Open Sauce Software

IBM has refused an online petition asking it to turn its venerable OS/2 operating system over and make it open source.

I'm perplexed by this story, reported on VNUnet. Firstly, I'm stunned to discover that there are apparently still lots of people using OS/2, and secondly to find that IBM, which is otherwise very clued into the open source mindset, thinks it's a good idea to keep OS/2 proprietary.

OS/2 was IBM's attempt to create a rival to Windows in the 1990s. In today's language we'd say it forked the code, turning a joint development with Microsoft into an attempt to compete - while Microsoft went ahead on a different route that became Windows version 3.

It failed, and we've all forgotten about OS/2, but apparently it still has lots of users, even though IBM stopped supporting it in 2006. The OS/2 comunity site OS/2 World delivered an 11,000-signatiory petition asking IBM to do the decent thing and turn the code over to them, to become open source.

IBM refused, for reasons that are a little vague: "for a variety of business, technical and legal reasons, we have decided to not pursue any OS/2 open source projects".

It could be greed - IBM says it can help migrate users onto other platforms (for money). But my bet is that there's still Microsoft code in there, and it simply can't open source it. It might be possible to root out all the Microsoft code, but that would be expensive and risky, and vulnerable to legal blocking.

And, appafrently, a version of OS/2 is apparently still for sale - under the name eComstation - so that's another company that wouldn't like it going open source.
`

Sun's acquisition of MySQL demonstrated a lot of things, but one of the main ones is that there's lots of money in "free" software.

You could put this another way if you're an analyst: "After several years of mostly one-off deals, open source deal flow took off in 2007, setting a new record for M&A [mergers and acquisition] in the sector," say Brenon Daly and Raven Zachary, analysts at the 451 Group in a report. And they were writing BEFORE the $1bn Sun deal.

And, by and large, this is a good thing. Open source companies mostly make components of various sizes. Once these have proved themselves, they become more valuable when they are owned by a big company that has the clout to impress IT managers, who may not trust a small outfit.

So, while the large parts of the financial world are in turmoil, there's value here. More open source companies are being bought for more money.

Previous booms have turned out to be bubbles but - in my opinion at any rate - there's plenty of value to underwrite a lot more open source business activity.

Red Hat and Firefox are reported to have more bugs in them than their Microsoft equivalents. But the truth is, as always, more complex. And once again, security is shown as a key point where rival approaches are bidding to distinguish themselves.

Secunia reported the discrpancy, stating in its 2007 Report that Red Hat had 633 flaws, compared with Windows' 123. However, Red Hat's Mark Cox quickly pointed out in a blog that a) the number was wrong, b) it counted flaws in all the third party products associated with Red Hat's OS, and worst of all c) it counted several bugs six times, since it added up fixes made for the same bug, on multiple Red Hat products.

Now, Secunia has a reasonably respected position in alerting the industry about security flaws, and promoting fixes, but it clearly isn't doing quite so well on the whole business of statistics and counting.

The interesting thing is why Secunia would push this story at all.

Even if there were a greater number of reported bugs on these open source products, that would not equal lower security. It could just mean that there is more publicity for known bugs in the open source world (as we saw recently, when code-checker Coverity announced it had found around 8000 bugs in open source projects, I commented here that this was actually good news for open source).

Obviously, whether or not Secunia deliberately got its sums wrong, it remains the case that "open source security flaws" is a much more arresting headline than "Microsoft security flaws" - for exactly the sam reason that "man bites dog" is more interesting than "dog bites man".

After all this time, we still know which way the the security argument goes.

I see that Infoworld has started a campaign to save XP. I'm sure the Linux world will feel a little wry about that one.

The XP-to-Vista transition is a pretty clear example of the way proprietary operatings fail users. Everyone is happy with XP, says Infoworld , and it's got plenty of life in it.

So, Infoworld asks, please can we keep it? The site is running a rather jolly Doomsday clock, a petition, and a blog at SaveXP.com.



Microsoft is pulling the plug (stopping selling XP) at the end of June, and resellers can only go on selling it till the end of 2008. But why is this, if demand is high? And what are the chances of "saving" XP?

Moving people on, is obviously in Microsoft's interest, as more new PCs, mean more profits (though we've seen elsewhere that in many circumstances, the familiar and trusted XP commands a higher price than most versions of Vista). Infoworld makes the parallel with windows Millennium edition (though the number of users and the amount of development cost was far smaller, in that case.

Unlike previous Windows versions, Microsoft could build a business model around keeping XP going more or less forever. It's got better copyright protection than previous versions, so XP users would be likely to have paid money that will reach Microsoft.

But fundamentally, Microsoft upgrade cycles don't fit well with what people want in PCs. In the early days of IT, new versions were necessary, because the technology changed rapidly and radically. If I remember rightly, I upgraded to Windows 95 so I could use a CD drive, moved on to Windows 2000 so I could use USB, and to XP for stability, roll-back and automatic online upgrades.

Now, new features can be added easily. A PC is like a VCR or a maybe a fridge, and I'd rather it stayed as stable as possible. I don't want to have to get a new VCR every year (though with writeable DVDs, Blu-Ray, and hard disks, htere's pressure to do so. We certainly don't want to replace our fridges.

It's extremely Ironic that, while killing it off for business users with one hand, Microsoft is thinking of allowing XP to live on the OLPC XO - a laptop which originally chose Linux to bail out of Microsoft's expensive upgrade cycles.

The challenge now is for Microsoft to create a model that actually delivers what users want, at a level of cost and disruption they are prepared to accept.

Linux - and the basic maturing of technology - has set this challenge, because finally, there are alternatives.

It's a story that probably got spun many ways, but the bottom line is good news for open source.

Code-checking maestro Coverity, working on behalf of the US government, has been checking open source code for security issues - we've covered it here. It's found some, and the open source projects concerned have, at different speeds, been fixing the holes.

This is a good thing and, in common with most things open source, beefits from being done in the open. How many bugs are found in proprietary code? We simply don't know.

So what's the news point? For some, fed by unrealistic expectations around open source, the newx is a Shock!Horror! headline - Security bugs found in open source". Which implies that bugs aren't found in priprietary code?

To be fair, that particular article was an accurate one, from our friend Mathew Broersma, and headline writing is a task more often dedicated to getting readers than spreading truth.

But once again, the "open" bit of open source is a mixed blessing.