oberman at es.net
Fri Apr 16 10:26:04 PDT 2004
> From: Jesse Guardiani <jesse at wingnet.net>
> Date: Fri, 16 Apr 2004 12:35:39 -0400
> Sender: owner-freebsd-mobile at freebsd.org
> Kevin Oberman wrote:
> > I have been running CURRENT on my everyday use laptop (IBM T30) for well
> > over a year and it's been running well (except for suspend/resume
> > issues) for some time. But I am now so used to any quirks introduced in
> > CURRENT that I may no longer even notice them. So, if you run CURRENT on
> > your laptop as a new user, you can provide valuable input on rough
> > spots BEFORE V5 is declared STABLE. (Hint, hint.)
> I've gotten a bit frustrated with submitting bug reports on -CURRENT. I understand
> that the developers want to spend their time developing NEW software and features
> and that they don't usually want to fix bugs unless the bug effects them personally
> or they have a particular interest in it. I think -CURRENT has changed too much
> from -STABLE at this point. I think it's too unstable.
> I long for 5 stable because I think developers (at least some) will switch focus
> and begin fixing bugs again. I submitted a ton of bug reports back in Jan 2004 and
> Feb 2004. Most of them were ignored. I think 5-STABLE will change that.
> Personally, I think there are three kinds of people who run -CURRENT:
> 1.) Developers
> 2.) People who want to play with or NEED the new features.
> 3.) People who think they are helping the project by submitting bug reports
> but who can't program, or can't debug hardware/kernel problems (this is me).
> I think people in category #3 are getting the short end of the stick and
> will continue to get the short end until 5-STABLE comes out. That's not
> necessarilly a BAD thing. The developers certainly need to make progress.
> But 5-STABLE needs to come out SOON (within a few months), IMO. Otherwise
> I fear that people in category #3 will start going elsewhere. I'm already
> starting to have visions of Linux on my laptop again...
> And please, I understand that the developers and almost entirely made up of
> selfless volunteers. I respect that. I'm just voicing my personal observation
> about the current state of the project.
Speaking primarily as part of category #3, I see what you mean. But there
is a reason that many problem reports have been ignored when they are
for laptops. The ACPI code needed to get many new systems to boot at all
(especially that needed to boot servers) was the top priority for the
developers. It had to be. While it applied to laptops, it left many
holes in laptop support where ACPI is utilized for may things not found
on servers. Suspend/resume is the most obvious issue. Power management
is related, but not quite the same.
Now that ACPI is stable for the basic functions, changes to support
power functions and suspend/resume are finally starting to show up. On
my T30 the improvement in many areas has been tremendous over the past
couple of weeks. Just look at the commit logs for stuff in acpica and
pci. Nate and Warner have been REALLY busy.
I still can't resume cleanly, although it's getting really close to
working. USB and sound are now working. Nate is looking at the odd
problem of losing all interrupts on IRQ 11 some time after I
resume. (This is REALLY weird.) And I still can't get my back-light to
turn off while suspended which really limits the value of suspending.
I'm keeping my fingers crossed that these last issues will be resolved
soon and that I will be able to suspend as a normal part of life
again. Even though I don't do much programming any more, I do plan on
trying to dig through the DSDT and see if I can hack together something
that will turn off the !@#$ back-light on this thing. If nothing else,
I'll learn a lot about reading and understanding ASL. Odds are that
someone will have a solution before I do, though.
And it is almost certain that 5 will go STABLE this summer. Only a few
significant issues remain. I think that early summer is even a
possibility (but I'm an optimist).
R. Kevin Oberman, Network Engineer
Energy Sciences Network (ESnet)
Ernest O. Lawrence Berkeley National Laboratory (Berkeley Lab)
E-mail: oberman at es.net Phone: +1 510 486-8634
More information about the freebsd-mobile