(from the CLDR-users archives; source: CLDR-users index for Mar. 2007)
From: "Don Osborn"
To: <cldr-users@unicode.org>, "'Deborah Goldsmith'"
Cc: "'CLDR list'" <cldr@unicode.org>, <duerst@xx.xxxxx.xx.xx>
Subject: RE: Concerns about relative dates
Date: Tue, 27 Mar 2007 18:04:09 -0400
I assume Deborah's question about relative date references is about proposed fields for locale data, and if so Martin's implied question in response is one that deserves explicit attention. Remembering a discussion a few months ago about adding fields for certain kinds of terms (it might have been gender designations such as one might have in a questionnaire - maybe someone remembers), and a response that locales should not become dictionaries, I still wonder about the perceived need to add more terms to locale data. And indeed about the dynamic at play: the number of fields is more likely to increase than to stay the same unless the purpose is strictly delimited.
Where exactly is the line drawn? How exactly would having "the day before yesterday" or shoe sizes (sorry Tex) defined in locale data assist in localizing software or loading a webpage? At what point should we expect folks needing to know how various languages refer to a particular thing or concept to look it up in something other than locale data? (Multilingual dictionaries?)
My original conception of the purpose of the locale data was as kind of a "linguistic boot file," so the software will know how to order dates, what character repertoire is necessary, and a limited number of other basic parameters. "Common, necessary software locale data for all world languages" per the CLDR Overview via http://www.unicode.org/cldr/ . Are other purposes now foreseen?
Sorry if this is just clueless and thanks in advance for any enlightenment...
Don
< RE: CLDR 1.5 Survey Tool beta | CLDR-users | RE: Concerns about relative dates >