Practical Technology

for practical people.

September 9, 2010
by sjvn01
0 comments

Last chance to get XP

Windows 7 has been around for almost a year now, but lots of people are still sticking with Windows XP. Now, Microsoft is getting ready to pull the plug on Windows XP sales. No! Really! They mean it this time. Would they lie to you?

I mean just because Microsoft has extended XP sales over and over again doesn’t mean that they’ll keep selling XP forever. Well, yes, they are supporting XP for years more to come, but this time, cross their hearts and hope to die, Microsoft really is killing XP sales on October 22nd, 2010. Some companies, noticeably Dell, are pulling the plug on XP even sooner.

Microsoft spokesperson Brandon LeBlanc has written that most Windows users may “not notice much change.” I beg to disagree.

I know lots of Windows users and even now they really don’t want any part of Windows 7. For some of them, it’s a matter of finance. They look at the cost of moving their business PCs from XP to Windows 7 and they get the shakes. In this economy, who can blame them?

More >

September 8, 2010
by sjvn01
1 Comment

The Internet is running out of IPv4 gas

If you listen to some people, businesses don’t need to worry about the growing shortage of Internet IPv4 addresses. Instead, most “network owners find it more affordable to just make do with the [Internet] addressing scheme they’re already using. This is so, so wrong.

When the Internet began, IPv4’s possible 32-bit 4.3 billion addresses looked like more than enough. Things have changed.

We’re running out of IPv4 addresses, the 32-bit numeric addresses that network devices need to connect to the Internet. All those mobile devices that we love so much like iPhones, tablets, and iPods are gobbling down IPv4 addresses like an elephant does peanuts. For the longest time, we managed to avoid running out of IPv4 addresses with the use of technologies like Network Address Translation (NAT) and Classless Inter-Domain Routing (CIDR), but those haven’t been enough.

More >

September 7, 2010
by sjvn01
1 Comment

Digg, Dug, Done

I didn’t like the Digg beta. But, then, I thought, “It’s a beta, they’re often rotten.” So I waited for the new version, Digg 4 to be released. I didn’t like that either. But then I thought, “Well, I’m just not used to it. I’ll wait a bit.” So, about two weeks later, I think I’ve given the new Digg a fair chance and, folks, it’s awful.

I deliberately avoided reading all of the “The new Digg is awful” stories. I also didn’t even bother to look at competing social bookmarking sites like Reddit for Digg news. I knew Reddit fans would hate it. One sample Reddit comment sums their view of Digg up nicely, “The new digg is embarassingly bad.”

After due consideration, I have to agree. Digg gained its popularity as one of the first social news sharing sites. The idea was that users would pick the ‘best’ stories on the Internet by their votes. Of course, any public voting system like Digg can be gamed by people with an ax to grind, and it was … over and over again.

That said, Digg, Reddit, and the other bookmarking sites at least give users the illusion that they were calling the shots on what was good and what wasn’t. Both sites, and their less well-known counterparts like DZone for developers, try to let the ‘people’ decide on what’s good and what isn’t.

Well, that was the case with Digg anyway. Now, and this is what has killed my interest in it, users really have almost nothing to do with what becomes ‘popular’ on Digg.

More >

September 7, 2010
by sjvn01
0 comments

Android/Linux kernel fight continues

YYou could argue that Google’s Android, so popular on smartphones now, is the most popular Linux of all right now. There’s only one little problem with that: Android has continued to be apart from the Linux mainstream.

People became aware of the Android and Linux split when Ryan Paul reported that “Google engineer Patrick Brady stated unambiguously that Android is not Linux.”

Brady over-stated the case. Android is Linux. To be exact, version, 2.2, Froyo, runs on top of the 2.6.32 Linux kernel. To quote from the Android developer page, Dalvik, Android’s Java-based interface and user-space, uses the “Linux kernel for underlying functionality such as threading and low-level memory management.” Let me make it simple for you, without Linux, there is no Android.

But, Google took Android in its own direction, a direction that wasn’t compatible with the mainstream Linux kernel. As Greg Kroah-Hartman, head of the Linux Driver Project and a Novell engineer, wrote in Android and the Linux kernel community, “The Android kernel code is more than just the few weird drivers that were in the drivers/staging/androidsubdirectory in the kernel. In order to get a working Android system, you need the new lock type they have created, as well as hooks in the core system for their security model. In order to write a driver for hardware to work on Android, you need to properly integrate into this new lock, as well as sometimes the bizarre security model. Oh, and then there’s the totally-different framebuffer driver infrastructure as well.” As you might imagine, that hasn’t gone over well in Android circles.

This disagreement arose from at least two sources. One was that Google’s Android developers had taken their own way to address power issues with WakeLocks. The other cause, as Google open source engineering manager Chris DiBona essentially said, was that Android’s programmers were so busy working on Android device specifics that they had done a poor job of co-coordinating with the Linux kernel developers.

The upshot was that developer circles have had a lot of heated words over what’s the right way of handling Android specific code in Linux. Linus Torvalds dropped the Android drivers from the main Linux kernel.

Google tried to do the right thing by hiring two new Android developers to work more closely with the Linux kernel development team to get Android back in sync Linux. At the time, it looked like Google and Android would quickly get back to the same page.

It hasn’t worked out that way. At LinuxCon, I asked the Linux kernel developers about this, and I got an earful.

Google kernel developer Ted Ts’o said that he didn’t think it was that big a deal that Android included some non-standard software. “I can’t think of any shipping Linux distro, including Red Hat, that doesn’t have some out-of-tree packages.” And, Ts’o continued, “No one ever said, oh my God, Red Hat or Novell forked the kernel.”

From where Ts’o sits, the real problem is that “Android has been so successful, and that has inspired many hardware vendors to write device drivers for Android. WakeLocks calls in device drivers become problematic when people want to submit code upstream.” The bottom line is that this forces chip vendors, like Qualcomm and Texas Instruments to maintain two versions of Linux, with and without WakeLocks. Needless, these companies aren’t happy with the extra work.

Chris Mason, Oracle‘s director of Linux kernel engineering, added that this kind of conflict is not new. While James Bottomley, a distinguished engineer at Novell and Linux kernel maintainer, added that getting Android to work smoothly with the rest of Linux will “Take a lot of effort, but it will be worth the time for the larger community.”

Unfortunately, according to Ts’o, time is not something the Android team has a lot of. They’re too busy running to keep up with hardware requirements. Ts’o said that, although, “There’s less than 64K of patch, there’s been over 1,800 mail messages of discussion.” Ts’o made it sound like the Android team is getting fed up with the process. “Android is a small team. They feel that they’re spending a vast amount of time getting the code upstream (to the main Linux kernel).”

On the Linux Kernel Mailing List (LKML), Ts’o suggested later that “You know, you don’t have to wait for the Android engineers to do this work. You (or others who want to be able to use stock upstream kernel with Android devices) could just as easily try to do the ‘bit more work’ yourselves — that is, do the open source thing and scratch one’s own itch.” Later, Ts’o also pointed out on the LKML that mainstream Linux distributions include their own non-standard code.  He summed it up with, “Can we please cut out this whole forking nonsense?”

In the meantime, of course, Google has other Android worries with its Oracle patent fight. In the end, I’m sure that Android and the mainstream Linux kernel will get back in sync with each other. I don’t see it happening anytime soon, though, and I suspect there will be a lot more heated words exchanged before it finally happens.


A version of this story first appeared in ComputerWorld.

September 7, 2010
by sjvn01
0 comments

Four ways to get the most from your 802.11n Wi-Fi

In theory, 802.11n can zip by your 100Mbps Fast Ethernet at a real-world 160Mbps, but the practice it’s usually much slower. No, the Wi-FI vendors aren’t lying; the problem is that you have to set 802.11n up just right to really get fast performance.

First, you need to make sure that you’re using up-to-date 802.11n hardware. Older 802.11n equipment, built before the 802.11n standard was finalized in late 2009, may not work and play well with your newer devices. There were many 802.11n draft access points (APs), network interface cards (NICs) and chipsets and each vendor used its own best guess on what the standard would eventually look like.

Thanks to all this older, not quite standard 802.11n hardware, we have two problems. The first is that some older hardware, unless the firmware can be upgraded, won’t work at full 802.11n speeds with your newer standardized equipment. The other is that you can be almost certain that older APs, switches, or routers from one vendor won’t work well with another vendor’s equipment. Oh, it may look like it’s working, but if you check you’ll often find that your Wi-Fi’s connection is only running at 802.11g’s 54Mbps.

More >

September 3, 2010
by sjvn01
0 comments

Mint 9: Minty fresh Linux

Linux distributions come through my office in a constant flood. Most of them stick around long enough for a quick try-out on one of my test systems or on a VirtualBox virtual machine. A few find permanent homes. I currently use OpenSUSE on my servers, and Ubuntu, Fedora, and MEPIS on my desktops and laptops. Now, I have a new resident on my desktops: Mint 9.

I’ve been fond of Mint for years, but my affection for it never quite reached the point where I wanted to use it on a daily basis: Until now. It’s not that Mint, an Ubuntu-based Linux desktop distribution, has made a great strides forward, it’s that the distro has continued to get better and better with each release.

Mint 9 is based on Ubuntu 10.04, so if you’re already a Linux user chances are you already know its foundation. On top of Ubuntu though Mint layers a much more attractive, to my eye, default GNOME 2.30 interface. The color scheme is green with a very clean and simple desktop. I find Mint’s desktop to be both restful and inviting.

More >