[fitswcs] [Fwd: Re: WCS Paper III comments]

[Pedro Osuna]

Dear Dr. Greisen, dear all

I did not pretend in my note to hold back the approval process of the
Paper III which I think is in a completed final version. I just wanted
people to know of our approach, as we still believe it is a very good
way to handle a very old and "ugly" problem.

Let me, anyway, make a couple of comments on your mail.

1. The suggestions are incomplete since there are fundamental units
> not described and since the suggestion is not fleshed out with an
> adequate description.

if fundamental units mean some units for which we have not worked out
the dimensional equation, this is no problem, as any unit will have a
DIMEQ which is very easy to calculate. However, if by fundamental units
you mean the universal constants, then you are right: not all are
included as normally they will not be needed in the handling of spectra.

Nevertheless, I am working together with my colleague Jesus Salgado on a
full description of the dimensional analysis approach, more rigorous 
from the mathematical point of view and with a proper description. We
expect to have a paper ready in some time, and would be very glad to
circulate it to your community for comments.

2. The suggestions cover the matter of units which are described by a
> section in Paper I not III and Paper I has already been adopted by the
> IAU.  The units description in Paper I is a minor deviation from the
> description traditionally used by the IAU, but, even so, I was amazed
> to get agreement on it between the major data centers.  This is a very
> difficult area with a long tradition.

I agree that it is more related to the fundamental problem of the units,
which as you say is described in paper I.
With respect to the difficulty and long tradition on this problem, this 
is what urged myself and Salgado to work out a solution that came out as
this dimensional analysis approach.


3. I do not think that we should delay Paper III for the time
> (it could, and perhaps should, take years) that it will require to get
> full agreement on Dr. Osuna's suggestions.  

Yes, I agree.

> Instead, I would encourage
> the VO people to pursue a complete description of their proposal,
> including FITS keywords, and then pursue agreement on them in both the
> VO and FITS communities. 

We sent a detailed note with our proposal to the VO community one year
ago. At the ADASS XIV (one week ago) I got agreement finally from
Jonathan McDowell to include it in our VO standards somehow. I agree
that we should work in close collaboration in this issue and come to
agreements in both communities.

 The FITS keywords should probably consist of
> two words per axis rather than combining the two strings in one in
> order to avoid parsing/interpretation difficulties.  Note that there
> also needs to be similar keywords to cover the tradition BUNIT keyword
> as well.

Our approach included two words per axis rather than one. Jonathan
McDowell, however, asked me whether I would agree to merge both into one
for the VO specification, which I agreed as it was not a fundamental
issue for us. However, I would prefer the first approach. I'm copying
this note to Jonathan as I am mentioning him so that he can comment. 


Thank you for your comments.

Best regards,
Pedro.


On Fri, 2004-11-05 at 16:57, Eric Greisen wrote:
> Dr. Osuna's suggestions seem to be a very clever and scientifically
> useful approach to automate the handling of units.  However, I think
> there are several essential objections to including them in Paper III.

> 1. The suggestions are incomplete since there are fundamental units
> not described and since the suggestion is not fleshed out with an
> adequate description.
> 
> 2. The suggestions cover the matter of units which are described by a
> section in Paper I not III and Paper I has already been adopted by the
> IAU.  The units description in Paper I is a minor deviation from the
> description traditionally used by the IAU, but, even so, I was amazed
> to get agreement on it between the major data centers.  This is a very
> difficult area with a long tradition.
> 
> 3. I do not think that we should delay Paper III for the time
> (it could, and perhaps should, take years) that it will require to get
> full agreement on Dr. Osuna's suggestions.  Instead, I would encourage
> the VO people to pursue a complete description of their proposal,
> including FITS keywords, and then pursue agreement on them in both the
> VO and FITS communities.  The FITS keywords should probably consist of
> two words per axis rather than combining the two strings in one in
> order to avoid parsing/interpretation difficulties.  Note that there
> also needs to be similar keywords to cover the tradition BUNIT keyword
> as well.
> 
> Eric Greisen
> 
> _______________________________________________
> fitswcs mailing list
> fitswcs at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/fitswcs
-- 
Pedro Osuna Alcalaya

 
Software Engineer
European Space Astronomy Center
(ESAC/ESA)
e-mail: Pedro.Osuna at esa.int
Tel + 34 91 8131314
                                                                                
European Space Astronomy Center
European Space Agency
P.O. Box 50727
E-28080 Villafranca del Castillo
MADRID - SPAIN