Re: [Tzdist] draft-ietf-tzdist-caldav-timezone-ref - timezone vs time zone

[Tobias Conradi <[e-mail]>]

On Mon, Sep 22, 2014 at 5:01 PM, Cyrus Daboo <[e-mail]> wrote:
> Hi Tobias,
>
> --On September 21, 2014 at 5:04:21 AM +0200 Tobias Conradi
> <[e-mail]> wrote:
>
>> "It has **already** been agreed with you to change "timezone" to "time
>> zone".  We got your point"
>
>
> My position is as follows: the correct term for use in prose is "time zone".
> However, when used in protocol elements, such as WebDAV properties or names
> of XML or JSON tokens, I believe "timezone" is the best choice. I believe my
> position is backed up by existing precedent.
Precedent that you created.

> Notably:
>
> 1) RFC 5545 uses "VTIMEZONE" not "VTIME-ZONE".
This is a component name. No component name contains a "-"

> 2) RFC 4791 uses the XML element CALDAV:calendar-timezone (yes I was a
> co-author on that spec so you can take this for what its worth).
Not worth much.

> 3) RFC 4833 uses tokens OPTION_NEW_POSIX_TIMEZONE and
> OPTION_NEW_TZDB_TIMEZONE.
Eggert and Lear messed it up.

> 4) Unicode consortium: as per
> <http://userguide.icu-project.org/datetime/timezone> class names are
> "TimeZone" and "SimpleTimeZone". They use "time zone" in the prose.
They use differences in case to indicate word boundaries.

> 5) Examples of URLs using "timezone": <http://www.worldtimezone.com>,
> <http://www.timezonecheck.com>, <http://www.timezoneconverter.com>,
> <http://24timezones.com>, <http://whatismytimezone.com> etc.
URLs are another world.

> 6) Java: <http://docs.oracle.com/javase/7/docs/api/java/util/TimeZone.html>
> similar to Unicode in its use of "TimeZone" for class names.
As in Unicode they use differences in case to indicate word boundaries.

> 7) Windows:
> <http://technet.microsoft.com/en-us/library/cc749073(v=WS.10).aspx> and XML
> element <timezone>
No, it is <TimeZone>, XML is case-sensitive. As in Unicode they use
differences in case to indicate word boundaries.


> 8) PHP: <http://php.net/manual/en/datetime.gettimezone.php> has "Timezone"
> used in the name of a method.
They messed it up.

> To be fair, there are some counter examples, e.g., SQL uses "TIME ZONE".
Unicode, Java, Windows, SQL are all counter examples.

Only Eggert, Lear, Daboo, PHP, URLs are sources for "timezone" or
"TIMEZONE" in sets where word boundary indication is available and
used.

> But
> for the most part, when "time zone" is used in the context of a protocol
> "token", it is most commonly seen as "timezone".
Does not seem so.

> Therefore, I would prefer
> to leave the spec as it is in that regard.
I would prefer word boundary indication.

> If there is WG consensus to make
> a change, then of course I will do so.

I quote one WG chair:
>> "It has **already** been agreed with you to change "timezone" to "time
>> zone".  We got your point"
Source:
http://www.ietf.org/mail-archive/web/tzdist/current/msg00014.html

Which was a reply to
http://www.ietf.org/mail-archive/web/tzdist/current/msg00012.html

which argued
----------------------
>  if this is purely one person's preference over
> another's,
As presented in the proposal
http://www.ietf.org/mail-archive/web/tzdist/current/msg00004.html
it is not.
-----------------------

The proposal 2014-08-12
http://www.ietf.org/mail-archive/web/tzdist/current/msg00004.html
said:
I propose to change their names to
- draft-douglass-time-zone-service-11
- draft-daboo-caldav-time-zone-ref-01
and to replace in the text bodies "timezone" with "time zone"

That means the proposal was not only about changing English prose.

So, if the WG chair was saying the truth with "We got your point",
then this including changes to strings that are not English prose.

Some IETF RFCs and PHP don't indicate the word boundary where they
could have done. But I see no IETF rule that one has to go on with
repeating such practice.

Precedence:
https://wiki.whatwg.org/wiki/Meta_referrer
is not repeating the spelling mistake of HTTP referer.

-- 
Tobias Conradi
Rheinsberger Str. 18
10115 Berlin
Germany

http://tobiasconradi.com