ml>> SOT/EOT
ml>> is not necessary... it is only someone's (yes i know who) way of
ml>> determining
ml>> the start and end of the user text. i see no problems with the current
ml>> methods other than lazy programmers who can't work up intelligent code
ml>> to
ml>> determine it one way or the other.

PE> Read the rationale and you will find out the answer to that. ESPECIALLY in
PE> your case, where you don't have a tearline.  It's technically IMPOSSIBLE 
PE> to
PE> determine the end of user-text, no matter how unlazy you are.  It is
PE> possible to take a GUESS, that's all.

ml> what is wrong with the origin line? =IT= is required...

You're completely missing (my/the) point.  If I get the following message from you:

*   EXEC SQL DELETE FROM tab1;
*   --- Whatever, you do, don't, repeat don't
*   --- comment out the above line, or the next job will screw up
*   SEENBY: etc


I want to determine which bit is user-text (in this case the user-text
appears to be an SQL program, which incidentally uses "--" as a
comment), and which bit is control lines generated in order to comply with
FTS-4. How do I do it?  Is that "--- comment out the above line"
etc a tearline (in which case I want to hide it, for the same reason I want
SEENBY lines hidden), or is it part of the SQL program?  What algorithm do
you have that gets it right for your program, and also gets it right for a
message from me, and 99.99% of the rest of fidonet?  It is TECHNICALLY
IMPOSSIBLE.  But it's the most fundamental REQUIREMENT of a mail system.
Just typical of fidonet ot make a complete dogs-breakfast out of it. BFN. 
Paul.
@EOT:

---