Re[2]: [zd1211-devs] CVS driver works with kernel 2.6.12-rc3 on ibook

[Roman K <rom...@ma...>]

> On Saturday 30 of April 2005 01:17, Roman K wrote:
> 
> > > There is some hope in ng driver but there is a long way before it reaches
> > > nice (mergeable with mainline) state. Keeping support for 2.4 sucks :)
> > > Lack of
> >
> > Do you think that 2.4 shouldn't be supported at all?
> Yes (for ng driver), 2.4 is a dead end if you thing about mainline merge. 

I am not so sure. But probably at the day we will get first release of ng
driver you would be absolutely right. And probably it is the time to make
a decision. It is hardly depends of our feeling of how long for the first
release we will wait.

> Keeping things compatible, creating wrappers and at the end there is a mess. 

The new driver has two versions of each file. One for each branch of kernel.
There is no mess, only some small headache for developer to not forget to
merge new changes to both versions.

> As example - zd1201driver is already in -mm kernel. It's quite clean code. 
> zd1211 could follow its direction.
> 
> > > _publicly_ avilable docs sucks even more - try to convince zydas to
> > > release
> >
> > What kind of docs do you need?
> All zydas has about the chip? :) Recently there were two questions where docs 
> could help - one mine about firmwares, other about registers.

That is almost impossible :-( The only way to become a member of a project.
 
> > No one will stop you if you write a couple 
> > of them and contribute them to the project. More over, if you will continue
> > to do that work you may be punished by doc developer status :-)))
> (some) docs are avilable but only to few zd1211 developers. Others have to 
> make guesses based on available code. The best way to change this is to try 
> convince zydas to become more open. Hard task I guess.

As you may see some of the last posts even project developer (currently) do not
have all needed docs. I don't know any reason to keep such docs as a chip description
from public access. This information do not allow to produce the same kind of
chip. You cant just copy it as a for example a product based on it. Product need
to be protected not the elements that were used to produce it.


rik

> > rik
> 
> -- 
> Arkadiusz Mi kiewicz                    PLD/Linux Team
> http://www.t17.ds.pwr.wroc.pl/~misiek/  http://ftp.pld-linux.org/