[rfc-i] hildeform

Peter Sylvester peter.sylvester at edelweb.fr
Wed Aug 1 03:22:40 PDT 2012


On 08/01/2012 11:39 AM, "Martin J. Dürst" wrote:
> On 2012/08/01 5:14, Joe Hildebrand (jhildebr) wrote:
>> On 7/31/12 10:49 AM, "Dave Crocker"<dhc at dcrocker.net>  wrote:
>>
>>
>>>> There's only one difference that is interesting, which is that current
>>>> best practice on the web is to not use<table>'s as a placement
>>>> mechanism,
>>>
>>> sorry, no.  my assertion of 'quite' was about appearance, not technology.
>>>
>>> The visuals of the xml2rfc document header are massively different from
>>> yours and, I believe, the differences are likely to have differences in
>>> some aspects of utility and possibly marketing friendliness. (But I
>>> said possibly.)
>>
>> I was unclear.  Other than the table-ness, everything else is easy to
>> recreate with the exact HTML we have in my proposal, modifying solely the
>> CSS.  Hence, the table-ness is the "interesting" part from an
>> implementation perspective.  I don't know how important that is to you.
>
> In the case of the HTML created by XML2RFC, the table used at the top is of course technically 
> totally obsolete as a way to put information into two columns.
>
> But not only that, but also design-wise, the table has lots of problems. There are clear dividing 
> lines e.g. between authors and their organization, which go together semantically. On the other 
> hand, the rows of the table visually pretend that there is some relationship e.g. between the 
> Intended Status of a certain document and one of its authors (or their affiliation), which is 
> nonsense.
table were often used in the last century as a pre-css method to do page layout
when web presence become something else than displaying a scientific article.

Today such a use is not really recommendable, especially when css is used anyway.
In our context <table> should only be used to encode the semantics of a "table".

The effect of not using table can best seen when css is turned off in joe's version.

I think that the result is not nice.

The head metadata should go after the first title IMO, and there could be some
kind of "identification section" which, when styled, gives the desired layout,
the metadata are best represented in a <dl>

This is of course a matter of discussion and evolution like for example between

http://www.w3.org/TR/2012/REC-xhtml-rdfa-20120607/

http://www.w3.org/TR/REC-html32-19970114

'branding' is nicely done with an <img> of the ietf logo.

as a strawman structure of the beginning of an rfc in html5:

<!doctype html>
<html>
<head> stuff, title, meta, </head>
<body>
<header>
<img towards ietf logo>
<h1>the title</h1>
<aside>
<h2>Identification</h2>
<dl>
<dt>Authors<dd>...
<dt>Date<dd>...
</dl>
</aside>
<section>
<h2>abstract</h2>
..;
</section>
<nav>
the toc   (no </li>)
</nav>
</header>
<section>
<h2>Introduction</h2>
...
</section>

I leave the appendix, references and footer as an excercise :-)











More information about the rfc-interest mailing list