Oblomovka

Microsoft's Open Source Strategy

>
>
> ----- Original Message -----
> From: "Dave Winer" <dave@userland.com> > To: <FoRK@xent.com>
> Sent: Thursday, May 03, 2001 9:56 PM
> Subject: It's IBM dummy
>
>
> > BTW, for those who don't have the liner notes, the Microsoft comments about
> > open source are almost certainly about IBM's marketing of open source
> > programming against Microsoft's stuff in the .NET area. I don't think they
> > dare name IBM this specifically, for fear of helping them market their
> > wares. Now when the analysts ask IBM about their strategy, they are armed
> > with the checklist of questions to ask. "Uhhh, how does the GPL effect your
> > IP, Mr. IBM?" Markoff zeroed right in on this, interviewing the IBM exec. He
> > says their lawyers understand how to build a firewall that keeps the GPL
> > from infecting everything that's not open source at IBM, which is almost
> > everything other than Apache. Microsoft's answer is "Pfui." (To quote Nero
> > Wolf.) Speaking of which I'd love to see Lou Gerstner and Richard Stallman
> > at a press conference explaining how their interests are aligned. This would
> > be one of those "I'm glad to have lived to see this" moments. Dave
> >
>
>

...I don't think it's IBM. Well, partly it is (because it's always partly IBM with Microsoft, like it's always partly about one's Dad when you're doing home improvement). But there doesn't seem to any point in these "the GPL is bad for business" lines if the target is IBM. Mundie's comments don't seem very well targetted to IBM's customers.

Here are his points, as I see them:

1. The good bits of open source (community, standards, access to source, fast response to bugs), we do already.
2. The bad bits of OSS destabilise the IP economy.

There's a bit in the middle about forking and a throwaway dig at "inherent security risks", but the main point is that OSS is no better than MS, and bad for business, in the most general way.

I can see how the "good bits" argument could be seen as a response to IBM's sponsorship of open source. I don't see why IBM's customers would particularly care about the second bit. I don't even know why analysts would make much fuss about it. Bleating about how much damage a company's policy does to the public good doesn't really effect its share price, sales or market share unless a) it might be against the law or b) somebody gets badly hurt in front a major network's TV cameras. Microsoft, more than anyone, knows this: people were whining for years about how shockingly evil it was being in the marketplace. But, heck, in business terms, who cares? Does their product look better than the competition for me? Cool. Sign me up.

I'm pretty sure that most people's moaning about Microsoft's business practices effect on the overall market had a pretty small (and calculated) loss on MS's profitsheet. I'm sure that Mundie and Allchin's complaints about OSS being "bad for the IP economy" won't have much impact on IBM (or anyone else's) OSS-based sales.

So who does care about "bad for the IP economy" comments? Who does care about discussions of the common weal? Who cares if Microsoft says that the GPL can infect software, and destroy IP?

The government does. And, I think, many small software developers.

Here's what I think is Microsoft's threat model for Open Source:

1. Open Source companies are going to go tits up, because there's no business model. We can't find one, therefore there isn't one. We can outlive this corporate dalliance with OSS.
2. We're losing developers to open source, because it's a comfortable, cheap, arena for them. In particular, we're losing fledgling developers, and we're losing academia to open source. This isn't a problem for our balance sheet, but it's bad for R&D, and it's bad for our future. The risk is always that the next killer app comes from outside MS, on a non-MS platform. We have to win back this group.
3. The free software argument has a strong appeal to the public sector. Choosing free software over proprietary is a political decision, and it's an easy decision for govt. departments to make without understanding the consequences. We need to dissuade policy makers from supporting OSS as a policy instrument. We need to explain that it's bad for the economy. We need to make it clear that companies like us are dependent on academic research being unencumbered by GPLish licenses.

I choose 1), a bit presumptuously, because it just fits what I know of the corporate culture at Microsoft (and for that matter, Sun , most of IBM, and any number of other proprietary code companies). Heck, they could be right at that.

2), I have to admit, is more what I would worry about if I was at MS. If there's anywhere that Linux/Apache/PHP/Perl etc is undeniably having a massive adoption rate, it's among young developers and in academia. Even people who disparage the whole "slashdot kiddies" public appearence of open source are acknowledging this. I don't know whether Microsoft feels this is as much of a problem for their future as I do, but a lot of their explicit anti-OSS initatives (code sharing with universities, cheap development kits for undergrads, fake windows "installfests" even) are certainly aimed at this market.

3), I believe that MS is overreacting here - but in a very characteristically MS way. They recently got badly burned by a new enemy, the government,which they now believe to be big, dangerous and - by far the worst in MS's vocab - seeminglyrandom. They've seen how a govt can affect them, and they believe it to be easily mislead by nice-sounding-theories and wacky hand-wavers. I think they genuinely fear that RMS might bend the ear of some influential senators, and introduce a "GPL for all Federal Software" law. Hence the clumsy "unamerican" dig by Allchin, and Mundies' obsession with protecting federal support for basic research.

d.