To read all comments associated with this story, please click here.
From the article:
It was made clear that there is still some work to do and that the driver is not ready for inclusion - yet. However, it was also made clear that it is not too far away, as realistic aims kernels 2.6.23 or 2.6.24 were mentioned.
The mm tree is kind of a testbed for new developments, that will probably find its way to the mainline kernel, so that driver developers can accommodate to them and more bugfixing/testing can be done.
Porting open source drivers over to the new environment until this stack hits the mainline should be a quite straight forward process (giving, that this development has been announced for quite some time and - again citing the article - most newer developments have focused already on the new stack) and wrt binary-loop drivers: Well, nobody ever guaranteed binary compability, right?
Since you can't have two network stacks at once, then we will have a lot a drivers for the old network stack that can't be used with the new one, right? So what's the point of this?
Which old stack? There is more than one other one out there. I guess you're talking about the in-tree stack, but I think most people now have wireless cards that don't use a driver that uses the in-tree stack. There are a lot of different wireless implementations floating around right now and by the time devicescape is in the mainline kernel it will probably support everything that that the in-tree stack supports plus a lot of previously out-of-tree drivers.




Member since:
2006-06-24
Since you can't have two network stacks at once, then we will have a lot a drivers for the old network stack that can't be used with the new one, right? So what's the point of this?