I can totally understand why Eric isn't a fan of Nikolai's paper. The paper sets up Eric as a straw man to be bounced around, and hey, I've been there. Having a cartoon caricature of your beliefs criticized is just frustrating and doesn't seem terribly useful, since one can't learn anything personally from such a critique.
However, I think that Eric misses the value in such a paper. Nikolai's paper answers the question "what are the risks in embarking in an open source development project?", and is the most focused and complete answer I've seen so far. That's very valuable from an advocacy perspective, because advocates need a single URL to point to for a thorough treatment of this frequently asked question. Otherwise, a properly skeptical skeptic won't believe that all of the homework on the subject is done.
Are there better treatments of this subject available that are as complete and focused on the question at hand? I think that this document provides a very good starting point if a better document doesn't exist. Even if a better document does exist, this points out some useful anecdotes and quotations worthy of inclusion in any answer to the risk question.
I'd like to put in a shameless plug for an email list that I run: the "election-methods-list". Details can be found at: http://www.eskimo.com/~robla/em
We discuss the messy mathematical ins and outs of various single-winner and multi-winner systems, in an effort to fully understand the spectrum of systems available and their relative merits. Not for the faint of heart, but if you are interested in that sort of thing...
(ahem) As one of the "lobbyists" involved in RTSP, I have to ask what you are talking about. If you look through the discussions on RTSP, most of the changes to the protocol (there were many) happened as a result of taking the original RealNetworks/Netscape proposal, and making it more consistant with other IETF protocols (HTTP, RTP, etc.) The end result isn't perfect, but it's an open and extensible protocol that's pretty easy to understand if you understand HTTP (since it shares a lot of HTTP's structure).
I'd really like to know what you think we "screwed up". Actually, I'd prefer if you posted your comments to "confctrl@isi.edu", which is where the discussion of the RTSP protocol design takes place. The archive of this list can be found at ftp://ftp.isi.edu/confctrl/confctrl.mail.
Eric Sink's article is spot on here. A single point of contact from the "suits" to the freeware community was very useful for getting the gnu/penguin/daemon's foot in the door. Though ESR can continue to be very helpful, it takes all types, and we haven't begun exploring the different ways people can make themselves useful to the community.
The lesson that seems to be emerging for big companies is that dealing with a figurehead such as ESR is not the same as dealing with the community.
Slashdot Mirror
However, I think that Eric misses the value in such a paper. Nikolai's paper answers the question "what are the risks in embarking in an open source development project?", and is the most focused and complete answer I've seen so far. That's very valuable from an advocacy perspective, because advocates need a single URL to point to for a thorough treatment of this frequently asked question. Otherwise, a properly skeptical skeptic won't believe that all of the homework on the subject is done.
Are there better treatments of this subject available that are as complete and focused on the question at hand? I think that this document provides a very good starting point if a better document doesn't exist. Even if a better document does exist, this points out some useful anecdotes and quotations worthy of inclusion in any answer to the risk question.
http://www.eskimo.com/~robla/em
We discuss the messy mathematical ins and outs of various single-winner and multi-winner systems, in an effort to fully understand the spectrum of systems available and their relative merits. Not for the faint of heart, but if you are interested in that sort of thing...
I'd really like to know what you think we "screwed up". Actually, I'd prefer if you posted your comments to "confctrl@isi.edu", which is where the discussion of the RTSP protocol design takes place. The archive of this list can be found at ftp://ftp.isi.edu/confctrl/confctrl.mail.
The lesson that seems to be emerging for big companies is that dealing with a figurehead such as ESR is not the same as dealing with the community.