Subject: Re: Suggested change to README.html generation: tables for binaries
To: T. M. Pederson <>
From: Julian Assange <>
List: tech-pkg
Date: 11/10/1999 06:21:45
> ><dl>
> ><dt>alpha
> ><dd>
> ><dl>
> ><dt>1.4
> ><dd>
> ><ul>
> ><li>mpg123-O.59q
> ><li>mpg123-O.59o
> ><li>mpg123-O.59q
> ></ul>
> ></dl>
> ></dl>
> I use lynx heavily too, especially when checking the Packages collection,
> but this is the Wrong Way to do things.  We're not looking at a Definition
> List here, so <dl> should not be used.  Even if it generates the kind of
> layout you want on most browsers, it's best not to lie about what kind
> of data is there.

It is no lie. <dt> is the meta-concept, and <dd> is the sub-conecpt.


Stefan Kahrs in [Kah96] discusses the
   notion of completeness--programs which never go wrong can be
   type-checked--which complements Milner's notion of
   soundness--type-checked programs never go wrong [Mil78].