[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Syntax and semantics

• From: Sean McGrath <sean@d...>
• To: xml-dev@x...
• Date: Thu, 18 May 2000 04:53:35 +0100

Assertion: Putting all the semantics of a message
into a message is impossible.

Here is why:
I have a file here called foo.ps. It can be thought
of as a stream of bytes. It can be thought of as a
collection of tokens that can be summoned forth
with the aid of a grammar. It can be though of
as combination of executable code with embedded
chunks of data. It can be interpreted in
a multitude of different ways. In short, it
has many semantic views depending on what
a process is doing with the resource at a given
point in time.

The latter "view" of the foo.ps
resource is more commonly expressed in
English by the phrase "Postscript file".

Now, If I send foo.ps to you over the Internet,
does it contain semantics as well as syntax?
Yes. It is clearly more than just data. It contains both
data *and* the code to interpret that data in
the form of a Postscript program.

Is the semantic component of foo.ps self
contained? No. In order for foo.ps to
be "rendered" as is the life
wish of Postscript files, a processor has
to have a-priori capability to interpret
and execute Postscript programs.

No matter how you cut it, external semantics
are a requirement to get anything done!

Produce an utterly self contained message with
no need for externally agreed semantics is
impossible it seems to me.

What is the most basic type of self
contained message? Probably a stream of
1's and 0's to be interpreted as code
to be executed by a Turing machine. Can
a single message be constructed that
contains both the semantics of 
a Turing machine and the data to feed
to the Turing machine once constructed?

It just ain't possible, right?

regards,

Sean,




***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@x...&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************

• References:
  • RE: Syntax and semantics
    • From: "David Galbraith" <david@m...>
  • Re: Syntax and semantics
    • From: "Dave Winer" <dave@u...>

• Prev by Date: Re: arbitrary bytes/characters?
• Next by Date: land specific and international use of common labels
• Previous by thread: RE: Syntax and semantics
• Next by thread: Re: Syntax and semantics
• Index(es):
  • Date
  • Thread