Socks and Sandals

According to Pankaj Ghemawat and Ramon Casadesus-Masanell of Harvard Business School free isn’t good enough to destroy Windows . I found this piece pretty fascinating as it is the first formal model of the Microsoft vs. Open Source debate that I’ve seen and there was plenty in there that was outside of the current zeitgeist. I had heard that Microsoft secretly loved piracy but I hadn’t really thought about what that would mean for Linux making inroads into China (implicitly referred to in the article). China is the biggest battleground for the Microsoft vs. Linux battle and pretty much the last one of any consequence in the operating system wars (barring a new kind of computational model, of course). It is pretty telling, though, that only two of the biggest Internet success stories of today are running Windows: eBay and Myspace. Shows where today’s developers heads are. Definitely give the article a good read.

This weekend, I was writing a paper that required some mathematical equations be shown. Apple’s Pages does not have an equation editor like Word does. So I fired up Word. Oops, I forgot to install it and the CD is back at the office. Damn.

Then, I figured I’d try LaTeX, but it was a hassle to generate the images the normal way. So I whipped up this little browser-based LaTeX->PNG formula generator so I could just drag and drop the generated images into my documents. I thought it was semi-neato so I’ve posted it up for all to see. There’s even a screenshot ;-) Have fun with it.

...”Why Windows Vista won’t suck ”.

I think this is the kind of thing that should really concern Microsoft executives. The title says it all: people will expect it to be an overpriced, underperforming piece of __. The media and consumer market hangs all over ever word from Apple, but Microsoft has to run out and tell people why their latest product is worth spending money on.

I hear rumors that they will try to steal Apple’s thunder with an announcement of that Origami device on Thursday, but think about it: we already know what they are going to announce. We don’t know what Apple will say. This is a no-brainer. How can you steal anyone’s thunder without any element of surprise?

With press like this, you know that you are your own worst enemy. This is exactly the kind of article they were writing about Sun a few years ago when they clearly didn’t get it. Microsoft needs a makeover. Fast.

So what if Vista steals all of Mac OS X Tiger’s cool features? By the time Vista comes out, Apple will have already shipped another release of their OS, Leopard, adding yet more features that Microsoft couldn’t hope to have copied in time; without another schedule slip, that is.

More importantly, by the time the next Microsoft operating system after Vista comes out, Apple will have shipped at least 6 more releases of OS X. Extrapolating from Microsoft’s current and past release schedule, one would actually suspect something like 8-10 more years, not just 6. The added features in OS X by that time will relegate Windows to the 99 cent bin at Wal-mart.

I’m pretty surprised that people are getting worked up over Microsoft stealing Apple’s features again. They’ve done it so many times in the past, I can’t believe its news any longer.

What’s ironic is that this is exactly the situation Apple used to find itself in with respect to hardware: twice a decade, Apple would get some hot new hardware that would put it at the top of the performance heap for about 6 months, only to find Intel quickly regain the lead for the next half a decade. Now that this can no longer happen to Apple (since they’ve switched to Intel, making that effect work for them), they are affecting the same half-decade period of decline and obsolescence on MS software by their aggressive release schedules, pertinent and well-implemented new features and ability to get it right for most of their customers. Except that when Microsoft comes out with a new OS, its arguably still worse than what Apple has out at the time (at least lately)...

I tell you what: Microsoft better hope that Apple really doesn’t want retail and whitebox hardware running OS X. If it does/can (legally), that will be the single greatest incentive to dump Windows, and with Google absolutely destroying them on the Internet front, they will be left with only their failing Xbox business to save them.

This post is in reply to this article by Tom Yager.

Tom,

Linux is just a kernel. The platforms you speak of are the myriad distributions that are built on top of that kernel. Right now, the Linux scene is too fragmented, with respect to the number of distributions that are out there, to capture the market share of any significant portion of business users. The choices can’t be differentiated by the average IT department and seemingly offer no real desktop advantage over what Windows provides. And there’s something to be said for having Gold or Platinum support, as well ;-)

However, the Linux platform has one advantage over Solaris and Java, and even OS X and Java. The desktop platform won’t last forever. More and more services are moving to the Web and other Internet transports. And you may find that Windows becomes irrelevant when Google and Yahoo (Linux and FreeBSD, resp.) and others like them power the majority of services of which the business user takes advantage. Salesforce.com (Linux), Gmail (Linux), Skype (Linux), Asterix (any *nix), Basecamp (Linux) you name it; its getting close. The users of the future will want their services wherever they are, and Windows can’t deliver that because people aren’t developing for Windows like they used to. The reason: Linux gives them the ability to create whatever they want, not just what Microsoft wants them to be able to. And, just like before, what business users use at work, they will use at home. It is only natural.

