[milters] Archive

Lists Index Date Thread Search

Article: 19
From: Anthony Howe
Date: 2004-09-15 02:25:09 -0400
Subject: Re: Mailing list for Snert Milters

Removal...........: milters-request@milter.info?subject=remove
More information..: http://www.milter.info/#Support
--------------------------------------------------------

Alex Tkachenko wrote:

 >>>Anyways. Nothing is final and the survey is intended to evoke arguments
 >>>for or against or find the balance.
 >
 >
 > I've got your point, thanks. Now tell me please, why would I invest my
 > time into fixing commercial software without being paid for it? Also


I'll answer that question with another: Why did businesses buy Unix
source licenses from AT&T when it was first developed? Why was Unix
traditionally been sold with source?

Businesses and people want independence where possible, but they also
want a supported producted. In the early days, having Unix source ment
you could create new drivers to supprot hardware, rather than wait for
AT&T Bell Labs to get around to it. It ment if you found a critical bug
   that would mean too much down time waiting for the official fix, you
could fix it pronto, ahead of Bell Labs.

Today, security and trust is more important, so having the source allows
for code review for back doors, buffer overflows, and other gotchas. OK.
Not everyone is a C programmer, but I'm trying to remain close to the
spirit of Unix and open-source, while pursing a living.

Some people get a buzz in finding and fixing bugs in commcercial code,
just to show off and prove themselves. Others do it because they need a
fix "now".  However, fixing a bug and sharing those fixes for a
commcercial source product, might not return you an income directly, but
did you ever consider that there might be future discounts on upgrades
for contributors?


 > that means that developing these milters is effectively a one-man show,
 > be that really one man or a company of a couple, so development process
 > and design decisions are concentrated in one pair of hands.


It currently is a one man show. I don't allow for published derivative
works of my software if you read the current license. I design it all. I
get suggestions, sometimes patches which I recode, but its all my code
after libmilter.


 > I do not mind to contribute by writing the code, testing, etc, but
 > buying (and managing) the licenses - I am fed enough with tracking all
 > this microsoft crap sitewide (won't ever write that using the capital
 > letter  :)


Another reason why I think "commercial source" is a simpler licensing
model. Selling a site license is easier for me to manage on my end. I
don't have to count machines and worry less about in-house copying. For
the buyer its easier, because they don't have to worry about how many
licenses they have, how many machines are installed, and proving such
when an auditor comes to control software licenses.

If I go a binary distribution route (not my prefered method), I have to
worry about licensing, locking code, machine counting, and priracy - in
other words I have think like Microsoft, the RIAA, and Hollywood, which
doesn't thrill me as an entrepeneur.

I think there has to be a middle ground.


 >>>>>I had a plans of Napoleon's magnitude but I might need to revisit
them
 >>>>>in the light of these recent news...  :(
 >
 >>>
 >>>Napoleon's magnitude? I don't follow the reference.


 > I meant "big". I wanted to build the comprehensive solution for mail,
 > using open source tools, which would address not only actual
 > delivering/filtering of mail, but also accounting, monitoring,
 > maintenance, etc. I am the only admin in our small company, so my goal
 > is to make it simple for one man to manage it (in addition to other
 > responsibilities).

I'm like that too. I work for a small ISP. I've migrate most of mail
system to open source solutions and/or implemented them myself. Great.
  From a sys.admin. view thats great. You've reduced the company's costs
for code.

However, software is a two way street, even free open-source solutions.
Most organisations Apache, FreeBSD, Mozilla, Linux (not sure) become
Foundations so that they can accept donations. Why is that? Because it
still costs money to develop code; at the very least equipment to
develop and test on, servers to host sites for distribution, archives,
bandwidth, electric bills, late night pizza...

Not everyone can find a day job to pay their bills with and work all
their free time on free code, not if they want a life and family.

Already looking many of the survey response, a lot of people have
indicated they would not pay for support as a business or an individual.
As an individual I can understand that support might be unrealistic, but
if a business (thats supposedly has a cash-flow) won't pay for
installation and/or support, then a "free source, pay for services"
model just won't support even just one salary.

That leaves either corporate sponsorship of some kind, like whats is
done for Linus T. and the full-time Perl developer; or your need a more
traditional business model barter/money in exchange for goods/services.


-- 
Anthony C Howe                                 +33 6 11 89 73 78
http://www.snert.com/       ICQ:
7116561         AIM: Sir Wumpus

            "Once...we were here."  - Last of The Mohicans


Lists Index Date Thread Search