Tuesday, August 28, 2007

Robert Love backs up my very simple performance experiment

It didn't take long for a member of the Linux kernel developer community to provide a more proper analysis of what I was able to observe on my own; that the Linux kernel, and Linux in general, is just so much superior to Vista when it comes to multimedia and network processing.

Robert Love, in a blog posting of his own, deconstructs the Vista networking problem by showing how to do it right, via the Linux kernel. Robert nailed it when he said:
Critical optimizations such as zero-copy aside, there is no excusable reason why processing IP packets should so damagingly affect the system. Thus, this absolutely abysmal networking performance should be an issue in and of itself. Unfortunately, however, the Windows developers decided to focus on a secondary effect:
Tests of [Multimedia Class Scheduler Service (MMCSS), a mechanism for the automatic priority-enhancement of multimedia playback,] during Vista development showed that, even with thread-priority boosting, heavy network traffic can cause enough long-running DPCs to prevent playback threads from keeping up with their media streaming requirements, resulting in glitching.
In other words, consuming half of your processor is (surprise!) detrimental to multimedia playback performance. At this point, it becomes clear that the process scheduler folks and the networking folks are bitter enemies and do not converse. Consequently, the obvious solution of fixing the abhorrent networking performance was bypassed for a quick bandaid:
MMCSS’ glitch-resistant mechanisms were therefore extended to include throttling of network activity. It does so by issuing a command to the NDIS device driver, which is the driver that gives packets received by network adapter drivers to the TCP/IP driver, that causes NDIS to “indicate”, or pass along, at most 10 packets per millisecond (10,000 packets per second).
Putting aside the larger problem for the moment, there are several issues with this solution. It prioritizes multimedia playback over networking performance, which, as the resulting clamor has shown, is not everyone's personal policy preference. It is almost assuredly a layering violation. It picks a fixed and hard-coded packet limit (ten per millisecond), which won't scale across different hardware—think significantly faster processors or substantially slower networking drivers. It ignores the commonality of GigE. And, finally, the solution is complicated, as the convoluted description and resulting bugs in the implementation demonstrate.
What I find so amazing about this entire episode is it exposes to the world how Microsoft's software development process works (not very well) and how they are going to add insult to injury by padding the band-aid rather than fixing the root cause of this problem; a poorly designed and/or implemented network stack.

If you'll recall, Microsoft announced during Vista's development that they were completely rewriting their network stack for Vista and later versions of Windows. The re-wrote it because they admitted that the then-current stack used in Windows NT up to XP was convoluted, complicated, and a hack. They chose to start over and redesign everything they needed, and then to reimplement the clean design. The goal was a networking subsystem that would have better security and (dare I say it) better performance. While the jury is still out on security, I think judgement has been passed on performance, and it's a big fat thumbs down.

As Robert says at the end of his piece, if you must wait for Vista then there are indeed viable alternatives to Vista while you wait for Microsoft to finally finish Vista. And who knows; you may find you like Linux so much it won't matter if Vista is every finished.

Congratulations Apple!

ComputerWorld reported today that Apple now sells one in six laptops in the U.S., and is in third place behind number two Toshiba and number one H.P. Apple beat out Gateway (my computer vendor for my two business laptops) for the third place spot.

I believe the major reason for the surge in Apple sales has been drive primarily by the market's disappointment in Vista, further exacerbated by Microsoft's poor treatment of the market with WGA. I've heard many a professional non-Apple user in Orlando express a poor opinion of Microsoft since January, saying that they would probably buy Apple, at least for home or personal use.

It will be interesting to see if Linux too can take advantage of this ever growing dissatisfaction with Microsoft in the marketplace, especially on notebooks.

Microsoft underscores the fallacies of Microsoft's distributed computing

The fallout continues from Microsoft's WGA failure over the weekend. You would think that a company as large and as rich as Microsoft, with over 30 years of experience in the microcomputer industry, would understand the eight fallacies of distributed computing. They are:
  1. The network is reliable.
  2. Latency is zero.
  3. Bandwidth is infinite.
  4. The network is secure.
  5. Topology doesn't change.
  6. There is one administrator.
  7. Transport cost is zero.
  8. The network is homogeneous.
It's point 1 that should draw the attention of the Softies. The network includes everything in the network infrastructure, especially the computers running the network services such as WGA at the head end. When you sit back and draw up a network diagram showing every connection point, router, and port from your PC to the WGA servers somewhere out in Redmond, it makes you wonder how it can work at all.

Don't get me wrong. From a technical perspective Microsoft's WGA across the Internet has worked quietly and reliably for some time now. The only complaints about it have come from old-guard reactionaries like me who have always questioned (and complained about) the need to validate (and re-validate) software I spent dearly to acquire just to make sure I'm honest. It's that process that underlines more than any other the fundamental change between Microsoft and its customers; the fact that it now considers its customers as adversaries rather than trusted and valuable customers. WGA to my way of thinking is nothing more than Bill Gates' ultimate solution to his 1975 manifesto about people who steal his software. But I digress...

Lest the Linux crowd get too smug (present writer included), we've also had our network-distributed gaffs, such as the xorg-server bug introduced during an upgrade of Ubuntu 6.06 a little over a year ago. It was quickly fixed, but the original update, distributed by Ubuntu servers across the network, left a lot of unfortunate users at the command line without a GUI. A good sized majority of them felt they had an unusable system, and rightfully so. The bug was found, quickly fixed, and distributed, and nothing of that scale has happened since, but you can appreciate how widely you can effect your community of users when, for whatever reason, you blindly assume the entire network is reliable.

The fundamental difference between Microsoft and Linux in general is due to motivation. Microsoft's was draconian, while Linux was honorable. Microsoft was intent on tightening the screws in their system making sure that you have rendered unto Microsoft your tax (the price) of their software, while the Linux problem was an honest (though very unfortunate) mistake where the developers were trying to improve the software, but failed. I don't know how you feel, but I take no great comfort in the fact that because of Microsoft's greed the full operation of my paid-for copy of Microsoft software is at the mercy of a less than perfect vendor on the other end of a less than reliable network. I can fix a bug like the xorg bug, but I can't fix a bug like the Microsoft bug unless I just replace it with an alternative and remove all those negative critical dependencies.

Monday, August 27, 2007

Democrats live up to their party symbol

As a long suffering Florida Democrat, my party's antics never cease to amaze me. Take, for example, the current mess we find ourselves in. As this story illustrates we can find more ways to snatch defeat from the jaws of political victory than Carter's has little pills. So here's the deal. Florida's Republican controlled state legislature, in a fine show of political manipulation, has moved the date of the Florida primary to January 29, 2008. The national Democratic party, in a grand and glorious show of righteous indignation over the matter, have threatened to strip us of our delegates to the Democratic National Convention if the state Democratic party doesn't change the date to something later in the year, Right Now.

What part of "Republican controlled legislature" doesn't the DNC understand? What do we do in Florida, launch a coup on the state capital and hold the Republican Governor hostage until the Republican controlled state legislature capitulates to the national Democrat's demands? I think the Governor and the legislature would die laughing first, thus foiling the state Democrats in their attempt to take over the government and correct the date.

What's happening right now in Florida should forever lay to rest any questions as to why we Democrats have an ass as a party symbol.

You can rest easy tonight...

secure in the knowledge that Microsoft's Windows Genuine Advantage is back up and running again! According to the latest from Ars Technica,
Update: Word from Microsoft is that this problem has been fixed, and all users affected should revisit the WGA site and re-validate. There's no explanation as to why Microsoft was originally telling people to wait until Tuesday, but the good news is that the problem has been solved.
So there you have it. You can now validate and re-validate and re-re-validate what you've already purchased, secure in the knowledge that Microsoft will let you use what you've already paid for. Isn't that wonderful?

Sunday, August 26, 2007

Linux works just fine, thank you very much!

In the last posting I made a few snide comments about Vista's documented network (un)performance while Vista was playing back something as simple as an MP3. I found a moment this evening to set up an extremely simple experiment where I streamed a movie and played an MP3 at the same time on europa. The movie, ripped earlier this year from a DVD using VLC, was streamed from an Ubuntu 7.04 system (rhea) while the MP3, ripped from my CD of the movie sound track using K3B, was played from a local file. The two multimedia files used in this test were:
  • Movie: "Ghost Rider", MPEG-4, 720 x 480, two channel audio, 1.2 GiB size
  • Song: "Son of Man", sung by Phil Collins from the Disney movie "Tarzan", MP3 320 KiB sampling
The two systems used in this test were:
  • Europa, a four-year-old 32-bit single core Athlon XP 3200+ system with an ATI 9700 Pro and 1GiB of DRAM running openSuse 10.2, and
  • Rhea, a three-year-old 32-bit single core Athlon XP 2500+ system with an nVidia 7600 video card and 512 MiB of DRAM running Ubuntu 7.04.
To run this very simple test I used a second system, rhea, to host the movie. Rhea and europa (running openSuse 10.2) are connected together with a Netgear FS605 five port ethernet switch. Both are connected into the switch with 100MiB ethernet. Rhea shares a folder with movies I've ripped via SMB. Europa sees this particular folder and mounts it. The test consisted of playing Ghost Rider via Kaffeine on europa from rhea across the simple network. I let the movie play for a while, turned down the sound on the movie, then played the MP3 with RealPlayer 10. I listened to the entire song, and heard no pops or distortion in its playback. The screen shot below shows the test as it was underway.

