[rfc-i] Character sets, was Comments on draft-iab-rfcformat

Ted Lemon mellon at fugue.com
Fri Dec 21 08:38:57 PST 2012


On Dec 21, 2012, at 11:21 AM, John R Levine <johnl at taugh.com> wrote:
> I toggled my share of programs in from the console switches, but I'm quite happy to limit my programming to devices with screens and keyboards these days.

I think it's a good reductio ad absurdum counter to Martin Rex' point, however.   Clearly it is not easier to read a document whose canonical representation is a sequence of RADIX-50, any more than it is easier to read a document whose canonical representation is some intricate binary desktop publishing format.   These are at opposite ends of the complexity spectrum, and yet both suck.

The point being that the question of whether there is some device that can't accurately present a document derived from the canonical format is a red herring.   Of course there is some such device, for any canonical format one might choose.   Since no canonical format, including 7-bit ASCII, could ever satisfy the requirement implied by this argument, we can safely ignore both the implied requirement and the argument.

However, there is a larger point which _is_ worth considering: that anyone who wants to read the document should be able to get a presentation, derived from the canonical format, which can be clearly understood and contains no meaningful errors introduced by the translation from the canonical format to the presentation format on some device that the reader possesses.

If Martin doesn't want to upgrade from his 3270 terminal (a truly wonderful and innovative piece of technology in its time, I might add!), presumably he can get someone to print out a typeset version of the document for him to read on paper.



More information about the rfc-interest mailing list