Just saw this on Channel 9:
http://channel9.msdn.com/posts/Charles/Garrett-Serack-Open-Source-on-Windows-and-the-CoApp-Project/

Apparently Microsoft is working on making their own open source repository with their own shallow forks of common open source software, much like most linux distributions.
This way you can download and build/install open source projects on a Windows system as easily as on a linux system (solving the dependencies and the lack of a build environment on a standard Windows installation etc).

I think it's an interesting move. Microsoft is just taking away another reason to go linux. One thing that is especially interesting is that Microsoft will be compiling and optimizing the code now. Up to now, most linux software that was ported to Windows was pretty much a second-class citizen. Projects like Cygwin make linux code work on Windows, but most of the time it isn't very optimized, and doesn't perform as well as on a linux system.
So let's get carried away with that idea for a while...
Wouldn't it be interesting if Microsoft managed to make things like Apache, mysql and php actually run FASTER than on linux, for a change?
The total cost of Windows servers may also go down, because eg unlike IIS, Apache won't have any limits on a non-server version of Windows.
So, in theory, just a simple Home Basic OEM copy of Windows would be enough to run a big Apache server.
And if all common open source software for linux would be available for Windows as well, then that is ALL you need to buy.

It will be interesting to see how far Microsoft will go with this project. But at the very least I think it's a nice illustration of the differences between Microsoft and the linux community.
Microsoft clearly acknowledges the benefits that linux has over Windows, and tries to bring them over to their side.
This reminds me of the times when Word Perfect and Lotus 1-2-3 were the main office applications.
When Microsoft introduced Office, they included support for importing and exporting these files, and they even allowed you to set up the hotkeys and various menu settings to behave like Word Perfect or 1-2-3.
The rest is history.
Posted on 2010-08-20 04:03:42 by Scali
Just another good ol' EEE initiative.
Posted on 2010-08-20 11:48:18 by SpooK
Embrace, yes... Extend, possibly. But Extinguish? I suppose the GPL will prevent that.
Posted on 2010-08-20 12:10:42 by Scali

Embrace, yes... Extend, possibly. But Extinguish? I suppose the GPL will prevent that.


Microsoft may not be able to control GPL'ed code, but they do control the Windows source code, and thus can control how "well" Windows works with certain applications.
Posted on 2010-08-20 13:47:17 by SpooK

Microsoft may not be able to control GPL'ed code, but they do control the Windows source code, and thus can control how "well" Windows works with certain applications.


Yes, but how exactly will they do that?
I mean, if they extend anything under GPL, they need to release the sourcecode too. So any Microsoft-optimized open source applications would be available to everyone for free. No need to bother with alternative forks/ports of the same application.
And the point for this exercise seems to be exactly to make things work WELL with Windows, since most open source applications already work on Windows, it's just much harder to set up than from a linux environment, and generally the performance isn't as good.
If Microsoft wanted things to work poorly and be hard to compile and install, they wouldn't need this project in the first place.
Posted on 2010-08-20 16:44:11 by Scali
To me this sounds like Microsoft's attempt at an AppStore for Windows, similar in concept to the Apple iPhone AppStore. Make a few highly visible open source applications like PHP, Apache, and Python available and in the not too distant future start charging for downloads. Revenue sharing would make a wonderful incentive I think...
Posted on 2010-08-20 19:58:47 by p1ranha


Microsoft may not be able to control GPL'ed code, but they do control the Windows source code, and thus can control how "well" Windows works with certain applications.


Yes, but how exactly will they do that?
I mean, if they extend anything under GPL, they need to release the sourcecode too. So any Microsoft-optimized open source applications would be available to everyone for free. No need to bother with alternative forks/ports of the same application.
And the point for this exercise seems to be exactly to make things work WELL with Windows, since most open source applications already work on Windows, it's just much harder to set up than from a linux environment, and generally the performance isn't as good.
If Microsoft wanted things to work poorly and be hard to compile and install, they wouldn't need this project in the first place.


Microsoft doesn't need to contribute to a GPL project directly. The can change Windows itself. They've done so before, but most of the time to maintain backward compatibility.

I'm not saying that they will absolutely use this as an EEE strategy, or spew the usual anti-Microsoft rhetoric that GPL zealots do... but Microsoft earned their reputation and general distrust by the open source community for reason... his name is Steve Ballmer.
Posted on 2010-08-20 20:00:05 by SpooK

To me this sounds like Microsoft's attempt at an AppStore for Windows, similar in concept to the Apple iPhone AppStore.


