Am 16.03.2018 um 15:20 schrieb David Crayford:
On 16/03/2018 5:16 PM, Itschak Mugzach wrote:
The purpose of RDZ was to save expensive TSO cycles and overhead.
That's interesting! RDz spawns a UNIX process for each connection and
the server is written in Java which makes extensive JNI calls. I would
suggest that RDz is significantly more expensive WRT CPU cycles than
ISPF.
Same for us. When we were in the RDz migration project
(which was cancelled in the end), we sometimes had severe
LPAR CPU consumption issues during RDz builds ... with our
"old" TSO ISPF based system, we never had such issues.
The motivation for the RDz migration was:
a) some problems of isolation of source codes which have been
checked out from other developers (this has been fixed partially
in the meantime, and it can be fixed totally by moving all sources
to the DB2 repository and making the checkout datasets private
and protected)
b) the hope that younger developers can be moved to mainframe
development by a more "modern" IDE (but they aren't interested,
anyway ... they simply don't want to learn PL/1 and such things,
which they consider hard).
BTW: I don't use the ISPF editor for everyday's work; I have all
my projects in PC or server directories and do the editing there
using KEDIT or THE (the Hessling Editor), and I send the files to
the mainframe only for compilation. I have also all tools for
searching and comparing at hand that I need, in best quality.
This is the best environment I can imagine. For my private
Stanford Pascal compiler project I use a GitHub repository.
Many of my co-workers do the same. So for us there is no
need to have a more "modern" IDE. For me and my co-workers
it would have been a pain to be forced to use "RDz only",
especially if there are no powerful import facilities.
Kind regards
Bernd
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
