At least it's late here. Morning in Blighty ...

I've just re-read this article for the third time in five hours. From my perspective, it's all over the place. I'm not sure if it's the author, or the folks being written about ... but I've never seen an article on ElReg that was as orthogonal to both itself, and to reality, as this one.

I'll try reading it again over coffee in the morning ...

Companies can absolutely function fine by hiring a bunch of people and teklling them to code up extensions to their favourite open source OS.

You get immediate and tangible benefits by paying a bunch of guys to rustle up some nice hardware drivers so you can run it on your latest shiny embedded system or whatever.

What on earth is in it for these companies, that they should foster community support? It sounds like a lot of time and effort, with no guaranteed return. *Maybe* you'll get a lot more testing and debugging, *maybe* you'll get patches from contributors. But if you don't waste their time playing nice to the community and just post a bunch of kernel source patches on a website somewhere, they still have a perfectly functioning system with none of the hassle.

I need to get this off my chest:

Why on earth would IBM act in any other way?

It is also possible that some/many OSS contributors do so because it will enhance their careers or prospects of getting a good job once out of college.

It is also possible that some/many OSS contributors do so because they see it as cool or simply enjoy it?

Are any of the above superior to any other?

Is a contributor who acts solely in order further the cause of Linux/OSS/World {peace|health|prosperity} superior to any of the above? If so, how?

Good and wholesome things can be a by-product of self-interest. Beware of the sanctimonious occupiers of the moral high ground who claim to they do what is right for its own sake. Some will be genuine, others will be self-deluding and/or fraudulent.

Today's word is "pragmatism"

"Wade said companies had to learn that they couldn't apply the closed-source model of software development to open source by hiring people and then telling them to code. He called this stacking a project by hiring all the people to make sure the programming goals are met."

Why not? If the interests of the community and the company providing the resources diverge, then the project forks; Because of OSS licensing, this is possible, the company does not get to control anything it does not own, and everyone continues to benefit.

If the company makes its fork closed source and markets a product then the FSF (or someone) has to prove that license terms have been violated and act accordingly. This is true of the general case in which a company takes some OSS and covertly makes a closed source product out of it. That is, it's something that the community has to deal with anyway, potentially.

What is it with these people?

Hammers, shovels, screwdrivers, you name it. Lots of them. Shiny ones, matte ones, large ones, small ones.

Never used them to build a damn thing, though. I just like to show them to people and let them be impressed that I can make such nice tools to sit on shelves.

I once knew a guy with a database cluster that was the same way.

I can understand the idiots at Novell need to distract the community after their Microsoft collusion, but why the hell did anyone see anything more in this little outburst? No story here beyond IBM learning there's advantage in going beyond their own direct self interest. Something Microsoft's buddies over at Novell still don't understand.

Of course they are in there to further their interests, however HP and Dell benefit more out of it without contributing much ...

Every vendor contributes drivers, and code that makes Linux run best on their hardware, there is not much interest in contributing something that will benefit everybody (like your competitors).

You only have to look at the "alliance of the willing" developing System TAP where HP and DELL are not present.

Technologies like DTrace and System TAP are amazing profit making technologies. How? they allow the support engineers to resolve customer issues faster. Before these technologies, custom debug kernels needed to be shipped to resolve issues, and that could take days ...

I'm not particularly prone to "dealing with the way things are", but there really is a case for it here. I remember reading an article reporting a similar winge from the Swing team some months ago, and having the same reaction.

People contribute to projects to meets their own needs. From the pimply 45 year old in his mom's basement to Big Blue themselves it's the same story. FLOSS projects can do themselves a big favour by adapting to take what benefits they can from this situation and minimise the damage it can cause.

World domination to the first project to get that right :)

Well Linux installs are easier, way faster and less problematic on all my IBM hardware than any MS OS. And that's not just the servers either. Everything just works.

If that is the end result then so be it.

I've read this article probably nine times. It makes no sense.

Why continue to try to make sense of it after mulling it over for more than 30 hours? Because there are some big names involved. There might be some there there ... maybe.

Yes, there are fully open FOSS solutions out there.

Yes, there are proprietary solutions that run on FOSS out there.

Yes, there are niche solutions, both proprietary & open, that run on FOSS out there, and I'm willing to bet that for the most part the big-boys don't really care about any of them.

And??? WTF am I missing here? What's the story supposed to be about?

@Jacqui: Eyeball Sourceforge sometime ... Rolling it out inhouse without the resources to support whatever the end result might be suggests a lack of familiarity with FOSS ... which in turn kinda goes along with your company's results.

"BSD ,what is/who is that."

It's what's on the servers.