Re: Latest CVS and a couple other tidbits



On Sat, 2005-12-24 at 08:45 -0500, Steev Klimaszewski wrote:
> Hey all, great work on all of the progress over the past 2 weeks.  I have been watching, and testing with great interest.  There are only a couple small things that I have noticed.
> 
> 1. nm-gconf-wso-private.h  not included with make dist
> 
> Not a huge problem, I just add it to the tarball manually (cp into gnome/applet)

I'll fix that.  Thanks.

> 2. It doesn't compile because of a test - If I simply remove the test directory, and all references to test, then it compiles and will install just fine, but I prefer not to push that out to the users - When I attempt to compile, I get the following...
> 
> make: *** No rule to make target `../../test/libtest-common.la', needed by `test-ciphers'.  Stop.
> 
> I suck at autotools, and have tried autoconf, automake, as well as running autogen over and over, and simply cannot get it to make libtest-common.la.  Not sure what I am doing wrong there, but any help/guidance is greatly appreciated.

My fault, my autotools skills aren't good enough to figure out how to
make the subdirectories post-processed rather than pre-processed
relative to the current directory (where libtest-common is).  I'll
probably move libtest-common to a subdir just to work around this.

> 3.  New madwifi drivers create a wifiX interface as well as athX.  This is similar to the aironet driver, for Cisco AiroNet cards, but NetworkManager sees wifiX as a wired interface, and (obviously) doesn't grab an IP - manually bringing the ath0 interface up, does not cause NetworkManager to notice it either.  Any ideas about that?  Or is it a driver issue itself?

Can you post the 'lshal' output sections for both wifiX and athX on your
system?  NM relies on HAL properties to figure out whether the device is
wireless or wired, so I'll need to see what HAL reports about both
devices.  AFAIK, the wifi0 on Cisco cards doesn't show up in HAL at all,
so NM ignores it.  So either the driver needs to get fixed, or HAL needs
to get modified to ignore the interface or not tag it as having "net"
capability.  The 'lshal' output should tell us what needs to happen.

Thanks!
Dan






[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]