At the Linux Collaboration Summit, held last week in San Francisco, an interesting panel discussion took place about Linux’ position in the wider operating systems market. Included were Jim Zemlin, Linux Foundation executive director, Ian Murdock, Sun community and developer vice president, and Sam Ramji, Microsoft platform strategy director. Titled “Why Can’t We All Just Get Along?”, the discussion focussed on Microsoft’s somewhat dubious relationship with the open source community.
Ramji explained that his role at the company is to educate people in and outside of Microsoft about open source, and how important open source software is for the Windows platform. More and more people are interested in running open source software like Apache and PHP on Windows servers, and this has led Microsoft to tighten its bonds with the Apache project. As Ars notes, “Microsoft joined the Apache Foundation last year and has already contributed some of its code to an Apache interoperability effort relating to service-oriented applications. Microsoft also adopted Apache’s permissive open source license for its Web sandbox initiative instead of using one of its own two OSI-approved shared source licenses.”
Ramji further noted, as paraphrased by Ars:
There is no dichotomy between Microsoft and open source software, he claimed. He said that open source software offers an opportunity for Microsoft and should exist across all platforms. He also emphatically distanced himself from Microsoft’s anti-Linux marketing campaigns and said that he doesn’t support that approach to competition.
He also made an interesting remark regarding competition with Linux. He said that Linux should not be conflated with open source, and that competition with Linux should not be seen as opposition to the open source development model. This is more or less an acknowledgement of the fact that competing with Linux simply isn’t the same as competing with any other closed source piece of software; the rules are different. You can’t sue Linux, you can’t buy it, you can’t stop it.
Microsoft still has a long way to go when it comes to working with the open source community, but there’s no denying that lots of progress has already been made. Still, I’d love to see more openness, especially on the interoperability front, between Microsoft and the open source world. As it stands now, Microsoft is often ambiguous about potential patent issues, which really hinders open source developers in working on interoperability. Let’s hope this can improve in the future.
Wow.
In the UFO community they call this counter intelligence.
In Open Source circles we just call this poo.
Which this article is sitting on the worlds largest pile of poo I have seen so far.
-Hack
That comment is true enough – there’d be nothing contradictory about Microsoft encouraging a strong open-source ecosystem on their platform while discouraging use of rival platforms. Whether such an approach actually works is a different matter – from experience, open-source projects tend to spread cross-platform if it’s useful for them to do so. The new Linux video driver frameworks, for example, are at least partly supported on FreeBSD as well…
Of course, there’s also the question of whether Microsoft are genuine about that statement. Talk is cheap – we’ll see what their actions have to say…
Well, not mine but steveb’s.
As long as Steve Ballmer is in charge (which is still some years to go by), Microsoft won’t be really interested in interoperability. They might _seem_ to be trying to get on board…. but the truth of the matter is that it requires gorillas even bigger than Microsoft (like the European Commission with its fines) to make them start _dragging_ their feet (with the corresponding spin of Microsoft’s PR machine to make it look like they invented interoperability).
What a world we live in.
Linux and open source software has eaten Sun’s Unix lunch and commoditised the hardware that they thought they were always going to sell. The odds are also very good that Linux and open source software will eat Microsoft’s cash cow lunches, albeit taking quite a bit longer as it is far, far more difficult to commoditise software. Those are the only reasons why Sun and Microsoft are there.
Microsoft certainly has no problems with open source software, as long as it stays running on their platforms. Effectively, open source software on Windows becomes free development for Microsoft, and they’ve never been one to turn that down since it bolsters Windows.
Interoperability? As long as Microsoft continues to play standards organisation that many of us rely on, like ISO, as political table tennis then their words count for zilch:
http://www.groklaw.net/article.php?story=20090412131523897
Edited 2009-04-13 23:54 UTC
Anyone linking to Groklaw articles have lost all credibility, not that you had it anyway.
Hmmm, you don’t get this, do you? If you’re going to allege a lack of credibility then you’re going to have to go through that article and say why.
You can’t and never have done with anything you’ve ever tried to comment on, ergo, you have no credibility? Ironic, eh? That’s why you get modded down as well.
