On Sep 17, 2006, at 11:58 AM, Roy Williams wrote:
> Alberto Conti wrote:
>>
>> The adoption of a well known ISO standard or specification seems
>> the most appropriate thing to do, instead of coming up with yet
>> another IVOA recommendation that is just a poor compromise.
>
> Alberto:
>
> (1) I have tried to find existing standards, but ISO seems to have
> nothing. The word "sexagesimal" in Google does not show a
> recommended notation that I could find. Other information is from
> "people like us" (Iraf, Starlink, Vizier, etc) and recommends
> various output formats. In other words, I don't think there is a
> Sexg standard representation.
>
this is true, but what you could do is look at "time standards" from
ISO (8601? I think) and extend them to IVOA standards.
> (2) The IVOA is a standards organization. Our job is to make
> standards. Obviously, our standards should not be poor.
>
perhaps poor was a bad choice of words. My concern is that we propose
something disjoint from what the users are using at the moment (see
FITS) and will be using. I do agree that some standardization is
indeed require...
Supporting all these format
Decimal Degrees
185.63325 29.8959861111111
Hours, minutes and Seconds
12 22 31.98 29 53 45.55 12h22m31.98s 29d53m45.55s 12:22:31.98 +29:53:45.55 12h22'31.98" 29d53'45.55" 12h 22m 31.98s 29d 53m 45.55s
12h 22' 31.98" -29d 53' 45.55" 12h22'31".98 -29d53'45".55 12h22m31s.98 -29o53m45s.55 12h 22' 31".98 -29d 53' 45".55 Hours/Degrees and Minutes (no seconds) 12 22 29 53
like MAST and most archives do is rather impractical.
> (3) I suggest that many formats be "acceptable" to IVOA
> applications with wide inclusivity. If that effort works, perhaps
> we can think of a tighter format that is "recommended".
>
indeed, you should accept input in various formats, but you must
recommend one from the IVOA.
> (4) Part of the effort here is to allow our customers to put
> multiple things in text boxes. If it parses to 2 floats, we assume
> long/lat in degrees, if it parses to six numbers, it is Sexg, or
> else perhaps it is an object name.
>
again, perhaps the focus should be to get a clear understanding from
users (archive searches for example?) on the most common format(s)
and then move to the definition of an IVOA standard.
Just my $.02
-A
>
> California Institute of Technology
> 626 395 3670
>