Why would MS care?
It's bad enough that Microsoft is still using 1990's operating system technology. Why would they want to own 1970's operating system technology?
357 publicly visible posts • joined 29 Apr 2007
"Microsoft built C# and the Common Language Runtime (CLR) to kill Java, yeah?"
So did SUN invent Java to kill Visual Basic? Microsoft developed VB and JIT compiling technology years before Java was even imagined. Java was developed with the idea of undermining Microsoft, and was one of many business moves and counter-moves. But there were also customer requirements factored into this, there are a lot of developers who want to write managed code for Windows, and CLR provides technology superior to and more general than Java for that purpose. I guess if you start with a world view where these companies are heros or villains, then it makes sense to write about Microsoft's common runtime with moralizing weasel words.
Wow, do you know anything at all about software? Do you think porting UNIX to the PC (Linux) was more innovative that Dave Cutler writing a new operating system (NT)? Do you even know what SQL Server does? And Azure (the Red Dog OS) is an amazing innovative operating system. Whether their cloud computing service will be successful I don't know, amazon is really the current leader there, but the azure technology is way more sophisticated than the buggy Linux infrastructure amazon is using.
It says something when a program is free and it still has almost no market share. And of course in the wonderful culture of open source, you can never discuss the problems or show the benchmarks. One of the main websites that talks about open office carefully segregates its benchmark results so you never see a straight-up comparison -- because open office is three or four times bigger and slower than MS Office. The open office spread sheet program frequently just crashes after a few minutes of manipulating a big excel file. If they want open office to gain market share, there has to be more than just a disinformation campaign...it has to actually not suck.
Offering choice is good, even though Bing is not usually better than Google. I use it occasionally because it gives different results.
The only things Bing is better at is maps. Probably because MS has been doing that for a long time with their old Streets product. Bing maps are nicer than Yahoo or Google's.
Cloud computing has been a big busines success for Amazon, but it's hard to predict how Microsoft will do. The Red Dog operating system is an amazing piece of technology, probably Dave Cutler's last great work (he's in his 70s now). He did a research on the requirements of large geographically distributed data centers -- and grasping requirements is really the hardest problem in inventing an OS. Furthermore, on multicore processors, Red Dog is reported to hit its Amdahl-Law limit at 128 cores (compared to 4 cores for spin-lock-addled LInux and 8 for NT). It's also likely to be as stable as NT, while large Linux enterprise sites (like Amazon) have to expend a lot of money and energy keep Linux -- with its race conditions, unsafe file systems and broken entry points -- patched and functioning.
That said, Microsoft could fail by being late to the market, and offering a less flexible service than Amazon. We know Google fell flat by limiting the capabilities of its so-called cloud computing. Oil companies are not going to buy blocks of CPU time to run seismic analysis rewritten in python or java. Amazon sells exacty the computing and storage resources that people need to do almost anything. Microsoft falls somewhere in between, with its strategy, but I think they could still be erring by requiring .NET API programming instead of general Kernel 32 API. They are forcing a lot of people to rewrite code before it will run, and that can be a serious cost to users.
it seems like Microsoft is losing it, under Ballmer's non-visionary leadership. I can imagine them winning a share of the pad market. The real shame for Microsoft is they had an eBook team more than 10 years ago developing a pad computer, and they pioneered the multitouch interfaces we see on Apple's products, with Dave Kurlander's work on MS Surface. And they just dropped the ball. it's particularly hard to imagine them winning back the phone market andgetting people to write apps.
Microsoft might be able to get back into the pad market, but it's hard to imagine them recovering from their loss in the cell phone market. Here is a company that was working on the eBook more than ten years ago and just dropped it. The multi-touch interface you see on the iPhone and iPad was pioneered in Dave Kurlander's research on MS Surface ten years ago. And Ballmer failed to take the risk. This is what happens when accountants run companies.
Hotmail works fine for me on IE 8. Facebook on the other hand seems to do something different with every browser.
I don't like the fact that all webites (including hotmail) keep making their pages more and more complex. It's the web 2.0 syndrome, where they feel compelled to replace every default behavior with script that does something slightly different. Suddenly there are drop downs and mouseovers you can't select text you are typing...and god only knows what will happen if you hit the back button. It's not just hotmail that is doing this, it's almost every damn site you visit these days.
Luther, I think Carl Sagan knew a bit more physics than you, when he proposed the greenhouse theory for Venus. But then the average household pet probably knows more physics than you. Compressing a gas will raise its temperature, but being continuously under pressure does not continuously generate heat. Your suggestion that the high pressure is the cause of the high temperature of Venus displays a conceptual misunderstanding.
When is Microsoft going to put its JIT compiled javascript engine into IE? Everyone else has done this. And the hilarious thing is that Microsoft had the first JIT'ed js engine -- available in .NET, but not in their browser. Maybe javascript speed isn't that critical (IE's display code is respectably fast, and that's really what people should be benchmarking), but this just seems like an easy fix for them. They;re the pioneers of JIT compilering, from VB to the new common runtime, and it just makes no sense that they are still using their old javascript engine in IE.
Stallman has nothing to do with Berkeley UNIX (BSD). He didn't get programmers to start sharing code, he just politcized the practice for his own opportunity.
It is a myth that the NT networking stack is BSD code. At one point, Microsoft adapted some utilities like ftp and rcp from BSD code. The network code in NT was initially bought from a company called Spider Systems, but Microsoft was not happy with it and began to write a new network stack from scratch. In NT 3.5, the new network stack replaced the Spider Systems code.
Dell says Linux is safer because hackers are not targetting it as much. I'd agree with that, but it doesn't imply anything about the intrinsic quality of LInux. For those who think Linux is stable and reliable, I wish they could spend a few minutes talking with engineers at a large enterprise site. I've heard endless horror stories from friend at amazon about the instability of LInux.
Thanks for posting the pdf to that famous report. The Russians also studied it carefully, their rocket engine genius A.M. Isaev found a copy of it Peenemunde in 1945. The giant rocket engine he draws would never have worked -- the Germans didn't understand the problem of combustion instability back then, and in fact they were just plain lucky that the V-2 engine worked without high-frequency oscillations. But many ideas in the report were very forward looking.
Actually Dennis Ritchie was more of the inventor of C, but he and Ken worked together on UNIX and probably on the C compiler.
UNIX is not a version of Multics. Oh my god, is that wrong! UNIX was a reaction to Multics which was a complex and inefficient system that Ken worked on before we went to Bell Labs. This of course was in the days when UNIX was actually a simple system designed by a few visionary people, not the sprawling buggy bloat that it has become today.
It will be interesting to see what happens when Microsoft deploys its online office tools. They've been slow to get that product out the door. I think most serious users don't want ersatz versions of office tools (like Open Office or Google Apps). Ultimately, Microsoft has the expertise and spends big bucks doing research and user studies on office tools, while Google and Open Office have been able to simply reverse engineer the features of MS Office, But someone needs to be advancing the state of the art, and so far that has only been Microsoft.
Open Office is a reverse engineering of MS Office, so it certainly must violate a ton of patents. But Microsoft is not going to start an exchange of patent suits with SUN and certainly not with Oracle. That would be mutual assured destruction. Fortunately for Microsoft, OO is bigger and slower than MS Office and has not eaten their market, as SUN had hoped.
Open Office is typical of a lot of open source software, just a poor copy of something that a company spent a fortune researching and developing. Why would anyone with talent invest in new software creation if they are going to be ripped off by zealots with no talent?
Furthermore, OO uses far more CPU and memory resources. A few years ago, Zdnet did some measurements that showed how bad Open Office was. Writer used four times as much memory as Word and used more than 10 times as much CPU time! (reference: http://blogs.zdnet.com/Ou/?p=120)
Just recently, a pro-OO site has published benchmarks that show OO is still slow and bloated. They did one study of MS Word and a separate study of Open Office, but of course they never display a side-by-side comparison of the two. That's because their measurements show Open Office is still four times slower than MS Office. (reference: 2009 benchmarks on oooninja.com)
Folks talking about MS Office being bloated compared to Open Office. Look at the benchmarks. OO is huge and slow, three or four times slower than MS Office at many of the benchmarked operations. This was true when ZDNet measured performance several years ago, and it still true today if you look at the benchmarks on oooninja.com.
For all the probems Micorosoft is having with innovation, you can't hold the open source movement up as a good counter example. They have been totally monkey see monkey do. Copying Office, copying UNIX, copying NT features into Linux. The innovation today is coming out of corporations like Intel, Google, Apple; where you have professional engineers, systems management culture, and the ability for indivdual architectural vision to be realized. The answer is to innovation has never been to collectivize.
I worked briefly in Dick's e-books group (on ClearType), and I think he's right. Not that this problem of burocracy is unique to Microsoft or even to corporate culture, but they need to address it. Innovative projects like e-books, terraserver, photosynth have languished and failed to reach their potential.
I don't think I would hold the LAMP stack up as an example of how to be more innovative, every one of those components is inferior to the corresponding MS product. Linux is not as strong or sophisticated as NT. And when people say Microsoft has never innovated, it's pretty funny to imagine that porting UNIX to the PC is somehow more innovative than creating NT.
If you dig around for the TPC-C benchmark results for MySQL/InnoDB, you can see why those guys don't want anyone to do that test -- on comparable hardware MS SQL Server is in the same ballpack as Oracle or DB2, and MySQL is about 1/3 that performance. It gets worse for MySQL if you look at multiprocessor/multicore scaling.
And comparing python and PHP to .NET. The P languages (and lets pretend Ruby starts with P) are all pretty amateurish and poorly designed. Which is unfortunate, because they are used so extensively. But here is an example of Microsoft failing to follow through on innovation -- they pioneered JIT compiling (with Visual Basic years ago), and .NET contained the first JIT'ed javascript engine -- so why isn't that JITed javascript engine in IE?
