From WysiwygWiki, GroupWare:

Web pages are normally fetched by a browser using what's called an "HTTP GET" request: it looks like this.

     GET /cgi/wiki?HttpPut HTTP/1.0
     User-agent: mozilla/4.0
     Host: www.c2.com
     ... ''more headers here''

Forms are submitted by an HTTP POST, which sends data back to the server.

One of the less commonly used commands is HTTP PUT, which sends a page back to the server to replace the existing page. The command normally requires some kind of authentication.

----

So...  "HTTP PUT" is a "copy this file to the server" function, overwriting any existing file.
Thus, the web server acts more like a file server than a web server (generally speaking).

''Right?''

That's right - but the web server can handle the incoming information however it likes, it need not just write it into a file. Typically it will do access control but it might also for example insert it into a VersionControl system.

----

See also WebDav, which takes this idea further.