To answer your earlier query about servers, Apache is on 77% of web-connected servers. I think the percentage overall is lower because Windows Server is a popular intranet server.

So, how do you propose we solve the situation of the Linux community (a large distributed collection of individuals mostly contributing in their free time) not being able to drive the hardware innovation? We would love your help in developing solutions to any problems that we may face.

airdrik wrote:So, how do you propose we solve the situation of the Linux community (a large distributed collection of individuals mostly contributing in their free time) not being able to drive the hardware innovation? We would love your help in developing solutions to any problems that we may face.

As was said before, the Linux community doesn't. If it were the case that Linux was the most popular distribution, then companies would do what they currently do with Windows and provide drivers. The difference is that, with Linux, companies would be able to alter the kernel directly so that it's specialized for their hardware (and provide a repository with kernel updates, for example in Ubuntu). Things like Project Natal would be developed by people like Logitech then support would be provided for the OS by the companies.

The difference is, everyone can benefit from the additional hardware if the drivers were made open-source, or installed from repository systems/programs like Jockey in Ubuntu. This would make installing drivers much more convenient then going to a company's website to look for them.

Dante Ashton wrote:'We' refers to the Linux community.

Well, I don't see the Windows community or OS X community or actually any other community doing that either.
Hardware is still mainly built by companies, not individuals.

Actually the Free Software/Linux community is ahead in this sector with initiatives like OpenPC http://open-pc.com/

Dante Ashton wrote:I'm not referring to a company making hardware then putting Windows or Linux onto it.

You were referring to companies putting Windows and OS X onto hardware, so I assumed you were kind of seeing a problem for companied doing that with Linux.
Actually Microsoft does not make most of the hardware it is selling Windows for, other companies do that.

Dante Ashton wrote:I'm talking about hardware like the Xbox 360's Project NATAL, an add-on bit of hardware running their software that they will be selling under their name.

Ok, but I don't see any community doing that. Looks like a company (Microsoft) doing it.

Dante Ashton wrote:We do not have the capacity to do a similar thing, Canocial could not easily come out with a Ubuntu-ready webcam, for instance, as they do not have the capacity to manufacture hardware, nor deals with hardware manufacturers.

Not sure why you put a company into the context of 'Us' (the community), but Canonical could definitely do that if they wanted to.
There is no magic involved in developing hardware and pretty much everyone outsources manufacturing to companies specialized in that.
But the market for peripheral hardware is pretty much saturated and diveded among the already exisiting players, so they most likely see no value in getting into it just because they can.

Cheers,
_

@Madman: my question was directed primarily at Dante Ashton, to see if he has any viable solutions to that which he sees as being a problem, which I'm sure most of us don't really see as a problem because it doesn't really fit the domain (the problem of a software community trying to drive hardware innovation).

@anda_skoa: Here, Here!

airdrik wrote:@Madman: my question was directed primarily at Dante Ashton, to see if he has any viable solutions to that which he sees as being a problem, which I'm sure most of us don't really see as a problem because it doesn't really fit the domain (the problem of a software community trying to drive hardware innovation).

I realized. My, "answer" was kind of directed at Dante Ashton.

Well, it's more innovation in general, not just hardware.

The Linux Community is unique regarding development, just two months ago a few researchers at a University gave code to the Linux kernel to allow multi-touch capability. A person can work on their own projects without really having to be hassled by budget, time and marketing.
However, this same strength also means bigger projects are harder to do; harder to start, harder to keep going and harder to reach the finish line. How many Sourceforge projects have failed simply because they couldnt reach their grand aim? Quite a few....

Which is why projects like KDE are run by multitudes of developers.

If you're going to start a project with a large goal, you're better off inviting other people into it. That's how projects like Debian started: just one guy set the ball rolling, then he started asking everyone else to jump aboard as well.

That may be so, to further elaborate;

I mean, in the term 'big' the same way Netflix, DVR's, Spotify, NEPOMUK...Google are big.

We don't have the capacity to run such huge projects because of our fragmented nature, There is simply no way a huge bit of R&D can be done by a community of people and know it's going to be successful.

That said, I think something like plasma, which attempts (quite succesfully, in 4.3, so I've heard) to redefine the desktop model is really the innovation I'm looking for.

I guess I just wish more people were out there creating things on the same scale as Plasma instead of maximising the potentional of current apps.

If (and I think it's quite probable) Linux does come out on top, then I think Moore's Law, at least how it relates to the desktop user, will slow down.

Dante Ashton wrote:However, this same strength also means bigger projects are harder to do; harder to start, harder to keep going and harder to reach the finish line. How many Sourceforge projects have failed simply because they couldnt reach their grand aim? Quite a few....

Sure, but doesn't that apply equally well to projects aimed to extend Windows or MacOS X beyond what Microsoft or Apple envisioned? If the scope is limited, it's easier to accomplish. If the scope is more ambitious, it's more likely to fail. Which, I would say, it's more likely that the more ambitious projects are not very likely to be a one-man-show whether the project is aimed at a primarily F/OSS system like Linux or a primarily proprietary one such as Windows.

Then again...I would argue that no software project, regardless of license, is EVER going to reach the finish line. There are always things to develop further.

True, but there is a diffrence between the person in front of their computer doing the work and someone paid to do the same. In the former, the job will be done with far more...enthusiasm, you might say. This is what I feel to be the deciding factor between open vs. closed development. People aren't hired to work on something, (at least, most of the time, they arent).

If a team of computer whizzes figures out how to do the next big thing, then great.

But I can assure you the computer whizzes sitting in the labs with five other departments feeding them info about their product and customers will have a headstart.

Dante Ashton wrote:But I can assure you the computer whizzes sitting in the labs with five other departments feeding them info about their product and customers will have a headstart.

I'd say that's a oversimplification. For starters, it pretty much guarantees that things like plasma and nepomuk would have been given the axe before the projects had even got of the ground. They were far too buggy and added too few positive features in their early stages compared to the predecessors for most companies to take a chance on them. Some would to be sure, but I'd be willing to bet that if plasma development had been entirely dependent on funding from a company it would at best have been scrapped as a "interesting idea but unprofitable and not what our customers wants" even before KDE 4.0.0 had been released and thus, its current usability and flexibility would never have been unleashed.

NEPOMUK was not built by us, rather, it was taken from an EU-study. If I remember, MS had an interest in it as well, but more for comercial exploitation (which such a system is very useful for)
On that note, that's why I first gave up on Windows, the MSN Messenger service put attention distracting ads on a client. I'm almost certain it'll be part of the wallpaper in later versions of Windows...

Anyway, yes, I didn't say that was adequate method either, both have their ups and down, and the closed source world is only slightly less unlikely to innovate (and even that's debatable what with the scale of technologies invovled, cloud computing was once something only a few companies could pull off, now if I wanted to I could setup a cloud computer for my family and friends.)
Regardless of budgetting, I've just noticed that the larger projects here either die from lack of interest or die before their time for any particular reason.

In the closed world, you need to simply get the job done, I don't think I need to point out their standerds are lower by comparsion when it comes to good software.

We can do things they can't do, they can't do things we can do. We function on a more individual-centric level, where the developer and it's audience is key; they function on turning a profit.
This brings up my eariler point about hardware, if any particular company decides to develop hardware that would take some serious hacking to get going outside of their closed software, then we're just going to be playing catch-up instead of innovating.

I'd love to take interested people away with me to a secret lab, and work on KIIM, NEPOMUK and all sorts of wonderful tech. I can't do that, neither can many other people. Comparing KIIM's programming to a calculator would be like comparing an Ameiga with an IBM Roadrunner.

I've been toying with the idea of trying to get interested people more informed about certain tech that is lacking within Linux, and hopefully to direct them to work on it...I don't know, if that could work, that could change a lot of things...but still.

Dante Ashton wrote:NEPOMUK was not built by us, rather, it was taken from an EU-study.

No.
Nepomuk-KDE is the desktop implementation part of the EU funded research project. So, yes, it was built by "us" (i.e. one of "our" developers), based on work done by other Free Software projects such as Redland, Sesame2, etc.

IMHO being part of the research process is even more effective than building products on the results like Microsoft or Apple have to do (or they have to do their own research and hope it will yield results).

Cheers,
_

Yes, but the original idea and code was not built by us.
We simply used something publicly avaible. If I remember, we played a large part in it's construction...but we didn't yeild the original design and idea.

It's a big project and I find it unlikely to really be a useful major feature until a ton of developers and their time have been dropped on it. There is only one dev really doing much worth noting...there may be others but I've been informed they mostly test code or add a few snippets...