Quelle: Slashdot, Stand: Juni 2001
http://slashdot.org/comments.pl?sid=01/06/15/003248&threshold=2&commentsort=0&mode=thread&cid=82
I'm a little surprised and a lot concerned that people seem to have forgotten the main reason Stallman holds the views he does. Check some of his earliest stuff and you will see it.
Stallman wants to be able to fix the software he is using when it breaks, and add things the original author missed.
Simple, isn't it?
Next level: how much software have you had to abandon because the author(s) and publisher
- went out of business and
- the source wasn't made available?
You then had to move on to something else, go through another learning curve, maybe even spend a lot of time converting old data to the new format. That takes time, brothers and sisters, and time is money. And that's where the word trust really comes in: are you willing to bet your livelihood that Whiz-Publisher and Bow-To-Programmer will be around in the future?
Then there is the other aspect of "trust," that the publisher and the author(s) will continue to maintain the software, fix the bugs you find, and extend the functionality in ways that are useful to you. A couple of examples will serve to illustrate my point:
EXAMPLE 1: Remember troff, the typesetting program developed for Unix? It was a great piece of work, and did things incredibly well. Unfortunately the author died, so much of the incredible work had to be scrapped because no one else could begin to understand the code (not even typesetter manufacturers -- I watched one guy at Varityper try). Now, if the source had been released widely (fat chance, being a bastard child of a utility regulated by the FCC) there might have been enough of a brain trust developed to fully understand the workings of the original program. Instead, some people wrote a work-alike that serves us today, but loses some really nifty code.
EXAMPLE 2: Microsoft WORD has an interesting history, being the first massed-marketed pieces of software whose beta was bound into a mass-market magazine. (PC World, for those who care about such trivia.) Since that time it has become a definition of bloat - yet there are features professional writers have requested of Microsoft that have not been included ... Because Microsoft does not make the source available, there is no way for the technically-minded professional writer to add any of those features that would REALLY make life easier. One of those features, a phrase dictionary, is one reason the legal profession sticks with WordPerfect.
We trust vendors to "do the right thing" but they are under no obligation to do so. Those who say "if you don't like it, go write your own" should be aware that the entry cost for writing a word-processor package is very, very high. Indeed, one reason for the Open Source Movement in general and the GNU Public License (not Virus) in particular is to lower the cost of entry by building a collaborative effort to accomplish a task. Divide and Succeed.
And so we now get to the bottom of why Microsoft and Stallman are at odds. Microsoft wants to hold your productivity hostage, so that THEY can release stuff under THEIR terms and to THEIR schedule. Microsoft has no significant competition in a number of markets, so competition won't keep them in line. (Remember the anti-trust suit?) The ONLY significant competition currently in place is GPLed software, because Microsoft can't "embrace and innovate" something that requires they show their cards for all to see.
The BSD and similar licenses are flawed in that Microsoft can "embrace and innovate" to the point that the original code is lost in the jungle of proprietary extensions that Microsoft would add.
By the way, Microsoft isn't the only company that plays the grab-and-obfuscate game, only the most obvious one.
What Microsoft fears most is that other corporations are beginning to "get it," that the large proprietary corporate model is not the only model for ensuring viable support for software products. The distributed development model, specifically OSS protected by the GPL, provides the same advantages as the corporate (or centralized) development model without the "bottleneck effect" of corporate management prejudice and the cost of "buying" 30,000 programmers.
And what about all those programmers? Banished to the bread lines? Guess again. Some of the most lucrative programming is in applications for specific industries. Corporations are looking to combine off-the-shelf components in ways that improve corporate productivity, and are willing to spend the bucks to make that happen. Look at the insurance industry. Look at the food-supply industry. Banking. Finance. Even waste management.
Want to work on something a little more generic? Try embedded-systems programming. There are still microwave oven controllers to be programmed, not to mention metal-forming presses and the like. Who do you think programs the firewall appliances we use on our cable and DSL feeds? Who do you think creates the new gambling machines now showing up in Vegas and Atlantic City? Even my furnace has a microprocessor in it.
And not to worry, e-commerce isn't dead, it was just overblown. There are lots of jobs there.
So stop crying about loss of jobs for programmers with the GPL. If anything, it will increase the number of programming jobs because the tools will be cheap enough to lower the barrier of cost of entry.