Mel,
        This tape is more or less Purdue's  contribution  to  the
next  UUG  distribution.   There are probably some missing files,
etc, and some things not documented.  The  kernel  and  adventure
sources  are kind of large and you may not want to clutter of the
distribution with them, however it is ok if  you  have  room  and
want to send them out.  I think everything on this tape will pass
a V6 source license.

	To load this tape (approx 10000 blocks):
# cpio -id </dev/mt0 (800 bpi, 9 track, 512 byte records cpio format)

        The file "ped.c" in xasm/c directory has some code stolen
from  the  V6  UNIX  ed.c,  and maybe should not be passed out to
"binary-only" licenses (without ed.c src).   Everything  else  in
xasm  should  be  ok  for a binary license.  Users to receive the
"src" directory should probably have a V6 source license, however
there  are  some new (Purdue written) files in there which can go
do binary licenses.  I  am  not  sure  which  ones  are  mods  to
original  Bell  files  and  which were created here, I will leave
that up to you.  Isn't all this new licensing a mess!!!!!.  (also
probably ok to give out adventure source)

        The APL directory is probably ok for binary  licenses  on
up.

        The  "sys"  directory  definiately  needs  a  v6   source
license,  except maybe the "syslog" file may be able to be passed
out to binaries??

        "usr" is ok for binary-only as well as "usr2"  (adventure
support files).

        About our System Industries disks,  SI  is  scheduled  to
come  out for a week or two anytime to test new controllers (both
unibus and massbus 9400 controllers on 11/70 and 11/45).  We  are
the  first 11/70 to run on a 9400 controller with 300 MB CDC 9766
drives. The drives have been ROCK SOLID, much superior  to  DEC's
RP04's.   Our current controller configuration runs now, but both
us and SI are not sure why, there are many Purdue custom mods  to
the  controller  (unibus),  but in the next week or two SI and us
should  find  out  why.   Currently  a  standard   "stock"   9400
controller will not run on our configuration due to bus protocol,
etc.  I should also mention that the SI  company  has  bent  over
backwards working with us solving problems and giving assistance.
I have never seen anybody so helpful.  I feel we  are  far  ahead
despite the current problems, than if we went with DEC or another
company.  I will send you a followup letter in 3 weeks or  so  (I
hope)  after  the  new  unibus  interfaces and massbus (cachebus)
interfaces are up and running.

        Bill Croft (bc) would like you to print a 2 page  summary
(enclosed) about Purdue's network.  I have also enclosed the full
copy for your use. (see file src/mx/paper) for  nroff  input.   I
have  also  enclosed  a letter to be printed containing bug fixes
for buffer pool problems and exec sys call problems.

        The file "sibugs" in this directory probably shouldn't be
passed  out  in  general,  without  Purdue's  permission,  I have
included it for your use.  The information in it is only accurate
for  around  the first of 1979 and I am sure many of the problems
will disapear in the near future.  If anybody wants specific info
about  SI  performance,  have them call me so I can give them the
current status of things.   I  don't  want  to  damage  anybody's
reputatio caused by out of date info circulating around.

	The file "crash_exec.out" is a Fortran program binary
compiled via CULC's F4P/linker and will crash v6 system if exec'd
due to large text size. Can be used to test exec fix.

If you have any questions, feel free to give me a  call.
--Geo. Goble
