Xref: utzoo comp.binaries.ibm.pc.d:1782 news.admin:4600
Path: utzoo!censor!becker!bdb
From: bdb@becker.UUCP (Bruce Becker)
Newsgroups: comp.binaries.ibm.pc.d,news.admin
Subject: Re: Perils of IBM/Bitnet and other communications channels
Message-ID: <256@becker.UUCP>
Date: 30 Jan 89 17:45:22 GMT
References: <2686@looking.UUCP>
Reply-To: bdb@becker.UUCP (Bruce Becker)
Organization: G. T. S., Toronto, Ontario
Lines: 36

In article <2686@looking.UUCP> brad@looking.UUCP (Brad Templeton) writes:
>As a gift to the net, I have written a new improved binary encoder which
>is going to be released soon.
>
>It uses non-whitespace, printable characters only, but that's not enough.
>I am told that sometimes travelling through IBM machines (bitnet etc.)
>maps tilde ~ to ^.  Are there any other perils like this I should be
>wary of, before release?  Is the tilde problem real or isolated?
>
>Thanks.
>-- 
>Brad Templeton, Looking Glass Software Ltd.  --  Waterloo, Ontario 519/884-7473

	Try using the 64-byte charset ".,[0-9][a-z][A-Z]".
	Some countries use different representations for
	the values, but the translations are consistent.
	IBM mainframes mistranslate at least "!^[]{}~`\|"
	and different parts of their systems mistranslate
	them differently...
	The 64-byte charset doesn't have this problem -
	probably one could get a valid 85-byte charset
	(I understand "btoa" etc. needs this), but that's
	probably pushing it. Also I've noticed that some
	versions of "uuencode" put a "table" entry in the
	output which gives the table values used for
	mapping. This works as long as each character
	translation value is unique going from machine to
	the next - sadly, in the IBM world at least,
	this is not always true...

Cheers,
-- 
   _  _/\	Bruce Becker	Toronto, Ont.
   \`o O|	Internet: bdb@becker.UUCP, bruce@gpu.utcs.toronto.edu
    \(")/	BitNet:   BECKER@HUMBER.BITNET
---mm-U-mm---	"Moderation in pursuit of justice is no virtue" - Oliver North