Lee (?) wrote:
The first thing I'd like to do is eliminate the non-bracketed external URL syntax. Specifically, I'd like to keep the single- bracket syntax [url text] for external URLs, and double brackets for internal page links and images, and eliminate entirely the interpretation of URLs in plain text. It would simplify the syntax, eliminate the problem of punctuation after URLs, it would make all the image links report correctly (ones that look external don't show up as links on the image's page), and eliminate a whole bug-prone pass in the interpreter. I's also eliminate the auto- numbering thing and make a bracketed simple URL without text just show the URL like an unbracketed one does now. Finally, I would not interpret external links to images as inline images, so /only/ the [[image:]] syntax would work for images (though you could still link to external images; they just won't be rendered on the page).
All of these seem OK as changes to the way new articles work. I only worry about backwards compatibility but, in this case, that's not really much of a problem, since a URL in running text, while less helpful if you can't click on it, does the job.
If numbered links are good for sources at the bottom (and I've rarely seen them used for this), then that's still possibile, using syntax like "[http://www.com/page.html 1]". Anybody that writes that out deliberately will probably remember to put the URL itself at the bottom under == Sources ==, so that will prevent the misuse (probably accidental) that I see now.
-- Toby Bartels toby+wikipedia-l@math.ucr.edu
wikipedia-l@lists.wikimedia.org