∂08-OCT-74  1620		HAL,RHT
 HAL mtg rescheduled for Thursday, 2:00 pm

∂07-OCT-74  1446		H,RF
 HAL meeting postponed until after meeting with Brewer Tues 1330.
 CC: @HAL:TOB,RCB,RF,LOU,RHT,[HAL,HE],JMC,JH,BO,BES

∂01-OCT-74  0814		HAL,HE
 HAL MEETING AT 2pm TODAY OCT 1

∂20-SEP-74  0923		D,LOU
 There will be a Hal meeting today Friday at 2pm in Lou's office.
 Agenda items include: schedule of regular meetings, current state of
 work, Jerry's comments of the document.
 				Lou.

∂11-SEP-74  0346		HAL,RHT
 Well I got SMALLB partially debugged:
 	The small block allocation & deallocation routines work
 	(IE   MOV #id,r0
 	      JSR pc,getblk
 	and
 	      MOV blk,r0
 	      JSR pc,freblk )
 I haven't really looked at debugging the garbage collector yet,
 but it should be pretty easy, too.  
 HAL.LST[HAL,HE] is a CREF listing of the program.  Take a look &
 see what you think.  I should be in in time for HE mtg.  See
 ya then.
 	Russ
 (ps,  if you use /C to palx, you can get a cref listing by saying
 	R PALX
 	FOO/C/M		;THE /M TURNS OFF MACRO EXPANSION IN LISTING
 	↑C
 	RU CREF[11,JBR]	;AN MIT SPECIAL
 	FOO.LST←FOO.CRF
 	↑C
 )

∂11-SEP-74  0006		HAL,RHT
 I recieved your messages.  Since I had not finishing writing
 SMALLB (to say nothing of debugging it) it isn't terribly surprizing
 that you ran accross a number of typos & undefined labels.
 Sorry about that.  My intent in calling it to your intention was 
 that you might want to look at the overall method.  I'm afraid
 that it may not have been very clear, judging by your comments.
 The reason you couldn't find a bit mask is that there wasn't one
 to be found.  Instead, I associate a 1 word tag with each small
 block (this is the time-space tradeoff I mentioned before).  Currently,
 this word is broken up into two bytes: a mark byte (could be made smaller,
 though a full byte is useful) & a "space id" byte (which is very
 useful both for debugging and for supplying garbage-collector intrinsics)
 In any event I dont think your suggestion about having an extra
 4-bit field for every pointer is a very good idea, for several reasons:
 (1) It will be an infinite source of bugs, since people will forget
 to update it. (2) it becomes much harder for routines to return
 pointers as values. (3) The space saving over a tag word is negligible,
 especially in those cases where you have several pointers to the same
 cell. (4) There still remains the problem of finding the bit mask
 containing the bit associated with the record.  
 
 I have gotten SMALLB pretty much written now & will try to debug the
 fixed allocation routines tonight. I think the design has a few advantages
 over a simple bit mask. These include:
 	(1)Each block is marked with what it is
 	(2)the same routines are usable both for collectable
 	   and uncollectable areas
 	(3)allocation of a record is faster
 	(4)fixed release is possible (though dangerous for a collected
 	   area, of course)
 	(5)The compactification scheme will readily generalize to
 	   include blocks that themselves contain pointers.
 Anyhow, take a look.  I may not have time to write up much documentation
 (especially on how the compactification works)  Perhaps we can go
 over that together, using a blackboard for all the spaghetti.
 
 On a slightly different matter.  When you changed my use of XX, you
 seemingly adopted a convention that is a bit different from the one
 I thought we were using.  Macros like XX are generally assumed to
 have only "local" uses (as with II & NNNN) and may be redefined at will.
 If you want, I'm perfectly happy to make XX mean what you say it
 does, but in that case perhaps we should choose some more meaningful
 name for it (eg INDEX, or some such).  
 
 Also, I made a grandmother file called HAL.PAL.  -- It might be 
 xxxxx 
 
 Well I guess thats about it for now.  See you ins morgen.
 Russ
 CC lou,he

∂29-AUG-74  1023		network site MAXC
 Date: 29 AUG 1974 1025-PDT
 From: TESLER at PARC-MAXC
 Subject: hal,he question about SLIDETOP1 in pub pass 2
 To:   he at SU-AI
 
 PUB is out of synch in its calculations
 concerning the centering or right flushing of a text line.
 Did the output look OK, esp. with regard to center and right flush
 lines?
 Were there any error messages earlier (pass 1) like:
 LINE TOO LONG, or anyhthing else??  
 I would like to get rid of bugs like that, so if you give me clues I
 will try to do so.
 -------

∂27-AUG-74  0821		OK,TES
 TO RUN OLD PUB, ALIAS PUB,TES AND THEN "RUN PUB".
 
