Re: Beta Validator usability suggestion

Good context-sensitive help followup suggestion for errors by Lars.

If this is adopted (which would ideally probably require some flatfile
database of sorts), I'd like to see links to references that are given
to explain error messages split into two categories: normative and
informative. (This is the approach already taken by the IETF for
references in its standards-track RFCs).

Normative references are to the relevant standards, encouraging people
to go read them. These are to stable urls at standards sites (e.g. the
relevant parts of W3C recommendations), and the things that the
validator is dependent on. Normative references should be immutable.

Informative references provide context. These are the tutorials
provided by third-party sites; they're less likely to be stable, more
likely to be break as urls (as cutting-edge webdesigners redesign
their websites yet again, or have their hosting services go kaput on
them), and it's the reader, not the validator, that really comes to
depend on them.

That is the ideal. Given manpower, what I think we'll get is an
automatic 'Please see if your error is listed in _the Validator FAQ_'
link in the Validator error output.

L.


On Tue, 17 Jun 2003, Lars Holst wrote:

> Date: Tue, 17 Jun 2003 15:26:29 +0200
> From: Lars Holst <lholst@robotics.lu.se>
> To: www-validator@w3.org
> Subject: Re: Beta Validator usability suggestion
> Resent-Date: Tue, 17 Jun 2003 10:00:22 -0400 (EDT)
> Resent-From: www-validator@w3.org
>
>
>
> Jeffrey has already explained the manpower situation at W3C
> (http://www.zeldman.com/daily/0603b.shtml#w3needsyou). With this in mind, I
> think the W3C validator team is doing  a great job. As a happy amateur that
> hand codes all his xthml and css, I can say without a doubt that the
> validator services have benefited my work in several ways, most importantly
> by saving me precious debugging time.
>
> This is not to say there isn't room for improvements of the kind Jeffrey
> suggests. I too have stared in disbelief at some of the error messages
> thrown at me, and even given up on at least one occasion (the
> "org.xml.sax.SAXException: Please, fix your system identifier (URI) in the
> DOCTYPE rule" is one of my favorites :).
>
> I would therefore like to suggest the following "intermediary" approach,
> which shouldn't add too much to the current workload:
>
> Seeing as there is already a number of well-written and highly informative
> Websites dealing with the very issues that most often lead to validator
> errors, why not link to those sites when displaying the error messages?
>
> I'm assuming that it wouldn't be hard to map the validator's error
> categories to certain external URIs. So if it's a DOCTYPE issue, link to an
> article or tutorial explaining the proper use of the DOCTYPE (e.g.
> http://www.alistapart.com/stories/doctype/), if it's a basic XHTML error,
> link to a site explaining XTHML in simple language (e.g.
> http://www.nypl.org/styleguide/), and so on. Finding the best resources
> (with succinct and correct information) might take time, but I'm sure there
> are loads of people who'd be willing to share their bookmarks. I know I
> would. Plus there would be an added benefit of saving people the time to
> hunt down all the good sites.
>
> Overall, I'd like to see the W3C embrace to a greater extent all the
> fantastic work that is being done out there in regard to informing and
> educating developers on the use of standards instead of doing everything
> inhouse.
>
> Thanks,
> Lars

<http://www.ee.surrey.ac.uk/Personal/L.Wood/><L.Wood@ee.surrey.ac.uk>

Received on Tuesday, 17 June 2003 10:29:44 UTC