David Brown wrote:
On Wed, Aug 20, 2008 at 09:09:17PM -0700, DJA wrote:
Windows works best for reasons we're all familiar with (read: market
share). Most everyone else is just doing a lot of research (usually
including looking to see what Windows does, and duplicating that
behavior right or wrong), tedious testing, or just guessing. Most do
all three.
I don't think Windows works because of market share, I think windows
works because the vendors take the effort to make the exact install
image it comes with work.
I didn't think I'd have to explain: Since Windows owns the OS market,
OEM's and ODM's take more pains to make sure that their products work
(however passably) with Windows. They test their products with Windows,
and they assume that the biggest chunk of their income will come from
people using Windows on their products. It is rare for them to take the
same pains to see that their products run on Not-Windows.
Not-Windows developers rarely get the same access to manufacturers'
design processes, implementation tweaking, testing, technical support,
and documentation.
If Not-Windows accounted for 9x% of the OS market, you can be sure that
most products would be developed with it in mind.
At my work, the laptops run an IT image instead of the vendors image.
There are an awful lot of people walking around with their laptops
propped slightly open, and usually if you ask they will tell you it is
because the machines don't suspend reliably.
No one has suggested that Windows never breaks. However, when it comes
to power management and ACPI-related features, Windows still tends to be
the more stable OS on the average motherboard or laptop out of the box.
Linux developers in particular know this and are working hard to at
least match that average reliability. We can cherry-pick special cases
all day to show that OS N is better, but if you want to get a real feel
for what Not-Windows developers are up against, feel free to subscribe
to a dev mailing list covering this area.
A fresh windows laptop usually works fine until you do silly things
like install software or extra hardware on it.
So does Linux. Especially when upgrading kernels and other fundamental
non-userspace software.
But, as far as networking and/or USB and suspend Windows (at least XP)
is just absolutely stupid. I've never had a USB device recover from
suspend without having to be removed and reinserted after wakeup.
These seem to work fine on OSX and Linux. Windows also seems to have
a habit of some network thing preventing a suspend. That usually
results in the very hot, still running laptop in the bag with the
nearly dead battery.
Certainly data points.
[I can add my own as well. My laptop's power, special key functions,
sound, display, modem, etc. all work fine in Windows. They work only
partly in Linux. The WinModem and VCR-buttons are not supported at all
in Linux. Other functions, especially special buttons only partially
work, and then only because of my tweaking or writing scripts to make
them work.]
But I'm basing my argument upon several years of following several Linux
dev lists. From that, my observation is that this is one of the areas
where Windows still tends to work better out of the box for the average
user under typical usage.
My attitude about suspending is that if something like a network
connection is still up, tough, it is going away. I'm suspending
because I'm about to go somewhere else, and I need the computer to be
off.
The only problems I've ever had with suspend on the Mac is with
VMWare.
David
But your expectations, while maybe more realistic and informed, are not
in line with the average computer user whose expectations are not always
so rational.
--
Best Regards,
~DJA.
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
