Slashdot Mirror
MS Anti-Trust Litigation - The Case For Standards
Ken Krechmer writes "Microsoft Anti-Trust Litigation - The Case for Standards is the title of an article which won first prize this year at World Standards Day. Since it offers a somewhat different proposed resolution of the Microsoft litigation, you may find it interesting. See http://www.csrstds.com/WSD2000.html to read and post if desired (it is available for free republication with attribution shown)." Not sure I agree with all of the conclusions, but the piece is very thoughtfully argued and constructed.
> frankly, i think the 'article' raises some interesting points, and most /dotters don't understand them.
Frankly, your response is exactly what any Microsoftie would say in response to criticism of MS. Let's analyse your arguments.
> let's see. most of the comments here immediately take the 'force M$ into O/S' position, others keep more in line with the article regarding 'standards'.
>well, how about crediting M$ with making the computer revoltion possible in the first place? how many of us would have our type of job today if it
> wasn't for windows spreading computer use globally? sure, standards and lower hardware prices play an important role as well, but if M$ hadn't
> set a standard all by itself [Windows], then we'd still be battling with interfacing VAX and NIXes and apples [mcintosh platform] would be
> sparesly spread in some well to do households.
The usual argument is that by creating *DOS*, MS created a standard. This poster points to one of the GUI interfaces for DOS as being the standard-setter, which is odd: by that point, the de facto PC hardware standard had already been created, based around an Intel processor, ISA bus, either IDE or SCSI peripheral intefaces, & the limit of 16 IRQs. This was not only MS's choice, but also a certain computer company known as IBM.
Then again, the software APIs did change somewhat between standalone DOS & Win 3.1, then again with Win 95, & again (most noticeably) with Win NT. But it has been documented that every time MS changed the API, another group of independant software vendors writing to the DOS/Windows API went out of business, unable to make the port to a new API & compete with MS at the same time.
Whether this was the intent of MS's upgrade has been hotly debated.
And what's so bad with VMS or UNIX, anyway? Both are reliable, robust OSes that banks & industry have used for years. And people with minimal training use every day.
> give M$ some credit, for chip's sake.
For what? MS does not innovate, it merely sees which way the industry is trending, then runs hard to make it look as if they are leading -- but leading in their own, twisted way. Remember artificial intelligence? Back in the early 1990's, this was supposed to be the Next Big Thing. MS got into this in & produced . . . Microsoft Bob, which was later salvaged as the paperclip in MS Word that everyone hates.
MS Kerberos & the brain-dead implimentation of DNS in Win 2000 is only the latest in a long line of puzzling design choices.
I find your typo ``computer revoltion" in that sense very apropo -- MS has made this indeed a computer revulsion!
> that said, i honestly think that the whole case and associted discussions should be marked 'redundant'. the government was too late, too slow, too
> assuming and the whole affair is [score 0, flamebait] at best.
Another typical MS response. ``This is boring, nothing insightful, nothing to see, it's all past history, let's move along with our lives."
I find it hard to debate an issue with someone who is too arrogant to even acknowledge you have a right to a different opinion on the matter.
> think about it. if the purpose was to shake M$ and scare them into rethinking their strategy, then the mission is partly accomplished. but breaking
> them up now, where new technologies and the internet give them a run for their money anyway, is somewhat childish.
Another typical MS response. ``Okay, we know we shoulda listened to Big Government, & now they have our attention. So let's knock off all this talk of punishing us. We just want to innovate & create good software."
[Yes, I know this poster didn't write that last sentence -- but Bill Gates did say it in response to the Antitrust case.]
> will the o/s-linux community come up with the next big thing? will they spend billions on developping and supporting [especially supporting] a
> standard setting innovation that is available to non-nerds at a low price?
MS has never innovated. Gates was asked point-blank once what innovation MS ever created, & he could only answer feebly that only their business model was an innovation. I guess if your business model is to gut your competition over creating a better product, this might be an innovation. But governments have done this in times of total war since history was first written.
As for supporting a standard, Linux & *BSD both support the next generation of TCPIP & SMP already. MS is just making promises that they will.
And I won't touch on how MS implimented Kerberos or DNS.
> M$ certainly has it's flaws and bullying is never a good thing. But now, the public/states are using M$ strategies to force changes that are not
> necessarily beneficial to consumers. Or how would you call the use of state monopoly on power to inflict changes on a company - and millions of
> users?
