- From: Daniel Barclay <daniel@fgm.com>
- Date: Thu, 19 Aug 2004 14:36:30 -0400
- To: Michael Kay <mhk@mhk.me.uk>
- CC: public-qt-comments@w3.org
Michael Kay wrote:
>>That comma isn't correct (it can't join independent clauses);
>>it should be something else.
>
>
> Good writers of English know when to break that rule.
Oh, great. I try to help clarify the specification and you start
off with an ad hominem attack.
Okay...in this case:
This mapping produces an Infoset, it does not and cannot
produce a PSVI.
how do you argue that the rule should be broken?
Specifically:
A semicolon (normally) signals something to the reader: that the
preceding clause is finished and that a related but independent
clause follows.
A comma (normally) signals something different: that more is being
added to the preceding text, and that what comes next is the next
item in a list (with an "and" or an "or" coming later), is an
appositive (a parenthetical expression), or is whatever else a
comma can signal.
Signaling the second case with a comma when what's actually coming
next is the first case is confusing. (The comma narrows down the
possible parsings, and when the reader hits the period, no valid
possible parsings remain, and the reader has to back up and re-
interpret something (e.g., the erroneous comma).)
So, how do you argue that the comma isn't wrong and shouldn't be a
semicolon (or other alternatives)? You can't very well argue that
writing symbols (words or punctuation) that mean one thing when you
mean another is good.
Good writers might know when to break certain rules. However, you
have presented no evidence that you understand the rules and their
effects well enough to advocate breaking them, let alone putting
down someone else.
Daniel
Received on Thursday, 19 August 2004 18:37:02 UTC