Slashdot Mirror
What Should Microsoft's Open Source Strategy Be?
JWinterboy asks: "I'm guessing that everyone here has a valid criticism of Microsoft's attacks on, and approach towards the Open Source model. To me, that begs the question of what we think would be an "appropriate" reaction from Microsoft towards the Open Source model. It doesn't have a service arm, so IBM's approach isn't really viable. At the same time, non-service related business models haven't fared very well.
What would we like to see Microsoft do? How can it work with the Open Source community, leverage its resources, and still make a buck?"
I think Apple has proved Open Source's usefulness for businesses and the general consumer market. Yes, their license is strictly controlled, but look at the innovation that has come out of it. They have the first and only viable "Unix for the Masses(tm)".
There is no longer anything that can be done with computers that is nontrivial and clearly legal. -- Paul Phillips
Check it out: Microsoft Consulting Services.
They built GAP.com, among other things. Operations in 30+ countries and all that stuff...
Kevin Fox
If Microsoft's products are worth the money, then people will buy them without being coerced to by incompatible file formats, protocols, and APIs. Their strategy should be good citizenship in the software community (open AND closed source), by making a good faith effort to make interoperability possible.
I think a lot of the animosity toward Microsoft comes from the obstacles they put in the way of fair competition. Standards are the means by which software can compete on the basis of merit, and Microsoft takes advantage of the fact that pragmatically, a market leader's de facto standard speaks much louder than any written document.
What MS should do to work well with Open Source:
a) Document API's thoroughly, and keep the docs up to date
b) Standards: Microsoft is frequently the first one to implement a standard or to make it mainstream. As an example, XSLT comes to mind. AFAIK, IE was the first browser to support XSLT. As the first big boys there, they usually claim the right to make modifications to a standard or to fill in details in the standard. They could win a lot of goodwill merely by consult other companies and open source developers before as they implement the standard. This will greatly reduce (though probably not eliminate) the feeling of railroading that we all feel when MS' software doesn't follow standards, and we all have to deal with it.
c) Document and admit mistakes and bugs. One of the most infuriating things about Microsoft software, is that it either doesn't do what it says, as in undocumented behavior and bugs, or cryptic error messages saying things don't work unless the OS is configured right (which is true ipso facto, but somewhat accusatory, and certainly not helpful). I think this happens mostly because they can get away with it, and writing thorough documentation for your programs is not nearly as satisfying or financially rewarding as designing and writing the code itself. They could again improve goodwill if they were responsive to outside developer's questions about these bugs and behaviors, rather than being dismissive.
I'm sure there are more, but these sure would make it easier for an outsider to like Microsoft.
The principle difficulty with using Microsoft products is that they seem barely capable of communicating with anything but other Microsoft products. I'd like MS to consider putting all libraries useful for interoperability available in open-source (without the useless licence) form. That way, well, if their software was better than the free version one could use them, and MS and non-MS software could be used together...
Basically it doesn't seem that Microsoft can totally change to an open-source strategy now. Even if they weren't too embarassed/unrepentantly monopolistic to want to.
I don't really see that they would open-source the entirety of Office, but it'd be nice if Microsoft were to make owning Office an option rather than a restrictive locked-in technology (yeah, I know. Word viewer available, inconsistent specs available. Not quite the same as working source code).
In any case, if the arguments about Linux's unsuitability for the desktop are correct, they have nothing to fear - if Linux users were to create Word documents or WMV or whatever with the code they were graciously permitted to use, the average human being would prefer to buy a nice user-friendly copy of Windows and view them on that.
Of course, if somebody were to create a piece of word processing software that happened to be better than Word and utterly interoperable, they'd lose out, but we all know that'd never happen (yeah, right).
They have a whole friggin open source OS (Darwin) which they have grafted their own closed source technology (displayPDF, QuickTime, CoreAudio, etc), and are selling for $130, or bundling with their Macs.
They also have an open source Darwin Streaming Server, and a complementary closed source QuickTime Streaming Server. They bundle Apache as their HTTP server, as well.
What can Microsoft do that would be similar?
How about release the DirectX library as open source? However, use their own in house optimization-compilation technology to ensure that their own DX libs are 10% or 15% faster than anything out there... IE, outinnovate the competition, themselves?
Or release their older Office programs as open source? Sell newer, more advanced copies, but allow the general public to self support and modify their older versions? Of course, again, the key is to out innovate yourself to convince people to buy the newest version instead of incrementally updating and fixing the older, free source version.
Or rather, release a Office Core, which allows you to compile a very basic Office devoid of nifty features... though this might backfire, as people don't generally use 80% of the features in Office, do they?
GPL Deconstructed
Well, it is. Now, whethor or not a threat to capitalism is a good or bad thing is left to the reader to determine.
