Tuesday, January 30, 2007

OSDL and FSG merge to create the Linux Foundation

Okay, I really like the name "The Linux Foundation". It reminds me of Knight Rider's Foundation for Law and Government, or maybe Isaac Asimov's Foundation series. It just sounds cool. And Important. Well, I don't know if it is really *that* important overall, but it will serve a necessary function in the overall Linux ecosystem. The Foundation's goal will be to promote, protect, and standardize Linux. Promotion mostly means marketing and awareness of Linux's existing capabilities, although I believe over time it will also include identifying gaps in capabilities which are inhibiting Linux adoption. Protection is an obvious and easy reference to the education of the legal impacts of Linux and Open Source, but also has a strong focus on Licensing and patents/intellectual property protection. And standardization refers to the existing LSB efforts which will likely continue to grow (slowly) over time. LSB standardization is a slow process and technically challenging, as well as limited by adoption by distros based on their willingness to claim the latest and greatest level of support.

Overall, in some ways it looks like the little fish is trying to swallow the big fish. FSG had a pretty small budget and a pretty limited scope. OSDL had a huge budget and a huge scope. But FSG tackled a small, hard, well defined set of problems. OSDL tried to tackle a very large, nearly unbounded set of problems. The merging of the two extremes should be good for both groups. Both are going through a little bit of culture shock, I think but all seem to be working hard to find a good working balance. I think overall this will work out for the best for Linux and Open Source in general.

The most interesting thing was a recent meeting in Palo Alto which brought together a few community members from the Linux kernel community, a bunch of vendors, and several large, corporate end users. The culture clash was apparent, even though all were on their best behaviors. But the goal was to bridge those culture gaps, make the groups aware of each others pain points, and ultimately lead to a positive working arrangement between the extremes. We'll see how that works out over time, but the first painful little step has been taken. I'm curious to see how this meeting of the extreme poles of Linux community and end users works out over time. I have high hopes, though.

powered by performancing firefox

Mozilla, IceApe, Firefox, Iceweasel and Open Source re-branding

Recently I was a victim of one of the "strengths" of open source. Specifically, if you don't like the direction a particular project is going, CLONE it. Go your own direction. Become your own Maintainer. Pray that you have the user's to validate all that work you are about to do.

I'm a, um, loyal?, Debian user on my laptop. That loyalty is perhaps, a bit more laziness than true loyalty. I often believe that I should be transitioning at least to Ubuntu but really haven't had the urge to figure out how to gracefully get there. I know Debian as a project, I understand apt, I normally sigh and smile at the sociopolitical views of Debian and sometimes even believe that aligning with those views is to be one with the great purist views of open source philosophy. How Zen (not Xen, this time ;-)).

However, I spent a painful week or so as I learned about Iceweasel and Iceape. What are those, you might ask? Well, nothing more than rebranded Firefox and Mozilla, respectively. Now, I don't claim to really know about or care about the licensing issue that tripped things up here. But, despite the pain of a non-working Iceape and Iceweasel for nearly a week (I'm still not sure I understand why, although I think update-iceape-chrome was the magic trick at the end) I got to see open source in action.

Specifically, when a project no longer aligns with your particular goals, you can clone the source. You can modify the source, you can develop in a new direction. You can start from a position which leverages all the existing efforts in community development, modify to fit your needs, use it and/or make the results more widely available (including making source available!). You can do it without consulting anyone, you can do so without fear of infringement. The good licenses allow you to copy, modify, release, without the cost of starting from scratch.

Over time, this capability of open source will simply grow the ability to create or enhance applications at very little incremental cost to each person that adds value.

Now there is some real heavyweight development power.

powered by performancing firefox

Open Source Driver Development

So, Greg KH and the Linux community are offering free driver development. This is very clever, I think, and potentially a very cool thing. The is interesting on many levels. On one level, the availability of free drivers enables hardware vendors/IHV's to reduce their costs. On another level, it means that the must admit that their hardware sales are their revenue input and the software is merely an enabler for that hardware revenue. In other words, software has no real economic value, and usually has a large development cost that can be avoided. On a completely different level, it means that the IHV's must release specifications to their hardware to the public (yes, there are some ways to restrict that, although odds are they are a hassle which is really not worth the hassle) and the level of competition in devices shifts from "secret ways to make hardware work well" to "public specs which have more features with great hardware support". In other words, I think public specs will drive more direct competition in real hardware value. Of course that means that vendors will be forced to have real value in their hardware which could be a good thing for the consumer.

All in all, the real proof will be in the pudding, as they say. Until a few hardware vendors show that this is a viable option, there will likely be some skepticism. Going open kimono on hardware specs scares a lot of IHV's - it means that they show their (often small!) hardware value add to their competitors on their public specs now. And, there is the risk that the rather judgemental nature of the Linux community may still avoid the use of more advanced capabilities. For instance, even releasing specs on a network card that supports TCP Offload Engine (TOE) support may not get full community support.

But, really, I think this will be a great thing for the Linux ecosystem overall. Provided someone signs up....

powered by performancing firefox

Tuesday, January 02, 2007

Linux Desktops getting easier? Or Microsoft Desktops getting harder?

This was an interesting article given some of my past comments on the Linux Desktop. I'm beginning to wonder if it is a race for Linux to get better, or if it is just a matter of waiting until Microsoft hoists itself by its own petard with its convoluted and ever more restrictive licensing. I don't think Linux won here because it is perfect on the desktop, but it clearly reached that cross over point where it was better at *something* important.

To be honest, Linux has had some similar problems (Adaptec and the infamous SAS driver comes to mind) but now when people hit problems like this, they have to stop and think: Is it worth it? What are my options?

And Linux is now truly an option to consider.