There's no doubt that Microsoft Office is getting bigger, hopefully in part because of new functionality and not just "bloat". Some telling benchmarks of different office versions can be found, for example: http://www.oooninja.com/2008/07/benchmarking-microsoft-word-95-2007.html
However, their own benchmarks of Open Office show it to be vastly worse than MS Office. Opening a document on MS Office took 0.5 to 1.5 seconds. Opening a document with Open Office took about 8.5 seconds. http://www.oooninja.com/2008/05/openofficeorg-getting-faster-benchmark.html
Don't forget Amazon, they have long been running warm server centers. Ironically, it was Microsoft who did some pioneering work there, but then they went and built a generation of over-air-conditioned centers. Folks from MS migrated to Google and Amazon, taking the lore of warm data centers.
I think when the smoke clears, Google will still be in China, and this will be seen as mostly them attempting to shift browser share by generating a high profile news story. We will see.
The risk of using a browser depends on two factors, the intrinsic vulnerability of the program and its security holes, and the probability that hackers will attack it. It's hard to know the first factor objectively, I believe IE is probably the least vulnerable simply because it's been toughened by external attacks and the army of "penetration" engineers that Microsoft hires. The other browsers are relatively untested by this hostile environment, and if you believe they are invulnerable because they are "open source" or "better written", you are fantasizing. However, there is no question that hackers target IE, and that increase the risk of using it. If many uses shift to other browsers, at some point that will create a ripe opportunity for hackers to go after programs that are probably squishy soft in comparison to IE.
I don't have a high opinion of Windows Mobile. But what he is saying about Linux is true. You only hear this from people supporting big e-commerce sites, not from the fan boys who don't really have a deep knowledge of systems. I've heard many horror stories about the cost of getting Linux stabilized and patched on new hardware, problems with lost interrupts corrupting file systems, problems with how it deals with new CPU configurations and hyperthreading and NUMA, etc. After all this time, it is really remarkable that Linux is not more stable and simple. THe problem is that some smart people work on Linux, but a lot of bad programmers have their fingers in it too.
Companies like Google, Oracle and IBM use open source for purely propaganda value. They use the open source meme in their battle with Microsoft and Apple, to shift the market away from personal computing and more toward centralized server-based computing.
But when it comes to their real family jewels -- Google search software, IBM and Oracle's database systems -- they are not open source at all. This kind of hypocracy is widespread. Few people (including the many liberal journalists and college professors who go on about open source) will sacrifice their own intellectual property, but are quick to pontificate, feign altruistism, and point at the professional programming or artist communities demanding that *they* collectivize their labor and property for the good of society.
You're right about Microsoft not using .NET enough (eating your own dogfood, as they say in Redmond). The best example of that is the javascript interpreter in IE. .NET has outstanding JIT compiling technology, something Microsoft pioneered for Visual Basic many years ago, and it came with a JIT compiled javascript. But it's not in IE! Now years later, you have the other browsers showing off their new JIT'ed javascript and embarassing Microsoft in the benchmarks.
Talking with a friend who manages a large Java effort, they regret not having used C#. The real problem for them was the confusion of different class libraries. Perhaps if they had established some conventions at the beginning, but knowing their team of hackers, that would have been like trying to herd cats.
Cheaper to buy, not cheaper to maintain. Talking to some freinds at a large e-commerce company, I've heard nothing but horror stories about Ruby, Linux and some other open source software they use. The cost for them is the subsequent need to hire a team of programmers to constantly mix a cocktail of patches and fixes to keep these programs running. One manager told me it cost them around $60,000 to prepare a stable Linux kernel every time there was even a minor upgrade in server hardware. They've had to fix things like race conditions in the Ruby garbage collector. The impression is a lot of open source software contains amateurish and boneheaded mistakes even in fundamental functionality. They demonstrated Linux crashing just from running mulitple copies of the Linpack benchmark, its scheduler so defective that it started to lose disk interrupts and corrupted the file system -- from running a numerical benchmark!
The Government should cut costs, but they need to look past the propaganda and make sure they don't get saddled with ersatz software.
Eric Hoffer has claimed that "True Believers" often feel justified in immoral behavior. I think that is evident in the disinformation you hear from religion fantatics or any zealous follower of a political mass movement. Even the comments on The Register is littered with disinformation from computer fanboys of various sects.
That's unfortunate. It does not help that folks like Stallman are acting as the "Glenn Beck" of the software industry, with his anti-cloud agitprop.
Until Azure is up and running, there is really not a good alternative to Amazon's services, unless you can implement your work in Google's relatively limited cloud-like service.
Interesting looking product. I think Microsoft made a mistake in cancelling their e-Book project in the 1990s. A lot of innovative ideas (like ClearType and readability studies) came out of that project, but somehow they don't market something new until they feel compelled or someone else productizes it.