Again, I stress that this is no "real" test. No hard numbers (specifically throughput) were captured, nor was the experiment repeated multiple times. I simply ran the Gnome System Monitor at some point well after the movie started and watched the processor and network performance while everything was running. You'll note that somewhere between 40% and 50% of europa's processor is being used. I noticed that playing the MP3 had little or no noticeable effect, at least by observing System Monitor. Nor did the MP3 have any effect on the movie; no frames dropped, no stutter, or any audio distortions. It seemed to be pretty much all due to video playback. Looking at System Monitor and just playing the MP3 showed such a low usage that it was indistinguishable from an idle system.

These aren't state of the art systems by any stretch any more, and there is no way either could run Vista. But with contemporary Linux distributions, they are more than capable of soldiering on. Here is a screen capture System Monitor from rhea.

You can see at the bottom how rhea was pumping out the movie across its network connection until I stopped the movie. But note the processor utilization at the top. Processor utilization was nearly flat at 6% the entire time.

For development or consumer use, of what use is Windows any longer?

Defective by design

ZDNet has been following a problem with Vista's MP3 playback and network performance problem. It's the digital equivalent of walking and chewing gum at the same time. With Vista, if you play back and MP3 and attempt to move anything across your network connection, then your network connection takes a severe hit (up to 90% performance degradation) until the MP3 finishes or is otherwise stopped from playing.

Call me silly, but I put a higher personal priority on network performance than playing back MP3s. If you want to play MP3s, then buy an iPod. That's what they're for. Or buy a Mac. Or run Linux on that Intel hardware you're currently running Vista on. I've never seen any problems with Linux handling multiple tasks, especially tasks involving networking and multimedia.

I'm going to have to run some network and MP3 playback benchmarks on Linux, or else find someone who has. Stay tuned.

Yet Another Reason for Linux (YARL)

Ars Technica reports that Microsoft's Windows Genuine Advantage servers have been down since at least Friday, and that they will continue to be down until Tuesday of next week. This outage effects Windows XP and Vista systems that need validation as "genuine" Windows, i.e. it's not a pirated copy. Here's the skinny on what not to do until this is fixed:
If you use Windows, do your best to avoid anything that requires a ping to WGA. That means you should stay away from patches and add-ons until the coast is clear. WGA will not reach out across the Internet and deactivate your copy of Windows, but you should avoid talking to a WGA server for any reason.

For those of you doing installations and upgrades this weekend, we recommend that you avoid activation at this time. Remember that you can run Windows legally for 30 days without activating.

If you attempt a validation and it fails, your install may be marked as non-genuine, which could lead to several annoyances. First things first, do not reboot a Windows machine that has been marked as non-genuine. Once you do so, you will lose functionality and the Aero interface. It would be best to wait until this problem has been resolved.
Got all that? So how does it feel to be treated continuously like you're scum? You never notice until the system breaks, in this case quite spectacularly. This is what monopolies can get away with. Microsoft, back before it completed a total lock on the U.S desktop OS market, would have never gotten away with this, and Microsoft knew it. But now that there's nothing and nobody of real consequence to stand in their way, they've put in place WGA to make sure you've paid for everything, regardless of your receipts that show you paid good money for either the OS or the system it came pre-installed on. Microsoft's WGA is worse than guilty until proven innocent. With Microsoft, you're always guilty until you ditch their OS and their business model.

I toe the line and use Microsoft for business. My company has paid for my Windows notebooks and I am responsible for keeping them in excellent condition while they are in my possession. But for personal use I have become nearly Microsoft free. I still labor under the delusion that I need one copy of Windows XP for the little gaming I still play on the PC. I had a minor epiphany this past week that the last PC game I'd played was C&C, and that back in late March. The only reason now I boot into Windows XP is to run Thunderbird and check my RoadRunner mail. Ironic that I boot a non-free OS to use a free mail client because I'm just too lazy to move my account information from the Windows partition on europa to the openSuse partition. The only question now is when I'll finally show enough intestinal fortitude to completely rid all my personal systems of Windows.

Angels of mercy in America

In an earlier post ("The horrors of Iraq") I raged against the monsters who burned a five year old boy. A CNN story now reports that he's coming to America to be treated for his burns, and hopefully, to be given his life back. In a nutshell:
Shortly after Youssif's story aired Wednesday, the Children's Burn Foundation -- a nonprofit organization based in Sherman Oaks, California, that provides support for burn victims locally, nationally and internationally -- agreed to pay for the transportation for Youssif and his family to come to the United States and to set up a fund for donations.

You can make a donation at the foundation's site by clicking here. There's a drop-down menu under the "general donation" area that is marked "Youssif's fund."

The foundation says it will cover all medical costs -- from surgeries for Youssif to housing costs to any social rehabilitation that might be needed for him. Surgeries will be performed by Dr. Peter Grossman, a plastic surgeon with the affiliated Grossman Burn Center who is donating his services for Youssif's cause.
Please make a donation if you can. And pray for him and his family that they travel safely to America and to a brighter future.

Saturday, August 25, 2007

It couldn't have happened to a nicer company

I had to log this so I could gloat. According to the article "Wii Beats the Xbox 360 in Total Worldwide Sales", the Wii has just edged out the XBox 360. And the XBox 360 has been out a full year longer than the "lowly" Wii. I know why, and it's more than just a dislike for Microsoft. I own a Wii and have since mid-January of this year. It was a late Christmas present from the family. What do I like about the Wii? It's cheap and it's fun. It's ultra compact compared to the 360 and the PS3 and it fits in unobtrusively and it's totally quiet. And it's fun. But to continue with the article:
An unofficial report from VG Chartz, a web portal that tracks the sales of game consoles, has suggested that Nintendo’s Wii has surpassed Microsoft’s Xbox 360 in terms of sales worldwide.

The latest statistics from the website said that Nintendo Wii stands first with 10.57 million consoles sold, followed by Xbox 360 lifetime sales of 10.51 million units. Now, Nintendo is the new market leader in both the home and handheld videogame console business. Nintendo Wii was officially launched a year later than the Xbox 360.

According to VG Chartz, the revealed figures are based on sample data from retailers worldwide, till 23rd August. The data shows that Wii went ahead in the race by an average margin of 2.3:1 each week.

To calculate the sales of the Wii, the site segregated the sales of the Nintendo Wii between three markets, namely Japan (3.46 million consoles), America (4.24 million consoles), Europe, Australia, and others (combined 2.87 million consoles).
I've been a gamer for decades, spanning all the way from the original Atari 2600 through the original Nintendo (I picked up at Fries in 1988 while on a business trip to Intel), through the Sega and Nintendo 64 and Play Stations 1 and 2 and on through PC games such as Quake 1, 2, 3, and 4, Doom, etc, etc, etc. I like computer games of all sorts. But something got lost along the way in the console wars. The price went way up and simple enjoyability went way down as both Microsoft and Sony chased after the so-called almighty and very small Über Geek Gamer demographic. That, and the impact on the wallet, which is certainly lighter with the Wii than the XBox 360 or the Play Station 3.

Well. Like I said, excuse me while I gloat. Like I've been gloating over Microsoft's mis-cues with its Zune and Vista roll out. In the mean time I'm going to fire up the Wii and relax.

Just how many Linux machines will Dell really sell?

I've been reading lately about how Dell is slated to sell just 20,000 PCs with Linux loaded on them. I've seen that number thrown repeatedly into the faces of an uncaring blogosphere by folks who obviously have no love for either Linux or Ubuntu. The source of the number seems to originate from this posting, which was itself quoting in part from an email the poster had received:
We will be launching a Linux based OS (Ubuntu) on the E520, 1505 and XPS 410 starting next Thursday, 5/24. We expect these systems to be less than 1% of our OS mix for the entire year which is ~20,000 systems annually. Please cover the huddle deck below with your team by EOB Sunday. If any questions come up, please let me know so I can address them before launch.

