ÚÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿
³                        SOURCEWARE ARCHIVAL GROUP                         ³
³                       WHAT'S COMING IN THE FUTURE                        ³
ÀÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÙ

With the May 1994 release, SWAG has been in existence for exactly one year.
In that time, some pretty remarkable  things have happened. SWAG has become
a worldwide resource for the PASCAL community.  This is EXACTLY what we had
in mind when we started the project.

I  get  mail just  about everyday  from someone  telling me  that SWAG  has
become  an important  part of  their  programming  arsenal. As  I read  the
network mail each day, I'm amazed at  how far and wide SWAG has spread, and
how often it is referred to as a reference source.

It has been and continues to be a  work that I greatly enjoy, and I want to
express my gratitude  for ALL those who have contributed  to the success of
SWAG; the  many network administrators,  snipet contributors, beta  testers
and pundits who have made many suggestions to improve the project.

We are up to version 2.8 of the READER.EXE. To be honest, I had intended to
release a new  program with this version that  would compress, using PKZIP,
the .SWG files.  This would save about 4Meg  of disk space. As many  of you
appreciate, the  size of SWAG  is getting to  be a problem.  After a LOT of
thought  and consideration,  I decided  NOT to  release the SWAG compressor
program. One  of the reasons  for this  was  that users would  HAVE to have
PKZIP  and PKUNZIP  to use  SWAG, and  I didn't  want to  rely on  external
programs.  Second,  the  GREP  function  became  **SLOW** and the quarterly
update program was even slower. This was because each snipet file had to be
unpacked  and  repacked  after  use.  I  eventually  decided  that this was
unacceptable.

I  realize that  the problem is much  more basic that just  compressing the
data. What is  needed is a completely new approach  to saving the data, and
accessing it. The .SWG  files are based on the QWK file  format. I did this
for simplicity, and frankly, it was easy.  Therein lies a great deal of the
problem.

So, by  the end of 1994,  I intend to COMPLETELY  rewrite the SWAG project,
with a completely  new file format. Each snipet will  be stored as a packed
KEYED record in  the files, with each having an  UNLIMITED number of search
keys. This  will reduce the  overall size by  greater than 70%,  and should
provide  the speed  that we  all  require.  Additionally, there  will be  a
"WINDOWS"  version  of  READER.EXE.  There  will  definitely  be  other new
features as well.

I welcome ideas that you might have regarding the "NEW" SWAG.

Thanks again to everyone who has made this entire project a success.


Gayle Davis
May 1994