COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\\M0BDR25\M1BDI25\M2NGR30\M3XMAS25\M4NGB25\.
C00005 ENDMK
C⊗;
\\M0BDR25;\M1BDI25;\M2NGR30;\M3XMAS25;\M4NGB25;\.
\F2\CARTIFICIAL INTELLIGENCE LABORATORY
\CCOMPUTER SCIENCE DEPARTMENT
\CSTANFORD UNIVERSITY
\CSTANFORD, CALIFORNIA 94305
\F0

Dear Steve:

	Here is the understanding I imagine us to have reached
concerning Martin Frost working on PUB and APE.

1. We will hire him as a systems programmer starting right away, and
you intend that we will be reimbursed for his salary + the usual
overhead by ARPA as part of some addition to our contract.

2. He will work half time on PUB and half on APE.

3. The improvements to PUB will adapt it better to using the XGP.  Les
Earnest, who supervised Tesler in writing the original PUB, will
supervise them.  However, these improvements will include making
PUB runnable under TENEX.  We will make it produce output under
TENEX acceptable to a reasonable XGP service program.  We hope the
TENEX users will be ale to agree on a single reasonable format
for files to be XGPed.

4. The work on APE will include making a year's worth of A.P. stories
accessible through the net.  We will explore the use of the data-computer
for this.  Specifically, we will use it or give a reason why its
facilities are not usable for this purpose.

5. We will make a TENEX program that allows using APE at Stanford
without logging in.  The TENEX APE will do all the communication
so that the TENEX user won't know APE isn't on his own machine.
We don't commit ourselves, however, to support more than a few
simultaneous APE users.

6. We are free to use the grant we hope to get from A.P. to make
additional improvements to APE as we see fit.


					Sincerely yours,



					John McCarthy
					Professor of Computer Science
					Director, Artificial Intelligence Laboratory