Bug#265866: lintian does not look in the scripts called during postinst to check the existance of an "update-rc"
reassign 265866 eagle-usb-utils
thanks
On Sun, Aug 15, 2004 at 02:41:22PM +0200, mcoolive@free.fr wrote:
> When I launch lintian, I have the following errors :
> $ lintian eagle-usb-utils_1.9.8+cvs20040806-1_i386.deb
> E: eagle-usb-utils: postrm-contains-additional-updaterc.d-calls
> /etc/init.d/eagle-usb
This is because lintian didn't find the installation of the init.d
script
> W: eagle-usb-utils: script-in-etc-init.d-not-registered-via-update-rc.d
> /etc/init.d/eagle-usb
>
> I think it is an insufficiency of lintian.
>
> - /etc/init.d/eagle-usb is installed by the package.
> - The postrm script execute always the "update-rc.d -f eagle-usb remove"
> command.
> - And the postinst script can execute the "update-rc.d eagle-usb start"
> command, but indirectly. Configuration is long and complicated. So I use a
> script named "eagleconfig" that is included in package: the postinst script
> is a "front end ++".
This is a non-standard setup -- note, I don't say it's faulty.
Lintian indeed cannot really be made to check for this case, please use
lintian overrides to teach lintian that you know what you're doing, and
it isn't an oversight.
> To mute lintian, I have added the following lines in postinst script.
>
> if false; then
> update-rc.d eagle-usb start 99 2 3 4 5 . stop 01 0 1 6 . >/dev/null 2>&1
> fi
>
> But I don't believe it's satisfying.
Indeed, overrides are much more cleaner solution.
> Maybe lintian must look the scripts of the packet that are called, but I think
> it is too much complicated and/or too long.
> Maybe a good solution for futur is tags in comments to help lintian...
Known as 'overrides' :)
--Jeroen
--
Jeroen van Wolffelaar
Jeroen@wolffelaar.nl (also for Jabber & MSN; ICQ: 33944357)
http://Jeroen.A-Eskwadraat.nl
Reply to: