(fwd) article on open source

[jabr at blu.org (John Abreau)]

------- Forwarded Message

From: TonStanco at aol.com
Message-ID: <3b.2a83ebf.260f5397 at aol.com>
Date: Sun, 26 Mar 2000 06:50:47 EST
Subject: article on open source


 These are follow-up questions to the one I posed to the open source community
 for my proposed article in Internet.com's Boardwatch magazine on a royalty
 system to pay developers in open source. I'd like to get community feedback
 on these, too. I write an Internet Business Law monthly column for
 Boardwatch <www.boardwatch.com> and I'm writing a series on open source. [If
 you are searching my past articles, please note that the bio page has a 
 problem and only lists 6 of my 13 articles, the rest are in the search
 results themselves]. The responses to the first question came in from all
 over the world and prompted these follow-up questions. Thank you again to 
everyone who responded to the first. The responses were interesting, 
enlightening,
 enthusiastic, highly intelligent and ..... overwhelming in number-- though I
 read and appreciated each one. I've enjoyed this project more than any
 other, mainly because of the tremendous feedback. The open source revolution
 has a passion not seen in the world since the storming of the Bastille. It 
also has 
the best and brightest people worldwide in it, I think.

 >>>>>>>>>>>>>>>>>

 QUESTIONS

 Am I the only person who thinks that developers should be paid a royalty
 based on the number of accepted lines each developer contributes to a program
 that's shipped and sold? Of all the responses I got on the first question,
 only 2 discussed "how" any royalty system could be done and they both
 concluded that it couldn't. Isn't a royalty system that compensates on the
 basis of the monatary success of a program better than being paid a fixed 
salary or
 not paid at all?

 As President Clinton recently said, "don t make the perfect the enemy of the
 good." Even though a perfect royalty system may be impossible, one that
 imperfectly pays developers is still a very good thing, I think, as long as
 it keeps the core values of open source. More money to open source developers
 only means more developers for the movement. Who would work for proprietary
 software if they could get paid in open source? Open source is a superior
 development model, after all, that empowers all developers, because it allows
 them the freedom to see, copy and modify the code. Peer review both helps the
 experienced developers and mentors the new ones coming along.

 Also, aren't open applications better than closed proprietary applications?
 So why does open source tolerate anything closed?

 Will money ruin the open source movement? People originally also said that 
money
 would ruin the Internet. Has it? Why is open source different? 

How will open source fight the inevitable software company backlash? Does 
anyone 
really think that software companies will go quietly into the night? Survival 
is the 
most natural instinct and corporate law actually requires management to do 
what is 
best for the shareholders. Does anyone remember Halloween I? Do you really 
think 
that Microsoft (after the trial, of course) will not use copyright and 
patents against 
open source? How does open source wage such a battle without a revenue model 
to 
help finance the war? 

 My leanings are:

1.  Developers ought to be paid. Software is the most important product in
 the world today. Developers are a new nobility based on brains. This is the
 first chance the world has had to have a real worldwide meritocracy. This is
 especially important for people in the developing world, who could earn a
 place in the world economy by producing software for which they are paid
 equally to everyone else.
 2.  Only large and medium corporations and all governments should pay for 
buying 
 open source software. Small business, students and consumers should not
 be charged. [This is not a sine qua non, but only my egalitarian bias.]
 3.  ALL software should be open, including all applications.
 4.  The most efficient royalty system is based on the lines of code produced
 by each developer as a percentage of the total lines in a final version
 that's shipped and sold.
5. Property rights are fundamentally important. Getting them wrong caused the 
Russian people to lose 3 generations. Capitalism works best at least until we 
get to 
the point where no one has to work anymore, even if it is not perfect.
6. Software companies are dinosaurs and will be replaced by open source 
development because the power is in the developers (at the base of the 
pyramid) not 
in the companies (with the few at the top). This is different from 
traditional 
industrial companies that own the necessary capital assets to produce the 
joint 
product, so those companies have the power in those situations. With 
traditional 
industrial companies if people want to work, they needed access to the 
capital assets 
to do their job. This is why third world countries have to wait for 
industrial 
companies to invest before their citizens can work. This is the fundamental 
power 
shift between the old economy and the new economy.  In the new economy, there 
are very limited capital assets involved for people to do their jobs, so 
people don't 
need companies anymore. They can perform their jobs without the corporate 
eggshell. This is what open source has shown the world. It's quite impressive 
actually, if you think about it. Corporations have organized major human 
activity 
for a couple of hundred years now. The Internet and open source prove this is 
no 
longer necessary, that people over the Internet can organize themselves. 

Please read the two responses below that I received on the "how" issue. Can 
anyone
 give me any good reasons why in the first John shouldn't get 3/5 and Dave 2/5
 of the selling price of the MP3 player (before you  include Bob s recorder
 code)? Forget about the interim bloatware since it gets eliminated. 

 Only the code in the shipped version counts. Competition to get accepted into
 a shipped version will create a feedback loop that will eliminate the
 bloatware in the end so only the best, cleanest, leanest code is included and
 priced. Developers as a group would decide what is the best code for each
 program as they do now in open source, since only they are the experts 
(again the 
expertise is at the base of the pyramid, not the top. Remember Gates almost 
missed 
the Internet). The process for inclusion should be democratic, not monarchic 
where 
one person can be a dictator (nor oligarchic as with proprietary software's 
current 
many  dictators). Every developer should equally have a vote on anything 
he/she is 
interested in voting on.

 In the second, what s wrong with the first guy getting 10000/10005 of the
 combined market and the other guy getting 5/10005, since it is assumed that
 the 10000 lines is the guts of the program and the 5 is only a clever
 porting? 

 Maybe there can be some other objectively determinable criteria added as a
 refinement at some point. But don't make the perfect the enemy of the good.
 Lines of code is an easy, objectively determinable, and viable model right
 now. It will get the developers some money for themselves and for the 
movement. 

 By the way, ever calculate what a line of code is worth? Based on the market
 cap of Microsoft it may be thousands of dollars. Yes, thousands for each
 line. One calculation puts it at $9,000 per line, but that may be overly
 optimistic. Do some calculations yourself. You'll be amazed. Just because it
 s fun and easy for developers to produce it, doesn't mean software has no
 value. On the flip side, proprietary software has shown that just because it
 has value doesn't mean that developers will get much of it. Something in the
 middle is clearly needed in my opinion.

 Just my thoughts, though. Dissension is a whetstone that sharpens all our
 minds.

 Thanks and best regards,

 Tony Stanco

 >>>>>>>>>>>>>>
  Subj:    RE:Open Source Article
  Date:   Tue, 14 Mar 2000 1:17:13 PM Eastern Standard Time
  From:   "R.A.Fletcher" <PMA99RAF at sheffield.ac.uk>
  To:      tonstanco at aol.com

  Question: Why cant open source software be sold?
  Answer: In a sense it can, but its not those who wrote the software 
  who sell it. RedHat sell it, SuSE sell it. They make money from it. 
  The authors don't recieve any. Since I'm a mathematician I'll proove 
  the idea that each author gets a percentage of what is sold is 
  impractical, by assuming that it isn't......

  A man writes a program. Lets call him John. John write 10 000 
  lines of code for a program. To be topical, it can be an MP3 player, 
  to play audio files. By defenition the source to his software is 
  available for changing as anyone sees fit, if it is open source. Dave, 
  an MP3 specialist, likes Johns program, but finds bugs in it. And 
  being a clever programmer with a kind heart fixes them. Suddenly 
  10000 lines of code has become 5000 and the functionality of the 
  program has improved. Say 2000 lines of code of this final 5000 is 
  Dave's own work. Does that mean he should get a two-fifths share? 
  Who would work it out. John doesn't know Dave, so they might not 
  be able to agree a percentage. Before you know it Bob has added 
  a CDRecording facility.

  Obviously it would take some disscusions to sort out percenttages, 
  and then your assuming that someone wants to buy it The aim of 
  each author getting ther just deserves is obviously impractical. Bob 
  says his cd recording feature is what makes the product stand out. 
  Dave says that's true but since Bob used his work, which fixed a 
  lot of bugs, he wouldn't have been able to implement it. John says 
  this too is true, but it was originally his program Whereupon the 
  have a fight.

  Open source means software has to be free in order for it to be 
  practicle. 

  Assume that John puts a restriction on his work stating that any 
  work done upon his code must not be released, except to him. 
  Then Dave probably wouldn't have bothered in the first place. And 
  Bob's groundbreaking CD record feature never comes to light, or at 
  least it takes longer.

  I think that makes it clear why open source is the way it is. It 
  brings about rapid development, small efficient code, and benefits 
  users by being easily adapted if slightly unsuitable. Take for 
  instance if a program decides to send as much info about you to 
  an organisation. It wouldn't happen under the open source model.

  I get the idea that you feel there should be reward for those who 
  spend their time doing these things for free. But they are usually 
  hobbyists, the system is working and needn't be fixed.

  Yours 

  _oOO-Richard Fletcher-OOo_


 >>>>>>>>>>>>>>>>>>>>>>>>>>                    
  Subj:   Re: open source article
  Date:  Mon, 13 Mar 2000 11:05:11 AM Eastern Standard Time
  From: Izar Tarandach <izar at linuxqa.com>
  To:    TonStanco at aol.com

  If based on a royalty model, how do you quantify how much each
  developer gets ? For example, if I write a 10k lines system that works
  on 50% of the computers in the market, and someone sends in a patch
  with 5 lines that miraculously makes it work on the other 50% percent,
  what's the size of the pie that each one of us is taking ?

  --izar


 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>. 
 Disclaimer:

 These are my personal, private opinions. The Securities and Exchange
 Commission, where I work as a securities attorney, as a matter of policy, 
disclaims 
responsibility for any private publication or statement by any of its 
employees. The 
views expressed herein are those of the author and do not necessarily reflect 
the 
views of the Commission or of the author's colleagues upon the staff of the 
Commission. 

 >>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

------- End of Forwarded Message



--
John Abreau / Executive Director, Boston Linux & Unix 
Email: jabr at blu.org / URL: http://www.blu.org
ICQ#28611923 / AIM abreauj
-----------------------------------------------------------------------
"Working with NT is like trying to tune a watch wearing oven mitts.
 You can't get your fingers inside like you can with UNIX.
-----------------------------------------------------------------------


-
Subcription/unsubscription/info requests: send e-mail with
"subscribe", "unsubscribe", or "info" on the first line of the
message body to discuss-request at blu.org (Subject line is ignored).