Dear David,
[quote user="David Harris"]Looking at the code, there is definitely message-id generation going on in Pegasus Mail,[/quote]
That's correct (but not the point in discussion about Mercury/32 and creating MIDs) ...
[quote]although for some reason it's controlled by the setting of an internal system flag, and I can't see clearly what the default value of that flag would be. There also doesn't seem to be anywhere in the UI to change it, although it *appears* that it should always be turned on by default.[/quote]
... if you want Pegasus to create a MID. For some reasons we don't Pegasus want to do it (see other posting in same thread).
The default value of "general system flags" is 0. Via UI you may only change it to 1 by activating session logging. I suppose default is 0, because otherwise SYSLOG.PM will increase endless.
[quote]Suggest to him (not as a solution, just as a curiosity) that he look in his PMAIL.INI file and find the entry called "General system flags": it can have the values 0, 1, 2, or 3. If it's 3, tell him to set it to 1. If it's 2, tell him to set it to 0. This will tell the program to generate message IDs. On my system, it has the value "1", which turns on system message logging. If it was a value other than 1 or 0, could you let me know?[/quote]
I know all about that flag and it was 2 - as wanted!!! But that's not the point in that thread and I think, there is no need for you to change anything on that in the code of Pegasus. Works fine and as designed! The changes should be done in Mercury/32 for the case, that no MID is in a mail from a local user to a local user.
[quote]Should definitely be working, though.[/quote]
It does ... as designed by you. [:D]
@Thomas
[quote user="Thomas R. Stephenson"]I've set the flags to 1 and it's creating a message-id for mail to local users when mail is sent via MercuryS using the built-in mailer. Here's the raw view as received.[/quote]
I know ... but we don't want Pegasus to create a MID and some mailclients (even on smartphones) used here don't do it too. Mercury/NLM, which I used until some weeks ago, did create MIDs for the problematic case (local to local), but Mercury/32 doesn't do that any more. That's my point and I hope David is willing to change that for next version of Mercury (obviously 4.74).
Cheers
Olaf