W3C home > Mailing lists > Public > uri@w3.org > September 2003

Re: delimiters vs separators

From: by way of Martin Duerst <mike@skew.org>
Date: Thu, 25 Sep 2003 13:53:36 -0400
Message-Id: <4.2.0.58.J.20030925135323.06062510@localhost>
To: uri@w3.org




Al Gilman wrote:
 > Can you suggest specific edits where changing the current language would
 > make the sense of the specification clearer?  Such as you provided in
 > the case of the 'path component'?

OK, I'll grep through it again and suggest edits.

 > It is plausible to consider that both terminators and separators are
 > delimiters.

I look at it like this: a terminator or separator, especially in linear data,
does not necessarily offset a piece of data from all other adjacent data,
which is what I perceive the role of a delimiter to be; it just gives you a way
to distinguish between what data precedes and what data follows the separator
itself, which may or may not be enough information to identify a finite range.

Feel free to take this with a grain of salt. It's just that I've been scolded
in the past (by people with far more experience and knowledge than me, though
of course that doesn't necessarily mean they're right) for not using a stricter
definition of delimiter. IIRC, I had written "comma delimited values" and was
told that there's a reason it should be "comma separated values" instead, and
not just because there happens to be a popular spreadsheet export format that
uses a filename extension based on the initials of the latter phrase.

 > I think we would have to see concrete evidence of actual damage (to
 > comprehensibility) to get a consensus behind a change.

No one seems to be at each other's throats about it, but there are definitely
two camps: those who distinguish between delimiters & separators, and those
who don't or just don't care. When you publish something with the longevity of
an RFC, you are going to have an influence on how many people in the future
fall into one camp or the other -- language is dynamic and is defined by how we
use it, over time.

Instead of saying "yeah it's a little fuzzy, but what's the harm?", instead
ask "do we want to help define these terms through our usage of them, or do we
want to leave them more open to interpretation, come what may?" We've seen
what happens when a term like "resource" isn't nailed down. What seems like a
harmless ambiguity now could become a bigger issue years in the future.

 > But most of us have been re-reading this language too many times to
 > perceive where it is impenetrable.  Fresh eyes such as yours are very
 > helpful under the circumstances.

Well, I don't claim to "know" anything :) Take it all with a grain of salt.
Thanks for taking the time to hear my concerns, at least.

My 2c,
Mike
Received on Thursday, 25 September 2003 14:09:12 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:25:06 UTC