RE: Additional thoughts on text transfers

[Damien Miller <djm%mindrot.org@localhost>]

On Mon, 1 Apr 2002, Jeffrey Altman wrote:

> > On Mon, 1 Apr 2002, Richard Whalen wrote:
> > 
> > > Text files are still in high usage, even with new (often proprietary)
> > > formats that may not be converted.
> > 
> > How/why can filexfer be expected to know about all these?
> 
> The Kermit Project implemented in C-Kermit 8.0 algorithms for
> examining files that can determine if the file is 7-bit text, 8-bit
> text, UTF-8, or UCS2/UCS4 with various byte orderings.  These
> algorithms are highly accurate and do not rely on any proprietary
> knowledge of file naming conventions or operating system attributes.
> If the file is text it will be transfered as such, if it is not it is
> transfered as binary.  
>
> The only text files that cannot be transfered with conversions are
> those that include mixtures of text and binary.  These files must be
> transfered as binary as we do not translate only portions of files.
> I'm not aware of any other tool that allows automated mid-stream
> translation.

That kinda proves the point that I was trying to make - that the format
decision and processing can be built into the clients (or servers) and not
into the protocol itself.

filexfer's beauty is its simplicity as a binary-only file transfer 
protocol. IMO it should be burdened with application level cruft that 
can be better handled in the application.

-d