How does this differ from engineers leaving from proprietary software companies? Philippe's statement "The code, without the people, is worth nothing." isn't open source specific. And it is the norm after an acquisition for the best people to leave after their lock-up is over rather than the exception. At least in open source after engineers leave they may very well continue to contribute their expertise to a project and the company from which they left might receive some sort of benefit. Look at the case of Postgres - SRA International and EnterpriseDB. Bruce Momjian one of the core Postgres team members, recently moved from SRA to EnterpriseDB. In this case SRA still benefits by his contributions to the community even though he isn't directly on their payroll.

Thanks for all your efforts. But I must differ in saying that Open Source is just not the way you define it. Even from the Program Code prospective, things are different. Open Source is more based on community model. The project when it is design based on community model, is right from its infancy in demand by community and gets better support. todays true example is Apache, Netscape, Linux, Asterisk and many more.In fact Postgres SQL ... is very good example of this, which has been funded via VC.

"[...] Computer Associates, who shortly after acquiring Ingres, experienced a 'mass exodus' of developers immediately thereafter...."

The article correctly points out "'[t]he code without the people is worth nothing,'" and the acquisition of Ingres by CA is one of the best object lessons in this. As the Postgres history link describes, there was a "mass exodus of engineers the day it was purchased by Computer Associates."

"...THE DAY it was purchased..." (emphasis added),

From my memory of an UnixWorld article:

CA gave every employee a survey which included a question of whether they wanted to continue working under the new management. Everyone who answered "no" or "maybe" was fired.

Everyone who remained was presented with an employee agreement with terrible terms, and a demand to immediately sign it or be fired. This happened during a high point in commercial RDBMS development; according to the chart in UnixWorld, in all areas of core development but one, every employee left (I think networking had two remaining employees).

I had previously helped set up an Ingres system to run a university department; according to a friend who still works there, after this no one she knew was still at Ingres, and she could not get any useful support. That site took Y2K as an opportunity to change DB vendors.

The field of software development will be much better off when the truths of this article are widely accepted.

I'd say this is *less* of a problem in open source than proprietary software..

Despite the abundance of software patents, most are trivial and open to challenge and any that are valid in the US are likely unenforcable in most of europe.

Then there is copyright, which will be assigned to the company, even if it's under the GPL or BSD license, the company can still license it to individual companies as it see's fit and therefore it remains an "IP asset" the same as in a proprietary company.

So the difference between a F/OSS company and a closed source company is that if your core engineers leave, there will be others who know the codebase and design who can be recruited - as well as the core engineers being likely to continue to contribute despite leaving.

To me, that makes OSS lower risk that proprietary regarding this problem. Of course there are different risks with each business model, and open source has it's risks - this is just not one of them

Oh, I do agree with the comment. I don't know what's the news in this article?

The fact that understanding the program requires more than the raw code is a common knowledge, actually. And I don't know what it should have in common with Open Source thing.
Actually, Open Source projects are far safer, as the programmers of the acquiring company can learn the code before takeover what brings you a lot of safety. What is more there is always a chance that there are people among community who understand the code and can be hired as substitutes.

So Open Source start-ups are rather safer to acquire. Is this article some kind of FAD, or sth?

Well said, Andrew. As well, in an acquisition, the company is potentially buying more than the code and the core developers. The company is also buying a market position (customer list and relationships and brand), as well as the process and values (embodied in the rest of the people as well) that enabled the purchased company to be successful.

The "code without the people" isn't necessarilly worth nothing depending upon the organization and the vibrancy of the community if the company's products were based upon an open source project (internally controlled or externally engaged). Even in a closed source product world, the code asset is still valuable, just more difficult to work without the core developers. Acquiring companies have always been sensitive to keeping the staff around. Indeed, in today's world, it's possibly the best way to find the aggressive best rather than recruiting them if that's what you're looking for.

I agree with Andrew Aitken's assessment -- this problem is not unique to open source or service-based companies. There are nightmare stories about proprietary companies whose all-contractor workforce erased all the hard drives and backups after the CEO sold the company out from under them with no notice. Had that been an open source-based company, the IP would have been safe in plain sight.

I can undersand the perspective of the weary investor, though. One thing to make sure of is that there is a real community outside of the commercial entity housing the primary developers. How many unique commits to the source tree are there each quarter? With a real community, there is a possibility that you can hire from that community if things get rough with the primary contributors.

Also, just because a contributor leaves the company doesn't mean they aren't working on the project at all. Sometimes a company environment becomes less appealing over time, or people need more flexibility due to serious personal/family issues, etc. With open source based companies, this doesn't necessarily mean the end of useful contribution as it does in proprietary software companies. A great example of this are some of the snort developers who left SourceFire but continued to contribute to Snort.