[Date Prev][Date Next][Thread Prev][Thread Next][Search]
[HEASARC Mailing List Archives]
Re: OGIP/93-013 - A list of standard strings for HE missions,instr..etc
Barry Schlesinger listed a few questions/comments in his recent HEAFITS post
(1994 Feb 23) regarding the OGIP Memo OGIP/93-013 "Standard Strings for
Mission, Instrument, Filter & Detector Names for OGIP FITS files".
Below we give our response.
A number of relatively minor changes have been made to the memo resulting
from all the comments received, and a new version (1993 Feb 28) is available
via anonymous ftp on legacy.gsfc.nasa.gov from
caldb/docs/memos/ogip_93_013.tex (LaTeX source)
> The formsats given for the keyword=value statements appear to be
> missing blanks. A keyword of fewer than eight characters must be
> filled with trailing blanks. There must be a space between the "=" and
> the quote beginning the characters value. Also, the recommended form
> for character string values is that they be at least eight characters,
> with trailing blanks, if necessary. While this format is not required
> for optional character values, it should be followed unless there is a
> compelling reasong not to. A sample keyword=value statement would be
> as follows:
> DETNAM = 'SDB '
... Yes we were aware of the rules/conventions concerning character
string keywords. Yet, it had not occured to us that the computer-style
font used for these strings may have lead to confusion in the minds
of some readers. However in order to be 100% clear & true, the latest
version of the memo has the correct FITS syntax (inc trailing blanks)
everywhere appropriate within the text.
> Because of portability issues, underscore is preferable to hyphen.
> Hyphens are standard for some FITS usages for historical reasons, but
> the recommendation now is to use underscore rather than hyphen if
... We understood the recommendation to use underscores rather
than hyphens was primarily aimed at keyword names, rather than
character string values. Since there are no hyphens in the
keywords we quote, and since it would be almost impossible
to 'enforce' a re-naming of most of the instruments which
commonly use hyphens as part of their instrument/detector names,
we consider the suggested strings are acceptable.
It's worth reiterating that our primary goal in the memo was
not to invent a beautiful multi-mission naming scheme, rather
to simply formalize/list the common strings used by scientists.
> On page 6, there is a note that all strings in subsequent sections are
> to be considerered case sensitive. What is meant by this statement?
... We mean that the values of the strings should be considered
case sensitive, again in an attempt to stick to the strings
commonly used by scientists, or already in use within OGIP
FITS files. In fact you will see that upper case is used
throughout, with the exception of a handful of filter strings.
> Having two meanings for the values of IPC or SSS for INSTRUME for
> EINSTEIN can be confusing. The two non-recommended definitions should
> not be allowed, unless they are there to grandfather in existing files.
> If that is the reason, then the usages should be allowed but
> deprecated -- designated as permissible for existing data sets but not
> to be used in the future.
... We agree, but in many cases (not just for the SSS & IPC) this is
the result of there being several similar instruments onboard,
but with one instrument being used for the vast bulk of the
scientific observations. We think our disapproval is expressed, but
there are simply too many files out there already to do much more.
> Since data sets have been produced for a number of the experiments
> included, are the designated keyword values (names) for the different
> missions, instruments, detectors, and filters consistent with current
> usage by the experimental groups? In the same context, I note that
> indexing is inconsistent from usage to usage, sometimes starting at 0,
> sometimes at 1, sometimes at other values. Does this indexing reflect
> that used in by the teams in referring to their instrument? Normally,
> indexing should be consistent, beginning with 1, but the actual
> designations given by the experimenters would override such a
... Yes, we hope that the usage is that of the experimental groups,
and yes, this is exactly the reason for the inconsistent indexing.
... But have you ever tried to tell an experimental group how they
should number their detectors !!! :-)
Ian M George