I agree that a means of expressing coordinates without a list should be
available. But if the main argument for this is that lists are, in
general, not ready for prime time because Xpath2 is not ready for prime
time, then we should note that the following things are essentially of
type list and require tokenization:
NMTOKENS
IDREFS
xs:date xs:dateTime xs:time hh:mm:ss.s dd:mm:ss.s
Should RA and DEC be only in degrees or only in radians? Date and Time especially benefits from XPath2 functions because you certainly want to handle before and after decisions. Shall we burn all of our schemas and start again? This time my advice is to use DTDs only.
Secondly,
<values><value>2.18384839484334</value><value>1.423442434</value></values>
satisfies the requirement to not have a list
But, <RA>2.18384839484334</RA> implies RA is a simpleType (trivial
quantity), and therefore does not provide for an accuracy, units,
mapping, metadata, etc.
Of course we could make RA, DEC aliases for value.
Ed
Martin Hill wrote:
> Arnold Rots wrote:
>
>> ...we have three options:
>>
>> a. STC sticks with list types and VOEvent may do something else.
>>
>> b. STC temporarily converts to enumerating the vector components as in
>> my first message, and changes back to list types when list support
>> is firmly there.
>>
>> c. STC converts permanently to separate the components witin the
>> vector element.
>>
>> Comments?
>
>
> Yes. Don't use lists (in any form) for coordinates! You're only
> making it harder for yourselves...
>
Received on 2005-12-17Z16:31:05