COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	April 1978 state of LOTS
C00007 ENDMK
C⊗;
April 1978 state of LOTS


The initial hardware goals have finally been met.  Terman and CERAS
have their full complements of equipment.

LOTS has met its obligation to take over the teaching except for
370 software dependent courses.

LOTS has taken a considerable but so far unquantified part of the
unsponsored research load.  It is even less certain that these
resources have been used with the priorities that the faculty
would prefer.

LOTS has been rather freely available to students without formalities,
but the initial expectation that this would be very valuable has
not been realized.  Even the student members of the LOTS Advisory
Board have cheerfully gone along with greater formality in this
direction.

LOTS is generally preferred to SCIP from cards even with the
overcrowding.  Comparison with the use of WYLBUR seems to be
made almost entirely on the basis of speed of service, and at present
WYLBUR is less loaded than LOTS.

The WYLBUR $2.00 per hour night rate has been popular, but it should
be remembered that LOTS average costs are less than $1.00 per hour.

The four man staff has proved adequate.  I have heard no complaints
about inadequacy of the staff, although some complaints about
slowness in installing planned servces could be interpreted as such.

The options include $50K conversion to 2060, and about $125K additional
to replace some of the memory by MOS and increase memory by 256K.  This
would give the users 60 percent of the machine instead of 50 percent
and the machiee would be twenty percent faster.  The overall improvement
has not yet been fully estimated.

The $50K 2060 conversion is required to remain in the front line
of D.E.C. software improvements.  I favor it, because if we remain
in that front line, we are in a good position to influence D.E.C.
software as the leading university installation, and we have
more to gain by influencing D.E.C. than by putting our own money
and time into software.

The machine has not been adequate to make possible the hoped for
breakthrough to mass use of computer text editing for students
and faculty.  This was a long shot anyway.

One reasonable strategy is to make minor improvements and save
money for the next generation machine - which won't be for at
least three years.  There is nothing wrong with LOTS that five
times the overall speed won't cure, but that can't be obtained
cheaply at present.

D.E.C. would like us to pioneer 2060 dual processor, but it isn't
clear that they will make this an attractive proposition financially.

Another option is to make a text editing and document production
machine.  This would be a bold step into the inevitable future
way of producing documents.

Another 2060 would make a qualitative improvement at somewhat less
than linear cost per unit performance.

Most likely, other schools and departments will follow the example
of the Business School in the near future.  This is not to be
deplored.