You’re correct in saying that they most likely will not get along. Many of the things the open source community tries to do is in direct competition with Microsoft. Fostering compatibility between email clients and exchange is not in their best interests. Allowing alternative operating systems to seamlessly replace windows in an enterprise environment, is not in their interests. As mentioned in the article, the only thing that has moved them to interoperable with open source software is A) When the open source product has the largest market share, or B) when governments rattle the anti trust sabre.
“Linux and open source software has eaten Sun’s Unix lunch and commoditised the hardware that they thought they were always going to sell.”
(*Cough*) Take a walk (or invitation) to your local phone company (Verizon, AT&T, etc.) Central Offices and MTSOs, and you’ll see SUN (metal) servers everywhere, and you’ll be able to count (with one hand!) the number of PC’s running Linux.
Dont forgot to check exactly when those big-iron servers were bought. Most likely not in the last couple of years…
“Dont forgot to check exactly when those big-iron servers were bought. Most likely not in the last couple of years… ”
Sorry fella!, but they’re up to date, and with current maintenance contracts for a long time.
As I said. You can count the number of Linux servers with one hand on these markets. Maybe in the next decade we’ll see a different picture, but as it stands now, SUN is far from being dethroned on this segment.
I’m sure they are but that has little to do with this. We’re not talking about keeping systems up to date or under MA contracts. We’re talking about Sun selling new machines and there’s no doubt that their market share is shrinking.
Why! oh why, do people with sh!tty knowledge of basic statistics open their mouths so wide so often?
Marketshare is a *useless measure* when the overall market is growing. Sun could have doubled their shipments and still lose “marketshare” if the market was growing at a faster rate than their sales. In that case Sun would be losing market share but still increasing their shipments at an impressive rate. Ok, in the real world they’re not, but I’m giving an example. Point is marketshare is a measure for the stupid (“Oh, surveys say 93% of other people eat maggots for breakfast, so I’d better do it too”). Switch the brain from “idle” to “on” please folks.
Marketshare is a crappy metric in other ways. For example, the US F22 fighter has a negligible share of the overall aviation market, with around 187 produced versus tens of thousands of jet transports and light aircraft. Does this means the F22 sux, no! Does it mean the manufacturer of the F22 is doomed, no! </example> All it means is there is a lot more aircraft out there – but so frikken what, not all questions are answered based on mindless statistics.
The thing that matters most for business viability is *profitability*. Simple. Companies can continue indefinitely provided they remain profitable. Once a company is no longer profitable it is a matter of time before it must either close or be consumed by a profitable competitor. Note: Linux is not a company, so doesn’t follow these same laws.
Growth in profits mostly matters for *investors*, especially speculative investors, and less directly for the company itself. Therefore, the *masssive (apparently mostly-US) obsession* with profit growth figures is unhealthy and leads to the decade-long boom-and-bust cycles we’ve been experiencing, as the herd mindlessly gallops between speculations. “Marketshare” is not how Warren Buffet makes his decisions.
I’ve been using Linux since the early 90’s and it’s my daily desktop. We recommend to customers for their servers wherever we can. However, there is a lot of Sun (S5000) and IBM (Z10 and Datapower) gear still out there and still being bought (just did some work for a large bank and that’s what they have). that’s why Sun still generates more revenue than you do (just not enough to offset their costs and acquisitions).
The real problem with Sun in my opinion is that is it hard to order their gear if you only want small quantities of stuff (kaiwai and I have already ranted about this stuff before in previous OSAlert threads – shame Sun doesn’t ever listen). They’ve obviously decided that small orders are not profitable enough to bother with. Which is true based on their antiquated sales model that requires dealing with a sales representative. If you could actually order their stuff online I’d be getting more for our own infrastracture and customers, but at the moment it’s too slow and painful. Often speed is the essence when making smaller or ad-hoc acquisitions.
Why do some people get so upset as soon as someone says something negative about Sun? Sun isn’t doing well and their sales are falling no matter what you say.
[snip all the stuff I already know about profits, market share and running a company]
Really? You know where I work and what our revenue is?
And why do people who cant even quote the right post open their mouth so wide so often?
