Re: single-bit errors

[Greg A. Woods]

[ On Thursday, November 23, 2000 at 14:15:09 (-0500), Rich Salz wrote: ]
> Subject: Re: single-bit errors
>
> Hunh?  The only place I noticed hashing was in auth challenge.
> 
> I couldn't see anywhere that SSH puts an error-correction code on top of the
> TCP stream.

The de-compression will normally fail if there's any error in the data
-- not quite as quickly as the 

Furthermore since the compression happens first, on the raw data, and
then the encryption happens before the data is encapsulated in TCP, any
error in the TCP stream that's not caught by TCP's basic 16-bit CRC will
corrupt the encrypted data.  De-encrypting will then amplify the error,
which of course will almost never allow it to make it through the
de-compression.

> The CRC could let corruption happen, but I betcha all sorts of things break. 
> Everyone assumes TCP doesn't mess up the bytes in-transit.

More things break quicker with SSH and thus even a tiny error (perhaps
one caused by a flaw in some networking device) that's not caught by the
CRC will immediately cause an error, and a retransmission, before CVS
ever sees the data.

The only problem is that SSH's retransmissions, especially in version-1
of the protocol, are far from ideal and often the connection will go
catatonic if there's much corruption or loss on the link.

-- 
                                                        Greg A. Woods

+1 416 218-0098      VE3TCP      <address@hidden>      <robohack!woods>
Planix, Inc. <address@hidden>; Secrets of the Weird <address@hidden>