Still uses multiuser argument to update-rc.d

The nis package does not use debhelper and therefore does not call dh_installinit at all, let alone with deprecated arguments.

I'd be interested to know how this got marked as confirmed...

Hi,

Sorry for the confusion, I just re-used the text from another package
that I filed this against.

nis does still use "multiuser", and so still needs to fix this, as the extension
is deprecated, and will not be around for ever.

Sorry for the confusion.

Thanks,

James

Hrm, OK. Please also delete the request to forward any patches to Debian (I don't have sufficient access to do so) - the change to use multiuser is Ubuntu-specific and the current Debian package uses "defaults" as is standard and recommended behavior for Debian packages so a patch submitted to Debian wouldn't help.

Hi,

I'm happy to remove that request if you do not want these changes,
it will save us all time.

However, Ubuntu no longer uses multiuser, mainly because it was never
accepted in to Debian. The new scheme is compatible with Debian,
and I have been discussing it on debian-devel with no real objections
to the changes being submitted, with the maintainers of each package
allowed to use their judgement for their package. You can see
this discussion starting at

  http://lists.debian.org/debian-devel/2008/07/msg00005.html

As I said, it is at the package maintainer's discretion whether to accept
that patch, so if you do not wish to have you can say so now.

I added that line to these bug reports, as I am keen to get these changes
back to Debian, as one of the reasons for doing it is that we can potentially
get rid of the delta. I have notified debian-devel of a MBF for the change,
so the next step is to get all of the patches sent back. I wanted anybody
that fixes any of these bugs to be aware of that so that I don't have to do
another check to pick up any changes that weren't forwarded.

Thanks,

James

Here is the debdiff fixing also the FTBFS https://bugs.launchpad.net/ubuntu/+source/nis/+bug/259138 because of NM 0.7.
As there is no patch system (there are just some deprecated .patch files in debian/ which seems not to be used today) after talking to james, I directly patch the source file (cf changelog)

I don't attach it to upstream as requested by Mark (the debian maintainer of this package) on IRC who will directly deal with it.cd

  * Replace "multiuser" deprecated init mode in debian/postinst with manual
    arguments. Adapt the corresponding change in LSB init script header and
    runtime dependency in debian/control (LP: #254262)
  * Fix FTBFS because of NM 0.7 in intrepid editing ypbind-mt-1.20.1/src/
    ypbind_dbus_nm.c (LP: #259138)

Hi Didier,

I'm not sure that your fix for the NM problem is correct. I assume
that "StateChange" is what the constant is defined as pre-0.7?

In that case doesn't it mean that the if would never be true, as
NM 0.7 wouldn't generate it. Perhaps this would be harmless,
but it's worth examining.

Thanks,

James

Hi James,

As I didn't knew exactly how to test it directly (I just installed the NIS package on my intrepid VM), I relied on the pidgin developpers tests (look at the links provided in https://bugs.launchpad.net/ubuntu/+source/nis/+bug/259138)

From http://www.nabble.com/Re:-pidgin:-fe68f3ce:-A-patch-from-QuLogic-to-replace-the-libn...-td17164891.html :
"Apparently we can just replace it with "StateChange". They're still
emitting that signal for compatibility reasons, even though it's been
renamed to StateChanged (as QuLogic noted)."

If you have some advice for me to take a test on it, I can do it, obviously.

Taking for sponsoring.

This bug was fixed in the package nis - 3.17-14ubuntu2

---------------
nis (3.17-14ubuntu2) intrepid; urgency=low

  * Replace "multiuser" deprecated init mode in debian/postinst with manual
    arguments. Adapt the corresponding change in LSB init script header and
    runtime dependency in debian/control (LP: #254262)
  * Fix FTBFS because of NM 0.7 in intrepid editing ypbind-mt-1.20.1/src/
    ypbind_dbus_nm.c (LP: #259138)

 -- Didier Roche <email address hidden> Mon, 18 Aug 2008 22:03:20 +0200