On 25-jun-2006, at 22:41, Stewart Bryant wrote:
As an example, this .gif extracted from the Y.1711 OAM protocol
would be quite difficult in ASCII.
I'm not surprised, as it contains too much information to be readable
in a 925 pixel wide GIF. I think this supports Stephen's point that
if a diagram can't be expressed as ASCII art it's too complex.
On 25-jun-2006, at 23:05, Lyndon Nerenberg wrote:
However the text in that GIF is unreadable as rendered in my mail
client (MacOS Mail.app). When viewed with Preview and the Gimp,
the background turns gray with white boxes behind the (still
unreadable) text.
Go blame Apple if they can't even decode a decades old format. (But I
don't think there is a problem there, the change in background color
is probably due to a transparent background and the text is rendered
accurately as it appears in the file: unreadable.)
This demonstrates why I was arguing for a vector format, and why I
am adamantly opposed to a pre-rendered bitmap format, for diagrams.
Vector wouldn't be any better as you still need to render it to a
bitmap at some point. That works if you have a 2000 pixel wide
display, but guess what, not everyone does. With GIF we can at least
all agree that it's unreadable, with vector we'd all have to spend
extra CPU cycles and then some would still see garbage but others
would claim everything is fine and we'd have a discussion on
reasonable screen resolutions on this list every couple of months.
Let's please leave that particular disaster to the web designers.
Stewart again:
It would take a lot of words
to describe, which many people would then have to transcribe to
some sort of timing diagram - which then may or may not be
correct.
Human communciation is done with words. Sure, sometimes images help,
but the only way a diagram can convey a lot of information with
enough precision is to use a highly structured modelling language,
which we'd then have to assume every reader of the RFC in question
understands. Images really aren't as useful as you think.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf