Re: uri=referer

olivier Thereaux wrote:

> On 13-May-08, at 6:39 PM, Andreas Prilop wrote:
>> It is in general not a good idea to link with this address
>> using the "Valid CSS" icon.
>
> Not necessarily.

The "Valid CSS" icon and relatives are worse than useless, necessarily. 
When uses as a link to something like 
http://jigsaw.w3.org/css-validator/check/referer (as required by the W3C 
in their licence conditions, though I don't think they have any legal 
force, at least where I live) adds a further level of confusion and 
absurdity.

This is what I get when I open a _local_ page (e.g., a saved copy of a 
web page) that has the W3C-adorned icon and I click on the icon:

"Some Headers, mandatory for this resource, are missing.
  a.. Referer"
This is even worse than the Markup Validator's complaint. What is the 
simple user who just happened to click on the icon supposed to do now?

It is wise to use "W3C CSS Validator", but it is not wise to pollute 
one's pages with "Valid CSS" icons. Surely an author can find a better 
way to make his use of the utility simple without causing such 
disturbance to visitors. As we can see from the repeated questions about 
the uri=referer issue, even web authors who are educated enough to use 
this utility will often get confused.

> Of course there were changes from CSS1
> to CSS2 and then to CSS2.1, but the case of valid properties becoming
> deprecated and disappearing from one version to another is less likely
> to happen now that the process checks implementations before the
> Recommendation status.

CSS 1 and CSS 2 are W3C recommendations, though not recommended by the 
W3C; CSS 2.1 is labelled as not citeable except as work in progress; and 
CSS 3 is a collection of mostly dated drafts of sketches. So the 
situation is inherently unstable. We can, and we indeed have to, use the 
CSS 2.1 draft du jour as the best surrogate for an approximation to a 
wannabe CSS "standard", perhaps with some consideration of proposed and 
partly implemented CSS 3 features, but let's not lure ourselves into 
thinking that CSS 2.1 is stable (not to mention CSS 3).

Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

Received on Wednesday, 14 May 2008 11:57:01 UTC