Nathaniel Borenstein <nsb(_at_)thumper(_dot_)bellcore(_dot_)com> writes:
Interesting. It is precisely the richtext generated by newer
versions of Andrew that people have found so unreadable. I'm
sending this message in that format as an example.
This is a red herring. Andrew's richtext is particularly unreadable
because whoever wrote that richtext composer made no particular effort
to generate readable output. I'm sure it can generate text/enriched
that is just as unreadable using the same exact technique of wrapping
lines mid-word using quoted-printable.
That's not to say "<nl>" isn't ugly.
I've discovered the main flaw in richtext's newline model. If you
convert every line break or sequence of line breaks to SPACE, it's not
possible to have both an encoded and decoded stream which ends with a
line break. The only simple fix (besides the identity or
text/enriched line models) is to follow the sample code and ignore
line breaks entirely, in which case you end up having to put a space
at the beginning or end of a line, where it respectively looks ugly or
gets stripped by mailers.
I'm ambivalent between "fixing" richtext and going with
text/enriched-minus-verbatim. I haven't spoken with Chris lately--I
don't know what he thinks.
Removing verbatim would certainly remove this objection as well,
though, wouldn't it?
Removing verbatim (possibly replacing it with <nofill>) would
certainly remove my objection.
I'd like to see Tony's patches to the metamail richtext program. I
have a sneaking suspicion I'll find at least one of the three bugs I
found in the sample text/enriched decoder.
Come on, counting newlines is not really all that hard.....
I agree it's not all that hard. The sample parser botches it, though.
The specification says to convert sequences of N newlines to N-1
newlines. The sample parser converts sequences of N newlines to a
SPACE and N-1 newlines.
<center>hello</center><flushright>goodbye</flushright>
At least in enriched, everybody should be treating it the same, i.e.
putting hello & goodbye on separate lines....
The sample parser doesn't put them on separate lines...
--
_.John G. Myers Internet: jgm+(_at_)CMU(_dot_)EDU
LoseNet: ...!seismo!ihnp4!wiscvm.wisc.edu!give!up