LWN.net Logo

Not logged in

Log in now

Create an account

Subscribe to LWN

Recent Features

LWN.net Weekly Edition for May 23, 2013

An "enum" for Python 3

An unexpected perf feature

LWN.net Weekly Edition for May 16, 2013

A look at the PyPy 2.0 release

Printable page
Weekly edition Kernel Security Distributions Contact Us Search
Archives Calendar Subscribe Write for LWN LWN.net FAQ Sponsors

Freedoms granted by the GPLvX

Freedoms granted by the GPLvX

Posted Oct 18, 2006 9:22 UTC (Wed) by masuel (guest, #28661)
In reply to: Freedoms granted by the GPLvX by hofhansl
Parent article: FSF should separate GPLv3 changes (Linux.com)

... its a diffrent point of view

If I can have the changes you made, and understand them, I can benefit
from that. I can import the new feature you added onto my platform.

It is obvious that for the developer to benefit the product itself is
largely immaterial, so from that perspective, if the other party publish
changes etc they are "good" members of the comunity.

There is no need to manufactures to ship products user up-datable for the
relationship to be mutually beneficial between the 'open source'
and 'manufactures' developers.

The BSD licence is not comparable, doesn't insure this (vital?)
relationship.

"The spice must flow..."

The kernel devs are, mostly?, looking at the developer <-> developer
relationship. This is the reason they chose the GPL licence

The customer<->developer relationship is paramount to the FSF but not
everyone. It could be argued that the from the customers point of view
the "user update" feature could be a deciding factor when choosing which
product to purchase?

As for why the fuss, I think this steams from the fact that for your
printer, microwave oven, phone, aeroplane... etc this is not "new
technology" but has been the case, and accepted status quo, for the last
X years and now the FSF is looking to change the rules.

As LWN mentioned in an earlier article the previous position from Richard
has been stated as: "...I'm not really concerned with what's running
inside my microwave oven.", but that has now changed, as with the new
conditions he clearly dose.

This is a change, and one that is worthy of debate.

On the other hand the pattent provisions seem to be quite well received
by all(?) and this where the case for gpl 2.5 could (is) being made.

(Log in to post comments)

Freedoms granted by the GPLvX

Posted Oct 18, 2006 18:52 UTC (Wed) by nim-nim (subscriber, #34454) [Link]

> As LWN mentioned in an earlier article the previous position from
> Richard has been stated as: "...I'm not really concerned with what's
> running inside my microwave oven.", but that has now changed, as with the
> new conditions he clearly dose.

Oh, I doubt RMS' position on microwave ovens changed.

What changed is the FSF realised microwave oven business models were being extended to the computer space, and open computing devices couldn't be taken as a given anymore.

You may not realise FLOSS projects all over the world turn to the FSF for assistance whenever they have a legal problem (new law, "creative" GPL interpretation, etc), and as a result the FSF has a *much* more accurate view of legal trends and abuses than your average developper (exceptions being individuals like Harald Welte or Alan Cox who try to understand the legal landscape).

Would I trust RMS to write a successful OS kernel? No

But would I trust the people Linus herded in the Bitkeeper debacle on legal matters? God NO. The single most successful decision they did in the legal field was to use the GPL and let people who actually cared about legalities take care of them.

Freedoms granted by the GPLvX

Posted Oct 18, 2006 18:53 UTC (Wed) by Arker (guest, #14205) [Link]

You're mistaken, the FSF position hasn't changed in the least.

If the manufacturer can update the device, the customer must be able to as well.

If the manufacturer cannot update the device (as in the microwave) that's not the same situation.

What's changed is not the FSF position, what's changed is that more devices are being made which the manufacturers want to pretend are embedded, unchangeable devices, but really aren't. Tivo is, as usual, a great example. They can and do change the code on those things regularly, for their own purposes. They aren't willing to allow the actual owners of the devices, their customers, to do the same. That's where they're breaking the fundamental intention and spirit of the GPL, and that's why the GPLv3 is needed and is clearly in the same spirit.

Freedoms granted by the GPLvX

Posted Oct 19, 2006 2:33 UTC (Thu) by masuel (guest, #28661) [Link]

The FSF intention is clear.

