- From: Shannon J. Clark <shannon@jigzaw.com>
- Date: Wed, 6 Jun 2001 01:10:36 -0500
- To: "brian moseley" <bcm@maz.org>, <www-rdf-calendar@w3.org>
An additional thing to consider, besides translations potentially being offered for textual descriptions (presumably by the organizer of the event - though perhaps a mechanism should exist to allow for translators who might not be the organizer?) is that to fully deal with internationalization of time related information more than just translations of descriptions will need to be done. Much of this should be doable by the client side software, however currently iCal stores dates in a form that is derived from text strings - that is it is a set of strings that are interpreted (numbers in a particular order, optionally followed by a "Z" character to indicate UTC time, or with a corresponding TZID string identifying that a TimeZone should be applied to the time - for international purposes the representation of the dates should be adjusted to match locale specific formating (commas vs. periods as seperators, order of presentation of Month, Day, Year; abbreviations of month names, day of week names etc, even ordinal markers (1st, 2nd etc.) all of which vary from language to language) More specifically, this is all fairly simple when moving data between one-byte languages - most of the code that is written to manipulate the strings, break them apart, convert them to the locale specific abbreviations/names etc is not hard to write - and frequently system locale information can be used to get the correct names (or alternatively locale like string mapping can be loaded into an internationalized software). Where is gets much trickier is dealing with multi-byte languages (mostly Asian languages) where the underlying datastructures and hard coded assumptions about separators, how many bytes between separators etc will all be very different for double-byte languages. One issue to also consider is the USE that the underlying data schema will be put to - one consideration that iCal has tried to keep in mind is that while an iCal formatted file is intended for a software package to interpret and deal with, it is also written text with human understandable field names and formatting - thus it can be (and often is) also read by a human to look at the information (see also vCard's which are also meant for a computer software to read and process, but are also written with humans in mind.) However - both iCal and vCard (and vCal) were written mostly assuming English - keywords used to separate information out are English words, and the schema mostly assumes an ASCII format, single-byte, manner to depicting the information. It is assumed that content may have fields in multiple languages (indeed there is even a LANGUAGE property/parameter (have to look up which) within the iCal specs - but the keywords identifying each iCal section and the other underlying assumptions about how it will be dealt with assume English. This is not necessarily a bad thing - but worth keeping in mind. So - do we care just about the content - and about having the ability to store content in multiple languages (perhaps even in multi-byte languages) or should we also look at the assumptions and structures underlying the schema itself and not just the data that the schema might be called upon to store? Shannon -----Original Message----- From: www-rdf-calendar-request@w3.org [mailto:www-rdf-calendar-request@w3.org]On Behalf Of brian moseley Sent: Tuesday, June 05, 2001 10:27 PM To: www-rdf-calendar@w3.org Subject: Re: RDF calendar schemas On Tue, 5 Jun 2001, Aaron Swartz wrote: > Shannon J. Clark <shannon@jigzaw.com> wrote: > > > Five - will you have to handle multiple languages? Unicode and/or multi-byte > > languages? > > XML takes care of this for us. xml takes care of specifying data in unicode, but how does it signify that the textual description of an event is in french?
Received on Wednesday, 6 June 2001 02:12:10 UTC