Dear all, why is Mutt even reading and encoding attachments prior actual sending?
Not to say its wrong (other MUAs do the same with inline images and the like), just wondering. Other thing, why is mutt looking for atime instead of mtime? 'touch' or systemwide scripts (security scans, backups, whatever) shall modify only the former one. But I suppose portability is the keyword here... Regards, Martin V 20260603 0251, David Haguenauer napsal(a): > * Kevin J. McCarthy <[email protected]>, 2026-06-03 09:07:13 Wed: > > On Wed, Jun 03, 2026 at 01:17:18AM +0200, Vincent Lefevre wrote: > > > Attachment #1 modified. Update encoding for [...] > > Can you duplicate the issue? > > I've seen this message before, in circumstances where it was true that > a file had been modified after it had been attached, and before the > e-mail had been sent. > > I just reproduced this with: > > 1. `touch foo`; > > 2. within Mutt: attach `foo`; > > 3. `touch foo`; > > 4. within Mutt: send. > > → "Attachment #2 modified. Update encoding for /tmp/foo? ([yes]/no): " > > > But Vincent's case is different; it is about the "attachment" that > constitutes the body of the message. If I go and use `touch` on that > file, I get the "Attachment #1 modified" prompt when attempting to > send. > > Perhaps some background process on his system touched the file that > backs the attachment just before sending? This is probably difficult > to investigate after the fact. > > -- > David Haguenauer
signature.asc
Description: PGP signature
