Archive of UserLand's first discussion group, started October 5, 1998.

Re: Open Source -- a world onto itself

Author:Paul Snively
Posted:7/20/2000; 9:29:40 AM
Topic:Open Source -- a world onto itself
Msg #:18805 (In response to 18800)
Prev/Next:18804 / 18806

Dave Winer: Now what really spells doom for the Open Source guys is that they don't have very many users. So, can any new ideas permeate? How?

I think, with all due respect, this is a deeply flawed argument.

The Open Source guys have millions and millions of users because, in many cases, what they've written is server-side Internet stuff: Linux itself, BIND, Apache, PHP, sendmail, qmail, fetchmail, MySQL, PostgreSQL, OpenLDAP, OpenSSL, the plethora of Apache modules tying page-generation and database technologies together, and of course, some of our favorite scripting languages: Perl, Python, TCL, and others. It's a big and vibrant culture (and yes, it has been since at least the 1960's, long before anyone thought to formalize it as the "Free Software Movement" or "Open Source Movement;" I get that). Because it is almost totally interest-driven, it vacilates wildly between innovation and reproducing what's already been done, just with source code and a different license.

The Open Source community does not appear to have, at present, a user focus in the sense that those who develop for a mainstream desktop OS does, for what amount to historical reasons. The minor skirmishes between GNOME and KDE may have, as their most beneficial side-effect, an impact on that (but I'm not holding my breath). The work of the Eazel team is, frankly, the only thing that gives me any confidence of seeing genuine user focus on Linux anytime in the predictable future.

A good indicator of what's available in Open Source, and the breadth and depth of the offerings, is <http://www.freshmeat.net>.

Dave: Their 15 minutes of fame are over. The Next Thing is already here.

Napster isn't the next thing, if for no other reason than that they made the engineering error of centralizing their directory. That engineering error will likely also turn out to be their primary political error, as it's the thing that will get them shut down if they are shut down. But as has been pointed out already, Gnutella (Open Source) and FreeNet (Open Source) are waiting in the wings. Because they have no centralized anything--directory or corporation--there's no legal target. The fact that there's no corporation to sue follows directly from the fact that they are Open Source. The fact that there's no centralized directory a là Napster stems from the fact that these are true P2P systems, unlike Napster, which is a client/server connection-brokering system.

Dave: ...I would tell them that there's a big world outside. Take off the parachute, jump, and let's dance. You don't have all the answers. Same message to Microsoft, which heard it a lot better.

People have a lot of strange ideas about Open Source, like that the movement maintains that you can't charge money for software (not true even of RMS, let alone ESR). Another strange idea is that the movement feels it "has all the answers." While I can see how someone could get that impression from RMS, who flat-out doesn't believe, philosophically, in proprietary software, I personally don't feel it characterizes ESR accurately, and once you move beyond RMS and ESR, it's hard to even say who serves as an indicator as to what the movement believes. Certainly a great many of us who contribute in some way, large or small, to Open Source software do so strictly because we're interested, it solves some problem for us, and if it does that for us it's likely to do so for someone else, too, so we contribute back. That's all. It's actually pretty far from having all the answers; it's much closer to having a problem!

As for Microsoft hearing, we'll see. At the very least, it looks like they're moving SOAP in some important directions with discovery and the like--things that real distributed systems need. As we see more SOAP implementations on more platforms and in more languages, we'll presumably start seeing more .net services--from Microsoft and other sources--and I'm sure some of them will be Open Source, because to have them would solve a problem for someone who is interested enough to develop them.

So even if Tim O'Reilly, Brian Behelendorf, and some others seem to have missed opportunities to build bridges, I think it likely has more to do with just feeling the need to maintain a certain focus in their efforts than any particular set of blinders, let alone malice. But I could be wrong. In the meantime, there are the rest of the people--the non-famous ones--working on Open Source who would probably be interested in building some of those bridges.

There are responses to this message:

This page was archived on 6/13/2001; 4:55:45 PM.

© Copyright 1998-2001 UserLand Software, Inc.