The goal of launching Linux is to continue to give our customers more choices to customize their new Dell. Providing more options to our Linux Enthusiast customer group will hopefully create even more Raving Fans!!
Now there are a lot of questions raised by these two paragraphs. For example, the "OS mix for the entire year" includes what operating systems? Just Windows? Windows and other versions of Linux? And is it across all machines (in other words does number of machines shipped equal number of operating systems shipped)? And how much less than 1% are we talking? Is is 0.9%, or 0.8%, or 0.5%? Or something even far less? And the length of time, is it a calendar year or a fiscal year?

In an effort to determine just how many systems Dell might ship this year, I have attempted to bracket the initial conditions and to find some statistics with regards to Dell's past performance. I want to see where that 20,000 might have came from.

First, let's consider the length of time. Dell's official announcement about selling Ubuntu on their machines was made within the last week of May. My own posting on the subject was May 26, which I tend to believe was the day it came out. So, rather than say that Dell will only sell 20,000 Linux machines for the whole year, it's better to say they expect to sell 20,000 machines for the last half of 2007. If we stick with the 20,000 number then we could say that Dell could sell 40,000 of their machines loaded with Linux within a full calendar year.

Second, let's consider the reach of Dell's efforts. Dell deliberately restricted the number of machines they would ship Ubuntu on, first to three, then several months later to just four. And it was also limited to just the US market. That market was later expanded to Europe. Now compare this to the total number of models that Dell sells across the world. I have some statistics in that area from 2006. In a lone ZDNet article from April 2006, I read that 57 million PCs had been shipped world wide from 1Q05 to 1Q06. According to the article Dell shipped 16.5% of those PCs, or 9,405,000 machines. Assuming that Dell will sell the exact number of machines for the next several years, through 2007 on into Q108, and assuming they were to ship exactly 1% of those machines with Ubuntu installed on them, then Dell would ship 94,050 machines during a full year, or 47,025 from June to December 2007.

So, while my numbers are higher, the 20,000 is certainly within the ballpark. The questions all this raises of course centers around what Dell's starting assumptions where that they came up with the 20,000 number in the first place, and I don't think we'll ever really know.

Regardless of what numbers you use, all the numbers are embarrassing, starting with Dell's basic assumption of 1% or less. I have never really believed that Dell would be successful. I always believed, and continue to maintain that belief, that the effort is a bone being thrown out to the community. I don't fully understand Dell's motivations, but it sure isn't to promote Linux. Otherwise Linux would have been offered as a selectable option on every machine they sell, in every market they sell in. And to rub salt into the wound, consider Apple's performance during the same time period from which I quoted Dell's numbers. Apple, in the same period (1Q05 to 1Q06) had a world-wide market share of 2.3%. That means they shipped 1.3 million of those 57 million machines. I know it's unfair to compare world-wide numbers to a percentage of a single vendor's number, but really. No matter how you slice it, 20K in the first year is pathetic.

If Dell meets that miserable target of 20,000, and there is no guarantee they'll even meet that, then Dell will be completely within its rights to conclude that the effort was a failure and turn it off. For Dell's Linux efforts to truly succeed then Dell would literally have to show a 10-times greater number of computers shipped with Ubuntu. Not double, or quadruple, but ten times (an order of magnitude). Otherwise it really is a waste of Dell's time and money.

And if it does fail, then it will leave a black eye on the movement that no amount of time will ever completely erase. And that, I think, is what this whole effort has been about. Set it up, hobble its scope, and then sit back and see what happens. It will fail, and by direct association so will Ubuntu and Linux. And Microsoft will wind up with a huge PR bonanza in its continuing fight against Linux. This carefully crafted trap isn't aimed at the geeks, it's aimed squarely at the business community and general buyers who are wavering in their commitment to Microsoft Windows. This has nothing to do with merit. It's all about marketing and maintaining market share, even for a convicted monopolist. The coming public failure of Dell's Linux marketing efforts will provide Microsoft with all the heavily documented ammunition it needs to support its position that Linux can't be sold, certainly not as a viable alternative to Windows.

Nano-review of Ubuntu 7.10 alpha 5, part 2

After booting this release of Ubuntu on a machine with an ATI graphics card, I moved over to my other box that has an nVidia-based Gigabyte 7600 GS AGP video card. I had no idea how Ubuntu would handle this card in a live situation. When I installed 7.04 on the machine it had, at the time, an older ATI 9600 SE video card. I replaced the ATI with nVidia after 7.04 was fully installed to take advantage of the nVidia graphics drivers.

The earlier experience with 7.10 and the ATI 9700 Pro was surprisingly good. The free driver gave me full resolution, and the desktop graphics effects were enabled and working. I was expecting a similar experience with the nVidia card. Unfortunately, it was not to be. When 7.10 booted, the desktop was in stock 1280 x 1024 resolution, and even though the Desktop effects tab showed normal effects, there were no effects. What immediately showed up on the desktop after boot was a notice that restricted drivers were available (see below). I found it odd that the notice was on the left of the screen instead of beneath the restricted drivers icon on the upper right of the screen. But it is, after all, an alpha.

Clicking the icon on the upper right in the upper panel kicked off the following sequence of dialogs.

The final dialog (Restricted Drivers, see below) provides a little lecture about the evils and problems of proprietary drivers. As a user of Ubuntu and long-time dabbler with similar distributions, I have seen this type of dialog (and warning) for years. And I believe it's time to drop it and move on. This type of language does not motivate the general Linux using public to rush out and convince ATI (AMD) and nVidia to release free and open versions of their drivers. Instead, it makes them pause and wonder what's wrong. And many then move back to Windows or another distribution (such as Suse) that don't seem to have this problem. The time has long since passed for this nonsense to stop on our end. If we continue with the attitude that it's free or nothing with video drivers, then Linux will continue to remain a niche desktop system no matter who is pushing us, or how hard. And that would be a shame. Ubuntu and other contemporary Linux distributions show considerable polish and quality. The attitude we show towards binary-only drivers, especially video drivers which we need now more than ever, is the proverbial cutting off our noses to spite our faces.

After installing the video drivers the installation process suggests a reboot of the system. You can't reboot the live session, because you haven't really installed the driver on the boot device; the boot device being a CDROM. Instead, I took a chance and rebooted the X desktop with [Ctrl][Alt][Backspace]. That was enough to enable partial functionality, specifically the graphical desktop effects I'd experienced on the ATI desktop. But once again we get a "restricted driver in use" bubble on the upper right desktop, which just happens to hide the bubble telling us a reboot is needed to fully enable those restricted drivers.

And finally, some simpler applications showcasing the graphical desktop effects such as transparency in the window borders, shadowing behind the windows, and real transparency in the shell window.

Normal desktop effects with a decent graphics card has come a long way in Ubuntu. When enabled the look is nothing short of beautiful, and in my eyes begins to rival the look you see on a Mac OS X desktop. It exudes a visual quality that all desktops will benefit from once it becomes widespread.

I also like the subtlety with which the features are offered. To my taste I believe they should be a little more configurable, but as Apple has shown so many times, less can be better than more. Far better. In spite of the lecturing on the evils of closed drivers, Ubuntu still makes it easy and straightforward to install said drivers, easier than say openSuse 10.2.

One other surprising note. Being a developer of sorts, I was surprised to see that Ubuntu is sticking with gcc 4.1.3 instead of moving to the 4.2.x release (4.2.1 being the current release). Release 4.2 has been out for a while, so I'm curious to read about (or hear from someone) why they continue to stay with 4.1, and if they have plans to move up to 4.2.

Overall I believe the progress between Ubuntu 7.04 and 7.10 to be quite good. Ubuntu, like any other distribution, is both a stand-alone product as well as a focal point for many supporting software technologies. The 'ecology' underlying all distributions is broad, deep, and rich. Across the board I see a constant maturing and polishing taking place, leading to a wave of distribution releases over the next six to twelve months that I believe will make a lot of folks quite happy. This strong diversity in the Linux ecology helps to create real differences between distributions to support real choice, rather than the stilted artificiality found with Windows. Putting aside the rancor and the politics of the first half of this year, everyone should stand back and truly appreciate what has been accomplished to date. I know that I do.

Friday, August 24, 2007

Nano-review of Ubuntu 7.10 alpha 5

Downloaded, burned, and booted into Ubuntu 7.10 alpha 5 just to kick the virtual tires. Impressed over all with the quality of this alpha release, and I can see (and appreciate) where Canonical is going with Ubuntu. Looks like lots of work under the hood, with subtle but very important changes to the desktop. Tools are also beginning to show maturity. I am truly glad that there is now an Appearance Preferences application that combines Theme, Background, Fonts, Interface, and Desktop Effects all in one spot. After using it for five minutes all I can say is why did it take so long? It might seem insignificant but it sure helps to quickly tweak and tune the look of the desktop.