But the users of the GPLv2 do not all have the same intention. It is a valid point of view that the licence should enforce the developer <-> developer relationship, not the manufacture <-> customer one.

The manufactures have been making devices for many years, from there point of view, as developers, this is a change.

It may well be argued that this was selective blindness, the aim of the FSF is about the customers freedom, so the licence should be read from that point of view.

However enforcing this explicitly is a significant change.*

Also the draft currently goes further and tries to control the behaver of the device after the gpl portion of it is changed. Not just parts covered by the licence. Is this morally and legally OK?

All in all its worth a bit of fuss anyway, don't you think?

*)by some readings the current LGPL already has this requirement...

ps I think the definition of embedded products (disallowing updates) is very, urm, odd? Apart from a few rom based products in my experience everything can be updated one way or the other.

Freedoms granted by the GPLvX

Posted Oct 19, 2006 4:03 UTC (Thu) by Arker (guest, #14205) [Link]

But the users of the GPLv2 do not all have the same intention.

I don't think that's ever been in dispute. Nonetheless, the FSF wrote the license, their intention and the spirit of the license was always clear, and the version 3 draft language clearly follows that intent and spirit, as it should. Those who don't like that remain free to use a different license.

It may well be argued that this was selective blindness, the aim of the FSF is about the customers freedom, so the licence should be read from that point of view.

I'm sorry, what is 'this?' Tivoisation is a relatively new phenomenon, and the FSF certainly haven't turned a blind eye to it for even a moment - they've been vocally opposed to it since it first became an issue.

Also the draft currently goes further and tries to control the behaver of the device after the gpl portion of it is changed. Not just parts covered by the licence.

Could you quote the language you're referring to? Because your description of it rings no bells.

ps I think the definition of embedded products (disallowing updates) is very, urm, odd? Apart from a few rom based products in my experience everything can be updated one way or the other.

I don't think it's odd at all. I've never heard of anyone updating the software in their microwave, for instance. All that's changed is that more and more general purpose devices are being sold in locked-down configurations where they pretend to be unmodifiable to their owner, but someone else still gets to change them at will, and often to the owners detriment. And I don't think that's something that should be encouraged in any way. Allowing the use of GPL software in such situations would certainly count as encouragement.

Freedoms granted by the GPLvX

Posted Oct 19, 2006 6:58 UTC (Thu) by masuel (guest, #28661) [Link]

>>It may well be argued that this was selective blindness, the aim of the
>>FSF is about the customers freedom, so the licence should be read from
>>that point of view.

>I'm sorry, what is 'this?' Tivoisation is a relatively new phenomenon,
>and the FSF certainly haven't turned a blind eye to it for even a
>moment - they've been vocally opposed to it since it first became an
>issue.

I was referring to the other developers of GPLv2 code.

>>Also the draft currently goes further and tries to control the behaver
>>of the device after the gpl portion of it is changed. Not just parts
>>covered by the licence.

>Could you quote the language you're referring to?

from Section "1. Source Code"

"The Corresponding Source also includes any encryption or authorization
keys necessary to install and/or execute modified versions from source
code in the recommended or principal context of use, such that they can
implement all the same functionality in the same range of circumstances.
(For instance, if the work is a DVD player and can play certain DVDs, it
must be possible for modified versions to play those DVDs. If the work
communicates with an online service, it must be possible for modified
versions to communicate with the same online service in the same way such
that the service cannot distinguish.)"

Now granted it all depends on what you consider "work" to be in the
above. If a "decoding" non-free proxy checks the signatures of the gpl
components and then doesn't decode just passes though the data... is this
in violation? I think so. But maybe you do not?

I *don't* really think the big picture (DRM bad, User Control Good) is
questionable, but I am yet to be convinced this is the way to do it.

Also sticking to my embedded point, as a general rule every device that
is connected to any kind of network is up-datable. These updates are
normally validated in some way and its not these devices that are
changing... they are and allways have been "embedded" devices to me.

Freedoms granted by the GPLvX

Posted Oct 19, 2006 17:38 UTC (Thu) by Arker (guest, #14205) [Link]

I was referring to the other developers of GPLv2 code.

Then I don't see what point you were making earlier, at all. What, they were supposed to forbid folks like Linus from using the GPL? How? Why? You're just not making any sense to me.

from Section "1. Source Code"[...]

Sorry, still not following your argument here.

After some pondering, I *think* the key may have been in an earlier post though.

Not just parts covered by the licence.

The license covers the terms under which permission to take actions otherwise forbidden by copyright law are forbidden. I think you're looking at it as "the license covers software, not hardware, but then it's trying to dictate something hardware related." But this is irrellevant. The license extends permission to do things otherwise forbidden by law, as long as the conditions of the license are met. Those conditions could be just about anything. In the case of the GPL, the idea is that if you want a license, you have to pass on all rights you received to those downstream of you. In a particular case that might limit your choices on hardware... so what? It's legally permissable and ethically required.

Also sticking to my embedded point, as a general rule every device that is connected to any kind of network is up-datable.

They'd better be, if they're directly connected. Which is exactly why you shouldn't have bricks directly connecting to public networks.

These updates are normally validated in some way and its not these devices that are changing...

There's no problem whatsoever with validating them, as long as the owner of the device has a way to control that validation process. The devices do change - every update changes them. They're general purpose computers outfitted with specialised software - not bricks.

Ever play rogue on a Cisco router?

Freedoms granted by the GPLvX

Posted Oct 20, 2006 1:10 UTC (Fri) by masuel (guest, #28661) [Link]

> What, they were supposed to forbid folks like Linus from using the GPL?
> How? Why? You're just not making any sense to me.

Some developers use the GPLv2 to ensure that other developers using there
code give the changes back to them.

The FSF uses the GPLvX to ensure the users freedom, ie that the user can
become a developer.

The changes from v2 -> v3 draft 2 make the distinction much clearer (or
actually important) hence some are saying they want to stay with v2.

Hope that helps a bit.

> In the case of the GPL, the idea is that if you want a license, you
> have to pass on all rights you received to those downstream of you. In
> a particular case that might limit your choices on hardware... so what?
> It's legally permissible and ethically required.

This is a good couple of sentences :)

The new bit is the limit on the overall "system/hardware"... it wasn't
present on the previous version of the licence.

Hence the discussion and quesion about the licence being "in the same
spirit" as v2... If you are the FSF its a no brain-er, but
for other people it may give pause for thought. No?

> Ever play rogue on a Cisco router?

... no sorry ... but I have taken features written for a platform I have
never seen and added to the box in-front of me, You must have done that
too :-)

The developers clearly benefit from the availability of the code without
the associated platform.

The FSF point of view that is not enough, some people say it is.

I hope you recognise it is worthy of debate even if your point of view is
that "enforcing code sharing" alone is naive/wrong.

ps I actually haven't made up my mind yet...

Freedoms granted by the GPLvX

Posted Oct 20, 2006 3:20 UTC (Fri) by Arker (guest, #14205) [Link]

Hope that helps a bit.

I don't disagree at all with what you're saying, but I don't see what it has to do with the earlier allegation of 'selective blindness' however. I suspect that those, Linus for instance, that have a real problem with the spirit of the GPL had ample warning and opportunity to omit the 'or later versions' clause, which was explicitly left out of the license itself to permit that, in fact.

I hope you recognise it is worthy of debate even if your point of view is that "enforcing code sharing" alone is naive/wrong.

I can see a worthy debate on some issues here, but not on whether or not v3 is in the same spirit. That question seems absolutely beyond question.

Freedoms granted by the GPLvX

Posted Oct 20, 2006 6:41 UTC (Fri) by masuel (guest, #28661) [Link]

> ... I don't see what it has to do with the earlier allegation
> of 'selective blindness'.

sorry for the confusion

As we have agreed the FSF view the GPL is about users freedom and some
developers view it as about enforcing sharing.

If the second party are not aware of the first parties intentions or
assume they are the same as there own then the "clarification" in the new
version could come as a bit of a shock. It is this mistake on the
*developers* part that I was classing as "selective blindness", the FSF
are far from blind...

It could, therefor, I feel be argued that "in spirit" is only true of the
FSF.

By changing the licence, any developers just meant "tit-for-tat sharing"
on there code but included the "...or later version" cause risk having
there code taken and extended under a licence that, from there point of
view, not in the same spirit.

Does this make sense?

(I know the bsd lot must be use to the feeling but its a new one to GPL
people...)

Copyright © 2013, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds