Could you file a bug report for this issue and attach your patch?
You have got to be kidding.
A bug that causes the device to crash. A one line fix that obviously does nothing other than convert a crash (kernel oops) to a nop, in a module that is only supposed to save a little power, so the nop results in only a small increase in power draw in some very specialized situations. Nokia's reply, after two months, is "fixed in harmattan".
Nokia's reply, after two months, is "fixed in harmattan".
If you do file a bug, that'd be the likely outcome but if you don't they can just say "no one told us". It's more about sending the ball in their court than anything else.
Oh well, it's not like the patch is even needed unless one compiles the netfilter module into the kernel, and in that case one can patch the module too.
titan already included that patch to his kernel-maemo packages and anyone else compiling custom kernels should be smart enough to use the search.
guys is there any progress on getting this into the repos. I see that the joikuspot app is now only a paid app. I'd prefer to support the community apps :-)
guys is there any progress on getting this into the repos. I see that the joikuspot app is now only a paid app. I'd prefer to support the community apps :-)
Haven't really thought about it yet, however there is a new thread and a real possibility of doing the NAT in userland.
Using custom kernel has a bit of a chicken vs egg problem with the kernel flashing that is difficult to handle sanely via package dependencies. Titans kernel-maemo is in extras-testing and should be usable with the mobilehotspot package (we agreed on a scheme of Provides-info so any kernel providing the required features will just use that and hotspot will work).
In theory I could just upload the current version to extras-devel but I'm not planning on promoting it to testing in a while (maybe after the userland NAT stuff works).
I was noticing that, on executing the backend, dnsmasq was complaining about being unable to bind the interface, yet manually executing the same command afterward would succeed.
After popping a small delay in just before the execution of the command in backend.py, it started working for me. Has anyone else run into this issue? rambo, thoughts?
I was noticing that, on executing the backend, dnsmasq was complaining about being unable to bind the interface, yet manually executing the same command afterward would succeed.
After popping a small delay in just before the execution of the command in backend.py, it started working for me. Has anyone else run into this issue? rambo, thoughts?
I think I saw something like this at some point but have no clear recollection. In theory when the IcD connection request has returned the gprs0 interface should be fully up but I guess there is no harm in adding a small delay there anyway.
my n900 was overclocked with litho's kernel. after i set up mobile hotspot with the projects kernel OC is gone. what should i do.
reflash with litho's kernel ?? will it affect the hotspot application
Thank you for this.
Installed from devel and it's working great. So far no hicups.
Oh I had to install ipconfig from apt-get,
look it up using apt-cache search ipconfig
I vote to put this up in testing!
Hey guy great work with that app!
But i have the problem that my ipod touch isn't able to connect to the internet. My netbook and Nokia n95 works fine but Safari (ipod touch browser) connects and connects and connects and nothings happening!
Also the other ipod applicaytions can't get a network connection!
