nemesis <nam### [at] gmailcom> wrote:
> I don't think so.  Most enterprises either build their private systems
> in house or pay a third party to build it and even hand over the source
> code anyway.  Private enterprise commercial systems are hardly portable
> or fittable to other organization needs so there's really no point in
> open-sourcing a one-fits-all kind of system and even if there was they
> would still hire software developers to customize it.

Okay, we're talking about two different worlds here (yet again): You're thinking
of custom-tailored corporate inhouse software; I'm thinking of stuff that is
typically distributed as shareware (say, for example, tools like UltraEdit).

Then again, there's companies like the one I currently happen to work for,
trying to do stuff with embedded Linux; it's an awful hassle, as we found out.
How can you build a Linux box that will be built into a car?

It basically requires you to (1) use GPL'ed software, because you can't write
*all* the stuff from scratch, nor can you buy it - there's no true commercial
alternatives out there; because of that it requires you to (2) release *all*
the source code of the final product under the GPL as well, and provide
interested users all necessary tools to - well, let's just call it by name -
basically "hack" the box and flash their own software into it; but at the same
time you must (3) keep the automotive company's confidential information (CAN
bus protocol details and such) secret, *and* make sure that nobody can convert
the car into rolling danger by "hacking" the box.

So it's no surprise the embedded Linux idea has not really reached full speed
yet.

While clearing away the "legal landmines" of software patents, the FSF is laying
their own "legal mines" in the territory gained, with the aggressive GPL'ization
of free software.

An old african proverb says, if elephants (here: big companies and the FSF)
fight, it is the grass (here: small developers) that suffers.


> I see open-source primarily as a great way of diminishing production
> costs by sharing such cost of development and maintanence with other
> interested parties:  everyone benefit from it and the more interested
> parties, the better.  It's a very good development model geared
> primarily for general purpose infrastructure software, but not so much
> for user applications which are meant to be specific.

The problem with it is that the FSF doesn't see it this way; their GPL is geared
towards making *ALL* user applications free. Including, by the way, web
applications custom-tailored for an individual website.

Their agenda leaves absolutely no doubt about it: Their aim is to make *every*
single piece of software "free" and open-sourced. They're just not going full
gear for strategic reasons (read for example their statements about when to use
the LGPL and the GPL to see for yourself).

Honestly, I'd liken this to communism or socialism: The basic ideas might have
been good ones, but they have been bred in academic brains a bit too far from
reality (e.g. too optimistic about the nature of man), then pushed by
enthusiastic masses too far from the academic ideas, to ultimately fail - after
having caused a great mess that will not be so easy to clean up.

Post a reply to this message