Another typical MS argument: ``Private industry makes mistakes, but would you trust the government to make the right decision?"
I find it interesting that frequently, after finding private industry has mismanaged delivering services to the public, the state has had to step in time & again to enforce standards. Or would you like to go back to the days before the FDA was created & use medicines laced with opiates, milk of unknown purity, & other products misrepresented as being good& beneficial for the consumer?
And I find it ironic that all of these apologists for Microsoft are avowed Libertarians when their boss Bill Gates is clearly a Limousine Liberal.
> finally, who is holding car manufacturers responsible for the fact that things start to break down and need replacement just as the warranties run out?
Before the government cracked down on them, car manufacturers offered no warranties at all. It wasn't that long ago when if you bought a new
car, drove it off the lot, & the front wheel fell off, all you got from the dealer or manufacturer was a lot of sympathy.
In case you don't understand, a warrantee means the purchase will last for THAT set amount of time. It will start to degrade after that time. you are arguing over a very stupid point.
> shipping software with bugs and making money on service packs is what pays the high development costs in redmond. nobody forces anyone to run
> with faulty new M$ software.
I haven't heard that argument from any MS apologist before. Probably because it is true: if MS released reliable software, they would make less money. It is to their benefit to write substandard software, & make the public debug it for them. No apologist would unwisely make a statement that could be turned against her or him.
> I convinced management at our company to run Win95 OSR2 on desktops and NT4 IIS4 on servers.
> no blue screens in months.
I guess you avoid that by rebooting & reinstalling frequently. And reinstalling when some cracker roots your webserver.
> given development times, all you gotta do, is wait for M$ to release SP3 or 4 and then upgrade and you got a reasonably stable platform to work
> with. our three linux nerds in the trial department are too busy messing with the o/s to really produce anything else for them company.
This is a non sequitor, at best a troll to get people to respond with how reliable Windows is over Linux. I should ignore it, but I can't help myself.
Yes, there are incompetent Linux admins. Yes, it can be installed wrong, insecurely, & run inefficiently. And there are inappropriate implentations for Linux -- sometimes Solaris or *BSD are clearly better solutions. Given the number of idiots out there, msiuse of Linux somewhere is inevitable.
But in my experience, it is far easier to diagnose & fix problems under Linux or any UNIX variant than under Windows. Vague error messages, incompletely documented interfaces & applications, & millions of lines of spaghetti code that functions thru amazingly weird kludges only lengthen the diagnosis process. And MS's own support people are either $10.-- an hour sweatshop workers who have no idea of what to do beyond ``Reboot, reinstall, blame the hardware" -- or MSCEs who know their books better than their computers.
If you haven't had Windows crashed on you, you haven't pushed it at all. And I speak as someone who spent several hours getting _The_Sims_ to work on my wife's computer yesterday -- about as long than it took me to get sendmail to function, for God's sake. And sendmail hasn't crashed my Linux box once, where this game locked her computer several times.
And if you are happy with MS solutions, then you are in a minority. Studies have shown that Linux & BSD installations are growing in both number & market share. And Apache has 60% of the web server market share to IIS's 20%. All of this without any organized marketing push.
Geoff
I think I see a trend here. Maybe for them it really would be easier to muzzle the entire internet than to produce p
The most important point the paper makes is this:
Hence the popular argument that splitting up MS may accomplish nothing. MS is a monopoly, but the solution is not the traditional solution. A new type of monopoly requires a new type of resolution.
...and a very well written paper.
It could have done a better job addressing the primary Microsoft strategy towards open standards:
embrace, extend, destroy.
1: Embrace: fund the standard (perl, kerberos) and make it a vital part of your OS. Get a large user base to follow you.
2: Extend: create "extensions" to the standard which are trade secrets, which no one else can implement.
3: Destroy: Leverage your userbase, using your new extensions, to destroy the old userbase. The standard is now closed.
This is a very difficult strategy to implement, and Microsoft is a master at it. Don't say they don't know how to innovate- they are brilliant criminals. Look at the AARD detection code! A self-modifying, debugger-defeating virus built into an OS. Now that's innovation!
