On Wed, 20 Sep 2006, Francois Ochsenbein wrote:
>
> Hi VOTablers,
>
> The attendence at this IVOA session is rather sparse -- and we were
> just 8 persons at the VOTable session. Quite easy to take decisions,
> but a larger consensus is wished...
>
> The viewgraphs are available from
> http://www.ivoa.net/twiki/bin/view/IVOA/InterOpSep2006VOTable
Francois et al.,
Sorry I couldn't make it to Moscow. I have one or two comments on issues mentioned in the presentation on the wiki:
As I understand it, both of these are suggestions to change the
VOTable definition in order to accommodate the behaviour of
some specific software items (automatic code generation tools)
which are buggy or incomplete. I do not support this course of
action. Changing the standard in response to vagaries of
specific software items is unlikely to lead to a clear standard,
and is likely to be overtaken by events as the implementation
status of existing software changes or as new software arises.
3. Empty TD
Allowing <TD>NaN</TD> instead of <TD/> for integers doesn't
fix anything; there's still the problem of how such an
element is represented if you want to convert to the BINARY
or FITS representation.
Providing a magic number to stand for integer null values
can already be done by using the "null" attribute on the
VALUES element. Is your suggestion of using the fixed
0x8000* values intended to replace this?
4. Schema changes
"&32;" should read " ".
Mark
-- Mark Taylor Astronomical Programmer Physics, Bristol University, UK m.b.taylor@bris.ac.uk +44-117-928-8776 http://www.star.bris.ac.uk/~mbt/Received on 2006-09-20Z13:39:40