What Is Open Source (O'ReillyNet)

Here i go again... sorry for this tendency of beying so irreverent!

In general the primary difference and almost only one, is the source code being open, available to anyone, and licenced in a way that not allows anyone to abuse the work of others for personal or particular profits. So provisions in the licencing are made for that to be enforced.

Home/general users.

No apparent difference at the source level because this users are not generaly interrested in it. What they are interested is in good software that works, specialy in an intuitive visual way, and open source can deliver that, and has most of the Windows Counterparts for them. General polish is as good, sometimes better than the more popular used solutions. And addressing the *free as in beer* sub culture feature of the Windows ostensive strategy for prevasive dominance, that downplays the software for the hardware turning every computer into an appliance, is automatically achieved without making every user as a potencial pirate. What is more required is for them to be a little more pro-active concerning choice, because an inexperienced user at his own, like the large majority with Windows, can easly pick up a buggy development package and them blame the whole for a tiny part. They in general intuit this very truth and in consequence trend to go with what the tide offers and stay awhay of the less knowned. But THEY CAN HAVE AN OUTSTANDING ADVANTAGE IN SECURITY AND RELIABILITY, and be free of many of the annoyances that plague the Windows environment for ever.

Business/general users.

No difference at all at the source level because this *users* are obviously completely uninterrested in it. General polish is as good, sometimes better than the more popular used solutions. Lack of some regional or more large area used applications, generaly only available in native format for the Windows platform and with no open source counterpart, is the handicap for the solution. Virtual Machine technology, and others, could be used to close this gap for some fat Desktop centered businesses, and in a future if something like ReactOS could be served on top of Xen, many businesses could be freed from restrictive licences forever and MS top executives exploding with rage. Any business/institutions upon hiring a "good enough" linux technician, with some server compositions in their IT deploiments, can be more at ease and run transparently *most* their Windows and Linux applications thanks to the technology of "Graphical Remote Terminals". Example: http://www.theregister.co.uk/Design/page/linux_case_study...
Business users CAN HAVE AN OUTSTANDING ADVANTAGE IN SECURITY AND RELIABILITY, and be free of many of the annoyances that plague the Windows environment for ever. The cost of deployment and maintnance or TCO, contrary to much publiciced FUD can me "astronomical", as easly seen in the example .

In general for *users*, it can be said that many combinations of configurations are possible, where more security, reliability and substancial savings, pondering all factors, can be achieved in almost every single one of them, if open source is joined to the solution.

Developers and Engineeres
Agreed.

IT professionals
Gush... Besides developers and engineers these are most of the other techies that are adopting it at a relative large scale. But they are more or less neutral, if not MS centered, in pushing it into their managed deploiments. For the large majority it represents having to gain other skills and certifications, so they are affraid to lose their jobs. I've encountered some, even with their own tiny businesses, that play the hipocrit, running for themselfs on desktops and servers open-source solutions, and being *TOTALY* MS centric at their jobs or with their clients(fear without doubt).

Commercial software vendors.
Agreed in large extend. Only that for local applications it has no real difference, since those with capable Development Departments can easly adopt and adapt, and those with not, sonner or later will die, more rapidly for a rival on the same Windows(r) environment, than for a Linux Venture. For big sellers it is a big opportunity because if it wasent for open source, perhaps none of them had the money to endure staying at the top of the "State of the Art", and sonner or later had to be bought *on the cheap*, by the bigger sharks like MS, until only stayed one or two.

Entrepeneurs, investors and such
General agreed with the much reduce exposition. Add that for web-services, technicaly and financialy, is clearly *THE CHOICE*. Because all and much more of the traditional "data center" can be replicated and offered to the world as a globally accessible proxy + gateway, engulfing in the path the traditional telecoms as more services move into IPv4 or IPv6, and good enough bandwidth is delivered. I belive future ISPs wil be telecoms in a proxy + gateway topology like that.
Other industrys like the hardware one, for technical and financial reasons, it is also clearly *THE CHOICE*. But they are the jocker of Microsoft the big trumph of M$, playing along with them, with fear of retaliation, and giving them next, most probabily, the change of having the source code for every device driver they'll produce. More secure, reliable, versatile, powerfull and inexpensive OSes on any computer BOX, any at all, *SHOULD SELL MORE BUT THEY DONT*. The key for this paradigma is the grand masses adoption, the first and most neglected item on the top of the list.

" And IMHO, as H as it can be, this last paragraph is the most important issue, that the Unix/Linux world has so far failed to understand properly. "

IMO "grand masses" adoption, is more or less very difficult to obtein, and once achieved it goes naturaly entrenched. Shoal liminal behavior or something like that. I'm most probabily not the one to explain it right.

So and since Microsoft is preparing to lock a hardware platform with certification and DRM features, and with it most probabily the "grand masses" of users, it will be **WISE** to do something in the strategic plan, that could very well not be the most receptive or accute technical feature.

And since in my opinion is much more easy to offer an escape path to the hardware industry than to gain substancial masse adoption in relative short time, i propose, as i did before, as possible solution the adoption of something like in http://www.cl.cam.ac.uk/Research/SRG/netos/papers/2004-oa...
That is an isolated safe, hardened environment for drivers. Though it is based on paravirtualization, but as Intel Vanderpool and AMD Pacifica will be pervasive both for sure, the performance overhead can be reduced. And thought it is isolated detachable from the "main" kernel by an API or an API/ABI it dosent mean it has to be the Xen format. It could very well be in *LINUX FORMAT*, sharing the schedulars, interrupt handlers, memory allocaters and many other features, as more or less they are now in the main kernel with ease. If this is done, than the performance overhead could be reduce even further to a completely negligenciable value.

The ability that a proprietary module pops in, more or less freely, into these interfaces dosent mean the layer is not for Linux drivers. It is for Linux drivers, because as i, many will prefer an open source version of a device driver, for the majority of places, than a proprietary one. And never the less proprietary modules will never stop to pop in, but wrapped in outdated, many times messy, interfaces and for so crucial devices as graphic adaptors of wireless chips.

With a safe or "crash freely that i dont mind", reliable, secure, hardened interface is the only way we can tell that a hardware device driver is better than another be it proprietary or open source. I'll bet on open source where appliable.

Rest my case

Cheers.