cbp is right, but to answer your deeper question about how it works:
The link: function looks for all pages that link to the page you specify. This includes links from other websites (external) and links from other pages on the same site (internal).
The discussion could end there except that for some reason Google does not display all of a page's inbound links when doing that kind of search. Many speculate that they do this to "throw off"
SEO attempts; the reasoning being that most link: users are trying to gauge their
SEO attempts, attempts to game G's algo.
While not all links are shown, many believe that even the IBLs not displayed still count toward the page's
PR and get benefit from the "keywords-in-anchor-text-thing".
If you are looking for a better method of counting a page's IBLs, you can try Yahoo, where it appears they display most, if not all of the IBLs they are aware of. This is done similar to Google's method by searching Yahoo for "link:http://www.webproworld.com/", for example.
There are other ways to try and trick google into displaying more information about IBLs, I don't really bother with them, but if you are interested, do a site search here and you'll find a few.
I use Google's link search to do a quick check of a site's IBLs. If google displays 2 links for a site, it is safe to assume that that page doesn't have too many IBLs, on the other hand, if Google shows 100,000 links, it is safe to assume they aren't doing too bad. It can also tell you what kind of links they are, check a few and you can see if they are mostly internal links, if they look like link exchanges, if they are getting keyword rich anchor text links, what keywords they are targeting, if the site is part of a larger "network" of site's, etc.
So while I don't think G's link: search is worthless, it is important to to use it with an awareness of its quirks.