RE: [WEB400] get vs post...

[web400@xxxxxxxxxxxxxxxx]

Hello,

> I'm still not finding what i'm doing wrong here...If anyone can help, I'd
> more than appreciate it...here's a quick refresher...if my method = "get",
> my submit button does nothing (doesn't call my cgi and error out or
> anything)...however, if i change it to method = "post", it calls my cgi
> program.

Have you looked at your logs?  I suspect that this is a problem with your 
HTTP server configuration.

Some background:

The major difference between GET and POST is the way that data is 
submitted. Back in the good ol' days, web servers only served out static 
documents. You'd say "I want a document called /path/index.html" (or 
whatever) and it'd send back that document.  This is called the GET 
method. A browser sends the following to the server:

GET /path/index.html

The server responds with the contents of index.html.

Later, someone said "hey, here's a neat idea... what if we designed a way 
for a web server to call a program so that the contents of a page could be 
built with program logic!" the problem was, there was no way to send input 
to that program.  So they came up with the idea of encoding data into the 
URL itself.  This way, they could continue to use the GET method, but it 
would provide information to the program.

For example, if the web browser sent the following request:
GET /cgi-bin/myprogram?purrate=A&wcommcd=MC

> From the browser's perspective, this is still just a static HTML request. 
The browser and the network protocol don't have to change. The server is 
modified slightly to know that "?" separates the program name from the 
variables, but that's all.

As things got more sophisticated, they started having length problems with 
URLs. Most browsers & servers had a limit to how long the URL on the GET 
request could be, and that limits how much data can be sent this way.

So they fixed that by actually changing the protocol. They added the 
ability to upload data to the program, like a file transfer, rather than 
encoding the data into the URL. This is called a POST request.


So, that's the difference between GET and POST.  GET tacks stuff onto the 
URL where POST does a file transfer.  The symptom of your problem is that 
your program DOESN'T GET CALLED when a GET request is done.

Hmmm... the URL is being changed and your program doesn't get called... 
that's interesting, isn't it?

The reason that's interesting is that HTTP servers allow you to specify 
expressions to determine how paths are mapped from the URL sent by the 
browser to find the physical object on disk. If the modified GET URL 
doesn't work with the particular expression you've put in your 
configuration, it would explain the exact symptom that you've described...

That's why I suggest looking at your logs. When a request fails and you 
have error & access logging turned on, you should be able to get a glimpse 
into how the server is interpreting that request. For example, maybe it's 
adding ".pgm" to the end of the request, resulting in

 GET /cgi-bin/myprogram?purrate=A&wcommcd=MC.PGM

That might not be the exact problem, but it might be SOMETHING like 
that... Check your logs to see what's happening.