Please don’t change the subject. I wasn’t disputing Sun was struggling. However, I was attempting to point out that the problem with your argument is that you said that “marketshare” was a problem when it is obviously “profitability”. Ok, you say you knew this was the case, well then please write that next time instead or carping on with the “marketshare” argument (which seems to be a particular affliction of Windows desktop users), which is actually irrelevant in many discussions.
Apologies, I wasn’t specific enough. I didn’t mean your company, I meant *you*. They have a high turnover, just not enough to cover their expenses and acquisitions.
I hope the two of you that bought such systems recently will be very happy. You won’t save Sun.
*Cough*, bull. What kind of precarious position do you think Sun is in right now? Where have you been?
It’s a very, very, very, very niche market, pretty insignificant and I’m afraid the expensive systems Sun sold many years ago aren’t going to save them.
Edited 2009-04-15 16:11 UTC
Sorry, but you have that the wrong way around.
It is very hard for software to commoditize hardware.
IBM made hardware a commodity. Open Source software just brought better use of that same commodity hardware so that the commodity hardware could take the place of the non-commodity hardware. There still had to be commodity hardware for that to happen though – and that’s thanks to IBM (for the IBM-PC) and Microsoft with its monopoly on DOS/Windows.
Software by default is a commodity. That is its nature. The difference is more market share – who has more software installations, or what standards are adhered to.
Microsoft tries to win on installs alone, screwing the standards in the process to grow their base.
Open Source Software, on the other hand, implements standards to the best possible degree and where possible implements all the non-standards too. This is what really irks the likes of Microsoft – the fact that a Mac, Unix, BSD, BeOS, Linux, etc. system could employ some software stack (e.g. an open source stack) and behave in the Windows environment without a Windows system knowing that it is not a Windows system.[1]
Why is software a commodity? Because while (yes) it does take time to write, debug, and maintain – the barrier to entry is very low – and only gets lower the more that commodity hardware could be used.
Thus, Microsoft will inevitably fall to Open Source or any other competitor once that competitor can break their monopolistic strangle hold on the market. Thus far, Open Source has been the only one able to do so because of how much further it goes into ensuring compatibility with both (a) Microsoft (in many many cases), and (b) ensuring portability across platforms. Add to that the low cost of initial development, the community behind it, and lucrative support contracts (as a result) for the players that join (e.g. Red Hat, Novell/SuSE, IBM, Oracle, Sun, etc.), and it’s more just a matter of time before Microsoft with either have to give up, or become more IBM-like and join.
More likely than not, once Ballmer leaves, the next round of management will make them more IBM-like and they’ll likely become more and more open-source oriented; likely even to the point of (in the long run) moving the Windows APIs to an open source community and becoming a distribution of sorts themselves where they provide the central, primary distribution (equivalent of Red Hat, Debian, Slackware, and Gentoo) and others provide similar distros (equivalent of uBuntu, Mandrake, etc.) only with the Windows NT Kernel instead of the Linux kernel; though with the full ability for either to be underneath.
[1] Why? Performance. Microsoft reserves their best performance in Windows-to-Windows interactions, even Windows-only APIs; and does their best to ensure that their Windows-to-Non-Windows and non-Windows-only APIs have lower performance. It’s how they justify the need to be Windows only, and thus grow their base.
I’m not too sure what you’re talking about, but you only need to look and see how software has commoditised hardware by looking at the same open source software you can run on different hardware platforms and how once people started using Linux on x86 the expensive Unix workstation market Sun and SGI had disappeared overnight. Look at Java – Write Once Run Anywhere. It doesn’t matter what hardware you use.
Software is not a commodity, yet, although open source software has gone some way. You only need to look at why people aren’t jumping ship from Microsoft Office to Open Office. Everything has to work just so, not to mention people opening their existing documents, and that takes a lot of time, effort and code.
Edited 2009-04-15 16:20 UTC
Oddly you seem to have gotten my point in some respect per software and hardware. But needless to say, it’s not purely software that made hardware a commodity. IBM, Intel, and Microsoft did that – by standardizing the components. That is what Sun and SGI lacked for many years – standards. Now that those same hardware standards (PCI-Express, etc.) have dramatically improved and as a result have pushed into Sun and SGI’s territory, thus commoditizing it as well. Software may have been one enabler, but it was one of many.
