First - if anyone feels that this board should not devolve into a technical discussion of Linux rather than a financial discussion of RHAT, say something and we can go underground. :-)
Eric -
Regarding the validity of the survey, I think it's just a "best shot" - I'd love to see one that could provide more accuracy. I doubt that anything was double counted, because IP addresses are unique. You are correct that nothing behind a firewall would be counted.
Referring to your argument of "bureaucratic overhead" in Open Source development, looking over the past 5 years of development, I just don't see it. In fact, I see the lack of it. One person "scratches her itch" for software she needs to do her job (raid controller, scripting language, whatever) and writes it. She publishes the source. 100s of people use, it, poke at it and make comments - peer review it, in effect. They send in patches to the source to fix things up. Where's the bureaucratic overhead? Granted, larger projects take more leadership. Usually there is an exceptional person who leads a project such as kernel development.
When you speak of the sheer volume of apps available for Windows vs Linux, I think that's primarily due to the fact that historically, the people using Linux were programmers, networking pros, and the like. As a result, there are lots of apps in these areas, but not many Create-a-Greeting-Card type apps.
Regarding open source failures - sure, every day. Search around freshmeat.net for projects that stalled at version 0.0.1. But there's nothing wrong with that. The interest and/or expertise wasn't there, and resources went somewhere else. At least Marketing didn't push out anything akin to Microsoft Bob(tm) :-)
You asked about very large scale projects, enterprise databases and the like. Oracle is available for Linux (I believe) but this is not an open source project. Linux is currently lacking in apps the very high end computing - I can't argue with that. Perhaps someone else can. :-) |
