Le Thu, 16 Dec 2010 20:01:46 +0100,
Dhaval Giani a écrit :
> On Thu, Dec 16, 2010 at 7:53 PM, Dominique Michel
What is the point to install a kernel feature without the corresponding
user space tools ? So, this is a kernel issue at the first place. And
this must be documented into the kernel help.
What is the point to install the user tools when its default
configuration will break existing applications ? So, the second place
to document it is into the user space tools, so that both casual users
and distribution makers can make the right choice and configuration.
> An application which uses
As project administrator, the documentation is one of my major
preoccupation. If anyone can understand it, I am happy. If not, I make
the needed changes it myself.
Many rt aware software was existing long ago cgroup and libcgroup. So,
b) must be addressed at the first place : the cgroup kernel help
(for advanced casual users) and into licgroup documentation.
And also, distribution makers must be aware of their choices : by
enabling new features that can break existing ones, they must
understand that they have a hard work to do : make sure than their
packages will work with each other, or they will be loosing users.
Most of the good applications cgroup and libcgroup are breaking has
been into linux for years.
> and if it were such a big deal, there would have been tons of bug
Wrong, users are not fools, and before reporting bugs, they done some
searching, or ask to some forum. Why do you thing they will issue
bug reports when it is only some good working software that is not
working anymore because of a miss-configured third party application ?
> Also, this problem is going to come up again once systemd
Why do you want to force third party applications to adapt their code
to your software when it is possible to 1) get ride of your software,
2) configure your software in a why the user can use its favorite
application without problem ?
Or do you gave a secret agenda where you want to make 1) impossible for
every one and 2) impossible for normal users ?
In one hand you say that it is easy to configure the user tolls, in the
other hand, it is no documentation than a casual user can understand
and follow. It is not a conspiration theory but a prod by the act: you
are not willing to make your piece of software affordable for the
casual user, and it will make life harder for them, as well than for
the distribution makers.
I agree with you, but the first change to make are to the documentation
of cgroup and its user tools. Again, why do you want to force to make
changes to their code when this is not needed.
Do I need a special list ? the LAPA Linux Audio Project Administrator ?
Why do you not explain what will be the benefit for a casual linux
rt user of cgroup and libcgroup ? If you can do so and I find that it
can be useful for the applications I am managing, I can and certainly
will change my mind.
As project administrator, the project documentation is one of my main
concern. If I huge than anyone can understand it, I am happy. If not, I
make the needed changes myself.
"We have the heroes we deserve."
Linux-audio-dev mailing list