User talk:SF

From A KoL Wiki

Did you know typing naughty sorceress quest into the search box and hitting "go" would have taken you to Naughty Sorceress Quest anyways? What's with the redirect?--Dehstil (t|c) 19:52, 2 August 2006 (CDT)

That requires two pageloads: loading the wiki, then searching, whereas the redirect requires one pageload. It also requires more system resources to search rather than to redirect. In general, pages that differ by capitalization bug me anyway. --SF 11:53, 3 August 2006 (CDT)

Actually, here's how it works:

With a redirect
  1. Hit "go" with "naughty sorceress quest" sending http request
  2. Server parses request does sql query in cached table (of words longer than three letters) for "naughty sorceress quest"
  3. Finds redirect
  4. Finds redirect's target
  5. Return page to client
Without
  1. Hit "go" with "naughty sorceress quest" sending http request
  2. Server parses request, does sql query in cached table (of words longer than three letters) for "naughty sorceress quest"
  3. Finds page
  4. Return page to client

Also, the clutter that is a useless redirect consumes more system resources for other things too. At Proposed Standards we decided capitalization redirects are infeasible and useless in most cases (except in very special cases because of the longer than three letters thing), especially because of the number of possible permutations of capitalization per page. *rant* Sorry about that; it's just that it keeps coming up.--Dehstil (t|c) 15:52, 3 August 2006 (CDT)


Ah, I must apologize then, as I did not see it in the Established Standards page. Anyway, finding a redirect and then looking up that page is a *linear operation*, it consumes virtually no resources - the redirect is hashed with the query term. And without the redirect there is fully a second pageload (*not* a linear operation), since the first time you try to load the page you get an empty article.

To most people this might not make a difference: they load the wiki, then look for the page they want. I access the wiki constantly with scripts, going directly to the page I want as the first pageload. Others do it the way I do as well, and the redirect makes a significant difference in this circumstance (on both user- and server-sides).

And - why not change the code to search automatically for capitalization differences in the event that no page is found? Though for that we'd have to ask Jin or Thrae.

So, in summary: in a case like mine the redirect replaces a polynomial operation with a linear one, which is good; two pageloads are more annoying for the user, and harder on the wiki resources. On the other hand, I'll defer to you to call the shots however you decide.--SF 12:06, 4 August 2006 (CDT)

  • Continue discussion here posting reasons why you like these redirects.--Dehstil (t|c) 16:15, 4 August 2006 (CDT)