Ernest E Vogelsinger wrote:

At 07:54 12.11.2002, Chris Shiflett said:

Anyway, thanks for your insight. Some of these slightly off-topic issues are more interesting than the on-topic ones. :-) Maybe our community needs a historian to collect some of this information for those of us who are interested (or am I the only one?).

It's slightly more than historical - in fact the behavior of GET and POST
is quite clearly laid out in RFC2616
( which represents the
"proposed standard" for HTTP/1.1.

You should probably read the original discussion. The thread was regarding naming conventions used to refer to variable types, and the reference to historical information was regarding Rasmus's choice of get/post for PHP as well as his reasoning.

Also, it should be noted that the HTTP/1.1 specification is a draft standard, not a proposed standard. See for the latest status of standards like that.

To start with, data passed within the URI is called "query"

Not exactly. This is like saying the path in a URL should be called abs_path, since that's what the specification uses. Most everyone uses the term query string to refer to the section of a URL between the ? and the # (if any). You're the first person I've seen to call it query. :-)

In plain english, developers SHOULD use $_GET variables only to _compose_
information, not to take some action to _create_ information. We should
always be aware that an URI can always be saved in a link collection
("Favirites"), or used as a link target, as opposed to POSTed data.

Personally I try to use GET (with or without a session identifier) only in
application environments that are not security related.

I spoke about the functional differences between GET and POST in a previous email. You might find it helpful:


PHP General Mailing List (
To unsubscribe, visit:

Reply via email to