>> In inverse order of importance:
>> 1. The UCD-like syntax is wrong: first the "quantity", then the "qualifier"
>> it should read: weather.seeing;instr.obsty.site
>
> Sorry - I confess that I haven't been worrying AT ALL about things like
> order...
That's why I said it's the least important thing :)
>> 3. My mind works in terms of "how to assign ucd-words starting from a
>> description, and then build an UCD" ...
>> It is impossible (in this framework) to go from the single keyword
>> "seeing" to an UCD composed of two UCD-words
>> (each one of these UCD-words needs its own description).
>
> Hmmm.... don't quite get that. My approach has been
> - I want to express something
> - look at the official UCD list to see what comes closest
> - if there's no exact/good match, look for related things to see if
> composing works
The above points correspond to the action:
which is followed by the action of
These two actions together are performed by the "UCD-builder" that
you can find, in its "beta-version", at
http://vizier.u-strasbg.fr/UCD/cgi-bin/descr2ucd
>> Let's reverse the roles: you jump on the chair when you see the description
>> of the src.class.* words.
>> You are perfectly right. What is not said is that I do not use the
>> description
>> field to assign words/ucds: I use a series of words (keywords
>> derived from the description) that are OR-ed,
>> which is exactly equivalent to the loosier description you suggest.
>
> Sorry - still don't get this..... Are you saying that
> src.class.richness DOESN'T mean
> Abell class (even though that's what the official list says) or not?
It actually means :
richness class e.g. Abell richness class
(by the way, ther is also a "Abell distance class" !! Try "Abell class"
as input for the builder...)
> If it DOES mean
> that, then we need to rename the word. If it DOESN'T mean that, we
> need to change the description. Either way, we need to change
> something, because leaving it as
> it is will forever prevent someone from using richness for anything
> else with a good conscience.
Yes, we need to change the descriptions of all src.class.*, and this thanks to the discussion that is going on!!
Andrea Received on 2005-06-15Z10:29:25