|
Authored by: Anonymous on Saturday, September 15 2012 @ 08:18 PM EDT |
"Clover Trail?" Is that like the "primrose path"?
(I use the phrase in the Shakespearean sense, though said
usage is disavowed by Wikipedia.)
cpeterson[ Reply to This | Parent | # ]
|
|
Authored by: mbouckaert on Saturday, September 15 2012 @ 11:10 PM EDT |
I really would like to know how they intend to do that -- technically. Linux
can be recompiled to run on any instruction set (with way fewer constraints on
the instruction set that Windows). And it has.
So does this mean that the chip will have some secret enabling key to it, and
that only Microsoft will get that key? But that would be restraint of trade,
would it not? And even then.
Or would Intel only sell to M$, with M$ putting the key in and re-shipping? But
that would be very costly.
I really wonder.
Same for AMD. Must cost M$ a pretty penny.
---
bck[ Reply to This | Parent | # ]
|
|
Authored by: ailuromancy on Sunday, September 16 2012 @ 01:51 AM EDT |
I tried to find the statement from Intel that
led to the idea that Intel
would make a Windows only
chips. The closest I could find was an article at theinquirer with the title:
IDF: Intel says
Clover Trail will not work with
Linux. The article includes a couple of
tiny quotes
with a small hint at the context. At a guess, Intel's
representative had just explained some of the new power
management features of
Clover Trail, then answered
questions from reporters. The quotes are:
Clover Trail "is a Windows 8 chip" and that
"the chip cannot run
Linux".
The article later explains that the quotes are probably
not
literally true, and why. The power management
facilities require extra support
from the operating
system. That support is not currently in any operating
system, so Clover Trial chips will always be power
hogs with Windows 7 and
earlier Microsoft operating
systems (if they run at all).
A more
accurate title for the article might be
something like: "Intel are not going to
create a fork of
the Linux kernel with Clover Trail specific power
management". I can understand Intel not leaping into
that barrel of acid.
Google did something similar
to make Android. Android has some excellent power
management features purpose built for mobile phones.
Those are not features
you would want on a web server.
Google are making an effort to make Android
specific
code less intrusive so it can be included in the
mainstream kernel.
Despite Google's major contribution,
there are plenty of articles about mean
and nasty
Google forking the kernel.
I think Intel tried to avoid
repeating that mistake.
There is no way that a technical description of
P-states is going to generate any page hits so some
journalists focused on
missing Linux drivers instead. [ Reply to This | Parent | # ]
|
|
Authored by: sgtrock on Monday, September 17 2012 @ 10:44 AM EDT |
Phoronix
has a pretty good article covering Intel's past history and present
development activity supporting Linux that also details exactly what is going on
with Clover Trail:
If you search the mainline Linux kernel commit
activity for Clover you will also see no hits from Intel on "Clover Trail", but
the only clover mentions are from the days of Clovertown. (In comparison, you
can search for Cedar there are hits for current-generation Cedarview and a
Valley query shows many hits for the future Valley View Atom.)
Clover Trail
hardware might end up working out on Linux based upon the contributions of other
developers, but don't expect any official support out of Intel or for any IHVs
to be shipping Linux-based Clover Trail tablets/hardware.
If there's any
good takeaway out of this for Linux users is that this is hopefully just a
one-off affair with Microsoft this time around. Intel is preparing great Linux
support for Valley View, which is a next-generation Atom SoC to be delivered in
months ahead.
Lots more in the link. [ Reply to This | Parent | # ]
|
|
|
|
|