Re: [gnu elpa] Adding disk.el
Stéphane Marks <[email protected]>
| Newsgroups | gmane.emacs.devel |
|---|---|
| Message-ID | <CAN+1HbqV8+7wNeAgEqVrPjkVck_p3R3i8QfAUuksCrUxoLS2xQ@mail.gmail.com> |
On Thu, Jul 16, 2026 at 3:48 PM Mario Rosell <[email protected]> wrote: > Hello. > > A few weeks ago I sent an Emacs patch for adding =disk.el= to the > upstream Emacs dist, since it didn't get accepted, I am now submitting > it as part of GNU Elpa. > > disk.el provides an indent-tabs-mode interface to block devices, plus a > couple commands to manipulate them. > > Repo is on SourceHut: https://git.sr.ht/~rosell/disk.el > > Since this is my first Elpa package, please tell me if I did something > wrong... > > Thanks in advance. > The comments you received in the original note about adding disk.el to Emacs proper still stand. Once it's in ELPA, assuming it gets there, it will still be a namespace and discoverability challenge as is. To amplify the concern even further beyond what I earlier said: Part of the challenge is that "disk" doesn't mean disk like it used to. A block device can be virtual and point to just about anything that implements the API including network devices and "FUSE" devices that aren't disks. Choosing a name that best matches the package's actual intent is better. To me, calling it "disk.el" suggests it should work on most platforms where people use Emacs like MS-Windows, macOS, FreeBSD, NetBSD, etc. So in addition to it being a misnomer as disk doesn't always mean disk (though it could be called block device or something similar), it should at least try to work on at least one other platform and present a thoughtfully uniform API under which other implementations can live. Otherwise, I'd call it "linux-block-device-mounter.el" or something like that to be clear it has no intention of being more (for example, I didn't see a way to format a block device just view/mount/unmount).