Slashdot Mirror
LinuxCertified LC2430 Laptop Review
Anonymouse writes "OSNews posted a comprehensive review of the made-for-Linux LinuxCertified LC2430 laptop. They found that all its components are fully compatible with Linux, except with ACPI in recent kernels (which actually affects many laptops recently). The laptop is a desktop replacement with strong performance and some good extra features: Firewire port, 3-1 card reader, combo drive, SXGA+ TFT screen and an ATi Radeon 3D card. Four Linux distributions were tested with it."
the P4 is the WORST possible cpu for a notebook. It needs massive memory bandwidth to perform somewhat acceptible, which most notebooks dont have. It draws a LOT of power, killing the battery rather quickly (90 min run time is pathetic), and all that dissapated power has to go somewhere (aka heat). The Pentium M or Athlon 64 series would of been a far better choice for a notebook.
Lawyers, MBA's, RIAA? A jedi fears not these things!
I recognize that laptop case from when we used to sell them at my last job. They were notorious for the hinges breaking very easily under repeated openings of the display. Be wary of weak components!
My Sager 5600matched this machine 2 years ago for ~$200? more. It's been running each release of RH (well, 7.2, 8, 8.1B, 9, FC1, FC2) since I've owned it, nary a problem.
The specs on this machine are underwhelming. The Radeon 9000 is so, um, 2001. Save for a few weeks more and pick up one of these to get your TV tuner, etc., too.
...if I wanted to read garbage like that, I'd go to \.
ahr, wrong...it's true that if the voltage does fall below a certain level copper would start deposit on the electrode, and the li-ion can never be recharged again, but all laptop uses Lithium-ion battery now a day, the "smart" battery tech built-in automatically takes that into consideration. So 0% doesn't not necessary mean the battery is completely empty, just mean it's near the recharging voltage limit. If i remember correctly apple actually recommends ipod users to drain the battery till the device "die" once every few months.
As for many other types of rechargeable batteries, it's actually better to discharge them completely before recharging, because of the so called "memory" effect.
This Sig is removed due to factual inaccuracy
The problem is not 100% with the laptop. MOST lapstops with recent kernel distros get BROKEN acpi, because of acpi/other-modules bad interaction. From a standpoint, it's the kernel that's broken lately, not the laptops.
all of the athlon models run at a much lower speed when on battery power. and they're equally as power hungry as a pentium IV.
Pentium M and Efficeon are the only sane choices for x86 notebook cpus.
Review says: sleep doesn't work, bad placing of the PCMCIA card slot wrt the optical drive, and a funny sound card. When the battery gets to 10%, it just shuts off, instead of sleeping--which I guess is related to the first issue. I have to say, that as noble an attempt as this is, if I purchased a new computer with any of these issues I would send it back. Is it right to cut them this much slack? Oh, and it's 7 pounds and get 1.5 hrs of usage. Let's compare to an iBook:
So the reviewed laptop costs $300 more, + wireless card, and sleep doesn't work? Plus the HD is smaller, weighs a pound more, and gets 1/3 of the battery life? You can put Linux on the iBook, even, if you don't like the UNIX part of OS X.
While there's a place for Linux on a laptop, I don't see this as an iBook killer. Get it below $999 and I'd be interested--if you're going to pay a premium I think this laptop has some competition.
--
$tar -xvf
The laptop's integrated Radeon chipset works with the GPL'd r200 drivers that come with Xorg.
Remember that before ATI began releasing any drivers at all for Linux they released the specifications of their chipsets and even gave cards to developers who wrote open source drivers.
Look in your kernel config under Direct Rendering Infrastructure.
I had one of these. At first I couldn't be happier - it was awesome that all the components worked with linux, and when ordering it you can ask them to partition the hard drive however you like (so you don't need partition magic to put something else on it - you can tell them to leave an unused partition for you to put something else on later), but then I discovered one really dissapointing problem with it: On the physical quality side of things, it is extremely fragile. For one thing, some screws came loose inside that part of the case that you aren't allowed to open without voiding the warranty - I could hear the screws rolling around in there, but to fix it I'd have to mail the thing back in since I'd void the warranty, and that would be a huge delay and I *needed* this thing daily. So I decided to wait until it was out of warranty anyway, and then open it up to fix it myself. If anything broke while it was under warranty, then I'd ship it in.
Well, the fateful day came, the warrany passed, and I opened it up. Inside I discovered, much to my dismay, that not just one or two, but an entire 9 different screw holes were missing their screws. I only ever found 6 of them in the case. Some were for the heat sink. The heat sink was held on by only the gooey thermal gell and pressure from the back of the case. None of the screw holes for the heat sink were still fastened. No wonder I occasionally got shutdowns with "hot cpu" messages in the logs (I had thought it was because I might have been covering the vent with my leg when I had it on my lap).
Also, Xfree86 had started behaving badly (dying at random times) and I found out why when I opened the case - the ATI video card (which was also supposed to be protected by that unscrewed heat sink - it covered video card and CPU) had a few spots where the soldering had apparently melted and I could see brown burn marks near it.