The point here is that Open Standards are not enough. Microsoft has a lot of brilliant developers, and even if you have a nice open standard, it doesn't help if Microsoft releases a new "open standard" - government approved, mind you - every week. This is called "churn", and it means introducing artificially new technology every two years (95, 98, 2000, blah blah, new paperclip) and forcing your users to upgrade (subscription model). If you can develop new open standards faster than the world can keep up, you've beaten this system.
Never underestimate Microsoft. The world is littered with the corpses of companies that did.
--
What happens when you outlaw guns
They must break ASCII or all those HTML pages written in MS Word would use the correct character for ' instead of ?. I've had a friend complain that connecting to an MS POP server was difficult from his UNIX box (Around the same time as the PPP issues) but I've never done that so I can't comment there.
They've got their dirty little fingers in many of the content streaming protocols coming out now too. With their patented IP in the protocol spec (Don't get me started on patented stuff in "open" specifications) they can guarentee that their OS is required to view any sort of streaming content in the future. I'd be willing to bet that they come up with a web page delivery system that tunnels through one of those proprietary "standards" too, making it increasingly difficult to view any content storted on a MS box using a non-MS client. That's just "Competitive" and "Innovative" behavior after all.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
Hmmm....
HTML - I think webstandards.org would have something to say about MS' support of HTML 'standards'. There are published specifications about how HTML should work. They don't adhere to those standards. Should they or shouldn't they is another debate, but they don't.
MARQUEE - 'standard' tag in HTML? Yes, NS had 'blink' and got yelled at for it to, from a 'standards' POV
SSL - ack - I can't find the URL. We were just tracking down SSL problems in the latest IE last week. OK, OK, maybe not a full 'standards' issue, but someone was monkeying around with something to make something as basic as encryption which USED to work in a product NOT work in an upgrade. (cheap shot, yes, but the sites we were reasearching this on were coming to the same conclusion).
HTTP - Our HTTP headers indentifying pages we were creating as 'gzipped' quit working in later versions of IE (something like IE 5.00.2610 and below worked - above didn't). Either they DIDN'T support the HTTP protocal re: GZIP before then, and fixed it, breaking our scheme, or they DID work, and broke it. Our code didn't change - just versions of IE. Which was it? They were broke and fixed it, or worked before and purposefully 'enhanced' it to not work with the same headers which used to work?
Kerberos - I need say no more than go read some old slashdot articles on this topic.
These are just a few examples of 'standards' where MS wasn't quite up to par. Whether or not they SHOULD be is a different point.
creation science book
I have been using PCs since the early 80s - the CP/M days and even before. The major drag then was 'no standards'. You could not transfer a document from one OS to another - forget it, you could not even move a floppy from one machine to another. Or a casette tape. Or address a serial port in the same way on two machines. Etc.
When Wintel appeared and became popular, this changed - big time. Suddenly we could all share commands, floppies, documents, code... it was great.
And this is still the case. At work, all our tech guys use Linux machines. But they run VMWARE on top, because everything you do in a business environment involves MS.
When someone sends you a resume, it is in Word format. A spreadsheet, in Excel. A presentation, in powerpoint. A database, in access format. I work with people all over the world and it's simple, this is the standard.
People want standards. Me too. Sure, I can accept a Wordperfect resume, and spend an hour cursing trying to convert it to some other format - inevitably this takes and hour and only partially works. So do I want to keep life simple, or spend that hour? The former: all resumes must be in Word or ASCII format.
In Linux too: do I want to stick with RedHat or use a 'better' distri? Go to deja and ask a question and 7/10 answers are about RH. So that is my Linux standard.
What I am saying is, standards are good, becasue they enable faster growth - and they are made 'de facto', not by committeees. Don't knock it: people will always take the easiest way.
Michael
---
---
BDOS ERR ON A:>
Do we really want our government defining and enforcing protocols and standards for operating systems, desktop software and networking protocols? I'm no fan of Microsoft, but keep the government out of protocols and standards. I don't want to "elect" the best operating system, the best network protocol, the best script language, etc.
Someone you trust is one of us.
The government decided that there had to be standards after the 1920s fall, AND they do it intelligently--they have the various people in the market get together and agree on the standards, and then they enforce them.
And in software, we wouldn't be choosing the best method, we'd be choosing the standard method. If you have a better method, use that and convice people to switch.