Publisher's Note: UniNews online thanks Michael Paddon for this stimulating and provocative editorial which appeared in the May 1997 issue of AUUGN, the excellent journal put out by the UniForum affiliate in Australia. We also want you to know that the AUUG97 Winter Conference is set for September 3rd-5th in Brisbane, Queensland, preceded by two days of tutorials September 1st and 2nd. For more conference information, and to learn about the work AUUG is doing, please go on-line to http://www.auug.org.au. Michael can be reached at michael.paddon@auug.org.au and I'm sure he'd be pleased to have your thoughts and comments on his views below. RRS
Opinion: Is the Future of Unix in Your Hands?
by Michael Paddon, President, Australia Unix Users Group (AUUG)
Just a day or so ago I ordered another server for my network. This time I needed a machine that could really kick some butt, loads of CPU, disk and i/o. Two years ago, I'd have picked up the phone and spoken to a friendly salesperson at one of the major Unix vendors. These days, as you'd expect, I ended up buying a fast PC with a SCSI controller, etc.
Naturally, I'll spin up a real operating system on the box (NetBSD; though FreeBSD, OpenBSD and Linux would do just as well). Then I'll load up a bunch of really useful software such as the GNU tools, Apache, nethack :-), tex, xv... you name it, it's out there somewhere.
This is quite probably a familiar scenario. We've all known for years that you don't have to pay for the best software in the world, and we've tended to try and spread the word to those who are ignorant and pity those who don't understand the new dynamics of the real information economy.
"It doesn't cost anything? That just doesn't make sense!", is the common lament of someone outside of our sub-culture. Some people just aren't equipped to understand that software can be worth more than money; that an ancient technology is being supplanted by new currencies of interaction and exchange.
I tend to be particularly amused when the aforementioned confused person tends to make the next obvious leap, and concludes "free software simply can't be any good, then". After all, the creators aren't asking for any of the paper survival tickets that I value so much in exchange for their work!
One of the true visionaries of our industry (where collecting paper tickets seems to be valued highly, at least in the short term), is Richard Stallman, a founder of the Free Software Foundation. Mr. Stallman saw, and more importantly believed, all of this years ago, and his vision has led to the creation of the GNU software suite amongst other things. It's probably impossible to quantify the value of contributions to our industry and hence the world at large, but he has inarguably made a gigantic impact.
What does this mean to you (apart from being able to get hold of really neat software)? Think about the long term effects of cheap as dirt hardware and cool free software...
Our industry is no longer going to support vendors. Unix hardware vendors are already in trouble, and have been for a long time. I've been hearing whispers that another major vendor may be just about to pack it in. But it's only a matter of time until the same fate befalls the operating system vendors. And I didn't use the adjective "Unix", there.
Most people's initial response to this is "Good, I hope the bastards burn. We'd still be running 7th Edition with ed as our text processor if it hadn't been for Berkeley, anyway". The oncoming troubles of certain other OS vendors is usually discussed with what can only be described as enthusiastic glee.
But in this brave new word where corporations can only afford to focus on applications, what happens to OS development? Who codes, tests, documents and supports the Unix of the future?
The answer is very simple. And you're sitting with that person right now. It can't happen without your involvement, and the alternative is an unattractive shade of corporate grey. This new economy, new community and new world that has started evolving needs your help, and it needs it now.
But in this brave new word where corporations can only afford to focus on applications, what happens to OS development? Who codes, tests, documents and supports the Unix of the future?
The answer is very simple. And you're sitting with that person right now. It can't happen without your involvement, and the alternative is an unattractive shade of corporate grey. This new economy, new community and new world that has started evolving needs your help, and it needs it now.
If you can code, this means writing something useful and making it available. The GNU and Berkeley licenses are good examples of different schemes of "giving" away your work without getting screwed over. Consider joining a development team of one of the numerous free projects, from systems to graphics to databases to word processors.
If you can't code, volunteer to test. Or maybe to write documentation... programmers need all the help they can get in that department. Or even just donate your ideas.
Look at what's been achieved by a small and hardy group of developers to date in the free software world. I doubt if they number more than one developer in one thousand users. Imagine what can be done if we all donate a little of our time similarly, and imagine the richness of the rewards.
This brings me on to a shameless plug for yet another piece of free software. The second version of the Berkeley DB Database is on the cusp of public release. I've been fortunate to get to play with a pre-release of this code and it is truly wonderful stuff.
Many of you will be familiar with the first version, which provided a choice of btree, dynamically hashed or sequential record databases, with a C API. The new version adds logging, locking, transactions and access for multiple readers and writers, amongst other improvements. In short, a fully functional free database. Check out http://www.sleepycat.com for more information.
My initial work has shown that I can exceed three hundred insert transactions per second on an ordinary PC, which leaves most commercial products looking somewhat anemic. When someone writes a schema and an SQL layer for this package, I'd start to worry if I were in that business.
I also found the Sleepycat copyright rather interesting. This software is free so long as you distribute, for free, the source of any program which uses it. Not a bad model for promulgating the concept of free code.