Sounds logical, a chess game between Microsoft and Apple.

Everyone knows this "change of heart" is about generating revenue, if only by assuring Windows remains the dominant platform, but the question is... what is the overall strategy for this arms race chess game?
Posted on 2010-08-20 20:06:48 by SpooK
Just because some software is GPL'd and open-sourced does not mean that companies like Oracle won't purchase companies like Sun and then sue companies like Google over their implementation of something like Java. You really need a specialist lawyer to read the fine print for you, and even then there is room for interpretation. With a track record like theirs, I don't trust Microsoft as far as the farm gate.
Posted on 2010-08-20 23:01:07 by Homer

Microsoft doesn't need to contribute to a GPL project directly. The can change Windows itself. They've done so before, but most of the time to maintain backward compatibility.


I really don't think that Microsoft is going to change Windows so dramatically that those half-arsed ports of open source software are going to work great as-is.
That's just not a very realistic option.
Besides, the presentation literally states that Microsoft is going to make shallow forks of the open source projects and do everything on a source code level.
So it's pretty clear to me that they're going to modify the source code where necessary, and then they'll have to give back to the community (for the GPL-ed projects anyway).
Posted on 2010-08-21 02:50:34 by Scali

Everyone knows this "change of heart" is about generating revenue, if only by assuring Windows remains the dominant platform, but the question is... what is the overall strategy for this arms race chess game?


As I already said earlier.... I think Microsoft just wants to make open source applications run best on Windows.
As long as the Windows kernel is more efficient than the linux kernel, properly ported applications should run faster too.
Posted on 2010-08-21 02:52:40 by Scali

Just because some software is GPL'd and open-sourced does not mean that companies like Oracle won't purchase companies like Sun and then sue companies like Google over their implementation of something like Java. You really need a specialist lawyer to read the fine print for you, and even then there is room for interpretation. With a track record like theirs, I don't trust Microsoft as far as the farm gate.


Well, in the case of Java, the source code was GPL, but some of the Java technology was patented.
Clearly at this point none of the linux-oriented open source projects contain any patented code.
Microsoft could slip in some patented stuff while porting... but I guess it's just up to the rest of the world to be careful about checking Microsoft's contributions. They don't *have* to include them in the main branch.
But I think that tactic is a bit too obvious. Besides, the linux community distrusts and hates Microsoft so much that they may not ever accept Microsoft's contributions in the first place... patents or no patents.
Posted on 2010-08-21 02:56:10 by Scali

Besides, the linux community distrusts and hates Microsoft so much that they may not ever accept Microsoft's contributions in the first place... patents or no patents.


... and now we've come full circle on the EEE strategy :)
Posted on 2010-08-21 03:29:22 by SpooK

... and now we've come full circle on the EEE strategy :)


Have we?
I still don't see how they could extinguish open source this way. Like I already said in my previous post.
To spell it out: in this case Microsoft doesn't have the dominant market share required, and if the dominant market (the linux/oss market, mainly LAMP at this point, apparently) doesn't adopt Microsoft's extensions, it's never going to work.
I don't really see Microsoft becoming the biggest competitor here any time soon, because they have two major disadvantages:
1) Unlike linux and various other open source platforms, Windows costs money.
2) Unlike linux and various other open source platforms, Microsoft is pretty much starting from scratch with these ports. They need to put in a lot of work before the Windows ports become competitive with the same software running on linux and other platforms.

Besides, even if Microsoft miraculously enough *will* pull it off (how unlikely is that, it requires the oss community to trade in their own coveted oss platform for their most hated, proprietary, commercial competitor), the code is still open source. So at any time, they can adopt Microsoft's extensions... either by using the code as-is (after screening by some patent specialists), or by re-implementing it in a free way (the very raison d'etre of GNU/FOSS/BSD and all that).
Various large commercial companies (eg IBM) have a common interest in maintaining linux, Apache and related open source projects, so finding the resources should not be that hard.
Posted on 2010-08-21 06:01:59 by Scali
I don't see adoption by the open-source community as a necessity or even a desire for the EEE strategy to work.

Microsoft only has to take the ball and run with it. Microsoft has the money and resources to out-pace and out-market the open source community at just about anything... if they are motivated enough... and not having to watch Ballmer prance around HQ like a fool chanting "developers... developers" is probably motivation enough :lol:
Posted on 2010-08-21 12:02:50 by SpooK

I don't see adoption by the open-source community as a necessity or even a desire for the EEE strategy to work.

