Linux on an HP Compaq 6715b Notebook

25 Aug 2007

Last week, some IBM ThinkPad T61p notebooks showed up at Guru Labs offices. There were 2 of them on Monday and another on Tuesday. I also know that there are 2 more coming and perhaps a couple of other co-workers will be ordering them, too.

I’ve been needing to get a new notebook for the past two years, but I kept putting it off because of time, money and that one more feature that’s coming out in a couple of months. With the arrival of so many new notebooks in the office, I decided to look again and dream about a new one of my own, so I made the rounds looking at systems of interest, including a couple of HP notebooks, the ThinkPad and Apple‘s MacBook Pro.

When I hit HP’s Small & Medium Business website, I noticed the one category of notebooks which I had always left unexplored (as they didn’t fit some of the criteria I look for) listed that there were models which had up to 16 hours of battery life. I was curious to see what they had in this “Balanced Mobility” category, so I took a look. Boy, am I glad I did.

I found the HP Compaq 6715b. They had (at this writing, I think it’s still on) a pre-packaged deal going for US$1,129 (Ed: The price is lower, now) with:

  • AMD Turion64 X2 (dual core) at 2.0GHz
  • 1GB RAM
  • 160GB SATA hard drive
  • ATI Radeon Mobility X1270 video chip (with 128MB dedicated RAM and using 192MB shared RAM)
  • 15.4 inch WSXGA+ (1680×1050) LCD
  • Broadcom Gigabit Ethernet NIC
  • Broadcom 4321 802.11 a/b/g/draft-n wireless NIC & integrated bluetooth
  • Fingerprint reader
  • 4 USB 2.0, 1 IEEE1394 (firewire), 6-in-1 card reader (actually, all SD type form factors), 1 Type I/II PC-card slot

That’s a lot of notebook for the money. So I put in an order. HP estimated that it would ship on the 30th of August, but it arrived on Thursday morning (2007/08/23).

In fact, I believe it’s around half the price of what any of the other guys have paid for their ThinkPad notebooks and it’s almost the same. They got a wireless USB 2.0 capability which I don’t have, but they only have 3 USB 2.0 ports (I have 4). Most (if not all) of their screens are 15.4 inch WUXGA (1920×1200) with nVidia graphics (256MB), an Intel Core 2 Duo (2.0GHz or 2.2GHz, I’m not sure which in all cases) and they have a nice “eraser-head” mouse which I don’t have, but really love. I hate trackpads, so I just picked up a Logitech bluetooth mouse, Saturday.

Overall, I think I got a better deal. My processor is as good or even a little faster than the ThinkPads’, and otherwise there’s very little difference in the equipment between the two, but they paid quite a bit more than I did for the HP. Thanks to that savings, I also picked up a 12-cell “Ultra Capacity” battery for my new notebook, which attaches to the underside at the back, causing the system to sit at a slight incline. The Ultra Capacity battery mounts in addition to the standard battery that came with the notebook and gives this machine up to 16 hours of battery life, with only a small increase in weight but a little more comfort and room for airflow underneath. We’ll have to wait and see just how much life I really get out of this setup, but I shan’t fear attempting to watch 3-4 movies on an international flight.

I’ve installed Fedora 7 on it. When I booted up the box to do the install, anaconda couldn’t get X to run, so it offered me the choice of using the text-mode installer or of starting VNC for me. I went with a VNC install. The resulting system had a couple of things to fix up. I checked on but this model isn’t listed, yet.

I believe there must have been a bug (I didn’t bother to go looking in Red Hat’s Bugzilla for it) in the version of YUM that shipped with F7 (32-bit) as yum update kept corrupting the RPM db and then deleting the errata RPM files as it thought it had installed packages but actually hadn’t. I simply edited /etc/yum.conf and set keepcache=1 before re-running yum again. That way, the packages stuck around and then I installed as many as I could using rpm instead (including an updated YUM package), which required me to fix the RPM DB, first. This was easy to do by simply running rm /var/lib/rpm/__*; rpm --rebuilddb and waiting for just 1 minute for it to finish. After installing the updated YUM package, all yum commands have worked perfectly for me.

To “fix” the X server configuration, I simply added the livna YUM repo to my new system and ran yum install kmod-fglrx followed by ati-fglrx-display enable as root (that’s not the command mentioned in the Unofficial Fedora FAQ for FC6, but the F7 version of the UFAQ wasn’t up yet) and the X server worked perfectly, even running the screen at it’s full, native resolution by default. I’ll have to see about running Cedega for a couple of games.

Next, I tried to get the fingerprint reader working, but so far, I’ve had no luck. Honestly, I haven’t really tried all that hard, yet. Some quick Google searches have only found references to people who haven’t gotten other HP notebooks’ fingerprint readers to work, but I also found some “hints” that others have. The output of the lsusb command showed Bus 003 Device 003: ID 08ff:2580 AuthenTec, Inc., which is the fingerprint reader.

I haven’t gotten the Broadcom 4321 802.11a/b/g/draft-n working yet. Linux does come with a driver that supposedly covers the chip in this Mini-PCI card, but I do not have the firmware for the driver to load. The tools for these cards come with a program called fw-cutter, but I haven’t found a file for this card that it will work on, yet. I suspect that I will have to wait for an update to fw-cutter to be able to get this working under the Linuxdriver . Perhaps I can find time to try to help patch it. In the meantime, my good old Cisco airo 350 card works fine, but I could also use NDIS Wrapper to run it with a Windows driver.

