On Mon, 12 Aug 2019 17:59:45 +0000, Seymour J Metz wrote:

>> It should be intuitive: add blanks by positioning the cursor at the end of 
>> the
>> line and pressing the spacebar.
>
>Be careful what you ask for; you might get it. At which point IBM will get 
>mobs armed with pitchforks and torches.
> 
Future shock.  But nowadays most of them would be more used to the
similar behavior of the echoplex terminals on their desktops.

>What you are asking for might make sense on an echoplex terminal; it doesn't 
>make sense on a 3270 terminal or TN3270 session.
>
Why not?  The host should transmit characters counted by the RDW, then NULs to 
the end
of the field.  Characters typed over those NULs would be returned to the host 
on Read
Modified; the NULs ignored.  Very close to the behavior of the echoplex 
terminal.

>OTOH, it might be viable to enhance the 3270 data stream to include a 
>collapsing blank (significant null) that would be deleted by an insert but 
>otherwise treated as a space.
>
Already, some competitors' terminals and emulators make trailing blanks
collapsible and/or convert NULs followed by non-null characters to blanks
on Read Modified.  I wanted a pitchfork and torch.  (But I believe the
behavior was optional.)

And some silly history:

Circa MVS/370, Data Management ref. stated that the minimum count
supported in an RDW was 5, and SPF Edit obligingly padded any empty
line with a single blank.  Nowadays, z/OS Using Data Sets states the
minimum is 4.  ISPF Edit continues to pad empty lines.  I can assume
only that this supports users who need to Edit on z/OS and run on
MVS/370.

And any 8-character line was padded to 9 with a blank.  I know no
explanation for this;  I don't know that the behavior persists.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to