Le 22/03/2026 à 11:00, Martin Steigerwald a écrit :

> Greetings.
>
> Didier Kryn - 22.03.26, 10:44:31 CET:
> […]
>>       rmount, smount tmount don't seem to have any meaning, umount
>> already does the opposite. qmnt?
> Just "mnt"?
>
> Otherwise can you describe a bit that your mount command does differently
> or what makes it so special compared to other alternatives? I got it does
> not depend on lsblk and mount. Was that your original motivation to create
> it? Or what else is your ambition?

    The original ambition was to overcome the impossibility to mount
exfat with pmount.

    Then came the policy. The default policy of pmount is fine for the
usual desktop/laptop, but, in general, it makes no sense. The policy of
hopman, OTOH, was to trust the kernel in defining what is removable and
what isn't, but the kernel fails to do that correctly. For my program,
just have a look at qmount's policy as explained in README.md in devuangit.

    On the contrary to the other qmount, my own program doesn't use any
other command to do the job: instead of spawning lsblk and parsing its
output to read /sys/devices, like the other qmount, it does it
programmatically with its own functions.

    On the contrary to pmount and the other qmount, my program does not
execute mount; instead it invokes directly the mount() system call,
because it finds by itself all the necessary parameters which the mount
command obtains from libudev and that pmount doesn't know.

    To summarise, my qmount is simpler and faster than the two others,
it has a carefully defined policy, and it can mount exfat partitions,
which are very common these days.

>
> Knowing more about the purpose of the command may help o find a suitable
> name. The other qmount has the following description on GitHub:
>
> "quickly mount a device in Linux, written in golang"
    Same here. The name came from a discussion on this forum with q
meaning quick.
>
> How about "quickmount" then? Might be a bit long. "qm" on the other hand
> is quite short and might be ambiguous with other programs. Maybe
> "fastmount" or "fmount"?
    quickmount and fastmount are definitely too long IMHO.
> From that GitHub I get it that the other "qmount" mounts without
> destination path given, inside "/media".
    Same for my program: you can give the mountpoint, but, if you ommit
it, it defaults to the label of the filesystem, or the device name if
there is no label. Like for pmount, you can use any fstab syntax to
denote the device.
> I remember such a command from
> earlier times. It was called "pmount" and could also be used as an user.
> But when its about reducing command line arguments, it could also be
> "easymount" or "emount"? Or when it is about mounting as an user without
> "fstab" entry "usermount" however "umount" as a shortcut would not work
> then?
    Yes, usermount is too long and umount already exists and does the
opposite.

    pmount officially means "policy mount". See above for its policy.
My program isn't configurable because it doesn't need to, which I
consider a quality.

    qmnt would mean "quick mount" exactly like qmount...

--     Didier