[thelist] 404s and URL strings, oh my!

[Daniel J. Cody]

Not a bad question at all.. but not sure if i'm following you 100% but 
i'll try :)

mod_rewrite and similar features sit on the webserver level, whereas the 
URLs themselves are formulated at the application server level(chich 
gets passed on to the webserver for further processing and sending to 
your browser)

lets say you had a URL that came in as

/dir/dir2/index.html where 'dir' and 'dir2' were actually URL variables 
that you needed to have parsed. if i'm following you correctly, you're 
saying, "why not just send /index.html?var1=dir&var2=dir2 " to apache, 
let mod_rewrite take that and process it as /dir/dir2/index.html and 
pass it to cold fusion for processing?" (if thats not what you meant, 
sorry :)

which isn't a bad idea
(and one i thought about doing for people that had our old URL scheme 
bookmarked: evolt.org/index.cfm?menu=8&cid=1232 would get mod_rewritten 
to /content/8/1232/index.html and passed to cold fusion and the person 
could still find the article they wanted with the old URL scheme..in the 
end though, walker and jeff ended up just writting a much simpler CF 
page that grepped certain variables and passed them to a new template, 
check out http://www.evolt.org/index.cfm?menu=8&cid=555 to see this in 
action)
except for the fact that spiders still don't like the query string, 
which brings us full circle to the original topic :)


if that has nothing to do with your question, lemme know and I'll 
rephrase the question alex :)

.djc.


John Handelaar wrote:


> This is possibly a dumbass question, but whilst 404 trapping
> is obvious for 'near misses', in general what's wrong with
> URL rewrites like mod_rewrite or one of the similar things
> you can do with ISAPI (I've got one around here somewhere,
> haven't deployed it yet)?