I
Isofarro
ManoDestra said:Incidentally, have you guys noticed that this newsgroup is called
alt.html.writers.
You may want to check the Groups: line in the header of your own post.
Does that mean Perl? ASP? JSP? No. I didn't think so.
It's about writing HTML which is what I have been talking about all along.
Ahh... you want to play topic-netkkkkkop. Fine, I can play that game too.
HTML is defined as a markup language, it is defined by the World Wide Web
Consortium in published documents called Technical Recommendations. Now,
the only other standards body that publishes a specification for HTML is
ISO. There are no other publically available and generally acceptable
definition of what consititutes valid and well formed HTML.
So, lets look at the W3 specification, and since the vast majority of
browser in use at the moment support the HTML 4.01 specification, lets use
that particular one. Kindly the W3 has a website where these documents can
be found (HTML is used to construct websites). So, the HTML 4.01
specification from the W3c can be found at the following URL:
http://www.w3.org/TR/html401/
The particular section we want is the specification of how to create a valid
and well-formed form (since this thread is about sending form data), more
specifically we want the specification about the form action. Here it is:
http://www.w3.org/TR/html401/interact/forms.html#adef-action
And it says the following about the action attribute:
This attribute specifies a form processing agent. User agent behavior
for a value other than an HTTP URI is undefined.
A HTTP URI is a URI beginning with http:// and it uses the HTTP protocol.
Because requests over SSL use the identical syntax (the request is merely
encrypted with a pre-agreed key before being sent), https:// also falls
within this definition. Mailto clearly doesn't - its not even a proper
protocol. Its what we call a pseudo protocol - it pretends to be something,
but in a number of cases it is useless beyond belief.
So what happens to URI's that are not http? That's what the second sentence
covers: "User agent behavior for a value other than an HTTP URI is
undefined."
Now what does undefined mean? It means that it has no definition.
What happens when there is no definition? We cannot reliably predict what is
going to happen. Clearly, within this definition, a browser that farts in
your general direction you when you submit a form using a non-HTTP URI is
perfectly complying with the specification. There is _nothing_ that
specifies how a browser is supposed to deal with a non-HTTP URI.
Again, further in the document:
http://www.w3.org/TR/html401/interact/forms.html#h-17.13.3
Step 4 contains the brief paragraph:
"For any other value of action or method, behavior is unspecified."
This is after dealing with GET and POST methods using HTTP URIs. No mailto
is covered, again pointing out that the behaviour is unspecified.
There is certainly no argument here. Only your inability to admit you've
erred is continuing this thread - even after two declarations of your
non-participating.
If you insist on remaining completely anal, then let's look at the ISO/IEC
15445:2000(E) ISO-HTML specification:
http://www.cs.tcd.ie/15445/15445.HTML
Here's the relevant part:
<!ATTLIST FORM
...
ACTION %URI; #REQUIRED -- Server-side form handler --
...
-->
Which part of "Server-side form handler" are you having serious difficulties
understanding?
Here endeth the lesson.
And all this is in the HTML Technical Recommendation, so it is well on-topic
for html groups, including html.writers. If you are not writing HTML as
laid out by either the W3C or ISO, then it is you that is off-topic, since
it is no longer HTML but some other non-specified bastardised markup
language.