411Irrelevant

Somewhere the ghost of Isaac Newton is extremely pissed off at Linus' dissing of alchemy.... BT

The ghost of Isaac Newton is too busy experimenting with quantum gravity to read blogs.

If it is a superior way to develop software why is the Linux desktop light years behind Windows and Mac? For something to be truly superior it must be superior to what it's competitors are currently offering. Linux isn't superior to Windows from the only perspective that really matters and that is the end user perspective. It isn't superior to Mac OS X from an end user perspective either. So how is this a superior way to develop an OS? There are some superior Open Source alternatives out there such as Mapserver, Apache, Firefox, and PostgreSql, but by and large proprietary software and systems lead in just about every metric. CAD desktop, GIS Desktop, Desktop documents, inter application communication and automation, fonts, UIs, general usability, SCADA systems, GPS Systems, Topo mapping software, routing software, ..... I could go on for a long list of what proprietary software has that is better than it's open source counter parts. I work on Open source projects, I support it financially as well because I believe in the underlying concept that information should be freely shared to improve the society in which we live in. I don't support it because I believe it currently creates better software at the same pace as proprietary software companies can. Developers have to eat and support their families. Until Open source can figure out how to coordinate large projects across hundreds of developers in a timely fashion, and pay them all good wages, Open Source software won't out pace the rate at which proprietary software puts out better solutions. The most successful Open Source projects have one person or a very small team of Core developers working in close communication towards a common goal. When projects get larger than that they fall behind their proprietary competitors or they fall apart completely. In the end Open Source the way it is currently done is a form of Communism and history has shown that Communism doesn't work on a large scale no matter how noble it's ideals are. Communism works in small groups like tribes, kibbutz's, etc. but it doesn't scale to a nation level. History has born this out many times over. Incentive based systems such as capitalism work better on a large scale by far than Communism does. Open Source today seems to be the same way. It works very well in small core groups and can produce outstanding results but when it tries to scale to encompass huge projects it falls apart. I believe the way to change this is to revamp the way Open Source is run so that it is no longer run in a communist fashion. The way to change this would be to have companies that develop software open up their source and take input on that source from the larger community while still paying their developers and generating revenue as a company from sales of their software and from the maintenance of it. Then you would have an incentive based system that still shared it's knowledge freely. This is the way AT&T Bell Labs did many of it's projects and it worked very well at the time. PGP also works this way. I think it could work on a large scale but I may never find out unless their is a radical shift in the way proprietary software companies start working. It is sort of a catch 22, the key to making Open Source really work is in the hands of those companies that fear it the most. If we can get them to change their mindset then I think the sky would be the limit and Open Source in it's new form would be a truly superior way to develop software. In it's current incarnation I don't believe Open Source is the best way to develop software on a large scale but we can change this over time and until we do get it changed it is still worth supporting because knowledge should be free and we should work to make our society better for everyone not just the ones who can afford it.

There will eventually be excellent open source alternatives in every software category you listed. We're just getting started.

The Eclipse IDE is currently the best desktop integrated development environment (well, maybe Visual Studio is better, but regardless we are talking about a very close race). How can this be? No one is selling it. But there are lots of traditionally paid developers working for big companies working on it. Lots of different big companies too. The "communist" label is just a big red (ha ha) herring designed to rattle Americans who have not gotten over the propaganda surrounding their previous Official Enemy. Open source seems to flourish once a significant part of the marketplace decides that a particular piece of functionality is no longer useful as a product differentiator. Server operating systems (Linux), IDE/application frameworks (Eclipse), scripting languages (Python/Perl/PHP/Ruby/etc), web servers (Apache). Desktop operating system interfaces have innovated enough in the last five years (thanks, Apple!) to keep marginally ahead of their open source followers, but if they slow down for too long, they too will feel pain as "good enough" and free alternatives catch up. Oracle is vacating the database market as fast as it can, and moving into areas where it can offer real value, like business intelligence and CRM -- they see the writing on that particular wall. There will always be a place for proprietary software in the niches, but this is a very long game, and the onus is on the proprietary companies to continuously improve their products to stay ahead of the game -- to deliver real value for money (like Apple does with OS/X). The days of locking down a customer base and charging monopoly rents ad infinitum are over.

"The days of locking down a customer base and charging monopoly rents ad infinitum are over." This I agree with completely.

I wouldn't characterize either as superior. Each has its' pros and cons. Certainly the collaborative aspect and low cost of open source makes it fun, accessible and provides a great deal of value and sustainability. However, end users have little control over QA problems and patches, little control on enhancements, have limited means of support, and either have to roll their sleeves up and fix/modify the product themselves or wait for someone else to deal with it. That is fine if your project has budget, capability and resources for scratch-building things, but otherwise for production end-users, it causes some concern and risk. On the other hand, some (but certainly not all) of that is alleviated with COTS products, however here you are stuck with proprietary code, formats, APIs, high cost and a host of other issues. In the long run, however, these things tend to follow a cycle of commoditization - where a piece of technology becomes less unique and more ubiquitous, and is relegated to a state of commoditization, at which point proprietary pieces become irrelevant as there are at that point many Open Source pieces which have evolved as stable and low-risk, to push the proprietary aside. At this point, the proprietary needs to turn to modularization and cutting loose the commoditized pieces to turn its efforts to other pursuits. It's a dynamic and continually-emergent process.