I’ve only been using this notebook for less than a day (and only a small part of the day, at that). Even so, I’m very happy with it already.

I’m thinking of installing Ubuntu (or Kubuntu, probably) on here alongside of Fedora. I’ve been wanting to learn more about that distro and now I have a hard drive that’s more than large enough for me to play with such things.

I also added vga=0x31a to the kernel line in the /boot/grub/menu.lst file (yeah, yeah, I know how Red Hat/Fedora only folks are going to say the file is “supposed” to be /boot/grub/grub.conf, but it really isn’t so; so, please, don’t add comments telling me about that). That sets up a framebuffer mode for text that’s 1280×1024. I don’t know if the kernel can support a 1680×1050 mode or not (so far, I’m not finding anything that would make thik it does). If so, I’d sure like to find out the right code for it. If not, I’d like to figure out how to add wide-screen friendly modes to the kernel framebuffer driver(s), as more and more systems are going that way.

I’m going to post this system on the website. If anyone else figures out how to get the fingerprint reader working under Linux on this or any other notebook that uses the same fingerprint reader chip/device, please, either TrackBack to this post or leave me a comment.

New U.S. Wiretapping Law

19 Aug 2007

On August 5th, 2007, President Bush signed S.1927 into law, after it passed a Senate vote (60 for, 28 against, 12 present but not voting) on the 3rd, and a House vote (227 for, 183 against, 23 present but not voting).

The new law amends the “Foreign Intelligence Surveillance Act of 1978 to provide additional procedures for authorizing certain acquisitions of foreign intelligence information and for other purposes.” It was sponsored by Sen. Mitch McConnell [R-KY] and Sen. Christopher Bond [R-MO].

I haven’t had time, yet, to fully read the resulting text of the bill (there are always amendments to bills as they pass through Congress), so I will reserve any specific commentary for a latter time. However, it appears that this new law could seriously affect privacy under certain circumstances in the United States.

Gun Shapped Notebook Battery

18 Aug 2007

Poor Ben Forta. The fact that he’s actually struggled to get through airports for a while before figuring this out seems rather strange to me. How could he be the only one? Why have none of my co-workers (at least, to my knowledge) not had similar troubles with their ThinkPad notebooks?

It seems that several TSA inspectors at several different airports were mistaking the laptop battery for a possible gun in Ben’s notebook bag as it went through X-Ray scanners.

Post Election Auditing Report

17 Aug 2007

The Brennan Center for Justice has published a report titled, “Post-Election Audits: Restoring Trust in Elections,” which is well worth a read to anyone who cares about liberty and election security.

IRS Security Troubles

16 Aug 2007

Simply, un-excusably Amazing.

The recent MSNBC story, “Computer security problems found at IRS,” discusses security problems found at the IRS. One of the more interesting items:

Sixty-one of the 102 people who got the test calls, including managers and a contractor, complied with a request that the employee provide his or her user name and temporarily change his or her password to one the caller suggested, according to the Treasury Inspector General for Tax Administration, an office that does oversight of Internal Revenue Service.

But even more disturbing:

Only eight of the 102 employees contacted either the inspector general’s office or IRS security offices to validate the legitimacy of the caller.

Commercial eVoting Security Problems Abound

15 Aug 2007

Recently, California’s Secretary of State was required to perform a security screening of the eVoting systems that the State of California is thinking of/planning to use. The California Secretary of State appears to have been highly opposed to this outside audit process, according to information found within the official reports (the site has lots of links to very interresting documents, most of which are well worth the reading).

Matt Bishop’s comments on the nearly total lack of cooperation from the California Secretary of State’s office gave to the review process are utterly amazing. It’s good to see that Debra Bowen (California’s Secretary of State), has now taken the step of decertifying, dis-approving all previously approved eVoting systems.

Avi Rubin has some excellent comments on the whole eVoting situation.

The State of Florida is getting into the act, reporting on their own security reviews of commercial eVoting systems (PDF). In this letter to Diebold (PDF) which the State of Florida has published, they give Diebold an ultimatum:

Based on the report, the Bureau of Voting systems Certification has determined that certain vulnerabilities outlined must be corrected by August 17, 2007, to continue this certification. Failure to do so will result in a denial of certification.

There’s 3 pages of required fixes attached to that letter.

The U.K. Electoral Commission recently released their report detailing serious security flaws in eVoting systems.

Electronic voting is a hard problem, but that doesn’t excuse Diebold Election systems, Inc., Hart InterCivic, Sequoia Voting Systems and Elections Systems and Software, Inc. from their demonstrated complete lack of fundamental understanding of how to secure … well, anything and in particular, they’ve all shown that they have no one with even the first clue of how to either implement nor apply cryptography correctly.

Applause go to both Florida and the U.K. for recognizing bad vendor crap in the first place. An extra-hearty ‘atta-girl’ goes out to Debra Bowen in California for throwing out approvals and certifications of these seriously flawed systems.

This topic is far too important to leave in the hads of the proprietary, closed-systems mindset crowd. It must be open. The code must be open and available to everyone. All systems must be thoroughly tested by reputable, recognized, outside authorities. I hope we’ll see an open source/free software implementation of an eVoting system that could be used for governmental elections. Such a system wouldn’t be limited to only government use, either, but I believe it would find place in many corporations and other institutions.