Hi,

 AG> Which echos are you talking about? And those using CP850 should use
 AG> LATIN1 instead (it's nearly the same as CP850)...tell them correct
 AG> their systems... 
Eg. LINUX.GER. There you can find such malformed messages. Indeed I used
this as a test echo. Regarding your reasoning: Why are there still Windows
users? The charset they use depends on their system. Eg. DOS or OS/2 users
(IIRC) won't have LATIN-1 but CP850. If their message editor doesn't allow
recoding they have no choice, which is no problem at all. Their is no rule
saying that you have to use LATIN-1.

 AG> I prefer correctness over userfriendliness...I know this saying
 AG> about being liberal in what you expect, but then it has to be a
 AG> runtime setting not a configuration verb which can be immitated by
 AG> a modified readmaps.dat ...
Your solution:
 - Pros:
    * correct
 - Cons:
    * inflexible because it's hardwired
    * "bloating" maps unnecessarily because the CP437-entry must be duped

My solution:
 - Pros:
    * correct (by it's default setting, which can be initialised in the code)
    * flexible (it allows the user to cope with the stupid programs of other users.)
    * slim since it's uses only some bytes of memory of the process.
 - Cons:
    hmm... tell me one.

Where is the problem here? Also the "incorrectness" does not pose
any problems since it never leaves the computer of any user. It's just for
reading echomails locally. (in contrast to "IBMPC" echomails with
wrong encoding)

The difference between "incorrectness" and
"flexibility" lies in exactly this point: to the outer world
timed behaves correct, but internally it is flexible enough to cope with
the incorrectness of other programs.
 
 AG> we should rather think about including the maps source...
The maps are taken from MsgEd. We can take the source from there.

Ciao
Meikel

--- timEd/FreeBSD 1.11.a5