RFP: ipp-usb -- Daemon for IPP over USB printer support
Package: wnpp
Severity: wishlist
* Package name : ipp-usb
Version : 0.9.3+34.1
Upstream Author : Alexander Pevzner <pzz@apevzner.com>
* URL : https://github.com/alexpevzner/ipp-usb
* License : BSD 2-Clause "Simplified" License
Programming Lang: Go
Description : Daemon for IPP-over-USB printer support
ipp-usb is a userland driver for USB devices (printers, scanners, MFDs),
supporting the IPP-over-USB protocol. It enables these USB devices to
be seen as regular network printers.
.
It is designed to be a replacement for the ippusbxd daemon, previously
used for this purpose. It has a greatly rethought architecture in
comparison with ippusbxd, and fixes all of its major flaws and issues.
Dear Prospective Packager,
Bug #909564 has been attended to in avahi 0.8-1 and Avahi now handles
the loopback device lo for local services. Therefore, local-only
IPP-over-USB printers now get advertised. Thank you to Trent Lloyd and
Till Kamppeter.
By all accounts, ipp-usb is a better ippusbxd. Please see slide #12 at
https://ftp.pwg.org/pub/pwg/liaison/openprinting/presentations/cups-filters-ippusbxd-2020.pdf
In
https://github.com/alexpevzner/sane-airscan/issues/29#issuecomment-629628416
Alexander Pevzner says:
ippusbxd has serious problems, and these problems are by design, not
by implementation. All these problems are fixed in ipp-usb, it uses
very different design, so it works absolutely reliable (this is Till's
words, not mine).
It seems to me that Chrome OS not accepting software in Go hasn't any
relevance when packaging for Debian. In any case, Alexander Pevzner
would dispute the claim:
Frankly speaking, I don't quite understand ChromeOS reasoning. I've
recently measured memory consumption of ipp-usb vs ippusbxd. To my
surprise, RSS (Resident Set Size, i.e. amount of memory really used)
of ipp-usb is even slightly less. If there are more that 1 devices
connected, ipp-usb will handle them all in a single process, adding
a small amount of memory per device, while ippusbxd requires a process
per device.
The remainder of his argument is at
https://github.com/alexpevzner/sane-airscan/issues/29#issuecomment-630438806
It would be good to see ipp-usb enter unstable alongside ippusbxd.
Regards,
Brian.
Reply to: