A day at the Open Source Business Conference

[Posted May 23, 2007 by corbet]

LWN readers will certainly be aware that your editor spends a fair amount of time at development-oriented conferences. In some ways all conferences are alike, but, still, all that experience was insufficient to prepare your editor for OSBC, which is a different sort of affair. Neckties, Blackberries, and Windows laptops are ubiquitous. There are booths for law firms. People wonder about whether customers should buy their "open source" software licenses on a one-time or subscription basis. The wireless network actually works, but power outlets are nowhere to be found. It's all very strange.

Red Hat CEO Matthew Szulik started off his talk with an effect-filled video filled with Gandhi quotes and related material, presumably to the end that open source is headed toward an inevitable victory. His talk, once he started talking, was a fairly general presentation on the value of open source software, standards, interoperability, etc. Lots of talk about how young the people pushing open source tend to be. He also dwelt for a while on the "social mission" of open source software, and discussed just how seriously the open source community takes intellectual property issues. LWN readers would likely not find much new there; it was more of a motivational talk for others building businesses in this area.

There was a session, led by Larry Augustin, on "downloads to dollars" - how to start making money once you have people actually downloading your software. Much talk on how to extract information from downloaders which can be used to "open a dialog" with them. When is the proper point to start requiring registration, with a valid email address, to download a software tarball? It was suggested that the source download is really the same thing as the free trial offerings from some proprietary vendors, with the same end: to lead to the "monetization" phase. It can all sound cynical and manipulative to ears more accustomed to gatherings of developers, but this is the sort of thing people building businesses in the open source mode worry about.

There was a lawyer-led session on reciprocity requirements in the GPL. Much worry goes into trying to figure out just when it might be permissible to ship proprietary components with free software. The presenter, Stephen Gillespie, thinks that GPLv3 might make the mixing of proprietary code easier in some situations.

Another session purported to explore "the future of open source," but seemed to be more about the present of open source companies. Much time was spent conducting polls of the audience by having everybody send their responses as cellular text messages to a special number. Eventually time got tight and the moderator came up with the new idea of having people simply raise their hands instead. Lots of talk about how customers should "buy" their open source software licenses. There was also discussion on how to build a community around software releases, though the main concern seemed to be keeping the download rates high.

In general, participants here are concerned with download counts. A large number of downloads is a crucial indicator of a successful open source release; prospective venture capitalists always want to know what the download rate is. Some participants seem to have concluded that there is a lot of useless downloading going on; people just collect software because it's out there for free. But they still want to know how to improve download rates.

The day ended with a keynote by Eben Moglen. It was a long, wandering discussion in classic Moglen style, well worth listening to. The core point, however, was that the way to build prosperity at any level - from nations to small businesses - is to stand up for freedom. At the business level, that includes using copyleft licensing for software. BSD-style licenses, he says, are "a really good license for your competitor to use." Any business which does not want to provide a free lunch for its competitors, however, should use a license which requires others to give back their changes.

In the question period, your editor asked about his statements that Microsoft would, by virtue of distributing Novell's coupons, eventually find itself bound by the terms of GPLv3. He answered that there was much he could not talk about because he signed a non-disclosure agreement to be able to read the terms of the Microsoft-Novell deal. He had expected that agreement to be moot by now, but those terms still have not been made public. Until then, he says, we can look at the terms which have been put into GPLv3 which require the granting of a patent license to all recipients of the software. We can also look at Microsoft's behavior, which includes "throwing coupons out of airplanes" and attacking the GPLv3 patent provisions, and come to our own conclusions.

Responding to another question (about the lack of terms requiring web service providers to give back changes they are running but not distributing to others), Eben had a fairly strong warning for Google. If the company continues to operate in a secretive way and not contribute back the bulk of its changes, there will be growing pressure for a remedy based on licensing terms. It is really up to that one company, he says, to determine where that aspect of the debate goes in the future.

The second (and final) day at OSBC will include a keynote by Marten Mickos, a panel on license enforcement actions, and a panel on the good (or not so good) effects of the Microsoft/Novell deal. Stay tuned for the report.

Index entries for this article
Conference	Open Source Business Conference/2007

(Log in to post comments)

A day at the Open Source Business Conference

Posted May 23, 2007 3:49 UTC (Wed) by kwink81 (guest, #33926) [Link]

"BSD-style licenses, he says, are "a really good license for your competitor to use." Any business which does not want to provide a free lunch for its competitors, however, should use a license which requires others to give back their changes."

Well put. Do we have a transcript of his talk anywhere? I would like to see exactly how he phrased this.

A day at the Open Source Business Conference

Posted May 23, 2007 4:38 UTC (Wed) by cventers (guest, #31465) [Link]

Screw a transcript, I want video! Here's to hoping someone was recording.

Moglen Webinar at Groklaw

Posted May 24, 2007 14:53 UTC (Thu) by southey (guest, #9466) [Link]

There is a Moglen Webinar at Groklaw within the story "Patent News from Novell and Moglen - Updated - OpenLogic Webinar in Ogg format" (there is an also earlier story with another format).

Choosing a free software license

Posted May 23, 2007 18:54 UTC (Wed) by mheily (subscriber, #27123) [Link]

Even though I prefer the BSD license for my own code, I think Eben Moglen is right to say that the GPL is a better choice for commercial software companies looking to reap the benefits of the open-source development model without taking the risk of having a competitor take the code and incorporate it into a competing product.

However, the fact that the GPL is a more corporate-friendly license does not mean that it is the best license in all circumstances. Many valuable contributions to the free software universe have come from academic, governmental, non-profit, and hobbyist developers. The BSD license, and similar licenses such as the Apache, Mozilla, CDDL, ISC, and MIT licenses, etc., should be considered whenever the profit motive is not the direct motivation for a project's development activity.

These more liberal licenses allow the code to be used in ways that provide the greatest amount of good for the greatest number of people. For example, imagine if Microsoft had been unable to import the TCP/IP stack from the BSD operating system into Windows due to licensing restrictions. They would have written their own buggy, incompatible implementation that only worked properly when two Windows hosts were communicating. Or imagine if Apple were unable to use the FreeBSD operating system as a basis for their proprietary UNIX system. They would have used the NeXT kernel and userland instead, spent a lot more time and resources maintaining the codebase, and the resulting Mac OS X would have been of lower quality and more incompatible with the rest of the UNIX family.

Hopefully my comments will not spark another round of the GPL/BSD license wars. Both licenses have their strengths and weaknesses, and developers should consider all their options and choose a license that meets their personal, political, and professional needs.

Choosing a free software license

Posted May 23, 2007 19:33 UTC (Wed) by kwink81 (guest, #33926) [Link]

>"imagine if Microsoft had been unable to import the TCP/IP stack from the BSD operating system into Windows due to licensing restrictions. They would have written their own buggy, incompatible implementation that only worked properly when two Windows hosts were communicating."

Yes, I agree that all standards of communication, such as Vorbis, ODF, and TCP/IP should be under permissive licenses that allow them to be implemented as widely as possible.

>"imagine if Apple were unable to use the FreeBSD operating system as a basis for their proprietary UNIX system. They would have used the NeXT kernel and userland instead, spent a lot more time and resources maintaining the codebase, and the resulting Mac OS X would have been of lower quality and more incompatible with the rest of the UNIX family."

If FreeBSD had been under the GPL, free software would have been given a significant competitive advantage over proprietary software?

For many of us, that is EXACTLY what we want. We want to help bring positive social change to the software world, not hand the old dictatorships stronger pairs of handcuffs.

Choosing a free software license

Posted May 23, 2007 20:53 UTC (Wed) by mheily (subscriber, #27123) [Link]

> "If FreeBSD had been under the GPL, free software would have been given a significant competitive advantage over proprietary software?"

Unfortunately, no. My point was that Apple would not have chosen to use a GPL-licensed operating system because it would have forced them to give away a lot of proprietary code that is critical to maintaining their market niche. They have a number of APIs (Carbon, Cocoa, Quartz, IO Kit, etc.) that are intertwined with the kernel and operating system. These APIs provide the "graphical user experience" that is at the heart of their business model.

I just ported a low-level system library from Linux to OpenBSD, and then to OS X. It was pretty easy due to the similarities between Linux and BSD. If Apple had used a proprietary UNIX OS (NeXTstep) as the foundation for OS X, porting would have been more difficult. In my mind, wasting developers' time is a far greater crime than selling a mixture of free and proprietary software.

Choosing a free software license

Posted May 24, 2007 11:38 UTC (Thu) by timschmidt (guest, #38269) [Link]

I think you misunderstand his meaning... He did not mean to imply that Apple would have chosen a GPL'd core for OS X, merely that Apple would have had to re-invent the wheel, and so, been at a disadvantage as compared to GPL'd OS vendors.

Choosing a free software license

Posted Jun 7, 2007 7:47 UTC (Thu) by hozelda (guest, #19341) [Link]

>> If Apple had used a proprietary UNIX OS (NeXTstep) as the foundation for OS X, porting would have been more difficult. In my mind, wasting developers' time is a far greater crime than selling a mixture of free and proprietary software.

Good BSD code allows proprietary companies to thrive, while good GPL code/competition forces them to join the competition and open up.

I and many that support GPL software absolutely care about not wasting developer time. We also care about not wasting end user's time. This is why we don't see why we should use licenses that help large proprietary companies to thrive, especially in a way where through the very powerful lock-in mechanisms, mixed with closed source, a few such entities eliminate a lot of business (and with it money to support even more FLOSS) for many others and put the end users in shackles, forced to accept inferior products than what they would otherwise have.

If all noncommercial entities benefit from BSD the same as from GPL, but if commercial entities (the competition not the incumbant) benefit more from GPL, then it seems to me that if the noncommercial entities join the commercial companies and adopt GPL, they will do two things to improve their position. One is that they will ultimately help to lead to a situation where there is more FLOSS code available for them (5 items of GPL vs. 1 of BSD, 2 of GPL, and 1 of proprietary (.5 of which reinvents the other .5); 5>3). Two is that they will have insurance so if they leave academia or whatever their nonprofit status is (especially true of students) they are more likely to have an interesting job because no very small number of companies dominate, setting the agenda on their closed systems for everyone else. And of course, I haven't even discussed that these non commercial developers are first and foremost end users, and have friends that are end users but not developers, etc. This is one more reason to help proprietary companies (like Apple) change to a FLOSS model and to make sure there is as much free (both senses of word) software as possible, ie, to add to the pool of good GPL code without adding to the pool of good BSD code.

Choosing a free software license

Posted May 24, 2007 10:07 UTC (Thu) by NRArnot (subscriber, #3033) [Link]

> "I agree that all standards of communication, such as Vorbis, ODF, and TCP/IP should be under permissive licenses that allow them to be implemented as widely as possible"

Danger!

This plays right into the hands of those who would "embrace, extend and extinguish". Or those who would insert a "poison pill" (a non-free technology) into an open standard (Rambus's ambush of the JEDEC memory hardware standards springs to mind, though ultimately Rambus failed).

If not the GPL, at the minimum one needs a license that makes sure that the license dies the moment that an implementation of an open standard is extended in an incompatible way. I don't actually know of any license that aims at BSD-openness but also makes sure that a standard (for a communication protocol, in the widest possible sense) is maintained as a single open standard rather than fragmented into mutually incompatible, proprietary, patent-encumbered variants. Is it possible?

If not, the GPL is best - it forces those who use and distribute the original code to give back their code and (GPL V3) patent licenses, but unless there is a way to force them to give back as a minimum full disclosure of their modified protocols and patent licenses to use the same, GPL V3 will have to suffice.

Choosing a free software license

Posted May 24, 2007 10:17 UTC (Thu) by mjthayer (guest, #39183) [Link]

What about LGPL?

Choosing a free software license

Posted May 24, 2007 12:00 UTC (Thu) by job (guest, #670) [Link]

Agreed. Note that even Stallman argued for the Ogg Vorbis code to be put under the (revised) BSD license, for exactly these reasons.

Choosing a free software license

Posted May 24, 2007 17:17 UTC (Thu) by njs (subscriber, #40338) [Link]

>Or imagine if Apple were unable to use the FreeBSD operating system as a basis for their proprietary UNIX system. They would have used the NeXT kernel and userland instead, spent a lot more time and resources maintaining the codebase, and the resulting Mac OS X would have been of lower quality and more incompatible with the rest of the UNIX family.

Or imagine if Apple were unable to use GCC as a basis for their proprietary compiler system. They would have used EDG or something instead, spent a lot more time and resources maintaining the codebase, and the resulting compiler would have been of lower quality and more incompatible with GCC and other compilers...

Well, except that they actually decided that they didn't want their compiler to be proprietary so badly that they were willing to put in that effort, and now are a major contributor to GCC proper -- so on net, way more people are getting way more benefit than they would have if GCC were under a BSD license, because we all benefit from Apple's code.

Their compiler and kernel are totally different matters, of course, and probably if the only kernels available had been GPL they would still have made the extra effort to keep it proprietary, because it's much more central to their business model. But the point is that you can't just say that BSD licenses provide the greatest good for the greatest number of people; that's only a rule if you look in the short term and ignore system effects.

Google ?

Posted May 23, 2007 5:12 UTC (Wed) by khim (subscriber, #9252) [Link]

Is there a reason to single out Google ? As far as I know Google did everything GPL requires it to do. For example when Picasa was "ported" to Linux all patches to WINE were made available. And most of the software is used in-house by Google so according to the FAQ they don't need to release anything (they often do this anyway since it's just easier then to maintain in-house patch) but there are no "pressure for a remedy based on licensing terms"!

Now I'm not sure it's good that GPL allows this kind of behavior, but it just look ridiculous to first allow private modifications and after that complain that patches are not returned back...

Google ?

Posted May 23, 2007 6:04 UTC (Wed) by xoddam (subscriber, #2322) [Link]

Moglen is arguing that private development of copylefted software for
public consumption is contrary to the spirit of copyleft, not that it
violates the letter of the GPL.

No-one is raising doubts that Google is in compliance with the GPL. It is
definitely a good Open Source citizen, making major contributions to the
free software ecosystem with Summer of Code, by directly employing key
Linux kernel maintainers, and even with some distinct projects of its own.

But it is also true that Google does a lot of in-house development,
possibly based on GPL code (how would I know?) which is *used* by the
public -- on Google's servers -- but not *released* to us. There isn't
much pressure to release it because most of the Google infrastructure
software is only of interest to people with a large distributed network of
storage servers. There is no doubt that such things as GoogleFS could be
useful to people outside of Google, but now if you want to play with them
you basically have to use them on Google's machines. The most attractive
option is to get Google to hire you.

The Affero Public Licence is something Moglen has developed with the
intention of keeping server software enhancements in the public realm. I
have no idea if this, or other legal measures, have any hope of bringing
Google to heel while the company itself is so magnificently
self-sustaining.

The public, no matter how freedom-loving, is not going to boycott Google's
services because the source code to the infrastructure is not available
for download.

Maybe Eben Moglen would do better appealing to Google's famous non-evil
conscience than sending "stern warnings".

Google ?

Posted May 23, 2007 11:07 UTC (Wed) by hpp (subscriber, #4756) [Link]

GoogleFS is developed and used only within Google, but do we have any evidence that it includes GPL'ed components? The same questions applies to other pieces of the Google world (map/reduce, monitoring code, search code) - it may run on Linux but I am not aware of it including GPL'ed components.

As such, there shoud be no pressure on Google to release any of this...

Google, yes

Posted May 23, 2007 14:41 UTC (Wed) by man_ls (guest, #15091) [Link]

As such, there shoud be no pressure on Google to release any of this...

Why not? At least the same pressure as on all other proprietary vendors that "do not get it". Maybe even more: other companies are openly "evil" and do not care about doing things right. Google does care, or at least that is the public line.

Google, yes

Posted May 23, 2007 17:51 UTC (Wed) by emkey (guest, #144) [Link]

I don't think it is the vendors who don't get it in this case. Nobody owes me a thing. I'm grateful when given a gift but I do not mistake it for anything else other than a gift.

This overly developed sense of entitlement by certain people in the free software community is just annoying and very counter productive.

Free software has gained the ground it has because it is a superior model. Odds are reasonably good it will continue to gain ground for the same reason.

Just to be clear, I'm not painting Moglen with this brush. He may in fact have a legitimate point in this case. I don't know enough to say at this juncture.

Following a superior model

Posted May 23, 2007 19:46 UTC (Wed) by man_ls (guest, #15091) [Link]

Sure, and it is out of the pure kindness of our hearts that we want Google to release their own software as free software. Mind you, we want Google to succeed, and the easiest way to success for such a good friend of free software is surely to free their own software. Since it is such a superior model, it is odd that they themselves have not seen the light, but whatever. A slight prodding is probably all they need to do it, lest something bad (following an inferior model) befalls them.

Now seriously. You are right in most things, but even Linus thinks that "GPLv2 is about tit-for-tat", and he is not suspect of ideological purity. Now let's zoom out a bit. In a general sense you could say that Google is what it is because of free software, so asking them to give some of their own back would not be so outrageous. All xoddam and others are saying is that freeing their own software would be what separates "good" citizens from "outstanding" citizens.

Meanwhile Moglen only speaks about modifications to software which is already free. Pure tit-for-tat, probably even Linus would be with him in that.

Following a superior model

Posted May 23, 2007 22:41 UTC (Wed) by i3839 (guest, #31386) [Link]

> asking them to give some of their own back would not be so outrageous.

Following the same logic, you could ask writers who use open source for their work to release their books under a free license.

Distributing local modifications to free software used is of course something different and a very good thing, but still not something that should be forced. Just imagine the pain when using a LGPL library with your own buggy program that arbitrarily scribbles memory!

Following a superior model

Posted May 31, 2007 12:06 UTC (Thu) by zotz (guest, #26117) [Link]

"Following the same logic, you could ask writers who use open source for their work to release their books under a free license."

OK. Here you go:

http://www.ourmedia.org/user/17145
http://www.youtube.com/profile?user=zotzbro
http://pc.celtx.com/profile/zotz
http://www.archive.org/search.php?query=%28creator%3A%22d...

If we can figure out a way for people to make a decent living using Free Licenses, why not seek to promote such?

all the best,

drew

Superior model ? Hardly...

Posted May 23, 2007 23:24 UTC (Wed) by khim (subscriber, #9252) [Link]

I often see this screams about "the superior model" and how everyone should only ever write free software. While free software clearly wins when moral and ethic is included when we are talking about practical viewpoint everything is not so easy. I think Craig A. James said it best: There is a natural "lifecycle" to software technology, which includes both commercial periods and FOSS periods..

If you'll think about successful FOSS projects - they are either converted former proprietary projects (Firefox, OpenOffice.org, etc) or reimplementation of proprietary projects (Linux, GCC, MySQL, etc). In rare cases FOSS dominates the niche from the start to finish (web-servers: from NCSA httpd to Apache), but it only happens when initial design is simple enough and can be done without massive efffort. Otherwise proprietary leads for a while but eventually FOSS overcomes it - and looks like that is natural lifecycle of software... I suspect most of Google's software is at the beginning of this cycle. But some are already at the middle and may be even closer to the end (whoever uses Picasa for Linux ? what for ?).

Superior model ? Hardly...

Posted May 24, 2007 7:20 UTC (Thu) by intgr (subscriber, #39733) [Link]

The vast majority of popular proprietary software is also reimplementation of other proprietary products; I think it's just a matter of luck depending on who is in the right place at the right time. But indeed, free software seems to be winning in the long run (or at least that's what I would like to think).

Superior model ? Hardly...

Posted May 24, 2007 8:00 UTC (Thu) by man_ls (guest, #15091) [Link]

Some proprietary products are also reimplementations of free products; like IIS from Apache or the Windows NT networking layer (some say directly ripped off BSD). Or version control software from CVS. Or even all Unices from the original AT&T Unix, which followed a model closer in many respects to free software than to proprietary products (patches and distributions) and which resulted in the free BSD.

Superior model ? Hardly...

Posted May 24, 2007 17:34 UTC (Thu) by njs (subscriber, #40338) [Link]

Proprietary software does seem to be better are opening up markets, though -- look at, oh, say, video editing software right now. The commercial development houses looked around at the market and could see that consumer video was coming down the pike and going to be a big thing, so they invested the capital up-front to build video editing suites, and it's partly the existence of cheap and usable editing suites that's making it possible for ordinary people to start playing with video and grow the market.

Traditional FOSS development is not at all forward-looking (and it's even a point of pride, for good technical reasons, YAGNI and all that). It wasn't until lots of people had access to hardware and were playing around with it that the itch grew to create free editing software, and so now the free stuff is still way behind the proprietary stuff. It's not clear that the free stuff would exist at all without the proprietary stuff having enabled hardware sales and development.

No reason that commercial investment on speculative markets has equate to proprietary software, though -- they could just as well invest capital up-front in developing a GPLed hunk of software, make money on support, and then when the market picked up ride the influx up new developers right past their competitors. Until that happens, though, this is a particular niche situation where FOSS development has fewer than usual competitive advantages over proprietary stuff.

Superior model ? Hardly...

Posted May 24, 2007 20:12 UTC (Thu) by oak (guest, #2786) [Link]

> If you'll think about successful FOSS projects - they are either
converted former proprietary projects (Firefox, OpenOffice.org, etc) or
reimplementation of proprietary projects (Linux, GCC, MySQL, etc). In rare
cases FOSS dominates the niche from the start to finish (web-servers: from
NCSA httpd to Apache), but it only happens when initial design is simple
enough and can be done without massive effort.

These are the new FOSS projects. There are a couple of good examples of
successful projects 10 years earlier (mid 80's); TeX typesetting system
and the X Window System. I wouldn't call neither simple, they were quite
massive by the standards of the time when they were created. At that time
there were no proprietary alternatives to them I think (e.g. TeX used
bezier curve strokes to describe its fonts in 1984). Both were created in
the university environment.

I'm personally still using X daily and TeX weekly, and both are also
used commercially. The TeX project had last known bug almost 20 years
ago (it hasn't change much since then). What proprietary software can
claim the same track record?

Google ?

Posted Jun 8, 2007 10:58 UTC (Fri) by hozelda (guest, #19341) [Link]

>> As such, there shoud be no pressure on Google to release any of this...

The pressure is not on Google to release something it won't have to under an Affero type license. The pressure is on every company (with special focus on one of the wealthiest businesses in the world) to make sure that if they are not using GPL, that they continue not to if they want to keep things secret. Otherwise, they are legal today, but things may change so (eg) Google should plan now. They would then need to open up or else reimplement the GPL components (that become GPL3,4,etc) to keep their secrets.

It's called "advanced notice -- just in case."

Google ?

Posted Jun 8, 2007 11:05 UTC (Fri) by hozelda (guest, #19341) [Link]

I forgot to mention, I do think Google has contributed. But it does seem it is disproportionate what Google has gotten from its use vs. what it has given back. In such a case, those contributing might want to adjust their future license to make such a disparity less likely.

Google ?

Posted May 23, 2007 13:56 UTC (Wed) by dank (guest, #1865) [Link]

I think the reason people are worrying that Google might not
be a great free software community citizen is that it's not
easy for outsiders to tell how much of Google's infrastructure
involves GPL'd code, and furthermore how many enhancements Google
is making to GPL'd code but is not contributing back to the public.
Maybe it would help for Google to invite an outside observer (Eben?)
to have a look and report back to the community.

Chris DiBona said in http://www.eweek.com/article2/0,1895,2132480,00.asp
"I can look at any end binary in the company and I can tell you what open-source software is expressed within thatbecause of the way that we manage our code base."
That might be a starting point for the outside observer: Chris could
give him or her a tour through the Google infrastructure and show
just how much or little GPL code is linked in.

It's worth noting that Google ships some of their infrastructure
as part of their Search Appliance, which means that even without
the service provider clause in the new GPL, they are forced to
be very careful about linking in GPL'd software.

Google ?

Posted May 23, 2007 14:33 UTC (Wed) by NigelK (guest, #42083) [Link]

I think it's a very dangerous path to go down to suggest that those users of GPL software who provide little or no new code as "suspect" and deserving of a code audit.

Smacks too much of Microsoft and BSA intimidation tactics.

Google ?

Posted May 23, 2007 16:09 UTC (Wed) by anselm (subscriber, #2796) [Link]

The GPL says that any recipient of GPL code is absolutely entitled to do with it what they want on their own machines, including changing it until its own mother wouldn't recognise it and not sharing the result, as long as they do not distribute compiled binaries of the code in question to others. If the folks at Google want to grab GPL code off the net and rewrite, enhance, ..., it for their own use, then more power to them. This is how the GPL works, and how it is intended to work.

When did that change?

Google ?

Posted May 23, 2007 17:03 UTC (Wed) by amikins (guest, #451) [Link]

Seconded. It's always been held that if someone wants to combine their proprietary code with GPL'd code in-house, without distributing it to anyone else, this is just fine. It's their codebase, and they can do what they like with it.. It's only when they go to distribute other people's code that they need to worry about the license.
The definition of 'distribute' has gotten mightily strained lately, between this and the nonsense with Microsoft's coupons.

Google ?

Posted May 23, 2007 18:43 UTC (Wed) by gnu_lorien (subscriber, #44036) [Link]

"When did that change?"

It hasn't changed, but it was a considered change for the GPLv3 that provided web services and the like should count as a form of distribution.

Both sides of this debate have valid points, and it seems to revolve around what we mean by "in-house." I would define "in-house" as meaning software that only affects your company. This could include custom modifications for your processes, internal bug tracking, etc.

Where some minds differ is when you develop "in-house" software that is marketed and sold as a public web service. While the software lives exclusively in-house, you are allowing outsiders to utilize it. Many feel that it violates the spirit of free software to not provide these people with a stab at the changed sources to set it up on their computer and do what they want to with it.

Copyright law defines what constitutes "distribution"

Posted May 23, 2007 19:38 UTC (Wed) by tseaver (guest, #1544) [Link]

The wish of some (e.g., in the "Affero" license) to extend that definition
to include situations not covered by copyright law are likely to end
up unenforceable, at best.

Copyright law defines what constitutes "distribution"

Posted May 23, 2007 23:32 UTC (Wed) by MathFox (guest, #6104) [Link]

Copyright law also covers "public performance" and a case can be made that running software on a publicly available computer (for instance a webserver) constitutes public performance of the program.
IANAL and I know that the details of the law vary by country... Affero might be stretching the GPL a bit, certainly not recommended for every application, but it is founded on the letter of the law.

Copyright law defines what constitutes "distribution"

Posted May 23, 2007 23:56 UTC (Wed) by xoddam (subscriber, #2322) [Link]

That's true, but copyright law (in some countries) covers things other
than distribution; for instance the right to prepare modified works or (as
a previous reply mentions) to perform the work to a public audience. The
point is that it is illegal (in those jurisdictions) to make "in house"
modifications except as allowed by the copyright licence (and by "fair
use", if such a right also exists in that country).

The GPL v2 gives blanket permission to modify the code as the licensee
pleases. The Affero Public Licence withholds that permission in specific
circumstances, namely it prohibits the removal of a prominent "download
the version of the code which provides this service" button from a public
web service.

The new notion of "propagation" in the most recent draft of the GPL v3
allows the extension of copyleft protection into areas not normally
considered by copyright law, by selectively withholding permission to do
things which *are* the provenance of copyright. This applies, in
particular, to the GPLv3's patent and anti-DRM provisions, and to the
no-more-MS-Novell-deals clause.

It might apply equally effectively if "propagation" were defined also to
include providing a service to the public based on the covered work. The
current GPL v3 draft does not do this; a later version of the Affero
licence probably will.

If copyright gives holders the right to demand payment per copy or for a
licence to prepare derivative works, it certainly gives them the right to
impose conditions on the way in which those actions are permitted.

Nevertheless I don't think applying those terms to the GNU body of code
would be sufficient to force a company the size of Google to publish its
in-house code. I am certain they are very well aware of what code they
own and what they have licensed from elsewhere, and of the terms of each
and every license.

Such a move might be so counterproductive as to move Google from the "free
software is a good thing and I'm pleased to contribute even if not all my
products are free software" camp into the "FSF are communists who want me
to give away my golden goose" camp.

Copyright law defines what constitutes "distribution"

Posted May 31, 2007 12:35 UTC (Thu) by zotz (guest, #26117) [Link]

I don't see why people seem to thing this. Can you explain the thinking for me?

I can deny the right to copy my work even in house unless you agree to the conditions set in the license. Yes? No?

all the best,

drew

Google ?

Posted May 24, 2007 2:30 UTC (Thu) by lysse (guest, #3190) [Link]

> It hasn't changed, but it was a considered change for the GPLv3 that provided web services and the like should count as a form of distribution.

By that standard, then if I provide someone with a shell account on my machine, then I am distributing every program they can access to them; even if I only provide execute access to the programs, I would still have to provide them with a means of downloading all source code. I can't see any sensible way of defining "distribution" that would include web services but exclude this; one might draw a distinction between public distribution and limited distribution, but that would make the desirable qualities of the GPL easily circumventable.

So such a restriction would have to be defined as a restriction on use, rather than on redistribution... and then you'd have to define "use" in a way that includes "present to the world as a web app" for Perl scripts, but excludes "compile a program" for gcc, or "produce a Postscript file from this text file" for groff - or even "run this script" for Perl itself - and you'd have to be absolutely sure your definition was both precise and enforceable.

Er, eek?

Google ?

Posted May 31, 2007 12:39 UTC (Thu) by zotz (guest, #26117) [Link]

"By that standard, then if I provide someone with a shell account on my machine"

Could be. Anyone agree? Disagree?

What about putting together a system, hardware and modified software, and renting it out.

all the best,

drew

Google ?

Posted May 24, 2007 8:11 UTC (Thu) by ekj (guest, #1524) [Link]

It didn't change.

But the *spirit* of the GPL is that every single user of a piece of software should have the freedom to improve, change, redistribute, adjust that piece of software, if he/she so chooses.

This breaks down with web-services.

If you took a GPL-program and adopted it so that it could run as a web-service, you could then (currently) sell access to this web-service per-seat, and the users would be unable to improve, change, redistribute or adjust the program they are using.

Which is contrary to the spirit of the GPL, but perfectly legal according to the *letter* of GPLv2.

In effect, if you use GPL-software only for a web-service, you're allowed to treat it like BSD.

Google ?

Posted May 24, 2007 9:17 UTC (Thu) by anselm (subscriber, #2796) [Link]

Let's stipulate for a moment that you are right about the »spirit« of the GPL, as applied to web services.

The question here is how far down the stack this should extend. Presumably Google are running their own proprietary search engine code, on which the »spirit of the GPL« has no bearing, on top of a possibly-patched Linux kernel. Should they be morally obligated to publish their Linux kernel patches just because the machine in question is running (or even merely talking to another machine running) a web server that is accessible by the public?

Google ?

Posted May 24, 2007 22:57 UTC (Thu) by giraffedata (guest, #1954) [Link]

No, something changed.

The GPL didn't change, and legal requirements didn't change, but the thing that Stallman sought to regulate with GPL has changed. Much of the computer use that formerly happened via distribution of software now happens by offering a service on the Internet. For example, this year I filed my income tax using a web service, whereas in the past I would pick up a CD at a store. And it looks like that trend will continue. Essentially, we're moving from the Microsoft era to the Google era.

So to continue the freedoms GPL hoped to give to computer users, there needs to be something new.

Google ?

Posted May 31, 2007 12:27 UTC (Thu) by zotz (guest, #26117) [Link]

"When did that change?"

When software as a service came into the game? (Spirit wise, not letter wise at this point perhaps.)

all the best,

drew

Google ?

Posted May 23, 2007 20:26 UTC (Wed) by BeS (guest, #43108) [Link]

>But it is also true that Google does a lot of in-house development, possibly based on GPL code (how would I know?) which is *used* by the public -- on Google's servers -- but not *released* to us.

I don't care that much about it, because even if Google would release this software probably most of us wouldn't be able to use the software because we don't have the servers Google has.

I care much more about software which Google develops and releases as non-Free Software (e.g. Google Earth). That's where Google looks really bad as a "Free Software citizen".

Google, why not

Posted May 23, 2007 8:50 UTC (Wed) by man_ls (guest, #15091) [Link]

Now I'm not sure it's good that GPL allows this kind of behavior, but it just look ridiculous to first allow private modifications and after that complain that patches are not returned back...

The GPLv2 was released in 1992, which was a different world: no internet, no HTML, client-server largely meant in-house... Nowadays "private modifications" might even have a different meaning, as "something you modify exclusively for your internal use". Google modifications are quite public in the sense that many people feel their effects.

I don't know either if it is good to embed this feeling in a license. As xoddam says above, an appeal to their conscience might be a better strategy. Google is a good citizen, but compare them to Sun who is an outstanding citizen: Sun takes the initiative to make lots of their own software free software.

Google, why not

Posted May 23, 2007 9:38 UTC (Wed) by NigelK (guest, #42083) [Link]

The recent arguments about the GPL3's Tivoization and patent clauses will seem like a picnic compared to what will happen when you say "if you use modified GPL'd software to produce a report for someone, then you *must* publish those modifications."

Thankfully the Afferro licence is a separate licence, and so reduced to niche status.

Google, why not

Posted May 23, 2007 17:08 UTC (Wed) by amikins (guest, #451) [Link]

Hasn't it always been held that the output of a program generally can't be considered a 'derived work' unless it somehow contains some portion of the original work? If it's not a derived work, then how can the license of the original work hold any relevance to the output?

Google, why not

Posted May 23, 2007 17:26 UTC (Wed) by man_ls (guest, #15091) [Link]

Nobody says that the output is a derived work. But since you are giving a license to your copyrighted work, and distribution is essentially forbidden in copyright law, you can limit distribution to whatever conditions you want. That is how the Affero Public License can do its job. Quoting from it:

You are not required to accept this License, since you have not signed it. However, nothing else grants you permission to modify or distribute the Program or its derivative works. These actions are prohibited by law if you do not accept this License. Therefore, by modifying or distributing the Program (or any work based on the Program), you indicate your acceptance of this License to do so, and all its terms and conditions for copying, distributing or modifying the Program or works based on it.

So, can you request in a license that distribution be performed only by specialists in Sanskrit Literature while they dance the Conga? Essentially, yes (although that would probably remove it from all free software or open source definitions). Can you request that programs run in private machines but accessed by people outside your organization be distributed only with source code? By all means. Whether you should do it is a different question.

Google, why not

Posted May 23, 2007 17:47 UTC (Wed) by amikins (guest, #451) [Link]

I have no arguments with you on distribution, I agree fully with what you've stated in reference to that. My issue is, if the program is run on a private machine accessible by people outside your organization.. Are you _distributing_ anything? To me, it looks like the only thing distributed is the output, not the program itself.
My question is not whether you should require source distribution in such a context..
My question is whether you _can_. Requesting is not the same as requiring, and I'm really curious as to where the law stands on this. Does 'copyright' come into play when making modifications that don't involve distribution? To me it seems it would not, making the license irrelevant if the software itself stays within the organization.

Google, why not

Posted May 23, 2007 18:37 UTC (Wed) by AJWM (guest, #15888) [Link]

It's not a question of distribution. As soon as you modify a program you have created a derived work, which is prohibited by copyright without the owner's permission.

Google, why not

Posted May 23, 2007 19:23 UTC (Wed) by man_ls (guest, #15091) [Link]

Against your argument it could seem that modification without distribution is different from distribution, and does not apply here; that is why I can sing my own version of "My way" (my "my way", if you will) in the shower without violating any laws. Well, I don't know about US law, but here in Spain software follows a completely different set of rule than artistic works.

The GPL FAQ is very lax about distribution "within your organization"; you can distribute a program without restrictions within your corporation, non-profit organization, and so on. You will find that copyright law is not so lax; you cannot make "private" copies within your organization, and even the number of personal copies for backup is limited. It has even been argued (this time in the US) that "running" a program is "distribution", since you are making a "copy" in memory of the software. Few people associated with free software want to tread that rute, but it is quite clear that distributing within your organization without a license is a big no-no.

Engineers would be forced to distribute code + patches and modify the original locally, which would not be the most convenient arrangement one can conceive -- even if Google is sure to find a way to automate it complying with the license at the same time. Gentoo could probably use some of those tools... if they ever see the light of day. Sigh. Too circular even for my liking.

Fair use

Posted May 24, 2007 0:44 UTC (Thu) by xoddam (subscriber, #2322) [Link]

> I can sing my own version of "My way" (my "my way", if you will)
> in the shower without violating any laws.

I am *certain* that this is covered by "fair use" in any reasonable
jurisdiction -- even where fair use does not formally exist :-)

Your neighbours may complain about noise pollution, but that's
a separate matter!

Fair use

Posted May 24, 2007 6:23 UTC (Thu) by man_ls (guest, #15091) [Link]

I don't think we have this "fair use" in Spanish law, although I'm certainly not a lawyer. The law regulates private performances:

The communication will not be considered public when it is performed in a strictly domestic scope which is not integrated or connected to a network of any kind.

So if I had a webcam in my shower (even if it is strictly private) a guy might show up and require payment.

Your neighbours may complain about noise pollution, but that's a separate matter!

Hey, I do what I can! :D

Fair use

Posted Jun 1, 2007 8:52 UTC (Fri) by forthy (guest, #1525) [Link]

> Your neighbours may complain about noise pollution, but that's
> a separate matter!

It's not. Public performances are controlled by copyright. If you intentionally give a public performance (i.e. not in your shower, unless you put the video on YouTube), you need the copyright holder's agreement.

Putting a program for everybody's use on a web server is probably the equivalent of "public performance". The law possibly doesn't say anything about "public performance" of programs (because literally, this would mean "read the source code out aloud" or something like that), so it's not a well-defined legal term. Therefore, the Affero license goes through the "modification" clause. After all, if you run an unmodified program, the source is available, anyway.

Google, why not

Posted May 25, 2007 13:40 UTC (Fri) by sepreece (guest, #19270) [Link]

"The GPL FAQ is very lax about distribution "within your organization"; you can distribute a program without restrictions within your corporation, non-profit organization, and so on."

Actually, I think this is inherent in the GPL. Since the corporation is providing copies only to itself, it only has obligations with respect to itself. That is, it has to provide itself the source, if asked, and grant itself any patent license it already holds. Both of those are, obviously empty implications.

I think the GPL FAQ you cite is probably wrong in its last paragraph (it says that providing copies to contractors for use off-site is distribution). I think a company that showed that it had rules requiring that such copies be treated as loans of the company's property and returned on completion of the contract, coupled with reasonable auditing to make sure that that happened, would be hard to attack in court under the terms of the license. Since ownership of the copies remained with the company, only the company would be entitled to the rights required under the GPL. [But, IANAL]

Google, why not

Posted May 25, 2007 14:38 UTC (Fri) by man_ls (guest, #15091) [Link]

Actually, I think this is inherent in the GPL. Since the corporation is providing copies only to itself, it only has obligations with respect to itself. That is, it has to provide itself the source, if asked, and grant itself any patent license it already holds. Both of those are, obviously empty implications.

I don't think so. A corporation that distributes or modifies software has obligations not to downstream receivers but to the owner of the copyright. This owner might set whatever requirements it saw fit, as for example "distribute source code upon request to any users, local or remote, of the software". In the case of the GPL the owner sees fit to set requirements for modifications only to outside distribution, but this is only in the FAQ and not in the license itself. (I didn't know this, by the way.)

So your rights for "internal distribution" are not as clear-cut as those in the GPL itself. In fact, the FSF might change the FAQ tomorrow if they see it fit and restate this paragraph, which would open a lot of entities to being sued. Basically, anyone who modifies and distributes a GPL program. Funny, isn't it?

I think the GPL FAQ you cite is probably wrong in its last paragraph (it says that providing copies to contractors for use off-site is distribution).

Imagine that you have a corporate license for a Microsoft program, and you want to distribute it to off-site contractors. Would you think it is wrong for the license to state that you cannot do it? Hardly, even if the company has strict rules and auditing and whatnot. Copyright law doesn't entitle you to internal nor external distribution, so your only way to do it is the license, in this case the GPL and the FAQ. Now that I think about it, I wouldn't even rely on the FAQ, but I guess you might argue you have "acquired rights". IANAL either.

Google, why not

Posted May 25, 2007 16:13 UTC (Fri) by sepreece (guest, #19270) [Link]

I hope I put this all in the context of the GPL, not general copyright law. Absent the license, copyright would clearly prohibit making internal copies.

In the terms of the GPL, however, I believe what I said was correct. The GPL establishes the grant of permissions and hte responsibilities associated with them and those seem to me to allow internal distribution because such distribution is already meeting the requirements of the license (and, thereby, your responsibilities to the copyright owner).

The FAQ has no legal force, so changing it would not in any way affect the terms of the license.

Again, note that I am only talking in the context of the current license terms. The GPL could be rewriten to limit internal distribution, since that is a right reserved to the copyright holder. I am simply saying that under the current terms, internal distribution does not need to be considered specially - it's just like any other distribution.

Google, why not

Posted May 25, 2007 15:09 UTC (Fri) by man_ls (guest, #15091) [Link]

On second thought, forget about my earlier post. You are right that from the current version of the GPL it follows that you can distribute a program internally, with or without modifications. A change in the FAQ would not mean anything.

The GPL itself would have to be changed a lot to close the "ASP loophole"; for example, distribution of modified copies would have to be coupled to distribution of source code to users. It would be a mess, and maybe it is why they haven't gone that way with GPLv3.

Google, why not

Posted May 31, 2007 12:47 UTC (Thu) by zotz (guest, #26117) [Link]

"Does 'copyright' come into play when making modifications that don't involve distribution?"

Isn't that all that is needed? Law wise...

all the best,

drew

Google, why not

Posted May 23, 2007 11:41 UTC (Wed) by jeroen (guest, #12372) [Link]

As far as I know Google actually never released any significant program they wrote as Free Software. Google Earth, Picasa, SketchUp, Google Desktop, etc. It's all proprietary and most of time it only runs on that proprietary OS. Their appliances are probably full of proprietary software too.