And FYI – Linux is not the only Unix-like OS on x86 hardware. Unix was ported to x86 hardware long ago (e.g. Minix, Solaris/x86, Xenix, etc.). There’s more factors to Sun’s and SGI’s demise than simple hardware commoditization.
Sorry, but software is. It is by default – part of the nature of software.
MS Office vs. Open Office is likely comparing the Apple Mac to a Dell or HP PC. Just because they are not 100% the same does not mean that they are not commodities. (Apple’s Mac very well is a commodity; a controlled one, but one nonetheless.) And even so, there’s more people moving away from MS Office to OpenOffice, Star Office, IBM Symphony, GNOME Office (Abiword, Gnumeric, etc.), etc. than you realize.
Market preference for one product (or a few products) does not negate the category the product is in as being a commodity.
Though I guess in the software space one could argue that no single category is commoditized until there are sufficient standards. Needless to say, even for productivity there are sufficient standards (ODF); so too in the OS and networking space.
Dude, you have to be kidding. Standards bodies are all about politics. The only thing worse then a standards body that accepts specs are the ones that create specs.
The best way to create a standard is write a spec, put it on the internet, and if it is good enough, people will use it. If enough people find it to be a good idea, the rest will use it for compatibilty, and you will have created a de facto standard. Thats how JSON came about, doug crockford bought a domain and put up a page with a good idea on it. JSON is awesome. The CSS spec is a steaming pile, and it was created by a commitee of competing interests.
Open Invention Network acquires patents and makes them available royalty-free to any company, institution or individual that agrees not to assert its patents against the Linux System. I think at this point to trust Microsoft more I would like to see them throw some of their 200 potentially infringing patents into a ring like OIN and help the I.T. community get some peace.
What does Microsoft gain? If an organization like OIN could gather enough members and patents it would probably prevent patent lawsuits. If all the members helped fund OIN it could (like Red Hat did early this year) buy and license patents for everyone. Until legislation catches up to reality it would be an excellent way for the tech industry to quit going to court.
Maybe he should go back and learn that the first most important thing isn’t open source but royalty free open standards – and for those standards to be properly documented and fully addressing all areas relevant to the specification in a platform neutral way.
Yes there is a huge dichotomy when executives such as Steve Balmer and Bill Gates lower themselves by resorting to the use of emotionally loaded jingoism’s in regards to their views of GPL/LGPL. Microsoft is entitled to hold the position that they believe GPL/LGPL is less than desirable but the resort to using language which is more fit for a USENET discussion rather than the kind of speak I’d expect from a highly paid executive – you really have to ask whether they’ve lost the plot.
Microsoft has many opportunities to sign up to the Open Invention Network but has decided that in their ‘fit of co-operation’ they run off an sue TomTom over the issue of a hack for a 20 year old file system to allow long file names.
The opportunity does exist but so far Microsoft has turned around and said, “no, I don’t want to be involved”. I can understand their arguments against open source but their arguments against open standards and submitting their technology to be an open standard and royalty free doesn’t make any sense.
So hang on, he says he doesn’t support it and yet he still works there. If I went to work for a company and found that all the work I did amounted to nothing – I’d leave asap. When you work for a company, not only do you get paid but there is the satisfaction that keeps pulling you back knowing you have made a contribution to something larger than yourself. When you fail to see your contributions being acknowledge – why would one wish to hang around in such an organisation?
As for the remaining conclusion; entering into dialogue without any change on the part of Microsoft is pointless. It might give some people the warm fuzzies thinking that Microsoft might have changed but the opportunity to prove that change would be for them to write up a contract that is a memorandum of understanding with the open source community (via the various foundations such as FSF, FreeBSD foundation, etc) where they agree not to sue companies who implement Microsoft patented technology – they have yet to propose anything like that. Not only that but also actively works with various groups to improve compatibility – where they compete not on the secret sauce but providing a better user experience.
Edited 2009-04-14 10:10 UTC
What if you want to change the company? What if you want to influence where the company is going? Denying the huge strides forward Microsoft has made in being friendly to open source is just plain nonsense. You can’t change a company as large as MS overnight – it takes time, dedication, and a high resistance towards frustration.
Change doesn’t happen with a flick of the wrist. It takes time and dedication, and a gaze towards the future – not the past, not the now. It is perfectly possible to work for Microsoft, with your goal being to change the company’s attitudes towards Free/open, but still disagreeing with how the company is operating now – you are the person trying to change that behaviour in the first place, so OF COURSE he distances himself from it.
Unless you have the executives onboard within the company – all the kind words and gestures ultimately will amount to nothing. What the individual people within Microsoft think about open source is ultimately irrelevant because the direction is decided by the executives and how much will they themselves have with pushing these ideas forward. Without their support – you will have no hope of ‘reforming’ the system over all.
We have seen over the last decade the kind of speeches from Microsoft executives that amount to the kind of trolling that one would expect on USENET with the usual flame wars between comp.os.linux.advocacy and comp.os.windows.advocacy. When I hear executives use the kinds of language such as labelling GPL as cancer, communist and unAmerican, I really have to ask myself whether these people are running a company or whether this is an ideologically driven personal crusade of theirs: Whether their decisions are made on the basis of what is best for the shareholders rather than their own phobia towards alternative business models and software licences.
Executives could have come out a decade ago and said, “we acknowledge that Linux through the various commercial distributors is a rising competitor, however, we believe that at Microsoft that through the control of the whole software stack that we can provide the best value proposition and assurance of stability for our customers for the long term” (interesting enough this is the same argument Apple uses to justify their vertically integrated business model).
Instead of taking the high ground, as said previously, they resorted to school yard style bulling that is unbecoming of the type of people I’d expect running a multibillion dollar organisation. These executives need to be won over so that there is a top down change in direction and so far Microsoft’s open source czar has failed to do so. The executives keeping the open source hobby horse around is akin to the political game of funding something – not because it makes sense but because the public feel better that it exists (not that they would ever use it). Its a nice way of certain people relieving themselves of the responsibility of actually having to find a middle way and open dialogue with the open source community and follow up on some of the issues that are raised. Talking without following up is nothing short of wasting a lot of peoples time.
Edited 2009-04-14 10:44 UTC
Again, you fail to realise just how complicated something like this is. Changing a company the size of Microsoft can’t be done in just a few years, by one person. There are various people higher up in the company now that are very friendly towards open source, and their numbers are growing. These people higher up then appoint and hire even more like-minded people, making sure the number of open source friendly folk is growing.
This is a process that can take years. Patience is a virtue. Executives will be replaced sooner or later, and the more open-source friendly people are within the company, the larger the chances one of them gets such a position. It’s a snowball effect.
but yes, Ballmer needs to go, but he will, eventually. Put someone like Sinofsky on top of the company, he seems like a much better person to lead the company instead of someone like Ballmer.
Personally, I’d say put Julie Larson-Green up there, but that’s for different reasons
.
I think what is important in the open source development is not how to generate the source code. The most important is how features compete, that is, which and what kind of features would survive, and in what direction they would evolution.
As thought of Linus and other major open source people, any before-hand design fails terribly in defining a feature that serve users the best. Instead, features must be thrown to user community and there must be many features functioning samely and then users select which to survive.
However, in a company business, this more or less render program managers useless. What if you tell them their intelligence in judging whether a feature will be welcomed by user is nonsense?
So by their nature, companies are not get alone with open source model. In order to resist this intrinsic, companies have to take something of open source that can be taken easily first. For example, embracing UNIX flavor and culture are many companies did. Being a company, you have to admit your intrinsic is anti-open-source, not merely in the aspect of intellectual property, but in the whole model of features survival and product design. So if you make your product has a great distance from what open source community accepted for a long while, and pretend you can embrace the inspiration of open source and get alone with it, you lie and lie to yourself.
But I think everyone know it is almost impossible to make Microsoft to accept some major open source culture like UNIX-flavor. Those non-UNIX things are plant to the core of Microsoft.