I use a local IMAP Server (dovecot) as my mail archive. I’ve used <F1> in Mail app to export some mail folders to EF and in parallel used the mutt mail client to export mbox files from the IMAP archive which I dropped into FE.
Oddly enough neither the size of the resulting mbox files nor the number of exported mails are equal in all cases:
I’m not sure why the sizes would be different. I guess you’d have to compare the mbox files and see what the specific differences are. As to why the message counts are different, this may be because EagleFiler skips messages that are marked as deleted. I don’t know whether mutt does that. I suppose it’s possible that Mail’s deleted flag (in the .emlx files) and the deleted flag on the server aren’t in sync.
Interestingly, I’ve been tracking the same problem. I decided to try importing fewer and fewer messages, until I found some that were not imported at all. (I hit “F1” on a given message for my mailbox, and it create an empty EF mailbox.)
I looked at the console to see if there was any message, but there was nothing.
Please drag the mailbox from Mail to the Finder, create a Zip archive of it, and send it to eaglefiler@c-command.com along with a description of which message you had selected.
The problem is that the source of these messages is in a different format that is confusing EagleFiler’s parser. The first line of the message begins with “>From” rather than being a valid header line. I’m not sure why that is. Did you import these messages from another program, rather than downloading them with Mail? Anyway, EagleFiler is storing the messages in the mbox file that it puts in the library, but it’s then unable to properly display the contents of the mailbox. I will fix this in the next version of EagleFiler.
Ah, I see. I might have used formail and procmail to convert a mbox into maildir. But I’m not sure it’s the issue here.
I could fix it on my side (running a script that removes these extra first “From” lines), but “strict in what you produce and lax in what you accept” is often a good software design principle
(I’ve just found that I might not have sent this message, sorry if it is sent twice.)