∂20-FEB-75  0900		network site RAND
 JOHN,
 CAN IVAN SUTHERLAND AND I SPEND SOME TIME WITH
 YOU NEXT TUESDAY AFTERNOON TO DISCUSS ISSUES
 WITH REGARD TO THE COMPUTING PROCUREMENT COMMITTEE.
 WE CAN MEET YOU AT THE AI LAB OR THE UNIVERSITY.
 REGARDS,
 PETER WEINER

∂19-FEB-75  1939		DOC,TOB
 I heard that you are about to hire someone
 for some purpose, and I would like to express
 my interest that hand/eye be able to replace
 some of the four people who have been lost
 in the last year; I would like to see those
 places kept available.  The lab is now almost
 entirely a theoretical lab.  The three areas
 MTC,etc		Hand/Eye	Natural language
 JMC		TOB		TW
 CCG		PDQ
 DCL
 MG
 WG
 ZM
 RWW
 
 JRA
 AC
 RF
 

∂19-FEB-75  1547		1,BES
 THERE IS A NEW INSTRUCTION THAT CAN BE TYPED IN YELLOW[WAVE].  THE VICE
 CAN NOW BE OPERATED WITH THE TWO INSTRUCTIONS:
 	VOPEN time  or VCLOSE time
 WHERE time IS THE DURATION THAT THE VICE OPERATION IS TO TAKE ( IN 
 SECONDS ).  THE FORM OF INSTRUCTION "DO VCLOSE TIME" OR "DO VOPEN TIME"
 DOES NOT WORK.
␈ CC: @HAL[HAL,HE]:TOB,RCB,RF,PAUL%SRI,RHT%SAIL,[HAL,HE],JMC,JH,BO,BES,VDS,DDS

∂18-FEB-75  1841		1,DEW
 I have rewritten my patterns using LOC.  They are on PRIM.FOL[1,DEW]
 if you care to see them.  I'll later remove some of my comments about
 pc not working so well and mention additions of conditionals, etc.
 I agree that pattern recognition should not calculate occupiability.
 I hope patterns will describe all the concepts that will be passed
 around in the tree search and used by the goal setting mechanisms.
 We might want a goal of "make Q5 occupiable by a knight" so these 
 functions will also be part of the language.  What will be a pattern
 and what will be a function will probably be mostly intuitive rather
 than well-defined.  I'll think abouta way for patterns to interact and
 then come and talk to you about it.   Dave
 p.s.  I dropped a Psych course I was taking and now only have 4 credits of
 courses so I signed up for 5 credits of r+r.  I'll also talk to you about this.

∂18-FEB-75  1041		R,AJT
 thanks for your note. It is not entirely clear to me how talk about
 computers &c is relevant to the problems that Chapter discusses. There is
 no a priori reason to assume that information processing language can
 solve EVERY problem under the face of the sun. I believe that there are
 `old fashioned' philosophy problems that are still interesting, even if they
 have been raked over quite a bit. I shall be glad to chat about it, of course!
 arthur.
 p.s. Hintikka is out of town this week - how about THURSDAY of NEXT week for
 committee meeting?

∂17-FEB-75  1827		BPM,BPM
 TO: SAIL blood donors
 
 FROM: BPM
 
 RE: Upcoming donation
 
 The SAIL Blood Donor Group is scheduled to give blood at the American
 Red Cross--Stanford University Blood Center  on Tuesday and Wednesday
 (25-26 February) of next week.  The Blood Center is open on both days
 from 9 am to 1 pm and again from 3 pm to 7 pm.  
 
 I need to know the following from  each person who said they would be
 willing to donate as part of the SAIL Blood Donor Group:
 
 (1) Whether or not you plan to give during our scheduled times (those
 listed above).   [If you don't,  you can go  on your own whenever  is
 most  convenient for  you.   Just call  ahead (493-1363)  to  make an
 appointment and tell them you are donating as part of the SAIL  Donor
 Group.  Also let me know if you donate.]
 
 (2) What your blood type is.  [Don't tell me anything unless  you are
 absolutely sure.]
 
 If you are planning  to give during the scheduled times,  I also need
 to know the following:
 
 (3) During which period you would like to donate (Tuesday am, Tuesday
 pm, Wednesday am, or Wednesday pm).
 
 (4) Whether or not you need transportation from the AI Lab and back. 
 
 If everyone who originally  signed up donates, we will  probably have
 only  two scheduled donation times  per year (the  minimum allowed by
 the Red Cross). 
 
 The Blood Center is located at 3330 Hillview Avenue, Palo Alto 94304,
 which is across from the VA Hospital.  The best way to get there from
 the AI  Lab is  to take  Page Mill to  Foothill, right  onto Foothill
 approximately  half a mile to Hillview,  left onto Hillview about two
 blocks, and left into the Blood Center parking lot. 


∂15-NOV-74  1440		ACT,REG
 FYI, Alan Kotok called today.  He reports that on a breadboard machine
 with a 25Mhz clock (as compared with 32Mhz intended for production)
 TOPS-10, with the same set of peripherals as a KI10 system, with a 
 "pessimal" cache flush algorithm, assembles the TOPS-10 sources 2.2
 times faster than the KI10.
 He quoted 92-93% cache-hit rate.

∂15-NOV-74  1434		001,MJH
 John:	In case you haven't already heard, I have accepted an offer 
 from Systems Control Inc, effective Dec. 1.		Jo.