The screen shot above shows both promise and a problem. First, the promise. Europa, the system on which this live CD is running, has an aging ATI 9700 Pro video card. The free version of the ATI video driver as delivered was able to enable graphical effects previously associated with Compiz and Beryl. In fact the idea of Compiz or Beryl is noticeably absent. Which is as it should be. The effects should just be quietly and reliably working away.

With the open driver included, OpenGL is enabled by default. That's the whole point of firing up glxgears. But glxgears illustrates a minor problem. You see two sets of gears not because two glxgears instances are running, but because the one instance was moved across the desktop. For whatever reason glxgears' drawing canvas is not properly synchronized with the rest of the application's window. Drag it around and it leaves artifacts on the desktop. It's certainly no show stopper, but it should be looked at and fixed before release. Oh, and glxinfo states that direct rendering is enabled ('yes'). And the frame rate from glxgears is an order of magnitude faster than when running on openSuse with the ATI binary-only drivers. I don't know how well Google Earth would operate, but in the past you needed to install ATI's and nVidia's binary-only drivers to get decent performance out of Google Earth. Which paradoxically killed glxgears' performance compared to the open version of the ATI driver.

Other little things I noticed are the jumping windows. Click on a window behind another, and the top-level window(s) jump up and then behind the window you clicked. It's cute, but the cuteness wears off real fast. Like the wobbly windows, which mercifully are only enabled when you select the Extra effects radio button on the Desktop Effects tab of Appearance Preferences. If I may offer a suggestion to the developers, would you please add the jumping windows to the Extra effects selection. The shadows and translucency are more than enough for normal effects, and produce a truly lovely desktop experience. Or better yet, allow the end user the ability to fine tune (select) individual effects in the Normal and Extra effects categories.

I believe Ubuntu's next release will once again raise the standard towards an excellent user experience on the Linux desktop. I look forward to updating rhea's current Ubuntu 7.04 installation with the next release, 7.10.

Oh. This post was made in the live session. It's slow when starting an application off the CDROM, but otherwise the experience is indistinguishable from an installed instance of any other distribution.

Thursday, August 23, 2007

The horrors of Iraq

There are stories so horrific in nature that they rivet you to the spot in shock for seemingly timeless moments until the shock begins to wear off. One of those moments occurred today when reading a CNN.com story about a five year old Baghdad boy who was doused in gasoline and then set on fire. And then seeing the pictures. An atrocity committed by adults.

I'm a Florida Democrat. Make no mistake; I despise the current administration. I despise the presidents smart-alec smirks, his choice of political chronies that help him dilute the powers of a Constitution he swore to uphold, his choice of staff members who help twist science to support a political agenda, to Karl Rove, one of the dirtiest political operatives to ply his trade in Washington for quite some time. And come 2008 I will work to sweep the trash of the Bush administration from Washington, to replace it with someone more qualified, regardless of political party.

But I have never doubted the need to invade Afghanistan nor Iraq. My belief may waver from time to time, but not this time. And not any more. The people who live in that part of the world, who could conceive of this act, let alone commit it, are no longer a people deserving of any consideration or respect as members of the human race. They are not even animals, for even animals, as vicious as they may get, can't hold a candle to such sadistic and evil acts committed by such degraded human beings as these.

There's only one solution to this. Take the fight to those who harbor and commit such acts. Provide our troupes with the material, good leadership and support necessary to hunt them down and kill them all. No surrender. No prisoners. No acts of mercy. There will never be peace in this world until such evil is completely expunged from it. And we should never stop until they are all dead.

Google Earth shows the sky on openSuse 10.2

Downloaded and installed the latest release of Google Earth, 4.2.0180.1134 (beta). Here's a simple screenshot of it running on openSuse 10.2 and algol.

The latest version has a Sky mode (YouTube product demo), which is selected via a switch at the top of the window. I've spent a few minutes just looking at some of the pre-selected destinations under the Hubble Showcase. The example above is Messier 101.

There's just something seductive about flying around interstellar space from star to star, galaxy to nebula. I feel like a kid again in the 60's when I bought my first telescope at K-mart in Atlanta, a basic Newtonian with three eye pieces and a sun filter. It cost $50 of hard-earned yard mowing profits. The place I lived in at the time had a high hill in the back yard that allowed me to look down and over the roof of the house and into wide open sky. I remember seeing Saturn's rings, Jupiter and its moons, Mars, our moon, and the transit of Mercury across the Sun. I even invited a fellow star-gazer and neighbor, Mike Langford, over to look through it. The telescope was bare boned, and I had to learn to constantly shift and focus by hand to keep targets of opportunity in view. But I did learn, and it quickly became second nature with time and practice. I hadn't really remembered it much until now, running this version of Google Earth.

Tuesday, August 21, 2007

SPAM ALERT: Mrs. Suha Arafat asks for my help

This one is a keeper for its comedic potential. The widow of Yassar Arafat wants my help in getting to her husband's "secret bank account and [c]ompanies." I reproduce the entire document, complete with original mis-spellings.
Dearly beloved,

This mail may not be surprising to you if you have been following current events in the international media with reference to the Middle East and Palestine in particular.

I am Mrs. SUHA ARAFAT, the wife of YASSER ARAFAT, the Palestinian leader who died in Paris.Since his death and even prior to the announcement, I have been thrown into a state of antagonism,confusion,humiliation, frustration and hopelessness by the present leadership of the Palestinian Liberation Organization and the new Prime Minister. I have even been subjected to physical and psychological torture. As a widow that is so traumatized, I have lost confidence with everybody in the country at the moment

You must have heard over the media reports and the Internet on the discovery of some fund in my husband secret bank account and ompanies and the allegations of some huge sums of money deposited by my husband in my name of which I have refuses to disclose or give up to the corrupt Palestine Government. In fact the total sum allegedly discovered by the Government so far is in the tune of about $6.5 Billion Dollars. And they are not relenting on their effort to make me poor for life. As you know, the Moslem community has no regards for woman, hence my desire for a foreign assistance. You can visit the BBC news broadcast below for better understanding of what I am talking about


I have deposited the sum of 20 million dollars with a security firm abroad whose name is withheld for now until we open communication. I shall be grateful if you could receive this fund into your bank account for safe keeping and any Investment opportunity. This arrangement is known to you and my personal Attorney.He might be dealing with you directly for security reasons as the case may be.

In view of the above, if you are willing to assist for our mutual benefits, we will have to negotiate on your Percentage share of the $20,000,000 that will be kept in your position for a while and invested in your name for my trust pending when my Daughter, Zahwa, will come off age and take full responsibility of her Family Estate/inheritance.

Please note that this is a golden opportunity that comes once in life time. In case you don't accept please do not let me out to the security and international media as I am giving you this information in total trust and confidence I will greatly appreciate if you accept my proposal in good faith.You are to contact me on this email address suha200@gawab.com and send me your full names and your contact telephone numbers,cell number for my attorney to contact your for us to proceed immediately.Please expedite action is needed.

Yours sincerely
Mrs Suha Arafat

I can imagine hoards of people from every corner of the Middle East descending on me like a Biblical plague of locusts if I were to so much as touch any of Arafat's money. And that doesn't even include every Federal agency, every police and security agency in Europe, and just about anybody else in the world who have any interest whatsoever. Yeah, I can see me trying to spend some of that cash. And in the process my life would become so complicated and involved it would make fictional character Jason Charles Bourne's life look like a walk in the park.

No thanks.

Monday, August 20, 2007

NetBeans 6 nightly build on openSuse 10.2 running JOGL demo

Here's a neat trick. I installed tonight's NetBeans 6 nightly build (20708201200) and then installed the NetBeans OpenGL Pack. From there I proceeded to create and simply run one of the JOGL demos, JOGL Vertex Program Refraction Demo. The demo is a slowly rotating glass bunny in the middle of a very (very) basic 3D scene.

Remember in the last post where I threw out some very lame graphics performance numbers using glxgears? In this example I have NetBeans 6 nightly builds installed on both algol (nVidia) and triton (ATI). I can't offer numbers as with glxgears, but in both cases the demos build and run flawlessly, and their performance is indistinguishable both in rendering as well as overall performance.

I have Java 6 Update 2 installed on algol and NetBeans is running with it.

Algol is a potent little Linux platform. Java 6 and NetBeans 6 are a combination ripe with possibilities on this platform.

