Xref: utzoo comp.unix.questions:25083 comp.unix.ultrix:4365
Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!uunet!aplcen!haven!mimsy!chris
From: chris@mimsy.umd.edu (Chris Torek)
Newsgroups: comp.unix.questions,comp.unix.ultrix
Subject: Re: EBCDIC to ASCII conversion under ULTRIX
Message-ID: <26264@mimsy.umd.edu>
Date: 29 Aug 90 14:17:14 GMT
References: <90240.091342HASKINS@MAINE.BITNET> <1694@quando.quantum.de>
Organization: U of Maryland, Dept. of Computer Science, Coll. Pk., MD 20742
Lines: 25

>In article <90240.091342HASKINS@MAINE.BITNET> HASKINS@MAINE.BITNET
 (Robert D. Haskins) asks for EBCDIC => ASCII conversion.

In article <1694@quando.quantum.de> omerzu@quando.quantum.de
 (Thomas Omerzu) suggests using `dd':
>	conv=ascii    converts EBCDIC to ASCII

Unfortunately, while there is only one ASCII (or two if you count the old
non-standard version with arrows where [\]_ appear in modern ASCII), there
are very many different (and incompatible) EBCDICs.  Take any two pieces
of IBM hardware that support more than the `basic' EBCDIC set (upper &
lower case, digits, and a few special symbols) and they will not use the
same table.  Square brackets on one will turn into hats on another.%
Backslashes will vanish.  Some IBM equipment can even display characters
it cannot generate!  (I know of no instances of the reverse, which is not
to say that they do not exist.)

So, the first problem with EBCDIC to anything conversion is to figure out
exactly which EBCDIC is being used. . . .
-----
% Actually, I am not sure what usually maps to what, but the characters []
  {} \ and ^ are the biggest trouble spots.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 405 2750)
Domain:	chris@cs.umd.edu	Path:	uunet!mimsy!chris