Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 10 Jul 2001 17:58:00 +0200
From:      Eivind Eklund <eivind@FreeBSD.org>
To:        j mckitrick <jcm@freebsd-uk.eu.org>
Cc:        freebsd-advocacy@FreeBSD.org
Subject:   Re: BSD, .Net comments - any reponse to this reasoning?
Message-ID:  <20010710175800.A77023@FreeBSD.org>
In-Reply-To: <20010710151059.A52201@dogma.freebsd-uk.eu.org>; from jcm@freebsd-uk.eu.org on Tue, Jul 10, 2001 at 03:10:59PM %2B0100
References:  <20010630174743.A85268@dogma.freebsd-uk.eu.org> <20010707160255.A18525@thinksec.no> <20010710151059.A52201@dogma.freebsd-uk.eu.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 10, 2001 at 03:10:59PM +0100, j mckitrick wrote:
> | More emotionally laden nonsense.   There are a bunch of reasons to contribute
> | changes back to the open source projects:
> | (1) You get much less integration work when you want to utilize newer version
> |     of the open source project.  Basically, the changes you have now made
> |     maintain themselves WRT the open source project for free, rather than
> |     needing more care.
> 
> That's one I never thought of or hear stated in that way.

Weird.  It's more or less the canonical argument for why contribute back to
BSD-licensed projects.  We obviously need a FAQ or good paper about this.

> | > Just because a thing like apache did not get seriously forked is no
> | > indication that it cannot ever happen. Why take risk? Why use the
> | > economical lever, when the legal lever is much more direct and more powerful
> | > in this case?
> | 
> | Because we see benefits in properitary extensions.  When somebody use our
> | codebase in a proprietary environment, they are working on the codebase.  This
> | means they likely will be producing beneficial changes to the codebase.  These
> | changes come in two forms - strategic changes, that are sellable and part of
> | added value, and tactical changes, that have value as levers for creating the
> | strategic changes, but no intrisic competitive advantage.   The latter usually
> | are more plentiful than the former, and have larger value when given back to
> | the community (buying goodwill) than when kept proprietary (costing money to
> | maintain).
> 
> Could you give an example of these 'levers' ?

A few examples (tilted in favour of work I've done and contributed back,
because that's what's easiest to remember):
- netgraph, developed by Whistle and contributed to FreeBSD.  This was viewed
  as strategic for a while, but when the Interjet was developed enough, it was
  degraded to being viewed as tactical and reasonable to contribute back.

- The CAM code in FreeBSD, developed by Plutotech for their embedded video
  editing system and contributed back.  This is the present basis for the
  FreeBSD SCSI subsystem.

- The original PnP support for the ed driver, developed by Yours Truly for Yes
  interactive and donated back to FreeBSD.  This also contained a simple but
  significant bug (one line difference) that was found by Bruce Evans after it
  was contributed back (thus demonstrating the point of getting multiple
  experts to look at it)

- Support in i4b (the FreeBSD ISDN code) for using userland PPP.  Developed by
  Yours Truly for Yes Interactive and contributed back.  My work on i4b also
  avoided the reproduction of a significant bug that used to be present in
  bisdn (the predecessor to i4b) in i4b.  This was a bug I found when making
  bisdn work for userland PPP, and which had blocked the tty driver there from
  working properly (instead giving sporadic crashes) through all releases of
  bisdn, and which I would not have found if I hadn't worked on that code for
  proprietary use (a week of debugging.)

- Support in libalias (the backing library for ppp -nat and natd) for punching
  minimal holes in an ipfw firewall for active protocols (ftp, irc dcc).
  Developed by me for Yes Interactive, and contributed back.

- A lot of the VM system work done by John Dyson for FreeBSD was sponsored by
  Network Computers (a subsidary of Oracle) because they needed a better VM
  system for the servers they used for the NCs.

And that's just a sample.

Eivind.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-advocacy" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010710175800.A77023>