-95" (Which means
Executive Committee case number 5, 1995"). I would like to see yet another 
attribute (or do something with type) to <ol>, where you can specify more
in detail how an item is numbered. <ul type="EC-#1#-95"> could for example
indicate the example above. And you should be able to enter start value,
if you don't want it to start at 1.

It is always nice not to keep the count, and I would like to see
LaTeX-style enumration. Guess some might argue it is against the
fundamentals of hypertext because you don't need numbers to reference
another section of the document. Any thoughts?


Like many others, I have been using background images with "something"
along the left margin, like a logo, or something fancy. And I have
been wanting to indent the text somewhat for most of the document. I
have been using <ul> for this purpose, but that is clearly not what it
has been ment for. MS Frontpage, I have seen, uses <blockquote> which
is no better, I think. I figured, it might be a good idea to introduce
something that has the its sole purpose to do this. I figure an
attribute to <div> might be a good idea. (And to body maybe.) 
<div indent=[number of pixels or a percentage]>. Just <div indent>
could cause the same indent as <ul> or <blockquote> normally does.


I was thinking about maybe adding keywords to hypertext
links. My main motivation for this is that I am about to begin making
a hierarchal directory of science and critical thinking links for
norwegian skeptics. And I am going to expand the site with a search
engine later, so I got to have keywords somewhere, and I thought, why
not store them in the html files itself? I thought I might store it as
a comment, but after reading the WD-htmllink-970328, I figured, it
might be convienient to have a keyword attribute to <a>
It may have about the same uses as the title attribute. I thought in
the future, smart browsers could use the keywords to load a page into
the cache while the user is still reading. Maybe that would increase
network traffic too much, I don't know. It could use the keywords
from the last pages read to guess which link is the next link the user
would click. The <link rel=next> tag would of course be a very good
place to look, but that is only obvious as long as the text is pretty
linear (isn't it what they call it?) 

A LANG-attribute to <a> would also be nice, I guess. I would have to
indicate somewhere what language the link points to has anyway, and it
would be nice to do it in the <a>.

I have seen, it is being worked with printing stuff, and separate
documents can be used for printing. I have been thinking for a while
about a <feet>-tag. I mean, you are supposed to have head, body _and_
feet... :-) I thought maybe, everything written between <feet> and
</feet> would be displayed, but not printed. Navigational buttons is
an example of stuff that could be put between <feet>-tags. But since
this tag could be used anywhere in the body, the name isn't such a
good idea... Maybe these are things that go into the stylesheet stuff? I
have admittedly not been looking at that at all.

Friendly Tiddely-pom,

Kjetil
-- 
Kjetil Kjernsmo
astronomy-student (B.Sc.)                         University of Oslo, Norway
mailto:kjetikj@astro.uio.no   WWW-homepage:http://www.astro.uio.no/~kjetikj/
President, Norwegian Association of Physics Students    Webmaster@skepsis.no