Linux will never beat Windows at its own game, that’s true. But the game is changing…

This morning, Microsoft sockpuppet Dave Massy posted a reply to an article with Mitchell Baker regarding Mozilla’s security relative to IE. I will reply to Dave’s egregious misuse of communication inline, with the quoted material in italics.

That’s an argument we can spend a great deal of time on and still not prove one way or the other.

This is proven over and over again all the time. Count up the number of security holes that were/are in IE versus the number in Mozilla, weight them by severity and IE is a clear loser by any metric.

Now I’m pretty confident that Mitchell doesn’t actually know the details of how IE is developed so I don’t fully understand the basis of the statement. As we develop IE we go through very thorough and stringent security reviews to ensure that every change is secure and does not expose the user to attack.

Who is this supposed to impress? The fact that Mozilla’s process is open to world review is a tremendous advantage over the closed-source development model of IE. As well, since it can be reviewed by anyone, security holes can be fixed much faster than that of IE’s, some of which are years old and will never be fixed, by Microsoft’s own admission. This would never stand in the OSS community. Also, on that same note, how can we be sure that the process is “thorough and stringent”? It might just be a guy in a room going to Yahoo! and if that works, testing is done… we don’t know. Finally, there cannot possibly be as many people reviewing IE’s code as there are Mozilla’s and therefore security bugs will be found quicker and fixed quicker than IE. Unless, of course, you open the source to IE…

The security of any browser is irrelevant to if it is part of the operating system.

What? In fact that’s extremely relevant. If Mozilla has a buffer overflow, it can’t affect the video driver or USB subsystem, for example. This is because its a normal process, confined to its own address space and nothing more. If IE, on the other hand, has some kind of overflow or corruption issue, the potential exists for a massive stability or security problem because some IE code runs in a IA32 CPU ring other than 3. Exactly how is that not pertinent to a discussion about IE’s relative security?

If we are to debate security of browsers then let’s bring in relevant arguments…

Yes, let’s. That includes you, as well.

...and accurate details about different possible attacks rather than rely on the irrational fear that because IE is part of the operating system it must be exposing OS functionality to the web. This is not the case as any software has access to the same set of OS APIs and can therefore expose the same set of OS functionality as IE.

So, let me get this straight: your argument is, since you expose kernel functionality for browsers for everyone, but you’re the only ones stupid enough to use them, somehow that’s better? Exactly how does the above statement make any sense??

Dave, please, before you post next time, make sure you’ve read what you’ve written. Other people will if you don’t and you’ll get more of this type of response.

It sure looks like Microsoft is up to their old tricks again; in this case, the old tricks are stealing the latest Mac interface innovations to put in Windows. I wish they’d at least try to come up with something original… maybe they’ll buy Apple by the time Blackcomb comes out.

John C. Dvorak recently posted an article entitled “The Death of Linux”. Just yesterday, a humorous rebuttal to John’s article was posted to LinuxWorld.

John failed to realize that Microsoft is, at its heart, a platform company, and the loss of that platform (i.e. relying on a platform outside of their complete control) transforms them into something else entirely. They will never move to Linux until they come to grips with what that means for them and their business model. Microsoft does not create the drivers it ships, for the most part, so for them to hack on the Windows driver layer to Linux and attempt to sell it would relegate them to the same market position as a Borland or a Sun without the hardware unit.

But that brings up the crux of the missed point: all of Microsoft’s moves indicate that they don’t know how to not be the 800-pound gorilla. They rely on their platform dominance not only to plan out their future but also for their very identity. John’s not talking about porting Office to Linux here, he’s talking about the complete and utter derailment of Microsoft as it currently exists.

Microsoft will never do this, at least not until its way too late for them. Linux is not dying nor can it be killed. By conventional means, that is… Linux seems, despite its communistic facade, very much like the United States of America in one key aspect. The USA is made up of a lot of diverse communities and interests, but when a major threat comes along, these merge into a (mostly) unified front. The Linux community is the same way, reacting to “threats” such as competition from other OS projects and FUD from corporate entities.

I believe that Microsoft will eventually be marginalized as a platform vendor; no one lasts at the top of IT, and they are clearly stagnating. But, once this happens, things will get very interesting for the Linux community. Without the galvanizing force that is Microsoft, the community may fragment due to internal struggles based on technology (unlikely) and/or politics, which I believe is much more likely, based on current trends w/r/t the Linux kernel (the purely political introduction of EXPORT_SYMBOL_GPL is the harbinger). The funny thing is, the Linux community pines for the eradication of Microsoft, but this could hurt them in the long run, unless they find another galvanizing force to rally around. Technology is not and can not be it alone; already its taking a backseat to politics in some areas. For their sake, I hope the Linux community starts thinking about these things now. And perhaps they can pick up a copy of The Watchmen for some inspiration…

Recently, Microsoft Research published a “paper” regarding the detection of hidden files on a Windows system, ostensibly for the purpose of detecting rootkits, Trojans, keyloggers and some forms of spyware. If you’re interested, the document is here (Office required, obviously). I can sum it up for you completely in one sentence: do a full directory listing, reboot off of a clean OS CD and do another listing and then compare the two listings, looking for files that didn’t show up in the first one. This completely obvious idea takes up 5 full double-spaced pages and is replete with popular movie references; the idea that one would need an analogy to understand such a blindingly obvious idea is actually insulting. It looks like the papers I used to write the night before they were due back in college. At any rate, I have a few thoughts about this paper.

The first is that I am glad that Microsoft is spending some effort in the anti-spyware realm. This is a much bigger problem than the average computer user realizes and even bigger than some IT personnel care to admit (to their customers, at least). I expect to see much more research to come out of Microsoft in the coming years with regard to spyware detection, prevention and mitigation, especially in light of their newly acquired Microsoft AntiSpyware product.

Having said this, I will reiterate that this “paper” describes an idea that, while currently effective, is so blatantly obvious that I can’t believe that they took the time to write it down, let alone take 14 screenshots of WinDiff in what was clearly an effort to legitimize the work. As well, they didn’t even take the time to seriously investigate how a piece of malware could circumvent this technique and what could be done about that (I have already thought of a way to circumvent this that I’m pretty sure they haven’t thought of).

The real issue with this paper is not that they wrote it, but that I believe that these Microsoft researchers will now attempt to patent it. It is no secret that the state of patents in the United States with respect to software has been seriously damaged for some time and this serves to bolster the frequency of frivolous patent application submissions. Personally, I would like to see Microsoft attempt to come up with something original in the anti-spyware space rather than waste my tax dollars in the pursuit of undeservedly extending their monopoly position.

When I was attending my first college, the entire university was decked out with DEC Alpha workstations in all the labs (except for Computer Science, which was using this little-known operating system called Linux). This was before the dominance of Windows and I am happy to report that this university has thus far resisted the move to Windows, having replaced the Alpha workstations with Linux boxes.

I took some time off of school to play pool and poker. When I returned, it was to another university; this one had a much better reputation (it is a private institution). It was also totally inundated with Windows. That is to say, except in the Electrical and Computer Engineering Department which used all Solaris machines on Sun hardware. I started working as a Solaris system administrator there and rekindled my lost passion for UNIXy-type stuff.

As well, the curriculum for Computer Science was switched entirely to Java-based programming the year I started there. No C or C++ (like my last school); just Java. They didn’t even cover C in my Programming Languages class. Forget about assembly. Luckily for me, I had been working with C and assembly for years prior to my college education. The administration felt that using Java in all Computer Science classes allowed them to focus on the more core aspects of the art like why O(n^3) is bad, m’kay?

Now, I missed the heyday of Sun Microsystems. I was learning to play pool and hanging out all night at truck stops while they were selling massive units of high-end hardware to dot-com startups and ISPs. I missed out on the feeling of superiority one felt at that time by staring down a rack of Sun iron every morning when you came in to work because you knew that was the best money could buy (this is, of course, assuming that you had never heard of the far superior Alpha processor). I also completely missed gaining the irrational feeling of Sun’s superiority, despite there always having been a more powerful system out there with another company’s name on it.

Given the position of Sun in the minds of IT professionals at that time, and their current position now, I’d like to say that I think Sun is finished in this industry. Sun no longer captivates any particular section of the market. Their hardware is too expensive for the college student or hobbyist and Solaris x86 is, has been and always will be a joke. Their operating system is nowhere near as open as Linux or any one of the BSD’s, and these platforms offer more for software development than Solaris ever did. And as far as their big iron goes, they are no longer even near the top. The Alpha was always faster than the SPARC, but now IBM and Intel have stepped up to clean Sun’s clock and totally remove them from the high-end server market. IBM’s zSeries servers are twice as fast for a sixth of the price of the SunFire 15K.

Now, given that position, one would think that Sun would be attemping to co-opt Linux as fast as humanly possible. This would allow them to immediately inherit hundreds of developers and applications, as well as step into a much stronger position in the minds of young developers (which is to say, against Microsoft). Except, they’re not doing that.

This is rather strange, to say the least. But it gets better. Microsoft and Sun just settled their long running lawsuit over the fiasco that was the Microsoft VM for Java. This effectively blinded Sun to what was really going on for years and Microsoft has finally delivered the killing blow by giving them just under $2 billion to fund joint development of to-be-announced integration products.

Now, Sun is fine with co-opting Linux’s desktops and applications and is now starting to talk the talk a little bit with the Java Desktop System, but this is way to little, way too late. No one is going to run out and buy much-slower-and-more-expensive Sun hardware just so they can run Evolution under GNOME when they can do that for a lot less money (not to mention faster) on the latest PC special from Best Buy or Walmart. And who but those most loyal to Sun would seriously consider them a competitor in the desktop market anyway, even desktops for workstations? I mean, they just moved off of CDE this year!

Lets get back to the settlement for a second. Microsoft gave them all this money for integration products. Raise your hand if you think that $2 billion is a lot of money to Microsoft. If you just raised your hand, you’re a winner. They have a lot of money yes, but no company on Earth can afford to piss away $2 billion on nothing and Microsoft had the legal firepower to draw that case out for a lot longer in order to just wait for Sun’s eventual death (which they seem to be rapidly approaching).

Sun is now infused with cash but cannot spend it freely. Microsoft knows this and will force Sun to work on projects that will benefit them or at the very least not threaten their existing position(s). Translation: no money for desktop stuff. Money spent on SERVER integration tools. So, in effect, Microsoft has taken Sun’s strength (the server arena) and forced them to spend time and effort on improving Microsoft’s own weaknesses (the server arena). In addition, Sun will also spend money on improving .NET as it will now interoperate with J2EE better than ever before, giving new projects even less of an incentive to choose Java as the development platform. All this plays right into Microsoft’s hands and leaves Sun spending found money improving someone else’s company.

Sun is going down fast. They need some serious help to get out of it. They need to drop the “Linux is Solaris’ little brother” attitude and get serious about Linux on their hardware (trust me, its the only thing that can save that tired architecture). They also need to get serious about Java on the desktop and possibly give up some control to see it gain wider adoption in that space. If that means opening the source to Java, fine. Do it.

One thing about that, though. The source for several Java VMs has been available for years. Blackdown has an excellent implementation of the Java VM and its totally open source. When people call for Sun to open source Java, they want Sun to give up ultimate control over the course of Java’s development, not necessarily the actual source code that comprises the VM and class libraries. IBM’s JDK has been superior to Sun’s on Windows and Linux for some time now; I have to believe that when IBM calls for Sun to open Java up, they mean “let go the reigns”, not “show me the source to something inferior to a product we already have”.

As well, Sun would do well to can its current management. They clearly have no idea what they are doing. They shift positions frequently and make just plain idiotic statements to the press and are clearly suffering from a delusion that makes them think they are the company they were 10 years ago. Earth to Scotty: Sun is sucking in every conceivable way now. Bad Boy Ballmer is not your friend. He wants to eat your lunch and stuff you into a trash can after he’s sucked the Java bones dry.

A friend of mine hit me on IM a few weeks ago, totally ecstatic. He had just been hired by Sun and was going to move there to work for them. Another friend of ours from college was hired by Sun right out of school and has always said good things about them (given that he was hired as an evangelist for them), so my buddy was looking forward to working for them and making things happen. Unfortunately, he never even got to really dream about it. His division was hit with layoffs before he even got to move up there (luckily) and now he’s interviewing like a madman at any place with an IT deparment. Wow. Bet he won’t be too excited next time Sun offers him a job.

Sun has so many problems its hard to pick the biggest one. Personally, I would say that its their management but others might pick their adherence to tired hardware, their unwillingness to move on OSS development and integration, their scattered public statements and image, their lack of an ability to fend off Microsoft now that they’ve taken Bill’s money. Take your pick. Microsoft will marginalize Java with this settlement and if Sun doesn’t move fast to the Linux side of the fence, they will go out like Netscape.

P.S. Sun, this postscript is for you specifically. Whoever decided to publicly propose the idea of opening the source to Solaris, fire them as fast as you possibly can. That is the single stupidest idea that could ever have come out of your company, ever. Should you do that, any key features that are in Solaris that are still not in Linux would be quickly absorbed and Solaris itself forgotten. I am normally a fan of Open Source, but in this case, it makes no sense for you to give up the only piece of software you have that is actually worth a damn in a half-hearted attempt to win over a handful of developers. Do yourself a favor and wait until you file for Chapter 11 protection before you open the source for Solaris.