resource identifiers

[Clive Page]

On Fri, 23 May 2003, Roy Williams wrote:

> And I have known about URL encoding since before you were born.  :-)

Ah, I wondered where that nice Mr Berners-Lee got the idea from.

> It just seems a shame to start with a syntax with this big "gotcha" in it.

Up to a point, Lord Copper.  But in practice URL encoding seems to work
pretty seamlessly.  I doubt if even 1% of frequent web users are aware of
it beavering away in the background, and even people like us who've known
about it for donkey's years probably notice its side effects only once in
a blue moon.  (oops, the data modellers will want me to have standardised
units there: let's replace those terms by gigaseconds and tens of
megaseconds respectively).

> Oops, your code doesn't work because you forgot to escape the = sign that we
> planted in there. Can't we find a URI syntax that will not trip up our
> developers? For example
> ivo://www.ncsa.uiuc.edu/nvo/registry?ADIL/SIA/targeted#95.DR.01.01.fits

Perhaps, but there's always a danger that some other separator will have
worse side-effects elsewhere.  The simple formula:
  name=value
is really very well established, in FITS, HTML, and XML, and readable.

One does have to be so careful with reserved character: for example not
long ago I knocked together a program to produce an SQL "CREATE TABLE"
statement from a FITS table (together with a CSV file of the contents).
Almost the first table I converted threw up an SQL error because I forgot
that the hyphen is valid (though actually deprecated) in FITS column
names, but completely verboten in typical SQL-speaking systems.


-- 
Clive Page,
Dept of Physics & Astronomy,
University of Leicester,    Tel +44 116 252 3551
Leicester, LE1 7RH,  U.K.   Fax +44 116 252 3311