I sometimes enter a page in wiki, rather at random, perhaps on the recent changes page. Often the page is pretty empty, so I am curious as to who made it and who refers to it.
How do you find out what pages refer to this page?
Answer: Put your pointer over the title of the page. Note the URL notice ends with ?...search=<page title>?. Go ahead and click on it. Voila! You get a list of the pages referring to the one you were reading! Now you can go read those, and eventually work your way up into a set of conversations.
Is this broke? I only see ?...search now. -- DavidGreen
It probably uses the HTTP referrer header so that it knows which page you want without it being mentioned in the URL.
The search is done on a copy of this database that is updated only once a week.
Though true, that isn't what I think DavidGreen
was asking. However, it does lead to another problem in that you may not find all hits. In particular, for new pages you may not even get the page which linked to it.
Just a little problem, the search is quite slow...
It needn't be. CvWiki has each page store and maintain its own backlinks, so clicking on the title to see them is just as fast as clicking on any other link. The price paid is that saving a page is a little slower. Saving a new page is slower still because it requires a search of the whole PageBase?. In the ~500 page PageBases? I've used in my work that's still not bad; in a PageBase? as large as C2's, it almost certainly isn't acceptable.
The way to correct the performance is obviously to keep a little LinkText database instead of backlinks in each page file; other matters have intervened but one day soon I'll get around to doing this, and maybe try to use PageArchive too. I'd also remove Ward's non-freeware parts of the code and release the whole thing on CPAN at the same time. Or anyone should feel free to preempt me.
I'm not convinced that clicking the page title works. I keep getting "no search term" errors. However, I do use Opera v4, so perhaps Wiki has an incompatibility problem.
This works with Mozilla and not Opera v4 (both on Linux), so I think it's Opera that has the incompatibility problem.
Opera notices that it has already loaded the URL with search inside and gets the page from the cache. I personally think this is the RightThing
to do, but I don't know much about http.
I used to HaveThisPattern too (Opera v5). I was getting the HTTP_REFERER message, but I went to File->Preferences->History and cache and set Check modified documents = always, and now it seems to work. Ugly workaround.