Babyl Options:
Mail: DJ: SAZ; MAIL.TEXT#0
Version:5

0, recent,,
*** EOOH ***
Date: Monday, 7 November 1988, 14:15-EST
From: Keith Corbett <keith@GSI-CAM>
Subject: Errors from Initialize Timebase
To: cstacy@GSI-CAM, BUG-LISPM@GSI-CAM
In-reply-to: <8811060104.AA17573@gsi-cam.UUCP>
Message-ID: <[GSI-GLASS].7-Nov-88 14:15:53.keith>

    Message-ID: <8811060104.AA17573@gsi-cam.UUCP>
    Date: Saturday, 5 November 1988, 20:04-EST
    From: Christopher C. Stacy <cstacy@GSI-CAM>
    Subject: Errors from Initialize Timebase
    To: BUG-LISPM@GSI-CAM
	:
	:
    PS.  Is it really likely that the "update bit on clock chip seems to be stuck"
	 message in fact indicates a hardware problem with Boris Badinoff?
	 This happens every time I boot, and seems to involve some bit that is
	 supposed to be cleared within a second or so after frobbing the clock.
	 I haven't ever noticed the time on my machine being wrong.


Some SDU boards do indeed have the "update bit stuck" problem, and they
will exhibit it forever.  (I.e. until the clock chip is replaced.)  The time
gets sets correctly when the Lambda reads it over the net.



0, recent,,
*** EOOH ***
Date: Monday, 7 November 1988, 17:53-EST
From: Keith Corbett <keith@GSI-CAM>
Subject: Please fix your latest patch
To: saz@GSI-CAM
Message-ID: <[GSI-GLASS].7-Nov-88 17:53:11.keith>


  (DEFMETHOD (ZMAIL-WINDOW :AROUND :EDIT) (CONT MT ARGS &REST IGNORE)
    (SYS:%BIND (LOCF (TV:BLINKER-DESELECTED-VISIBILITY POINT-BLINKER))
	       :ON)
    (setf (fill-pointer (get 'ZMAIL-SUMMARY-MOUSE :WHO-LINE-DOCUMENTATION)) 0)
    (setf (get 'ZMAIL-SUMMARY-MOUSE :WHO-LINE-DOCUMENTATION)
	  (array-push-portion-extend
	    *edit-msg-documentation*
	    "Edit msg with normal Zwei commands; click mouse in summary window to save changes and jump to new msg."))
    (LET ((*SELECTABLE-MODE-LINE-ELEMENTS* NIL))
      (AROUND-METHOD-CONTINUE CONT MT ARGS)))))

This blows up -- the first setf, because the string doesn't have a
fill-pointer.  (Most do not.)  In fact, after commenting out the first
SETF, the second one gets a similar error - *EDIT-MSG-DOCUMENTATION*
also doesn't have a fill-pointer either, as it's a simple string.

(How did this ever work for you?)

And ARRAY-PUSH-EXTEND isn't what you want here; that extends the array in
place (i.e. it's destructive).  That means the array could get extended
every time.

A cosmetic suggestion: instead of "edit msg with normal Zwei commands;",
how about something like "Edit message in buffer, or ...".

  --Keith

P.S. I hope to be able to disk-save the new system 129 band tomorrow
(today, by the time you read this), and ZMail is the last system with
problems, this being one of the two problems.  So I'd appreciate your
quick attention to this.  (Blink blink.)



0, unseen, recent,,
*** EOOH ***
Date: Monday, 7 November 1988, 18:03-EST
From: Keith Corbett <keith@GSI-CAM>
Subject: My proposed fix
To: saz@GSI-CAM
Message-ID: <[GSI-GLASS].7-Nov-88 18:03:30.keith>


I think this does the "right thing":

  (DEFMETHOD (ZMAIL-WINDOW :AROUND :EDIT) (CONT MT ARGS &REST IGNORE)
    (SYS:%BIND (LOCF (TV:BLINKER-DESELECTED-VISIBILITY POINT-BLINKER))
	       :ON)
    (let-globally ((who-line-override-documentation-string
		     #.(string-append "Edit message in buffer; press "
				      #/end
				      " to save changes, or click in summary window to save and jump.")))
      (LET ((*SELECTABLE-MODE-LINE-ELEMENTS* NIL))
	(AROUND-METHOD-CONTINUE CONT MT ARGS))))

You should try it, and (re)patch.  (Unless, of course, you see something better.)