W3C home > Mailing lists > Public > public-rdf-dawg@w3.org > October to December 2005

Re: XSD decimal syntax

From: Dan Connolly <connolly@w3.org>
Date: Tue, 08 Nov 2005 11:21:25 -0600
To: Steve Harris <S.W.Harris@ecs.soton.ac.uk>
Cc: RDF Data Access Working Group <public-rdf-dawg@w3.org>
Message-Id: <1131470485.7185.469.camel@dirk>

On Tue, 2005-11-08 at 17:05 +0000, Steve Harris wrote:
> On Tue, Nov 08, 2005 at 10:58:15 -0600, Dan Connolly wrote:
> > > The turtle thing is a big deal, if 1.0 is a decimal, then:
> > > 	SELECT * WHERE { [] rdf:value 1.0 . }
> > > does not match
> > > 	[] rdf:value 1.0 .
> > > 
> > > That's a big drawback IMHO.
> > 
> > You don't think turtle would follow SPARQL and migrate to decimal?
> I dont know that. If I had an assurance that turtle would change inline
> with SPARQL syntax then I probably wouldn't object to this change. Its a
> pretty major change to the practical use of SPARQL though. I'd probably
> decide to change my engine to internally prefer decimals over doubles,
> which I'm not that keen on.

Yes.. hmm... well...

I don't see any really good way to make this decision. I'm moderately
inclined to do a straight straw poll, see whether decimal or float
gets a higher count, and put the question in that direction.

But this _is_ a technical design decision, and with attendance at
teleconferences subject to so many random influences...

Let's do this by email. If you have an opinion on this, please
send mail. Try to give a little technical rationale.

I'll probably allow for last-minute discussion at our next
teleconference, but when I put the agenda together, I'll
include a PROPOSAL based on counting positions in email.

Andy, feel free to write up the decimal design for this Friday;
the WG may instruct you to revert it, but we'll take that risk.

Right now, I count:

  Andy/HP: for decimal.
   "What gets written is exactly the value in the query itself " etc.

  Howard Katz: for decimal (7 Oct)

  DanC/W3C: for decimal (following discussion with TimBL,
   and in the interest of consistency with XQuery)

  SteveH: for float/double
   impact of change to big/uncertain

  dajobe: my preference is that as xsd:doubles are more commonly used
(typed) and should be the default. nobody wants to do 3.4e0 lots.
bignums are rare in data

That's sort of 3-2 in favor of decimal. I _really_ don't want to decide
on a margin that small. Please let's hear from others.

> - Steve 
Dan Connolly, W3C http://www.w3.org/People/Connolly/
D3C2 887B 0F92 6005 C541  0875 0F91 96DE 6E52 C29E
Received on Tuesday, 8 November 2005 17:21:43 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:00:49 UTC