Bugzilla Bug 270039

Every one of the packages listed below needs to be tested (see definition of
testing), and amended in one of two ways:
- DOES work with libusb-compat:
  1. Change your [R]DEPEND to =virtual/libusb-0* OR virtual/libusb:0. The slot
dep requires EAPI-1 or newer. The former works with EAPI0. (See note #1).

- DOES NOT work with libusb-compat, or you don't have time to fully test right
  now:
  1. Change your [R]DEPEND to dev-libs/libusb:0 (preserve any existing version
     dependency as well). You use at least EAPI1 for this.
  2. If it really doesn't work, leave a comment in the ebuild as well as on
     this bug.

When you are changing the package, please change ALL ebuilds, including the old
stable ebuilds, so that downgrades of the package are not computed to avoid the
libusb migration.

Note #1:
========
virtual/libusb-0 is set to bring in the current stable of dev-libs/libusb-0,
0.1.12-r1. If your package needs something in the 0 series newer than that, you
MUST use the slot dep variation.

How testing is defined in this case:
====================================
No recompile is needed in switching between libusb-0 and libusb-compat-0, that
was the entire point of the compat library. However by testing:
- You must actually USE the USB functionality of the package
OR
- carefully review it's code with regards to the libusb-compat limitations.

Of the limitations in libusb-compat (per the README):
- usb_device's "num_children" attribute is hardcoded to zero, and the
"children" pointer is NULL. Anything that tries to use these will thus fail.
- Running a transfer in it's own thread and simply killing the thread rather
than properly stopping the transfer will kill ALL libusb I/O in the program.
This is esp prevalent in multimedia code with isochronous USB transactions, but
can turn up in some bulk transactions as well.
- Proper error checking of opening devices, libusb-compat fails earlier when
there are insufficient permissions.

Packages (I've fixed :
app-accessibility/brltty
app-accessibility/gok
app-misc/acdctl
app-misc/digitemp
app-misc/ifp-line
app-misc/lirc
app-misc/rioutil
app-mobilephone/bitpim
app-mobilephone/gammu
app-mobilephone/moto4lin
app-mobilephone/obex-data-server
app-mobilephone/openmoko-dfu-util
app-pda/barry
app-pda/coldsync
app-pda/pilot-link
app-text/calibre
dev-embedded/avrdude
dev-embedded/ftdi_eeprom
dev-embedded/libftdi
dev-embedded/openocd
dev-embedded/pk2cmd
dev-libs/libhid
dev-libs/luise-bin
dev-libs/openobex
dev-libs/serdisplib
dev-util/usb-robot
media-gfx/gphoto2
media-gfx/iscan
media-gfx/sane-backends
media-libs/hamlib
media-libs/libdjconsole
media-libs/libifp
media-libs/libkarma
media-libs/libmtp
media-libs/libnjb
media-libs/libptp2
media-sound/ardour
media-tv/linuxtv-dvb-apps
media-video/isight-firmware-tools
net-dialup/umtsmon
net-misc/dahdi-tools
net-misc/zaptel
net-print/hplip
net-print/mtink
net-wireless/bluez
net-wireless/bluez-utils
sci-geosciences/gpsbabel
sci-geosciences/qlandkarte
sci-geosciences/qlandkartegt-garmindev
sci-libs/indilib
sci-libs/libticables2
sys-apps/usb_modeswitch
sys-auth/thinkfinger
sys-fs/owfs
sys-power/sispmctl
x11-misc/ifpgui
xfce-extra/xfce4-cellmodem

bah, there was a minor typo.
- Packages (I've fixed :
+ Packages (I've fixed 15 packages beyond this list already)

net-print/hplip is fixed.

For my packages please open bugs on a tracker rather than a multi-assigned bug.
Thanks.

app-misc/acdctl
app-misc/rioutil
dev-util/usb-robot

are maintainer-needed packages. I'm going to open new bugs for them.

net-misc/dahdi-tools
net-misc/zaptel

have been tracked. Removing voip alias.

By the way, rajiv and chainsaw, I think you can remove yourself from this bug
as those packages are probably why you were in CC.

x11-misc/ifpgui done.

app-accessibility/brltty and app-accessibility/gok are done.

robbat2 removing embedded@gentoo from the list cuz I don't see any packages
which are directly maintained by embedded. Also when you mass add people like
this. Please make a list of what pkg belongs to what maintainers. tia

solar: herd embedded is specified by dev-embedded/pk2cmd in my list of broken
packages.

app-pda/pilot-link is now fixed.

Done:
media-libs/hamlib
xfce-extra/xfce4-cellmodem

for whichever pkg I am listed here, anyone is free to fix it

Using your definition of testing, I am unable to test sane-backends. I can't
let all USB related code run (no hardware) and reviewing all this code is
beyond what I can currently do. Suggestions?

Patrick,

(In reply to comment #13)
> Using your definition of testing, I am unable to test sane-backends. I can't
> let all USB related code run (no hardware) and reviewing all this code is
> beyond what I can currently do. Suggestions?

Then I suggest you do the same as I did for hplip. Change [R]DEPEND to
dev-libs/libusb:0 (you need EAPI1 but it's normally harmless and no revbump is
necessary) and add a note saying your name, date and that you couldn't test
with libusb-compat.

As soon as I get my officejet back I can test that and update the ebuild if
you're OK with it, but that may be a few months away.

Denis.

(In reply to comment #14)

> Then I suggest you do the same as I did for hplip. Change [R]DEPEND to
> dev-libs/libusb:0 (you need EAPI1 but it's normally harmless and no revbump is
> necessary) and add a note saying your name, date and that you couldn't test
> with libusb-compat.

Done as suggested, thanks for the help.

*** Bug 273705 has been marked as a duplicate of this bug. ***

I've fixed packages belonging to mobile-phone and net-dialup herds:
app-mobilephone/bitpim
app-mobilephone/gammu
app-mobilephone/moto4lin
app-mobilephone/obex-data-server
app-mobilephone/openmoko-dfu-util
dev-libs/openobex
net-dialup/umtsmon

I really don't like being spammed by jumbo bugs such as this. IMO a tracker
bound to individual bugs is the way to go.

media-video/isight-firmware-tools done

if there is something left for us, please fill individual bugs. TIA.

(In reply to comment #19)
> if there is something left for us, please fill individual bugs. TIA.
> 

+1

fixed dev-libs/libhid