[comp.os.vms] Type-ahead while doing single character input via SET HOST

KARNEY%PPC.MFENET@NMFECC.ARPA (08/05/87)

I have observed strange behavior with type-ahead when doing single
character terminal input via SET HOST.  Two examples:

(A) I have a file FOO.TXT which contains 100 lines of which the first two
are:

line 00   line 00   line 00   line 00   line 00   line 00     
line 01   line 01   line 01   line 01   line 01   line 01     

I SET HOST to a VAX, run Carosso's pseudo TTY test program PTY.PAS (the
release of about a month ago), and execute
    $ type foo.txt
FOO.TXT appears unscathed on my terminal.

If however I type ahead while the file is being typed, the lines get
reordered!!!  Here is the order they came in when I typed a single letter
'a' on about line 10.
    00 01 02 03 04 05 06 07 08 13 09 14 10 15 11 16 12 17 26 18
    27 19 28 20 29 21 30 22 31 23 32 24 33 25 34 39 35 40 36 41
    37 42 38 43 52 44 53 45 54 46 55 47 56 48 57 49 58 50 59 51
    60 65 61 66 62 67 63 68 64 69 78 70 79 71 80 72 81 73 82 74
    83 75 84 76 85 77 86 91 87 92 88 93 89 94 90 95 96 97 98 99

Note, all the lines appeared, but in a jumbled order.  This was observed at
both 9600 baud and 2400 baud.  FOO.TXT is stream_lf.  The original SET HOST
is necessary to get this behavior.  If you connect directly to the machine
running PTY everything works OK.  I believe it is dependent on PTY doing
single character input.  Asbed Bedrossian's PHOTO program which does
multi-character input functions correctly.

(B) I have noticed a similar behavior in GnuEmacs via SET HOST.  If you
type ahead to it, while it is redrawing the whole screen, it displays lines
alternately in the top and bottom halves of the screen.  The final display
is correct, presumably because each line begins with a cursor positioning
command.  I used to think this was some quirk of Emacs.  Now it looks like
it's some DEC problem.

The crucial ingredients appear to be:
  (1) SET HOST
  (2) Single character QIO reads on the terminal
  (3) Typing ahead

Has anyone else noticed this?  Is there an explanation?  Is there a fix?

    Charles Karney
    Plasma Physics Laboratory   Phone:   609/683-2607
    Princeton University        MFENet:  Karney@PPC.MFENET
    PO Box 451                  ARPANet: Karney%PPC.MFENET@NMFECC.ARPA
    Princeton, NJ 08544-0451    Bitnet:  Karney%PPC.MFENET@ANLVMS.BITNET
CC:
        INFO-VAX@SRI-KL.ARPA
        KVC%ENGVAX@OBERON.USC.EDU
        ASBED@OBERON.USC.EDUubje64
    eno69