Duane Clark wrote:
>
> Mark Anderson wrote:
>
> >
> > End users should be given the option of what to do in this scenario.
> > Not to interpret the link as a relative one, when logic tells me that
> > this would be wrong (despite the 95-99% figure thrown about here, I've
> > yet to see a *single* link in my years of Web surfing that uses this
> > broken syntax to indicate a relative link).
>
> A bug report has been filed about this "bug" on bugzilla 14
> different times, by 14 different people. And I will bet that the
> archives of these newsgroups contain more reports, such as mine, that
> never made it into bugzilla.
>
> Likely the reason you have never seen such link in the past
> is because the browser you were using handled it in the way the web page
> creator expected. And that is because the web page creator used one of
> those browsers to test his pages.
No. I didn't say that. I said I haven't seen one. That means zero. I
do actually look where my links are being pointed to, most of the time.
Of course, I also tend to frequent pages where people actually know how
to use HTML correctly.
In any case, the way the Web page creator intended is what is absolutely
unclear. I contest your 95-99%, because I happen to think that the way
Mozilla does it now is more correct than assuming it to be a relative
URL. The only real solutions are to offer the user a choice or to just
simply not follow the link and display an error message. Implicitly
assuming it's a relative URL not only would be wrong in many cases
(assuming these links are actually as widespread as you claim, and I
doubt that), but it also violates an RFC.
I still wish HTML validation was required to publish a page. This kind
of thing would never have happened if browsers hadn't started letting
shoddy HTML through the cracks. Breaking the miniscule amount of pages
this will actually affect is worth it to promote people writing pages
correctly, IMNSHO.
