Bug#399970: successfull use of asterisk + misdn

[Simon Richter]

Hi,

Victor Seva wrote:

> now I have a version [0] that builds over 2.6.23-1. I've made changes from your
> 1.1.6-1 version diff attached.

I have explicitly avoided adding the mISDN init system to the debian 
package, as it is bedly designed and if packaged according to Debian 
policy, will have to be supported for a very long time (this is the main 
thing I want to see fixed before I think mISDN is ready for a stable 
release: you can get the kernel to crash by loading modules in the wrong 
order, and requiring an XML based init system to work around that is not 
a solution.

There are a few spots I disagree with (replacing ${source:Version} with 
${Source-Version} for example), and I'm not sure explicitly passing 
EXTRA_CFLAGS and the CONFIG_* variables on the command line instead of 
in the environment does much to keep the logs readable. I'm going to 
steal the 2.6.23 patch (the SVN version of the package already uses 
dpatch, yay), but I disagree with the device ownership hack (you cannot 
do anything with just read permissions anyway, so revoking the o+r 
doesn't do anything, and I do not believe that the "dialout" group 
should have low-level ISDN access).

> I've not tested the 2.6.23 modules. I have not a testing machine. Can you take a
> look over this changes?

Not easily at the moment, because I have the main testing box used for 
Windows driver testing at the moment.

    Simon