I disagree. Capitalist businesses will benefit greatly by not having to pay for restrictive software licenses.
Although I don't have hard data, I would venture that most people in software are not employed writing and testing closed source products that are sold, but making custom mods for internal use, supporting installed systems, doing system installation and integration and other services. These endevours can all benefit from Open Source.
Furthermore, the closed source companies seem to be doing OK. Microsoft is making record profits. Oracle, Siebold, SAP all seem to be unaffected, so far, from Open Source.
Open Source represents competition to the Closed Source companies, but I believe that everyone benefits from competition. For example, the improved reliability of W2K and WXP over earlier offerings is, IMHO, a direct reaction, to some extent, to Linux and FreeBSD. I think that MS has actually benefitted from this renewed focus on stability. You can actually learn your best lessons from your competitors, if you are listening.
All this speculation about how OSS will kill the software companies is, so far, just speculation.
Actually, it will have to be a lot better than windows
Interestingly, I was at an Entrepreneurial Conference put on by SEA (www.sea.org.au) in 1999, and a gentlemen pointed out that you'll never be successful making a better product, You're successful by making your product different.
To be quite honest, open source products are not going to be chosen simply because they are "better" - you have to show the consumer what's in it for them, what the product is going to give them over the competition.
One cannot think of Microsoft products individually, the difference/value that Microsoft provides its customers is a family of integrated/all work-together products. That's where Microsoft's success is: in it's product cohesion.
Cohesion/Consistency is what the consumer wants and ironically are willing to put up with a few BSODs every week (tho if you've used XP, this is a hell of a lot less...). Most "Joe Average's" I know associate "free" with "cheap/nasty". Until such times as Open-source products can get past this mis-informed attitude, then it will be relegated to the back office and those adventurous souls that actually know better.
As for Microsoft and Open-source co-existing? I think today that Microsoft would probably be happy as far as the consumer market is concerned... However, in the server arena they are more worried...
-- Dan "Maybe I should have done marketing instead of Software Engineering" Thomas =)
I think a great many people would be satisfied if Microsoft would simply keep their interfaces, configurations, and standards open and reasonably constant. It's the hidden stuff that makes my applets and programs break. It's the secret "upgrades" hidden in dll libraries amounting to only a few bytes code change but which also happen to completely break a competitors program, that irritates people.
Who really CARES about microsoft code? Get the API and hooks out in the open so we can SEE when they're deliberately forcing you to replace that "win95 only" application that still works fine but somehow doesn't run under win98 or XP. That's the "open source" I want.
No, this isn't flamebait. I keep a collection of system files archived because about once a year microsoft releases an "update" that breaks one program or another. I've seen this since MS deliberately broke netscape with a small dll file and Netscape support was forced to redistribute that dll file as a fix. Get the standards in the open and we'll be happier than we'd be with the actual code.
You haven't dealt with Sun lately, have you?
/.
They're the 'dot' in
rm -rf
Embrace and extend. What else? Or were you wondering what their strategy should be if they did NOT want to dominate the whole freaking world? That's kind of academic.
In fact, the strategy they have is a damned good one. It'll be even better if nobody clues to it in time, which is why I particularly delight in outing it here. This is my interpretation, and they may possibly phrase it differently- or not. Maybe in the NEXT antitrust fiasco this will come to light.
MICROSOFT'S OPEN SOURCE STRATEGY
This. Is. What. They. Are. Doing.
Note that it plays to their strengths, including the strengths they've learned in the antitrust trial, of barratrous lawsuits and dragging things out endlessly, and note the brilliance of embracing and extending, not the openness of collaboration, but the concept of a viral license. This is brilliant conceptual work on their part, it really is.
But it does not have to succeed- because they really need people who are KNOWN to have agreed to their license. They can't really go around suing everyone who writes open source and dragging them into court and saying, "You DID agree to the Shared Source license, didn't you? Everybody does! You had to have!". That won't fly- people who can legitimately say they've never agreed to that license are in a position of strength.
However, people who have in fact agreed to their viral Shared Source license, EVER, are fucked. And can never be allowed to participate in open source or free software development- because of the legal exposure.
Given this state of affairs, why would Microsoft ever need to find another open source strategy? This is unquestionably the best one for their goals. Yes, it's evil. And your point is?
It's true that Apple's end goal is selling more hardware. The particular way in which open source has done this, however, it to make their hardware more attractive by raising the quality of the software that it will run.
So, Microsoft could use open source in manner parallel to Darwin (and Apple's treatment of Apache, SSH, Perl, etc etc) to improve their software. Whether or not they're a hardware vendor, improving their software should make it more attractive to customers, and thus Increase Shareholder Value.
Actually, I suppose that competing on the cutting edge of quality is a novel strategy for MS. But heck, if they wanted to start doing that more more often....