VOunits draft

[Rob Seaman]

Hi Hervé,

> On my side, I don't understand why you're focusing on these two ends  
> of the distribution !

I'm just trying to write the email that gets me out of the  
conversation :-)

I'm mentioned yotta and yocto because they are part of SI and it keeps  
being asserted that we must support ALL of SI.  It seems more  
important to support table columns in units of parsecs (something we  
definitely will want to do) rather than columns in units of yoctohertz  
(something I am skeptical will prove of practical value).

> But will lead to revision in the future if new cases appear where  
> new prefixes are needed.

As with other VO discussions about maintaining enumerated lists (eg,  
UCDs, vocabularies), there must always be a process for revision.   
Astronomers frequently (meaning several times each century) create new  
units.

> the basic idea (that comes from the best practise of numerical  
> computation community)

Is there a URL pointing to a document discussing the topic of  
numerical precision in the context of SI prefixes?

> is the fact we can separated the magnitude (10^n here, not the  
> astronomical one :-) ) from the significant part of a real number  
> (mantissa). By chance (?), the SI use only power of ten to encode  
> unities...
>
> Let's assume a parcsec to be (I know this is an approx) :
>   1 pc =  3 085 677 581 282 10^{16} m
>   ...

Thanks for the discussion of floating point precision.  My two  
comments are that first, conversions may only be needed if the VO  
standards require them.  There are lots of tables containing columns  
of distances in parsec (and kilo and mega parsecs).  Are we going to  
require that these be converted to meters?

	1 Parsec = 3.08568025 × 10^16 meters (according to wikipedia)

So will folks have to convert 1 megaparsec to 30.9 Zettameters (Zm)?   
Bear in mind that the distance estimate may be accurate only to +/-  
50% and may scale with the hubble constant...

> IVOA standards already depend heavily on external standards as W3C  
> ones for instance. W3C seems to be such an external authority. FITS  
> is only heavily used, that is an IAU standard.

I'm not adverse to using standards.  The question is whether we are  
choosing standards solely because some authority has blessed them.

> Maybe I'm wrong, but what I've understood from the discussion is  
> that the defintion of a common unit system is needed for 'internal'  
> purpose (eg service-to-service communication). At the level of the  
> human client, any 'traditional' unit is allowed. Am I right?

The choice of internal units must still make sense astronomically and  
computationally.  But I think that VO facilities must also be  
considered to include their user interfaces.  If our internal systems  
are required to use zettameters and our external interfaces parsecs,  
some agent must perform the conversion.

Rob