 AS FOR THE PDL OV, DOESN'T IT SAY LINE/PAGE ... ... SO THAT
 YOU CAN SEE WHERE THE ERROR HAPPENED?  IF SO, THEN RESPOND
 ↑P
 AND YOU WILL BE ABLE TO INQUIRE ABOUT YOUR STATE.  READ
 22AUG.TES[UP,DOC].

∂15-AUG-74  2251		HAL,HE
 REMINDER THAT THERE IS ANOTHER HAL MEETING FRIDAY MORNING 9:30
 ... BRING THE LATEST COPIES OF YOUR WORK ... BOB
 CC: @HAL:JAF,RCB,RF,LOU,RHT,[HAL,HE],BBM,JMC,JH

∂14-AUG-74  2053		HAL,HE
 HAL raisers ... I have started to incorporate various files into
 	MEMO.DOC ... so far I have taken over 4R.DOC ... as I gobble
 	files, I'll mark them NOT TO BE UPDATED ... please make
 	corrections in MEMO.DOC which is now declared to be the
 	`original' ... I am intentionally leaving some of the files
 	separate (eg. 4RADD.DOC) so everyone won't be trying to edit
 	the same file at once. 
 	Check the `DETAILED OUTLINE' in MEMO.DOC to see what has
 	to be done yet ... feel free to comment on the outline.
 	... also DISCUSS.DOC is available for suggested topics for
 	Friday morning's meeting
 						Bob
 CC: @HAL:JAF,RCB,RF,LOU,RHT,[HAL,HE],BBM,JMC,JH

∂14-AUG-74  1328		HAL,HE
 MEETING FRIDAY TO DISCUSS THE PAPER.  9:30 AM.

∂09-AUG-74  1651		B,BBM
 Add ' to the list of characters used in the macroprocessor

∂09-AUG-74  1619		B,BBM
 I am currently using the following special characters for the scanner
 and macro-expander:
 ⊂ ⊃ $ % <Backspace>
 Please complain if you feel some of them will be needed for some other purpose.

∂06-AUG-74  1306		HAL,HE
 ALL MESSAGES DIRECTED TO [HAL,HE] SHOULD BE SAVED
 IN SAVED.MSG BY USE OF RCV AND THE T COMMAND.  THIS SHOULD
 BE DONE BY THE SENDER AFTER A REASONABLE PERIOD OF TIME.
 		RF
 CC: @HAL:JAF,RCB,RF,LOU,RHT,[HAL,HE],BBM,JMC,JH

∂06-AUG-74  1213		HAL,HE
 HAL raisers ... There are the following files on [HAL,HE] which
 you should be aware of, read, and make additions to:
 	DISCUS.DOC --- the list of topics to be discussed
 	MEMO.DOC --- the file where `finished' portions of the
 		     memo should be placed ... under the correct
 		     heading
 	XMPLES.DOC --- a file containing longer `complete' examples
 		       which hopefully will cover most of the features
 		       of HAL ... employed on "real" problems ...
 		       this is to keep the USER's coding style in mind
 Reminders:  Please try to tutor everyone on you latest syntax/semantics
 	    The first drafts should be in by Friday
 	    Add you references to the bibliography in MEMO.DOC
 	    SIMILARLY add any problems which we have discussed, but
 		put off to later generations
 						... BOB
 CC: @HAL:JAF,RCB,RF,LOU,RHT,[HAL,HE],BBM,JMC,JH

∂01-AUG-74  1022		HAL,HE
 Just a reminder that your first drafts to be included in the HAL memo
 are due Wednesday (Aug. 7) !!!    To motivate (&/∨ publically embarrass)
 the slower writers I suggest we have a meeting Tuesday at 9:30 AM ...
 (complaints about the time to RCB) ... WRITE ON
 CC: @HAL:JAF,RCB,RF,LOU,RHT,[HAL,HE],BBM,JMC