More testing with openSuse 10.2 latest install on algol

Previously I posted about my experiences installing openSuse 10.2 on algol, the Dual Core Gateway. For the kind of work I am involved in and based on past interests, I like to use and write applications that take advantage of the graphics hardware on the machine. One of the tests I perform is the build and installation of TrollTech's Qt toolkit on all of my machines. I always download and build the latest releases. In this case I've been installing and using version 4.3.1.

As you can see below, this version of Qt installs and everything runs, and runs quite speedily I might add.

Because this notebook uses an nVidia graphics chipset and its native drivers (100.14.11), it appears to support every OpenGL feature that Qt provides in its framework. What's more, OpenGL under the nVidia GeForce 7800 Go appears to be considerably more efficient than under the older Gateway running with the ATI Mobility X700 and its native drivers. For a very simple comparison, glxgears on algol (nVidia) gives over 12,000 FPS, while on triton (ATI) it runs at a paltry 300 or so. Even fgl_glxgears on triton barely breaks 500 FPS. Once again this is a very simplistic and crude set of benchmarks. I distrust these numbers instinctively; the ATI low performance may be due as much to how I set up and configured triton as apposed to any intrinsic differences between the graphics chip sets. It certainly needs more investigation on my part.

Compiz/Xgl Performance Note

I mentioned the FPS reading from glxgears above. I should note that with Compiz and Xgl installed and running on algol, that glxgears only averages about 5,600 FPS. That's right. Turning on the eye candy with the current release of those packages on this specific platform cut my graphics performance by more than half. I think I'll wait a good while, and then test extensively before letting it run my desktop in the future.

Sound Update

I thought I had working sound. Silly me. I shut down and then powered back up at home, only to discover that once again I had no sound. Out of frustration I fired up YaST | Hardware | Sound, and had the Sound configuration tool resense and recreate everything. I don't know what other files it touched, but it in combined the first two lines in the sound file into one:
options snd-hda-intel enable=1 index=0 model=ref position_fix=1

# u1Nb.iVO3S7Pf7gF:82801G (ICH7 Family) High Definition Audio Controller
alias snd-card-0 snd-hda-intel
Rebooted and sound came back. Rebooted again, and sound came back. Rebooted a third time, and sound came back. Logged out and in a number of times between root and my local account, and sound stays enabled.

Beagle Update

As promised I killed and removed Beagle from my system. I'm not going to have Beagle consuming one of my two processors full tilt, whenever it feels the need, especially if the notebook is on battery. The system runs a lot quieter without it. I'll wait to see if the developers fix it on openSuse 10.3.

Upgrading Yet Another Notebook to openSuse 10.2

It had to happen sooner or later. Over the weekend I finally upgraded my Gateway M685 to openSuse 10.2. I'd already done this with the older M680, and it had gone surprisingly well. This notebook, with its nVidia GeForce Go 7800, should have gone smooth as silk. It didn't. There were other issues as well, and I'll catalog them. Bottom line is that after more work than I expected I have the notebook updated to 10.2, and it's working more or less the way I want it to work. Notably different between 10.1 and 10.2 is the fact that Compiz is not working on 10.2.

System Hardware

Here's a list of the PCI peripherals:

00:00.0 Host bridge: Intel Corporation Mobile 945GM/PM/GMS/940GML and 945GT Express Memory Controller Hub (rev 03)
00:01.0 PCI bridge: Intel Corporation Mobile 945GM/PM/GMS/940GML and 945GT Express PCI Express Root Port (rev 03)
00:1b.0 Audio device: Intel Corporation 82801G (ICH7 Family) High Definition Audio Controller (rev 02)
00:1c.0 PCI bridge: Intel Corporation 82801G (ICH7 Family) PCI Express Port 1 (rev 02)
00:1c.1 PCI bridge: Intel Corporation 82801G (ICH7 Family) PCI Express Port 2 (rev 02)
00:1d.0 USB Controller: Intel Corporation 82801G (ICH7 Family) USB UHCI #1 (rev 02)
00:1d.1 USB Controller: Intel Corporation 82801G (ICH7 Family) USB UHCI #2 (rev 02)
00:1d.2 USB Controller: Intel Corporation 82801G (ICH7 Family) USB UHCI #3 (rev 02)
00:1d.3 USB Controller: Intel Corporation 82801G (ICH7 Family) USB UHCI #4 (rev 02)
00:1d.7 USB Controller: Intel Corporation 82801G (ICH7 Family) USB2 EHCI Controller (rev 02)
00:1e.0 PCI bridge: Intel Corporation 82801 Mobile PCI Bridge (rev e2)
00:1f.0 ISA bridge: Intel Corporation 82801GBM (ICH7-M) LPC Interface Bridge (rev 02)
00:1f.1 IDE interface: Intel Corporation 82801G (ICH7 Family) IDE Controller (rev 02)
00:1f.2 SATA controller: Intel Corporation 82801GBM/GHM (ICH7 Family) Serial ATA Storage Controller AHCI (rev 02)
00:1f.3 SMBus: Intel Corporation 82801G (ICH7 Family) SMBus Controller (rev 02)
01:00.0 VGA compatible controller: nVidia Corporation GeForce Go 7800 (rev a1)
02:00.0 Ethernet controller: Intel Corporation 82573L Gigabit Ethernet Controller
03:00.0 Network controller: Intel Corporation PRO/Wireless 3945ABG Network Connection (rev 02)
04:09.0 CardBus bridge: Texas Instruments PCIxx12 Cardbus Controller
04:09.1 FireWire (IEEE 1394): Texas Instruments PCIxx12 OHCI Compliant IEEE 1394 Host Controller
04:09.2 Mass storage controller: Texas Instruments 5-in-1 Multimedia Card Reader (SD/MMC/MS/MS PRO/xD)

The Sound Issue

I've complained in the past that sound was not working on this notebook with Suse Linux 10.1 nor with openSuse 10.2 Live DVD. Regardless, I decided to upgrade anyway in preparation for the eventual migration to openSuse 10.3 when it become available. In spite of my initial willingness to accept a soundless notebook, it wasn't long (about 30 minutes after the initial installation) before I wanted to hear coordinated sound with streaming video. So I started to search via Google for others with this issue. After chasing about a half-dozen leads, I finally came up with this simple fix. I added one line of text (in red below) to /etc/modprobe.d/sound:

options snd-hda-intel enable=1 index=0
options snd-hda-intel position_fix=1 model=ref # added by hand
# u1Nb.iVO3S7Pf7gF:82801G (ICH7 Family) High Definition Audio Controller
alias snd-card-0 snd-hda-intel

Sure enough, one reboot later and I had sound.

YaST2 Repositories

I have the following repositories in YaST2 (YaST Control Center | Installation Source). The allow me to upgrade a number of key applications to the latest version.

http://ftp-linux.cc.gatech.edu/pub/suse/suse/update/10.2 (default Suse updates)
http://download.opensuse.org/repositories/mozilla/openSUSE_10.2 (Firefox
http://download.opensuse.org/repositories/OpenOffice.org:/STABLE/openSUSE_10.2 (OO 2.2.1)
http://packman.unixheads.com/suse/10.2 (non-free codecs)
http://ftp.skynet.be/pub/software.opensuse.org/xorg72/openSUSE_10.2 (7.2 final)
http://download.opensuse.org/repositories/X11:/XGL/openSUSE_10.2 (Compiz upgrades)
http://download.videolan.org//pub/videolan/vlc/SuSE/10.2 (non-free codecs, libdvdcss)

Problems with Compiz

I was able to upgrade to the final release of Xorg 7.2, and I picked up the latest Compiz releases as well. I was very pleased with the latest Compiz. It worked with the Desktop Preferences | Theme engine. Any change I made in Themes was reflected with Compiz. The problem is that when I went in to try and modify the behavior of Compiz, the Compiz effects applet which resides in Desktop Preferences Look and Feel crashed each and every time. The latest Compiz has a new manager, but it appeared to do absolutely nothing. In particular I wanted to turn off the window wobble effect, and I couldn't. I also wanted to turn off the window-transparency-on-select effect, which I also found I did not like. Watching the windows flash from fully opaque to partially transparent and back again when released, along with the wobbly window, tended to make me sea-sick over time (no, I'm not joking). So after putting up with it as long as I could, I just uninstalled everything that had to do with Compiz, along with Xgl.