So here I was stuck with a computer that was broke, a day out of warranty, because it hadn't been screwed together very well, and since it's a laptop, trying to go to a third-party for a replacement video card is totally impossible - they're all unique.
Sigh. They're software setup was good, but my experience with their physical durability was piss-poor - and that's an important factor for a laptop.
Don't label something "offtopic" unless you know the topic well enough to tell what's on topic.
Get used to it, ACPI will always be a world of shit for anything non-Microsoft. This was by design. ACPI is complex. Remember that somebody at Toshiba or whatever laptop manufacturer is actually qualifying their ACPI implementation with a Microsoft OS. It is the vendor's responsibility to ensure that the ACPI system works...with Windows XP. The fact that ACPI works as well as it does in Linux and *BSD is due to a large number of volunteers.
Your choices are either to: a) fix it yourself and submit patches b) wait til someone else fixes it c) punt and use Microsoft Windows XP. I know a lot of people in the community that choose option c and I am not one those that would fault them (or you) for it if a and b are not practical.
Complaining on slashdot is definitely not going to help you though.
In 2002 when I bought a Compaq it was pretty much assumed that you would have to hack AML. Things have improved much since then, and the 2.6.x power interfaces and implementations are vastly superior to anything previous. The core kernel support for power support is excellent (the part that is not ACPI specific). ACPI sucks ass as a complete design, although the linux kernel implementation is decent and contributors are working to get a bug for bug compatibility with AML written for Microsoft Operating Systems.
When I first got by IBM T40 a few months ago the thing would lock up on power down and sleep (ie, if you powered down the screen would blank but the thing was actually still running). This required some troubleshooting and some code changes in the relaxed AML interpreter, and now it works.
I don't know about you lot, but I can't really extend the "linux certified" qualification to the radeon series, on account of those bloody drivers..
I can't help feeling that nvidia would have been a better choice here, on account of ease of use (I know, I know, both ATi and Nvidia drivers are binary-only, but the nvidia one is so much more reliable, and the Ati driver's a bitch to get working)
I've been searching for a windows free laptop with an nvidia gfx card, but they seem to be a mythical beast indeed. I'm not willing to pay for a GPU that I don't get to fully utilize.
True genius is grasping a situation like a peice of fruit, and peircing it just right so that it drains dry.
ACPI support is absolutely essential on a laptop.
I suppose, but ACPI is also an gawdawful piece of shit. I did BIOS development for a while. Imagine that (how to put this nicely) a lot of BIOS developers are, let's say, 'stuck in their ways'. Not all BIOS guys are like this, but to be a good BIOS person, you'd need a fairly large amount of experience, and it being BIOS and all, this results in a lot of people that think DOS is still a Pretty Neat Thing(tm). Assembler programmers who think that coding in C has too much of a performance hit. (again, this is a generalization, not all BIOS coders are like this at all, but still)
So along come a bunch of assholes who decide that the best way to get this power management thing going is to do it by creating an entirely new language/syntax. It's almost like a programming language but not quite. Bottom-line, my guestimate is that there are about one persons out there that really like the ACPI language. It's not very obvious, you probably really need some training before you should be allowed to mess with it.
Practically it goes more like this: the board manufacturer buys a reference BIOS for a certain reference board. The thing is, most boards end up being a little different from a reference board _especially_ in the power management section. So now the BIOS engineer needs to modify the ACPI code to match the board. Great, so (s)he goes ahead and does that. Then installs Windows XP, tests a couple of things like hibernate, or god forbid, stand-by. didn't crash? Ship it.
There are no proper regression tools to make sure ACPI is implemented correctly. It's very hard to debug/test/insure any changes you make to reference code, because it's like a bytecode language (imagine debugging Java with all you have is bytecode). Even worse, a lot of ACPI code runs in SMM mode, which is hard to work with and debug (even some of the hardware ICEs don't really support SMM mode properly)
And then managers tend to not understand why you'd all of a sudden have to spend 40 hours extra just to test this one little item on the requirements list called 'ACPI'.
In other words:
- management doesn't understand that ACPI support requires probably almost as much effort as the rest of the BIOS.
- BIOS engineers tend to not really like dealing with ACPI in the first place, so they are not going to bring this to managements attention.
Result:
ACPI support is just one big hack, in a lot of cases just copied straight from the reference design and the engineers are only going to work on it when they get dragged into it kicking and screaming (when the support people start to complain about ACPI related issues).
until ATI pulls their heads out of their asses and make linux binary drivers on par with nvidia they will continue to suck.
A binary only driver will always suck. But actually I have a computer with a Radeon 7000 that Fedora Core could use "out of the box". Some of their Mobility ones OTOH are going to cause lots of trouble. In fact I get better performance with a celeron 568 MHz CPU and i810 than on a 2.2GHz P4 laptop with a Radeon Mobility chip.
What I want to know about this laptop is if it works with all open source software, if not their certification is bullshit.
Do you care about the security of your wireless mouse?