[rfc-i] diagram issues

"Martin J. Dürst" duerst at it.aoyama.ac.jp
Sat Jun 28 01:28:57 PDT 2014


On 2014/06/26 03:56, George, Wes wrote:

> WG] conversely, your premise assumes that we even *need* to deal with the
> limitations of ASCII art. This is a question of the right tool for the
> job. I believe that the burden of proof here isn't "why ASCII isn't
> enough" but rather "why we should still be limited to ASCII in 2014". I
> still haven't heard any legitimate technical justification for that. And I
> don't think that an example will be useful, because the legibility of a
> diagram rendered in ASCII vs SVG is often subjective, and so there's not
> much point in trying to debate it.

There's one aspect of ASCII art that makes it quite illegible in places 
such as Japan, and that's the fact that many printers/browsers/fonts 
replace the ASCII 0x5C backslash with a Yen sign glyph. That's not too 
much of a hassle e.g. in string constants in programs, because any glyph 
is about as good as any other as an escape character. But it's terrible 
for ASCII art.

As an example (from RFC 6782),
                                        +--------+         -----
                                        |        |       /       \
                        Encap IPv6 Flow |  6rd   |      |  IPv6   |
                                 - - -> | Relay  | <- > |   Net   |
           +---------+         /        |        |       \       /
           |         |        /         +--------+         -----
           |   6rd   + <-----                              -----
           |         |                                   /       \
           |  Client |         IPv4 Flow                |  IPv4   |
           |         + < - - - - - - - - - - - - - - -> |   Net   |
           |         |                                   \       /
           +---------+                                     -----

looks reasonably okay (assuming you're not afflicted by the above 
problem), but
                                        +--------+         -----
                                        |        |       /       ¥
                        Encap IPv6 Flow |  6rd   |      |  IPv6   |
                                 - - -> | Relay  | <- > |   Net   |
           +---------+         /        |        |       ¥       /
           |         |        /         +--------+         -----
           |   6rd   + <-----                              -----
           |         |                                   /       ¥
           |  Client |         IPv4 Flow                |  IPv4   |
           |         + < - - - - - - - - - - - - - - -> |   Net   |
           |         |                                   ¥       /
           +---------+                                     -----
starts to use ugly, and if there are more backslashes, it gets worse, up 
to illegible (I used U+00A5, which should show up as a Yen sign anywhere 
around the world). As far as I understand, similar things happen in Korea.

Regards,   Martin.

P.S.: For the record, I strongly dislike and disapprove of the practice 
of showing 0x5C as a Yen sign, but it's difficult to get away from it 
quickly.


More information about the rfc-interest mailing list