Problems with the X Server

Before un-installing Compiz and Xgl I was having a problem with the X server when the system booted. I would get the following crappy text message on virtual terminal 8 right after boot:
There already appears to be an X server running on display :0. Should another display number be tried? Answering no will cause GDM to attempt starting the server on :0 again. (You can change consoles by pressing Ctrl-Alt plus a function key, such as Ctrl-Alt-F7 to got to console 7. X servers usually run on consoles 7 and higher.)
I would select 'No' and the X server would then start, and I would then log-in normally (as well as logout and back in to other accounts). I thought that uninstalling Compiz/Xgl might solve this problem (again via searches with Google), but it didn't. Instead I had to finally hack the file /etc/sysconfig/displaymanager:


That's right. I had to tell the display manager to stop trying to use Xgl and use Xorg instead. Since I was having this problem before I uninstalled Xgl, I'm wondering if there weren't issues with Xgl in the first place. Was Xgl crashing for some reason, and that's why the X server wouldn't fully start?

Final Thoughts

The system is working now, and it's quite stable. I selected the Gnome desktop this time, and to make Gnome work the way I wanted I selected all the DejaVu fonts:
Application font: DejaVu Sans Serif Condensed, 9 pt
Document font: DejaVu Sans Serif Condensed, 10 pt
Desktop font: DejaVu Sans Serif Bold Oblique, 10 pt
Window title font: DejaVu Sans Serif Condensed Bold, 10 pt
Fixed width font: DejaVu Sans Serif Mono Book, 10 pt
I also used the same fonts in Firefox, and told Firefox to use my fonts instead of those selected by the website. Lo and behold Firefox renders pages identically to the way Firefox does on Windows XP. In particular I use DejaVu Sans Serif Condensed as my default text for both sans serif and serif.

The only complaint I have right now is Beagle. Beagle starts up and consumes one of the two processors on this notebook (it runs a Core Duo) which is crap. I'm going to disable and uninstall Beagle just like I did under Europa.

Other than that I'm quite happy with openSuse 10.2 on the M685. Wireless networking, DVD playback, streaming media, USB peripherals, 3D graphics, the list of what works goes on and on. You always learn something when you install a distribution multiple times on different hardware. The tweaks and adjustments I've made to the Gnome desktop have made it quite usable for me. It rivals Ubuntu's default desktop I found so enjoyable when I installed Feisty Fawn. So far the combination of the Gateway and openSuse 10.2 makes for an excellent workstation, and it complements the Windows XP installation that I dual boot for work.

A Final Word to the Wise

I suppose folks looking for ways to slime Linux (and openSuse) would point to my problems as another example of Linux not ready for the desktop. Let me disabuse you of that tendency. I have installed Windows XP (and Windows 2000, and NT, and ...) over the years on DIY hardware as well as vendor machines. I have yet to install any OS without issues. With Windows it has always been, and continues to be, a driver issue. Drivers were either missing or else buggy, causing innumerable blue screens of death. It's great when an OS installs without any hassles, but sometimes you need to get in and just Fix It. Neither Linux nor Windows shines particularly bright in my book when it comes to a clean and tidy install. There's aways something that requires further attention. I'm not a fanboy of either, but a heavy user of both, and if need be I can be harshly critical of both. Don't use me to support your particular agenda. Just accept what I write for what it is; my experiences and some possibly useful data.

The Gateway running Compiz/Xgl before I finally (sadly) removed it. Another one of my favorite cheesy movies is playing in Totem.

Tuesday, August 14, 2007

Pokin' around Ubuntu 7.10 alpha 4 and openSuse 10.3 beta 1

Well, it's getting to be that time of the year when the next generation of Linux distributions are ripening towards release. So I went over to the openSuse site and grabbed a copy of 10.3 beta 1 and then over to the Ubuntu testing site and grabbed 7.10 Tribe (alpha) 4.

Trying to test openSuse 10.3 beta 1 turned out to be a bust, both on the Gateway notebook (running openSuse 10.2) as well as europa. openSuse refused to boot in graphical mode on the Gateway. After the grub selection the screen went black and stayed that way. On europa I got a lot further, and was even able to test the integrity of the DVD, which it passed. But after that the installation failed because it could not see the DVD. I've had this problem in the past (most notably on the Gateway M680 with Suse 10), but never on my desktop system. Two strikes and openSuse 10.3 beta 1 was out. Maybe the next beta will work...

Then I booted the Gateway with Ubuntu 7.10 alpha 4, and that worked rather spectacularly (see image capture below). Not only did it boot with the ATI X700 graphics chip in 1680 x 1050 resolution, but a number of 3D features were enabled, such as drop shadows around the windows and partial transparency on window borders for windows that don't have the focus.

The following is a closeup that shows shadowing and transparency.

One other feature in Tribe 4 was desktop switching. When switching using the keyboard ([Ctrl][Alt] left arrow or right arrow) the desktops slid left and right respectively, which was much preferable to the flipping as if spinning a cube. The only feature missing was the Expose-like feature that is currently available on regular Compiz and Beryl window managers.

Most other features seemed to work, such as wireless networking and USB mounting. I was able to plug in my Kingston 4GiB thumb drive and save off the screen capture to it. The only feature that seemed to have a problem was playback of video Ogg. I couldn't play back the Mandela sample video. I heard the audio, but the video did not display.

I think I'll wait until Tribe 5 before making a decision about installation. And from the looks of things, the good free version of the ATI driver seems to be spreading out to more and more distributions, which is indeed a good thing. Perhaps they'll re-spin Fedora 7 with the latest Xorg ATI video driver.

Saturday, August 11, 2007

Maureen O'Gara Disparages Judge Dale Kimball's Ruling, Tries to Prove the Bitch is Back

Maureen O'Gara, that intrepid American technology [sic] journalist [sic], took a swipe at Judge Kimball's ruling in SCO v. Novell. And she did it behind the ".NETDJ News Desk" byline on the cutting edge Sys-Con website, Open Source version. (If you want the entire article, Google for it. I refuse to provide a link.) How much braver can you get? So what was in the article that shows her writing fingerprints? Allow me to quote:
Judge Kimball dismissed the testimony of Novell management at the time Novell sold Unix to the Santa Cruz Operation, starting with CEO Bob Frankenberg who signed the APA without reading it, testifying that the copyrights were transferred; dismissed testimony from Santa Cruz management saying that's what they thought they bought and wouldn't have done the deal without them; dismissed evidence that Novell changed the copyrights in the code to read "SCO" when the deal went down; dismissed evidence saying Novell left the copyrights on the premises it turned over to the Santa Cruz Operation; and dismissed the now-famous 1996 Amendment 2 to the APA that was allegedly supposed to clarify that the copyrights were transferred, clarification the new SCO repeatedly sought again from Novell right before it filed suit against IBM in March 2003.

Instead he pretty much hung his hat on a declaration from Novell's former general counsel David Bradford, a last-minute witness that SCO couldn't depose because the time for depositions had passed, saying that he and Novell's outside counsel, a guy from Wilson Sonsini, changed the APA and omitted the copyrights to protect Novell in case the Santa Cruz Operation failed since Santa Cruz couldn't come up with the full price Novell contemplated for Unix. Novell submitted evidence showing that, as negotiations advanced back in '95, the APA was dumbed-down to exclude the copyrights and other IP.

Judge Kimball did not address testimony from a close friend of Bradford, swearing that Bradford had repeatedly told him he had nothing to do with the Novell-SCO deal.
And you thought Maureen had been banned from Sys-Con, didn't you? This is the kind of crap I've been reading from a multitude of pseudonyms on the Yahoo SCOX board for years now, especially over the last 12 months as the SCO v. Novell trial dragged on. Which tends to make me wonder. And this is the same writing style you can read on Maureen's very own G2News and LinuxGram websites (again you can go Google for them).

Friday, August 10, 2007

It's all over except for the crying

I've been on vacation on Amelia Island all this week, and even I found out about Judge Dale Kimball's ruling handed down in the SCO v. Novell trial today. Granted it was about four hours after the fact, but then I am on an island at the beach. Some things just take more precedence. But the bottom line is that Novell owns the Unix and UnixWare copyrights after all. That, and SCO owes Novell a considerable chunk (95%, I believe) of the monies it earned when it sold licenses to SUN and Microsoft. All that's left now is for the dominoes to fall down on SCO in the SCO v. IBM trial as well as the Redhat v. SCO trial. SCO is so fscked.

I'd offer up a link to the story on Groklaw, but I'm so bloody tired of error messages being offered up when you do link to Groklaw stories that I'll offer up this link to The Register's story instead. You can then navigate to Groklaw on your own.

As wonderful as this news is, what would really put icing on all this for me is, when SCOX finally implodes into its own black hole, that it pull the now-absolutely-insufferable Pamela Jones in with it. With this near-to-final resolution in SCO v. Novell, there is essentially nothing left to keep her ego (or her smarmy writing) in check.

A Subset of the Ruling

Here is a copy of a message with the conclusion of the ruling made by one Mr. Sizz on comp.unix.sco.misc. I copy it here because it will not be archived, and will be removed on August 17. Note that I have made bold certain portions, that that I have added the bullets in the last paragraph in order to make it easer to read.

Complete ruling on TuxRocks: http://sco.tuxrocks.com/Docs/Novell/Novell-377.pdf


For the reasons stated above, the court concludes that Novell is the owner of the UNIX and UnixWare copyrights. Therefore, SCO's First Claim for Relief for slander of title and Third Claim for Relief for specific performance are dismissed, as are the copyright ownership portions of SCO's Fifth Claim for Relief for unfair competition and Second Claim for Relief for breach of the implied covenant of good faith and fair dealing. The court denies SCO's cross-motion for summary judgment on its own slander of title, breach of contract, and unfair competition claims, and on Novell's slander of title claim. Accordingly, Novell's slander of title claim is still at issue.

The court also concludes that, to the extent that SCO has a copyright to enforce, SCO can simultaneously pursue both a copyright infringement claim and a breach of contract claim based on the non-compete restrictions in the license back of the Licensed Technology under APA and the TLA. The court further concludes that there has not been a change of control that released the non-compete restrictions of the license, and the non-compete restrictions of the license are not void under California law. Accordingly, Novell's motion for summary judgment on SCO's non-compete claim in its Second Claim for breach of contract and Fifth Claim for Relief for unfair competition is granted to the extent that SCO's claims require ownership of the UNIX and UnixWare copyrights, and denied in all other regards.

Furthermore, the court concludes, as a matter of law, that the only reasonable interpretation of the term "SVRX License" in the APA is all licenses related to the SVRX products listed in Item VI of Schedule 1.1(a) to the APA. Therefore, Novell is entitled to a declaration of rights under its Fourth Claim for Relief that it was and is entitled, at its sole discretion, to direct SCO to waive its claims against IBM and Sequent, and SCO is obligated to recognize Novell's waiver of SCO's claims against IBM and Sequent. Accordingly, Novell's motion for partial summary judgment on its Fourth Claim for Relief for declaratory judgment is granted, and SCO's cross-motion for summary judgment on Novell's Fourth Claim for Relief is denied.

Finally, the court concludes, as a matter of law, that the only reasonably interpretation of all SVRX Licenses includes no temporal restriction of SVRX Licenses existing at the time of the APA. The court further concludes that because a portion of SCO's 2003 Sun and Microsoft Agreements indisputably licenses SVRX products listed under Item VI of Schedule 1.1(a) to the APA, even if only incidental to a license for UnixWare, SCO is obligated under the APA to account for and pass through to Novell the appropriate portion relating to the license of SVRX products. Because SCO failed to do so, it breached its fiduciary duty to Novell under the APA and is liable for conversion.

The court, however, is precluded from granting a constructive trust with respect to the payments SCO received under the 2003 Sun and Microsoft Agreements because there is a question of fact as to the appropriate amount of SVRX Royalties SCO owes to Novell based on the portion of SVRX products contained in each agreement. Furthermore, because Novell has obtained the information that it would otherwise obtain through an accounting during the course of this litigation, the court denies Novell's Ninth Claim for Relief for an accounting. However, the court also notes that SCO has a continuing contractual obligation to comply with the accounting and reporting requirements set forth in the APA. Accordingly,
  • Novell's Motion for Partial Summary Judgment or Preliminary Injunction [Docket No. 147] is GRANTED IN PART AND DENIED IN PART;
  • SCO's Cross-Motion for Summary Judgment or Partial Summary Judgment on Novell's Third, Sixth, Seventh, Eighth and Ninth Counterclaims [Docket No. 180] is GRANTED IN PART AND DENIED IN PART;
  • Novell's Motion for Partial Summary Judgment on its Fourth Claim [Docket No. 171] is GRANTED;
  • SCO's Cross-Motion for Partial Summary Judgment on Novell's Fourth Claim [Docket No. 224] is DENIED;
  • SCO's Motion for Partial Summary Judgment on its First, Second, and Fifth Claims and Novell's First Claim [Docket No. 258] is DENIED;
  • Novell's Motion for Partial Summary Judgment on Copyright Ownership of SCO's Second Claim for Breach of Contract and Fifth Claim for Unfair Competition [Docket No. 271] is GRANTED;
  • Novell's Motion for Partial Summary Judgment on SCO's Non-Compete Claims in its Second and Fifth Claims [Docket No. 273] is GRANTED IN PART AND DENIED IN PART;
  • Novell's Motion for Summary Judgment on SCO's First Claim for Slander of Title and Third Claim for Specific Performance [Docket No. 275] is GRANTED;
  • and Novell's Motion for Summary Judgment on SCO's First Claim for Slander of Title for Failure to Establish Special Damages [Docket No. 277] is MOOT.

Tuesday, August 07, 2007

Comprehensive iPhone comparison to three other handsets

Information Week has an interesting comparison of the iPhone with the Palm Treo 755p, the RIM BlackBerry Curve, and the HTC Wing. David DeJean, the author, was looking at how these phones (the latest in their respective lines) handled Web browsing as compared to the iPhone. Along the way he came up with three rules to govern them all:
  • Rule #1: Does it have a full keyboard?
  • Rule #2: The more pixels, the better.
  • Rule #3: Browser features matter -- a lot.
I'll cut to his conclusions:
One thing that became obvious to me as I looked at these various Web interfaces is that data speed isn't as important as good software. Even with improvements like the upgrade of Windows Mobile to Version 6 and the spread of touchscreen support on smartphones, not much has changed. If you didn't like the PocketPC there's still a lot not to like about Windows Mobile and its Mini-Me version of Internet Explorer. Palm still hasn't delivered its next-generation operating system for the Treo line, and BlackBerry's browser remains essentially unchanged.

What the Treo, Blackberry and Wing have in common is a reliance on a menu-driven user interface (taken to an extreme in Windows Mobile), hardware that simply isn't up to the task of supporting a fully functional Web browser, and browser software that tries to do something like Web browsing with both its hands tied behind its back. Going forward these deficiencies will be even more obvious as Web services and Web-delivered applications get more sophisticated.

The good news, as you might expect, is the Apple iPhone. The genius of Apple is its ability, over and over again, to completely reinvent, from the ground up, the user interface for hardware, and to support it with brilliant software. Web browsing on the iPhone is a paradigm shift, a completely different experience -- just as the BlackBerry was, in its time, a paradigm shift.

The elements of the technology that makes the iPhone so different will find their way into other devices, just as the BlackBerry's thumbpad and push e-mail have become more or less standard on smartphones. Touchscreens and direct interaction with the Web page will become standards of their own sort because they've come along just in time as computing, both personal and business, moves to the Web.
Ignoring the overly lavish praise, the second paragraph stands out. Basically where Apple got it right and so many others get it wrong is a poor interface coupled with slow hardware.

The iPhone came out of the chute with a 620MHz ARM processor, while everybody else is half that speed. The processor rules, folks. It always has, and it always will. The last time I saw such as fast ARM in anything was Dell's Axim X51v with a 624MHz XScale 270 running Windows Mobile 5. And Dell has since discontinued that model.

And the interface. Love it or hate it, the iPhone's touch-screen-only interface means never having to use (or loose) another damn stylus. The industry's slavish adherence to the menu interface no matter what has simply transfered our frustrations from the dominant Windows platform to the mobile platform. And we, of course, have Microsoft to thank for that with its Windows Mobile platform. Palm attempted to modify that with its Pilots back in the day, but it was only a hybrid solution at best, with the menus hidden up on the top edge and still requiring a stylus. Apple blows all of that away, and in the process scares the bejesus out of the ultra conservative handset makers, particularly Microsoft. Is the Apple interface any good? Only time will tell.

Sunday, August 05, 2007

Reality sinks in (a little) with the iPhone

It's no secret I've coveted an iPhone since before its introduction. I, too, fell under the Jobs Reality Distortion Field, and I fell hard. The only thing that kept me from running out after an iPhone like a zombie after fresh meat was Apple's decision to use AT&T Wireless as its provider. That, and the fact I'd already been blinded by geek dementia over shiny new toys, and paid the price.

Now part of my willing acceptance of all things Apple springs not from the iPod, but from their more established computer lines of late, especially after Apple switched to Intel. I've watched half the Orlando SPARTA office use Mac notebooks in their day-to-day work, and I've seen the nearly-effortless way they've produced quality work, day in, day out. Of course part of the success is due to the quality of the people involved, not just the Mac. But what I've seen in this office just goes to reinforce what I've seen in the past (especially at Time Warner's Full Service Network before it was shut down), and that is highly creative people are most creative with the Mac. So shoot me.

But it appears that there may be a few flies in the iPhone ointment, at least with the first release. And after my absolutely piss-poor experience with the Nokia 770 (another first release), I have fostered a tiny but mighty skepticism about very portable technology, no matter who manufactures it. That means that no matter how much I stand there drooling over my shoes at the sight of the iPhone, there is this little daemon inside screaming that you can have my credit card when you pry it out of my cold dead fingers whenever it sees anything like the $600 iPhone.

So it is with some gleeful delight (schadenfreude, perhaps?) that I read Dwight Silverman's less-than-glowing review after he spent a month using the device. I didn't find his critique of the Safari browser on the iPhone all that surprising. I know that use of Safari is a mixed bag with the Mac crowd, with many of them prefering Firefox or Camino over Safari. My daughter, who has an iMac of her own for school, had Firefox installed on it within an hour of turning on her iMac for the first time. Firefox may not be perfect, but it works with a lot more sites than Safari appears to, whatever platform it runs on. I was surprised when Apple released Safari for Windows so that nascent developers for the iPhone could use it to test their Ajax widgets targeted for the iPhone. If Apple were really serious about using open standards, then anyone could use Firefox as well as Safari to develop content for the iPhone, not just Safari.

Another big minus against the iPhone which I have, and which Dwight touches on briefly in his article, is the sealed-in battery. Regardless of how long the battery lasts, the fact you have to send it in to Apple to have it changed (for $80) as well as pay for a loaner (another $30) is a ludicrous feature for a $600 device. Every major electronic device in that price range has a replaceable battery. Of all the cellphones I've owned over the last 17 years, every one had a replaceable battery and not a single one had problems with that fact. And there were a few times I was glad of the ability to swap out the battery. Like the original Mac, this is yet another example of design esthetic (Job's) outweighing engineering practicality.

But Dwight's review isn't all that harsh. He finishes his article with this piece of advice:
If you weren't one of the early possessors, but you're considering buying one, wait for the next version. The iPhone has a lot of potential, and it will surely influence what other phone manufacturers do. But for now, you're better off using something else if you're serious about getting your data on the go.
I'm curious to see how many folks with follow this advice. Apple has stated it intends to sell one million iPhones in this quarter, by the end of September. By comparison it took seven quarters for the original iPod to reach this number, and it took Microsoft from November 2006 to June 2007 (seven months) to sell that many Zunes (and Microsoft's Zune has dropped like a rock since then).

In any event my credit card is safe from the iPhone, at least until the next version comes out.

Saturday, August 04, 2007

Older Gateway M680 notebook upgrades flawlessly to openSuse 10.2

I've spent some pretty intensive work time on my M685 notebook over the last months. My first Gateway, the M680 has been wrapped up in storage since October of last year. The M680 is the Pentium M-based Gateway on which I installed Suse Linux 10.0 (before it became openSuse). In the past few weeks, tucked in between my primary job on OneTESS and proposal work with SPARTA, I slowly began to investigate if the M680 was worth the effort of upgrading the Linux partition beyond Suse 10. The answer is absolutely yes.

After installing and tweaking Suse 10, I had a pretty solid development system. I still had to add additional libraries and upgrade some of the default tools (such as gcc) to more current versions, but overall Suse 10 on the notebook was very satisfying and effective. Rather than automatically upgrade from Suse Linux 10 to openSuse 10,2, I decided to check out three distributions to see if I should upgrade to openSuse, or pick another distribution.

For testing purposes I chose Ubuntu 7.10 "Gutsy Gibbon" Tribe 3, Ubuntu 7.04, Fedora 7, and openSuse 10.2. All four of these have live CDs or DVDs into which you can boot and test before installation.

The first distribution I tested was Gutsy Gibbon. And it failed. It booted up into a black screen, and I was unable to access any of the text screens (Ctrl Alt F1, F2, etc). In spite of the fact that Gutsy Gibbon is an alpha release and this was to be expected, I was still a bit disappointed in that failure as I had really gotten on my feet with with Ubuntu 7.04 alphas on an older home desktop machine. I was hoping for similar success on the notebook, but that was not to be. The next distribution I tested was Fedora 7, and it exhibited the same failure. This was very surprising, since I was able to boot Fedora 7 on the more current Gateway M685. The only conclusion I can reach about this failure is that neither Ubuntu 7.10 Tribe 3 nor Fedora 7's version of Xorg supports the M680's ATI Mobility X700 graphic chip set.

The third distribution I tried was Ubuntu 7.04. This was far more successful. Everything worked, from video to hard drive to USB to sound, even the wireless network. It Just Worked. Well, almost. This notebook has 1680 x 1050 resolution screen. Ubuntu 7.04 would only support 1024 x 768. I wanted Ubuntu to support the video systems full resolution without any additional tweaks. It did on europa and on rhea, but not here.

The forth and final distribution was openSuse 10.2. I was able to download a Live-DVD of 10.2 and burn a DVD for testing. When the Live-DVD booted the M680 came up with everything working just as it had with Gutsy Gibbon, including the proper screen resolution. I was suitably impressed. The openSuse Live-DVD does not provide installation directly from it, but that wasn't a problem. I grabbed my 10.2 boxed set and installed 10,2 from its DVD.

Note that my idea of upgrade is a little different than most folks. I have three partitions dedicated to Suse; root (/), home (/home), and swap. When I 'upgrade' I perform a completely new installation over root. If there are any specific and special changes to files, I copy them off to a small save area in my home directory, and then merge or move them back as needed when finished. I have very few so it takes little time for meto do this. I've noticed that the need to do this grows less and less with each successive release of openSuse.

After installation I was further, pleasantly, surprised. Not only did I have full screen resolution of 1680 x 1050, but I was able to enable hardware acceleration. Even though openSuse 10.2 was released last December, its open driver's support of ATI hardware (at least on the various cards I've been able to check) is superior to other distributions, even those such as Fedora 7 that have been released later. All the distributions seem happy with the M685 which runs with nVidia video hardware, but only openSuse 10.2 seems capable of properly handling everything from the ATI X700 on back. And this is with the free ATI driver in Xorg 7.2 RC2 (which is what shipped with openSuse 10.2 in December 2006).

And because the wireless chip set was supported, even during installation, I was able to download the openSuse updates that have accumulated since 10.2's December release. The installation of 10.2 on the M680 was flawless and totally uneventful. It Just Worked.

In spite of the resounding success of the base installation, I still tweaked the system further. In particular
  • I upgraded the screen drivers to the ATI "non-free" versions according to these directions,
  • I upgraded Xorg to 7.2 final release according to the same directions,
  • I added the Videolan (VLC) repository to YaST, and
  • I added packman.unixheads.com/suse/10.2 in order to pick up other video codecs as well as the ability to play DVDs with Kaffeine.
The next few screen shots show the M680 at 'play'. openSuse 10.2 shows considerable polish over Suse Linux 10.0. I look forward to the final release of openSuse 10.3, and hope to upgrade all my openSuse boxes to this latest release.

In the screen capture above I'm playing back a DVD and running fgl_glxgears at the same time. This is a test of both the ATI driver installation as well as VLC's ability to play back DVDs. Note in the movie scene how the great director Peter Jackson (on the left) is expressing his intense displeasure with Simon Pegg's acting ability.

In the screen capture above, the same movie is still playing, but this time Google Earth has been started. As good as this version of the open ATI driver is, the primary reason for me to install the ATI video driver in place of the free version is that Google Earth runs horribly slow with the free version. With the ATI driver, certain applications that are key to me seem to run smoother, some considerably so. I've also noted how multiple graphics-intense applications appear to execute more smoothly together under the ATI video driver.

And finally, Trolltech's Qt 4.3.0. In an earlier test I conducted comparing Qt's execution on Windows XP and openSuse 10.2 on europa, I noted that not everything worked on europa. In particular the fancy visual effect with the flowing balls on the Qt demo's main screen. In the example above you can see it that at least the visual effect on the main screen is identical to the Window's execution. I was unable to find any differences between Qt running on openSuse 10.2 and the M680 and Qt running on Windows XP SP2 on the M685. I believe the reason for the limitations on europa is that europa is running with a much older ATI 9700 Pro, and there are graphic hardware features needed by Qt 4.3.0 that the 9700 can't provide.

The openSuse upgrade has breathed new life into the M680. I now use both machines for development and business work flow, with openSuse on the M680 and Windows XP on the M685. Both machines still dual boot between openSuse and Windows XP. And now that the M680 upgrade is finished, I intend to upgrade the M685 to openSuse 10.2 as well.