Problem with IRM and .htaccess ...

[zmax]

Im not sure if this is a known problem or oddity or what.
My main domain is www.glowsquid.com
I recently added an IRM called www.worldwidecat.com
I wanted to test how .htaccess works in an irm.

I created a test directory in the wwc area and copied a .htaccess file from another directory that I know is setup and works. I then created in the directory an index.htm file.

As you know if you specify a directory in a browser url such as http://www.worldwidecat.com/test where test is the directory, the browser will grab the default file it finds like index.htm and present it.

However, where the directory test is protected, when I enter the url http://www.worldwidecat.com/test, and I then enter the correct id and password, I get a 404 not found error and the url link in the location edit box is: http://www.glowsquid.com/test

I have a 404 file in wwc which works fine.
Then if I repeat this process but include index.htm in the url specifically, it then finds the correct file no problem.

Whats going on? Can this be fixed?

Thanks.

David

[Deb]

With IRMs it is very important to include the ending slash on the urls where appropriate.  

So
http://www.worldwidecat.com/test may provide an error
but http://www.worldwidecat.com/test/ may work as expected.

The IRMs require correct construction of the URL....

REF: http://www.aota.net/ubb/Forum3/HTML/000300-1.html

Deb
 --
  To slash or not to slash?

[Terra]

Can this be fixed?

In short: no

IRM's use a series of complex dynamic rules to make IRM's almost 100% transparent...

When the 'trailing slash' problem came up long ago, I thought to myself this should be easy enough to fix...  From there it was all a rude awakening...

Well, to make a long story short - this problem lives within a combination of 'Murphy's Law', 'Domino Effect' and the final clincher 'Cascade Effect'...

To sum up all the above Law/Cause and Effects, the problem becomes a cyclic 'Fix one thing, Break another' eventually completing a full loop back to the starting point...

In conclusion: IRM's depend on site owners to use 'Proper' URL paths instead of relying on ad-hoc hacks built into the Apache server to handle people not knowing/willing/lazy to append the proper 'trailing slash'...  This rule has been spec'd out in the HTTP 1.0 and HTTP 1.1 guidelines since it's inception years ago...

--
Terra
--Played musical chairs trying to solve this one--
FutureQuest

[zmax]

Oh my gosh youre right!
The trailing slash worked!!!!
Thanks!

And as for the long story short yea I work in QA at a software company and I see what development goes through and how little things can break a lot and sometimes its better not to touch it. But I dont work in web based apps so this is new to me. Im learning.
My first web page was one of the secondary domain types and I learned that you dont need a / at the end if you put a statement in the main page saying base href whatever etc and then promptly forgot about it. I have also referenced the discussion you mentioned.

Thanks again!

David

[Justin]

Actually you should *always* use a trailing slash at the end of a directory name in a URL. Otherwise this is what happens:

Browser: www.domain.com/dir
Server:  /dir is a directory
Server tells browser: go to www.domain.com/dir/
Browser: www.domain.com/dir/

This is pretty much transparent but it still adds overhead, as each request to /dir is checked, rewritten, and requested again. And of course in the case of an IRM, where the SERVER_NAME variable contains the main account's domain name, it is rewritten to an unexpected value...

---------------
Justin Nelson
FutureQuest Support
-- I can't believe I missed the first 20 minutes of South Park AGAIN - was on twice today and I still didn't catch it