Thread: Re: byte order and swapping, Documentation

Started: 2005-11-30 20:02:04
Last activity: 2005-11-30 22:52:00
Topics: SAC Developers
George Randall
2005-11-30 20:02:04
SAC developers,


I'd like to advocate for having SAC use the platform native byte order
and defer any conversion to sactosac. I've seen and used several
scripting packages to read/write sac headers and data (python, matlab,
and perl) that are useful in platform native byte order. The python and
perl kits essentially function as a shell like way to scan sac files
based on the header, and do any basic file level manipulation to
organize data sets etc. I would hate to see these efforts get complicated
by dealing with non-native byte order. SAC could take any approach if
well documented, but SAC files are now used in other processes, and the
impact on other systems of non-native file formats should have some
importance I hope.

My first experience with SAC was on Prime computers with a non-IEEE floating
format, and I had (now lost) a converter from Prime SAC files to Sun long
ago. We weren't dealing with NFS mounted file systems and multiple
architecures, so conversion to native format was the logical process back then,
and was usually done as part of a process to read Prime tapes onto the Sun.
I'd like to see SAC continue to use platform native formats.

On the issue of documentation, I remember a code that stripped comments for
the SAC source (fortan version) and formatted the output. I think it generated
the help files, but it also was used to generate TeX files which were used
to generate the now long lost manuals that were printed. I think this was
prior to Adobe PDF format. I think Bill Tapely at Livermore did this work,
and he moved on quite some time ago, so I doubt he still has the old code.
Perhaps a perl or python script could pull the comments from SAC source
and generate any format of choice? I am NOT volunteering, but might kibitz
with anybody who wants to tackle the project. I have an old and outdated
set of TeX files for the SAC Commands manual circa 1993 which would show
the TeX formatting used for the old paper manuals. I guess Livermore would
need to bless the release of the TeX documentation?

I'd like to see documentation as pdf and html packaged as part of the
released SAC kit so that documentation isn't tied to a web site, and can
be used while disconnected (or connected by low badnwidth) from the Internet.
That might give incentive to get the code and documentation in step for
release.

George Randall
EES-11 Geophysics
Los Alamos National Lab


  • Arthur Snoke
    2005-11-30 22:52:00
    I have the 1993 documentation in text and .ps for sac (and for map). I
    recall there were some problems because they were hard-wired for a LLNL
    printer, but I think I figured out how to get around that. My most recent
    perturbation was in Feb. 1995. Maybe not even latex209. Here is a start
    of one of the files:

    equake{snoke}642: head install.tex
    \vsize=12in
    \input docplain
    \setbox0=\hbox to
    1.00
    in{\null}
    \dimen0=\hsize\advance
    \dimen0 by -\wd0
    \dimen5=\wd0\advance
    \dimen5 by 1pc
    \hl{\brm
    equake{snoke}643:

    I have offered to work on the documentation -- for starters, updating the
    documentation on the Web site for some things like the header variables,
    etc.


07:46:57 v.01697673