In a followup in MVS-OE:
Linkname: [EMAIL PROTECTED] post from [EMAIL PROTECTED] :
URL: http://www2.marist.edu/htbin/wlvtype?MVS-OE.42223
an IBM employee has suggested:
If you want autoconversion in ishell (oedit, obrowse, rexx,
or anything other than LE enabled C programs) use AUTOCVT(ON)
in your BPXPRMxx.
To which the OP (and others) replied that setting makes some
things better, but others much worse, particularly some Java
operations.
Since SMP/E v3r4.0 will support Java v.2 as an alternative to
ICSF, I wonder whether SMP/E has been tested with this
environmental setting, and with various autoconversion and
other settings that the user might specify in his ~/.profile
(which I suggested earlier that SMP/E shouldn't read)? How
about RECEIVE FROMNTS with SMPNTS NFS mounted, with various
default translation settings? (I do this regularly, but only
when I load the SMPNTS from the same host, so the translation
should be the same in both directions.) How about RECEIVE
FROMNETWORK with server=localhost, and various settings of
autoconversion in the FTP server process space? (Well, that's
properly an FTP problem, but SMP/E will likely get the first
problem reports in some cases.)
It's regrettable that an IBM representative has suggested
a global configuration change with possible pervasive
consequences for what should have been deemed a local problem.
It's much more regrettable that there is no local alternative
recourse.
Allowing EBCDIC in the HFS was a colossal blunder; the pain
is spreading. HFS should have been pure ASCII from the start.
Autoconversion is an abomination that attempts to solve the
attendant problems, but largely adds complexity, reduces
predictability, and increases testing multiplicity.
-- gil
--
StorageTek
INFORMATION made POWERFUL
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