Is google a good citizen? If you look at the software they are using themself, they are using the freedom to modify the software and contribute back a lot. But it doesn't seem they want to give that freedom to the users of their software. And IMHO that's a bad thing.

Google, why not

Posted May 23, 2007 17:15 UTC (Wed) by amikins (guest, #451) [Link]

For some of that, there's some legal constraints outside their control. For instance, Google Earth makes use of data they don't control, and their own contracts prohibit them from allowing access to that data without prevention to ensure the data doesn't escape.
For others, I can't say offhand, but if it's their software in entirety, they have a right to do with it as they like. They're a service company, and that's how they make their money. That they provide so many services at effectively no cost and try to assist with the community at so many other points is, in my opinion, to be commended.

Google, why not

Posted May 24, 2007 12:45 UTC (Thu) by RobSeace (subscriber, #4435) [Link]

> 1992, which was a different world: no internet

Um, I recall being rather heavily addicted to the Internet in college in 1992 (and earlier), so I can assure you it certainly existed then... ;-)

Google, why not

Posted May 24, 2007 19:30 UTC (Thu) by leoc (guest, #39773) [Link]

Yeah, it was about 92 when I got my first internet address, found usenet and got into my first flame war. We have come so far since then!

Google, why not

Posted May 24, 2007 20:28 UTC (Thu) by jeffw (guest, #45419) [Link]

True, perhaps, but the general case of "application service providers" or "software as a service" has always been around. Arguably, ignoring PCs, it is the most frequently occurring type of software use.

H&R Block started Compuserve in the 1960s to sell computer time on their big systems outside of tax season. The Multics project was started in the 1960s specifically to develop software to run a "computer utility". Tangentially, even Centrex vs PBXs.

Assuming that the "ASP Crisis" exists at all - it is a direct result of the license. Yes the legal situation has changed, and the GPL needs to be updated to reflect the reality in 2007, and the reality globally. But the ASP problem has been around for ever - RMS diddnt need to foresee it back in '92. - it was there. That he diddnt see it is an indication that he isnt quite as perfect as some would like you to believe.

Google ?

Posted May 24, 2007 18:10 UTC (Thu) by JohnNilsson (guest, #41242) [Link]

See http://video.google.com/videoplay?docid=-7477852615698435519

Google Tech Talks March 27, 2007

ABSTRACT

The relation between Google and the free software movement is one of the most important ... all » diplomatic relationships in the 21st century. But it is largely invisible, even to the principals. In this talk I will try and make some of what we have taken for granted less implicit, so we can progress with mutual confidence and collective security.

Speaker: Eben Moglen, Software Freedom Law Center

A day at the Open Source Business Conference

Posted May 23, 2007 7:22 UTC (Wed) by mjthayer (guest, #39183) [Link]

I do wonder how many of the FOSS projects currently making money started out with that aim. Firefox was proprietary software which went free, and owes its advertising revenue to the fact that it was already successful. Wine was already a successful non-commercial project before a few of the developers were able to start making money from it.
My feeling is that most of the people making money are riding on the back of successful non-commercial projects (although if that helps those projects it need not be a bad thing).

Mozilla revenue

Posted May 24, 2007 0:47 UTC (Thu) by xoddam (subscriber, #2322) [Link]

> Firefox was proprietary software which went free, and owes its
> advertising revenue to the fact that it was already successful.

More to the point, it owes its advertising revenue to Google.

Registration

Posted May 23, 2007 10:48 UTC (Wed) by dark (guest, #8483) [Link]

"When is the proper point to start requiring registration, with a valid
email address, to download a software tarball?"

I would say... when you're tired of maintaining the software, and would
like someone to fork it.

Registration

Posted May 23, 2007 17:25 UTC (Wed) by JoeBuck (subscriber, #2330) [Link]

The problem with questions like this is to figure out what is really being asked. The questioner usually isn't interested in registration for its own sake. He or she is interested in statistical data: how many users are there? Who are they? Why are they interested?

Users can often be persuaded to provide the same information voluntarily. For example, when they download, they will probably be interested in subscribing to a low-volume list that would be limited to new release announcements and alerts about security bugs, but would not be used for anything else.

Registration

Posted May 23, 2007 17:58 UTC (Wed) by wblew (subscriber, #39088) [Link]

However, as a *user* I see registration another way.

I see it as being put on somebody(s) marketing list(s) to receive their unsolicited marketing materials (i.e. spam).

Thus I largely opt out on the "please register your software".

Yes, I put my right to privacy ahead of their right to marketing data.

If there are two things I *love* about open source, they are:

1) usage without fee or registration

2) the ability to fix a *really* annoying bugs

PS: Yes, I have made a career of writing computer software.

Registration

Posted May 24, 2007 21:39 UTC (Thu) by proski (subscriber, #104) [Link]

When I download software:

1) I want to try it. Chances are that I won't like it. I don't want to subscribe to anything yet.
2) I want to install it on another computer. Changes are I have already subscribed to everything I want.
3) I want to try the latest version (as opposed to the version that comes with my distro). Even in this case, it's likely that I just check the changelog and the mailing list archives.

Besides, I really don't want to give my contact information to any commercial entity that may have a stupid marketing department or may be forced to sell its e-mail lists in case of bankruptcy.

Even legitimate companies could misuse e-mail eventually. I got spammed by MontaVista recently. Apparently their marketing department just paid some spammers for the service. My position was strengthened by the fact that I hadn't subscribed to any of their mailings. If I had, they would have an excuse.

Eben Moglen just gave a recipe for prosperity.

Posted Jun 5, 2007 20:31 UTC (Tue) by mariomiy (guest, #45615) [Link]

The core point, however, was that the way to build prosperity at any level - from nations to small businesses - is to stand up for freedom. At the business level, that includes using copyleft licensing for software. ... Any business which does not want to provide a free lunch for its competitors, however, should use a license which requires others to give back their changes.

Any country that wishes to build prosperity without bounds to innovation should remove patents and adopt the GPL for all ideas that can improve society.

That is the way to reward creative people: recognize them. It is the fastest way for a country to improve as a whole. Unfortunately, the U.S. and Europe might have a hard time to adjust to this model. It is open for third world countries, and it is a great opportunity for them to catch up, and even surpass.