Re: lynx-dev lynx2.8.5dev.4 (EBCDIC multibyte)

[pg]

In a recent note, Thomas Dickey said:

> Date: Thu, 8 Nov 2001 16:45:36 -0500
> 
> yes - I thought about your change to simply disable multibyte characters,
> and it seemed that the real problem is that lynx was incorrectly deciding
> that your data was some CJK flavor - we should fix that.  But I also read
> up on the multibyte stuff and as far as I can tell, none would have
> embedded nulls, so I chose that solution.
> 
Probably a good idea in even the ASCII case, since it may prevent a
SIGSEV on invalid data which spoofs CJK.

But I suspect EBCDIC is highly likely to spoof CJK, and I haven't
the resources to do the coding to support multibyte characters
on OS/390.  So I'm resubmitting my patch to disable multibyte
processing if NOT_ASCII since this may affect far more than
LYLowerCase.  I've tamed the logic slightly, if that makes it more
palatable.  :-)

Thanks,
gil
-- 
StorageTek
INFORMATION made POWERFUL

lynx.patch.multi
Description: ascii text