Microsoft only has to take the ball and run with it. Microsoft has the money and resources to out-pace and out-market the open source community at just about anything... if they are motivated enough... and not having to watch Ballmer prance around HQ like a fool chanting "developers... developers" is probably motivation enough :lol:


So you think Microsoft can actually turn the LAMP userbase into Windows-lovers...
Well, I guess you have more faith in Microsoft's marketing than I do :)
Posted on 2010-08-21 12:40:42 by Scali

So you think Microsoft can actually turn the LAMP userbase into Windows-lovers...
Well, I guess you have more faith in Microsoft's marketing than I do :)


That's pushing into the realm of a false dilemma, much like your assertion that all of open-source needs to be extinguished for the strategy to be considered effective. Destroying open-source is actually the opposite of what Microsoft could possibly desire, from a business perspective... why would you destroy something that is not a threat and provides you something for nothing?

Also, to move forward in discussing this subject, we have to clarify what is meant by the term open-source . That term hasn't meant what many people still want it to mean, in a long time. In the last decade, it has meant anyone invested in open-source, and that includes the grass-roots code hackers as well as companies like IBM, Apple, Sun Oracle and IBM. With that clarified, let's continue.

Microsoft doesn't need to turn the LAMP userbase into Windows-lovers. Microsoft only has to ensure that Windows can run certain critical open-source applications, and ideally better than the equivalent "open-source" platforms.

Better can mean:
1.) Easier to install
2.) Easier to upgrade
3.) Easier to configure
4.) Runs integrated with Windows Services
5.) Runs integrated with Windows Logging
6.) Takes advantage of Windows-specific optimizations
7.) NOT HAVING TO COMPILE FROM SOURCE

You are also neglecting the huge potential in developing free and easy-to-use GUI front-ends for various command-line apps, including and especially for configuration. And I mean ones that actually work, not the millionth one-off sourceforge hosted project that attempts to do so and either is abandoned or is eaten up by a company and turned into a commercial product anyway.

In short, a minor shift in momentum is all that is needed or desired. Some inherently see this situation as Microsoft vs open-source in general, when in reality it is Microsoft vs IBM, Apple and Sun Oracle. In a way, this is just the next evolution of the now defunct Interix/SFU.

If you still have a hard time understanding where I am coming from, put yourself in the shoes of the primary target... not your average desktop user, but instead a typical company. A typical company treats the IT department as a cost center -- a drain on company resources that needs to be mitigated whenever possible. A typical company will have as few actual IT employees as possible and will invest in platforms that support that desire. What's cheaper than the controlled cost of Microsoft services contracts and dime-a-dozen "Microsoft Certified" sysadmins... especially with Google and Experts-Exchange being a click away? The lowest common denominator is usually in charge of running our infrastructure, not people who are capable of actually contributing to an open-source project or building their own servers.

If Microsoft can keep that typical company and their PHB's from being swooned by the "we'll saving you $$$" marketing from the competition, and firmly entrenched/happy with Microsoft-solutions, that alone is a victory well worth the cost/effort of this initiative.
Posted on 2010-08-21 14:04:44 by SpooK

That's pushing into the realm of a false dilemma, much like your assertion that all of open-source needs to be extinguished for the strategy to be considered effective.


I never said any such thing... I just said that I didn't see how Microsoft would extinguish ANY open source AT ALL. Else, it's not an EEE strategy in the first place. It would get stuck after the first two steps, and the crucial third step would fail.

Destroying open-source is actually the opposite of what Microsoft could possibly desire, from a business perspective... why would you destroy something that is not a threat and provides you something for nothing?


I wasn't the one who equated this project to EEE in the first place....

Microsoft doesn't need to turn the LAMP userbase into Windows-lovers. Microsoft only has to ensure that Windows can run certain critical open-source applications, and ideally better than the equivalent "open-source" platforms.


I never said *entire*... But clearly Microsoft would need to reach a certain 'critical mass' before they can get any control over the LAMP market.


Better can mean:
1.) Easier to install
2.) Easier to upgrade
3.) Easier to configure
4.) Runs integrated with Windows Services
5.) Runs integrated with Windows Logging
6.) Takes advantage of Windows-specific optimizations
7.) NOT HAVING TO COMPILE FROM SOURCE


I think 7) is exactly the opposite of what MS wants to achieve here, are you sure you bothered to look at the presentation? Most of it explains a build system.
Aside from that... this is easier said than done. Microsoft has a lot of catching up to do here.

