[TAP] data type for column metadata

[Norman Gray]

On 2009 Mar 25, at 20:13, Roy Williams wrote:

>> The point is: what if you don't find out that the data are not what
>> you expect them to be? That's where the trouble starts.
>
> The young astronomer gets data from VO and assumes position and  
> wavelength are heliocentric when they are really geocentric. Perhaps  
> the VO metadata was not filled in, and thus defaulted to  
> heliocentric. Young astronomer "discovers" a planet orbiting a  
> nearby star at one AU distance. But there is no planet, it is just  
> an artifact of the heliocentric mistake!
>
> Who is to blame for this snafu? I say the young astronomer should  
> have gone to the library before publishing.

Absolutely.  It is not part of the scope of the VO or of VO  
applications to remove the need for an astronomical education.

Warning: the following may contain some sweeping generalisations...

Before the VO, if you wanted to use data from an unfamiliar source,  
you had to read the archive's documentation, or mail them for advice,  
and then painfully work out how to use that data alongside others.

I believe it will always be the case that, for some perhaps subtle  
uses of data, you will at some point have to metaphorically pick up  
the phone to the archive specialists, and ask for advice or  
clarification.  The VO is never going to make that step go away  
completely.

Right now, you can at least do quick-look examination of unfamiliar  
data, or multi-wavelength alignment which is good enough to warrant a  
second look, purely with VO services and metadata.  As time goes on,  
the scope of such `no-phonecall' VO science will extend further and  
further.  What we can aspire to is this continuous qualitative  
improvement, not perfection.

It's often said that the reason the web was successful, where other  
hypertext systems were not, is that it accepted that the world was  
imperfect and often slightly broken -- that 404 was an acceptable  
answer.  It gave up the search for perfection, and instead designed  
its protocols, its applications, and thus its users, to cope with  
breakage, successfully.

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
Dept Physics and Astronomy, University of Leicester