∂08-JUN-73  0202		D,LES
 Something stange has happened to PUB.  I have compiled our ARPA proposal
 several times without incident, the last time a couple of weeks ago.
 Tonight it blew.  For example, on Page 13 of ARPA[D,LES], the text response
 defined at the top of the page now produces the error message "already past
 tab column 5" on lines 3300/13, 4900/13, etc.  Also, the counter REF is
 counting in lower case roman instead of integers as specified.  There are
 other mysteries further on.
 
∂31-MAY-73  2329		S,FW
 Even the "*" in the directory feature is not quite as simple as it would seem.
 Since the feature as you describe it would require writing the directory much
 more often than is presently necessary,  thus reducing efficiency,  I would
 prefer to make anything like that conditional upon some mode indicator in the
 directory to avoid degrading performance in the normal case.  I have already
 thought some about having features like that,  but it would be considerably
 easier to wait until all TV features have been incorporated into E.  Until
 TV is completely obsolete,  I am somewhat constrained by having to keep the
 directory format TV-compatible.  b  By the way,  the "generalized SWitch
 feature" in TV has nothing to do with interfacing to other programs,  and
 although that feature is not yet implemented in E,  the feature I think you
 want is.  See the description in TV2E.FW[UP,DOC].
 
∂31-MAY-73  1125		FOO,DBA
 LAST MEETING OF CS320 OUTSIDE POLYA 3.45 MONDAY JUNE 4
 *** WARREN TEITELMAN WILL TALK ON FUTURE PLANS FOR BBN-LISP,
 AND ANSWER YOUR PROBING QUESTIONS!  (WOT NO DATA STRUCTURES ETC)