You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I'm writing a long-lived SMTP server daemon (https://github.com/Ekleog/yuubind if you're interested, most information is in the mdbook).
As such, I start the daemon with openat on the queue directory. And then, I do all of the handling relative to this openat, so as to avoid any mounting changes disrupting the operation.
However, walkdir works only by taking in a full Path. Do you think it'd make sense to provide an API that takes in an openat::Dir (from the openat crate) -- or just a RawFd -- and walks the directory starting from this openat?
I see that #23 mentions openat (though more as a bugfix than as an additional feature), so I guess this would probably be blocked on #23 being solved so that the basic infrastructure of openat be present.
How would you feel about such a change?
The text was updated successfully, but these errors were encountered:
Yes, I consider this a dupe of #23. The API isn't too much of a concern. Building the infrastructure to support it is. I still haven't made up my mind on whether I will continue down that path. See #120 for more details.
Hello,
I'm writing a long-lived SMTP server daemon (https://github.com/Ekleog/yuubind if you're interested, most information is in the mdbook).
As such, I start the daemon with
openat
on the queue directory. And then, I do all of the handling relative to thisopenat
, so as to avoid any mounting changes disrupting the operation.However,
walkdir
works only by taking in a fullPath
. Do you think it'd make sense to provide an API that takes in anopenat::Dir
(from theopenat
crate) -- or just aRawFd
-- and walks the directory starting from thisopenat
?I see that #23 mentions openat (though more as a bugfix than as an additional feature), so I guess this would probably be blocked on #23 being solved so that the basic infrastructure of
openat
be present.How would you feel about such a change?
The text was updated successfully, but these errors were encountered: