Re: Faux Code

[holmgrn@xxxxxxxx]

> From: "Carl L. Distefano" <70154.3452@xxxxxxxx>
> To: "XyWrite Discussion Group" 
> Subject: Faux Code

> I want to stress that the caveat on manual editing is meant to be much
> stronger than "it's always better to double-check anything
> automatically handled...". In the nature of this exercise, it's
> guaranteed hands-down inEVitable that the output *will* contain errors!
> You could make the "conversion table" a mile long and still not begin
> to cover the possibilities. So checking the output is not optional
> here; it's absolutely mandatory in every case! I'd hate to lull anyone
> into complacency about that. The best we can hope for is reasonably
> accurate output most of the time. I do think FAUXCODE achieves that.

Carl, I haven't seen your PM yet, but I wonder whether the
potential problems don't outweigh the benefits. It seems to me
that in communicating XPL to another person or computer, one
wants to eliminate as many opportunities for misunderstanding as
possible.
If someone who didn't write the XPL, and very likely doesn't
understand it, is going to have to edit it manually to get it to
work, and moreover may be unable to fix it if it malfunctions,
then I think the question is whether it has value. I can think
of many exotic, but nowadays quite commonplace, strings that are
going to be darn difficult to fauxcode & transmit in the clear
over a 7-bit data link (assuming that's the purpose). So: why
not leave the XPL as-is, and just ship it with XXencode or MIME?
All you lose is the opportunity to eyeball crippled code in your
mail reader. Whereas you gain the real thing!


=======
Robert J. Holmgren holmgrn@xxxxxxxx
=======