You are also neglecting the huge potential in developing free and easy-to-use GUI front-ends for various command-line apps, including and especially for configuration. And I mean ones that actually work, not the millionth one-off sourceforge hosted project that attempts to do so and either is abandoned or is eaten up by a company and turned into a commercial product anyway.


I'm not neglecting anything, it's just that this is beyond the scope of the CoApp project at this time. Again, have you looked at the presentation? No mention was made of any kind of customization on this sort of level. It's purely about getting a linux-like build system and open source repository going.

If easy-to-use GUIs and good integration with Windows were most important, then Microsoft wouldn't need to bother, because they already have IIS.

In short, a minor shift in momentum is all that is needed or desired. Some inherently see this situation as Microsoft vs open-source in general, when in reality it is Microsoft vs IBM, Apple and Sun Oracle. In a way, this is just the next evolution of the now defunct Interix/SFU.


It's never Microsoft vs open-source... but only the Microsoft-side will understand that, I think.

If Microsoft can keep that typical company and their PHB's from being swooned by the "we'll saving you $$$" marketing from the competition, and firmly entrenched/happy with Microsoft-solutions, that alone is a victory well worth the cost/effort of this initiative.


I think Microsoft already owns that market. It's the more hard-core linux/unix market that they're now going after, I think. Which is more of a fundamentalist/idealist nature. None of the linux/unix admins I've ever met seemed to know much about Microsoft and its products in the first place, and had little desire in getting to know their products, so even if Microsoft's stuff really is better (whatever that means for them), they wouldn't know, and wouldn't want to know. They just use what they know, and that's *nix.
Posted on 2010-08-21 14:33:27 by Scali
I've looked at the presentation, but I am not convinced of anything. I'm not excited about CoApp essentially being the next evolutionary phase of the SFU initiative. I'm also not impressed by the disclaimer about Microsoft "not calling the shots" as the person funding you is always calling the shots... whether you realize/like/accept it or not.

What Microsoft says and does today is not indicative of what they have planned for tomorrow.

Moreover, the first stage of EEE is embrace, and you can't achieve that phase by being different or alienating the thing you are trying to embrace.

Your mention of the "hard-core linux/unix market" is what catches my attention more than anything. By all means, Microsoft should and would be concerned about reversing Apple's market-share, before they are ever worry about converting *nix geeks that represent an insignificant fraction of the total non-Windows userbase.

So, we are back to the beginning of this discussion... figuring out what the catch is, and I don't think any amount of speculation or guessing is going to help with that. It's definitely a wait and see situation.

Unfortunately, even time might not tell us the whole story, as things/intentions can change depending on the success/failure of any part of this initiative.
Posted on 2010-08-21 17:10:33 by SpooK

What Microsoft says and does today is not indicative of what they have planned for tomorrow.


Surely, if Microsoft gets any kind of opportunity to extinguish its competitors, they will take it (just like any other company... well you heard about what Oracle did after acquiring Sun for its Java IP... they sued Google and killed off Open Solaris).

Your mention of the "hard-core linux/unix market" is what catches my attention more than anything. By all means, Microsoft should and would be concerned about reversing Apple's market-share, before they are ever worry about converting *nix geeks that represent an insignificant fraction of the total non-Windows userbase.


Except when you realize that Microsoft is an Apple shareholder (and one of the reasons why Apple is still alive today, they would have gone bankrupt in the 90s if it wasn't for Microsoft's financial support)... Aside from that, although Microsoft doesn't have full control over Apple's platform, they can still sell major applications such as Office to Apple users (heck, that's how Microsoft started, by writing BASIC and business applications for other people's platforms. Guess who made AmigaBASIC?).
Apple isn't really a concern of Microsoft.

I don't really think that the open source world is a 'concern' of Microsoft either... the CoApp project doesn't strike me as a major undertaking with lots of resources directed to it, let alone that it is one of the main pillars of Microsoft's strategy.
I think at this point, Microsoft is just exploring the open source world, trying to map out what makes it tick, and trying to offer the same on Windows.
Their next move will depend on how successful they will be with setting up the CoApp environment and how people will respond to it.

Even the thought that Microsoft will extend the software significantly (let alone hacking their OS to 'extend' the software) seems extremely far-fetched for me at this point... Let alone that Microsoft will manage to maneuver itself into any position of control so that they could get such extensions adopted enough to frustrate the rest of the market.
Posted on 2010-08-21 17:38:52 by Scali