From: Helmut Coelfen <coelfen@castor.mpikg-teltow.mpg.de>
  To  : rasmb@bbri.harvard.edu
  Date: Wed, 9 Apr 97 19:11:24 +0200

Binary file format

Dear RASMB,

in the discussion about file formats raised by John Philo, I would aggree 
with the
points Jim Bloom, Jo Butler and Borris Demeler have made so far. In our lab, we 
are very happy with the current ASCII data output format as it is because it 
allows
easy access by all kinds of software on all computer platforms. I see problems
using binary files insofar as this would raise compatibility problems with a
certain probability at least if I want to import these data into various 
commercial
programs or the whole lot of existing programs which are only accessible as
compiled file and thus cannot be changed anymore (Maybe because the pHD 
student who has written them has left in the meantime).

Furthermore, ASCII is a common standard which has been established throughout
the last decades and which is compatible to every kind of software. Therefore
it is much easier accessible than any other file format and this is of advantage
for all people - if they know much about software development or not -

I see the point of a quicker reading in process of the data which is 
certainly of
benefit. But if I compare it to the loss of data compatibility by a change 
to a binary
format, I would highly recommend to leave the ASCII format as the output
of the ultracentrifuges. The people who are developing software are still free
to write any kind of conversion programs of ASCII into any binary or 
whatever format
is needed. This leaves the ultracentrifuge users with an accessible file 
format and
the software specialists can convert to their file format of desire. This is a
much better scenario than having the XL-A/XL-I data files as a binary output
with the option to convert them back to ASCII if one wants to.

In short words: Keep ASCII as output file format

Best regards

Helmut Coelfen