Coloring of replies

Jochen Bern bern@it-services.lu
Wed, 21 May 2003 21:54:23 +0200


laurent.duperval@microcell.ca wrote:
> Pan has a feature I find interesting: different repli levels
> are color coded. [...]  How can this be added to tkrat? Is it
> desirable?

It's much more suggestive than usual indentation - which is good
when everything goes A-OK, and bad else (e.g., someone sloppily
editing replies and leaving lines with too many or too few '>'s,
or people using indentation strings other than '>' or '> ').
Choose your poison. :-)

MUAs that reformat the quoted part, be it by just splitting:

| > > > > > This text started out fitting into 72 columns just
| > fine,
| > > but
| > > > > > the lines got reformatted to 72 cols by adding line
| > wraps
| > > > (and
| > > > > > 1st-level indentation prefixes immediately following
| > them)
| > > twice
| > > > > > while being re-re-re-...-quoted.

or (worse) by recreating the line breaks completely from scratch:

| > > > > > Again, this text originally fit well into 72 cols,
| > but the > > > > last MUA doing quotation saw fit to remove
| > all existing line > > > > breaks and create new ones on its
| > own, firmly embedding (its > > > > foot into its mouth and)
| > the old indentation prefixes into the > > > > middle of the
| > resulting lines. As soon as the prefixes "run > > > > off"
| > the end of the line, like, for example, right here+now: > >
| > > > , you get other apparent indentation levels than 1 and
| > the > > > > correct one, too. Also if *several* MUAs did this
| > %&*#!!! thing > > > > to the text.

will threaten to give you a headache. %-} Both seen occuring
"in the wild".

(I suppose that algorithms having a guess at "good" alternative
text colors, given a limited number of text and background colors
already in use, are floating around on the 'net someplace ...)

Please do NOT, however, have the differing color *replace* the
indentation - that'ld downright break "ASCII art", including, e.g.,

| > I'm trying to connect to port 123/tcp on IP 47.11.15
| Try with an octet "08" inserted here --------------^

Regards,
								J. Bern