Problems I see with the GNU General Public License (Version 2 and the Version 3 drafts)


1: Service providers loophole
	There has been much talk in the open source community about this loophole,
	it is widely known that companies providing services using open source
	software have been getting away with 'cheating' by simply storing the
	modified software on their servers and allowing users to interact with
	the software over the network.  By operating in this fashion they comply
	with the GPL without actually having to give back any modifications or
	even giving access to the unmodified original source code because technically,
	by the definitions given by the GPL, they have not distributed the software.

	This needs to be stopped IMO.

2: Use (and possibly modify) but don't distribute loophole
	Not all GPL software is related to providing services over the network.  In my
	experience, companies only make available their modifications to GPL software
	if they absolutely must (if they have distributed the software).  I believe the
	fact that not all GPL'd software is used by service providers to basically run
	applications remotely is an important one.

	For example, lets say there is a monitoring system, XYZmon, that is available
	for free on the internet under the terms of the GPLv2.  A hosting company, can
	download XYZmon, configure it, and use it to monitor everything in its
	datacenter XYZmon supports.  Soon, the company discovers XYZmon doesnt scale
	well as the datacenter continues to grow, so they fork XYZmon and put some
	developers on the project, effectively making it proprietary.  After a year or
	so of development, XYZmon has become an incredibly scalable piece of software
	and now supports almost twice as many device and service types for monitoring,
	and can even do it's monitoring at 10 times the resolution previously possible.

	The developers employed by the company realize how incredibly better their
	fork is than the current state of the original XYZmon, so they request
	permission from their management to send the local fork sources upstream.
	Management ignores all requests and the company shows no interest in giving
	back to the community which provided the shoulders to stand on with the
	original XYZmon allowing the company to get where it is now in such short time.

	What options do the developers have?  Were they "distributed" or "propagated"
	copies of the software or are they essentially part of the corporation working
	on a single copy the company downloaded?  Even if the employees believe the GPL
	gives them right to distribute their modified version because they have copies
	of it, even though this is all done on behalf of the corporation paying them,
	do they have the guts to just release it to the wild and probably lose (or at
	least seriously hurt) their jobs in the process?

	This is a problem, and it's not all that different from problem #1.

	It's becoming increasingly apparent that the license simply needs to require
	source availability of the original program and any modifications if you are
	executing the program, in any form.


There is an effort to plug the service providers loophole, it's known as HPL,
the Honest Public License.

HPL is simply the GPLv2 with minor modification, most notably:

d)  For the purposes of determining the right to obtain copies of the source code (as
well as the right to  modify and distribute such source code and object code), the
term distribution shall include the communication of the Program or work based on
the Program which is intended to interact with third party users (meaning anyone
other than you or if you are an entity such as a corporation and not an individual, that
corporation),  through a computer network and the user shall have the right to obtain
the source code of the Program or work based on the Program. This provision is an
express condition for the grants of license hereunder and any such communication
shall be considered a distribution under Section 1, 2 and 3.


This helps, it is an improvement because it will protect things like web
applications from being modified extensively and never getting redistributed.

But is this really a total solution?  What about programs that are very useful
and advantageous but dont get used by anyone outside the party who downloaded,
modified it, and continues to use it under the radar?

I don't think the GPL is doing a good job of protecting our freedom in these
situations, and I do believe they are quite common.  Hell, at Hostway I remember
two cases just off the top of my head that were major service provider loophole
exploitations:
proprietary blog app made from wordpress
proprietary ecommerce system made from oscommerce

Hostway would *go great lengths* to avoid installing copies of modified gpl
packages among the customers data, because it had no interest in giving away
code it developed even though it was all done standing on the shoulders of
the free community.

There were also plenty of applications all over the place that had gpl code
sucked in or inherited from legacy software that was being improved, but
there was no concern of gpl compliance because there was never any intention
of distribution.  But in most cases, the gpl software was being improved, and
the community as a whole would have benefitted from getting the changes.

Shady? sure, irritating? yep.  Illegal? nope.  Unethical? nope.  Hey, the
license set the rules, they are playing by them.

Do I want to release my software under the GPLv2 or v3 drafts as-is after these
experiences?  NO.  Why?  Because I won't be free to see what changes were
made to my software by companies or individuals that made it more efficient,
more scalable, more useful, or even better, more profitable, simply because
they avoided distributing it.

Distribution is becoming less and less important in terms of software usability,
as networks continue to get faster and bandwidth cheaper, it will only get
worse... imagine a time where nobody has software other than a core operating
system just to run *gasp* a web browser.  And this doesnt even have to do with
the cases where software doesnt provide service but simply performs tasks
(monitoring, trading, rendering...).


(C) 2006 Vito Caputo - <vcaputo@pengaru.com>