[rfc-i] What the text version is used for (was Re: The <tt> train wreck)

Julian Reschke julian.reschke at gmx.de
Mon Aug 23 05:35:01 PDT 2021


Am 23.08.2021 um 13:59 schrieb tom petch:
> ...

OK, I'll bite.

> I was going to pick up on this but the thread seemed to move on, so
> thank you for returning to it.  txt is simple, straightforward, easy to
> use, universal.
>
> txt is compact; not as compact as Shannon would allow but I have seen
> pdf in the IETF that are some 50 times the size of the txt version. Yes,
> computers get bigger and faster but not that much.

Example? (The only reasons I can think of are either scanned documents,
or documents with lots of graphics not present in the TXT version).

> txt can be processed on every device I can ever recall having used, give
> or take the ASCII/EBCDIC conversion and some awareness of code pages (my
> dollar is your cent) and that applies to display, to editing and to
> printing.  It is simple, supported everywhere, can be used for interchange.

Well, depends on "processing". How well does your smart phone "process"
plain text formatted for printing on A4/letter?

> It encourages people to be clear in their language.  I have seen
> diagrams lately of the kind
> I---I     I---I
> I   I --->I   I
> I---I     I---I

That is a problem, but the example looks exactly like it was taken from
plain text :-).

> with no explanation of the meaning of the arrow and, when I sought
> clarification, it transpired that the meaning was the exact opposite of
> what I had assumed!  And this is a general issue, the richer the
> language that is used in a specification, the more comprehension is
> assumed in the reader. At a slight tangent, the logic of AND and OR are
> widely understood, NAND or NEQ less so even if they are arguably simpler.
> ...

Best regards, Julian


More information about the rfc-interest mailing list