FSF should separate GPLv3 changes (Linux.com) [LWN.net]

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 7:25 UTC (Wed) by hofhansl (guest, #21652)
In reply to: FSF should separate GPLv3 changes (Linux.com) by svkelley
Parent article: FSF should separate GPLv3 changes (Linux.com)

Companies have to make a living. They likewise have restrictions on liability. I am *not* going to give you loader access to my embedded flight management system

Last time I looked nobody is being held at gunpoint and forced to use open source software. You use the work of others? Then respect the rules set by the authors. If you don't want to do that write it yourself.

As I stated somewhere below, TIVO is circumventing the freedoms granted by the GPL. They are taking work from others and denying the freedoms granted by the license. If they can't live with people recompiling and running new versions of the software, why don't they write their own OS or pay up to another vendor?

(Log in to post comments)

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 12:33 UTC (Wed) by svkelley (guest, #37299) [Link]

>>As I stated somewhere below, TIVO is circumventing the freedoms granted by the GPL. They are taking work from others and denying the freedoms granted by the license. If they can't live with people recompiling and running new versions of the software, why don't they write their own OS or pay up to another vendor?

Indeed, that is what will happen. Companies can and will choose WinCE and other commercial operating systems due to this restriction. Love it or leave it, understood.

We aren't making devices for hobbyist programmers. I am not an exception. In fact, just my interest in embedded Linux has been an uphill battle. GPLv3 essentially puts a knife in many interested developers like myself who want to see their company make greater use of embedded Linux. I am not making this liability and certification issue up. These are real commericial costs that have to be factored into the decision making for doing a device at all. I see an overall lack of understanding of consumer device development in this thread. Products come and go. They are changed quite frequently - sometimes they have a 3-6 month lifespan. We are not talking PCs with a standard platform for development.

Sean

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 13:27 UTC (Wed) by khim (subscriber, #9252) [Link]

I see an overall lack of understanding of consumer device development in this thread.

True. And that's good thing. I (and most of others) don't care about products "lifespan on market". I do care about usage. I f I do want to tinker with the box in the future - I'll buy Linux-based device. If I do not want to tinker with the box - I'll buy WinCE or VxWorks or whatever. My friends bought WTR54GL (and not WRT54G which is cheaper) exactly to have the ability to play with it later. When I buy the linux-based device and later (when this device is of course already fogotten by the manufacturer and is not supported) found out that there are no way to tinker with it - it feels like betrayal of trust...

Now if there are no way to bring Linux-based tweakable device on the market - bring WinCE's or VxWork's based one. This is where I truly don't care what's inside of the box: if I don't have the ability to change it what difference it makes if it's MS-DOS, Linux or even OS/360 ?

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 14:19 UTC (Wed) by cventers (subscriber, #31465) [Link]

I see it going one of three ways:

1. People get incredibly paranoid, over-react, and fork important
projects back off into GPLv2-only, leaving a large market of free
software available to be abused by manufacturers, who scoop it up rapidly
and DRM-lock every new device on the market, leaving the hobbyist
embedded developer _nowhere_ to go;

2. The kernel developers stand largely alone in sticking their nose up in
the air, some manufacturers get by with just the kernel, and the hobbyist
is still fucked;

3. Enough GPLv3 software exists that some companies pay for WinCE or
VxWorks in order to avoid the DRM restriction. Then their competition
comes out and builds a better, cheaper hammer on GPLv3 free software -
without the DRM.

By the way, you can replace 'hobbyist' with 'user' in these scenarios.
You don't have to be a programmer to enjoy the GPL's freedom to adapt the
software, just look at Rockbox for a great example.

And I hear that you're not making devices for hobbyist programmers, but
we hobbyist programmers are not making free software for you to abuse.
Would you like to meet at the same table and work out a mutually
beneficial relationship in which you freely use our free software to
build a better hammer, so long as we reserve the right to use it how we
please once we've payed you for it?

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 21:30 UTC (Wed) by RareCactus (guest, #41198) [Link]

I don't know which way each particular project will go, but I suspect that most of the important ones will stay GPLv2.
In particular, it is certain that we will have at least a GPLv2-only kernel and libc.

None of this, of course, has anything to do with whether hobbyists will be able to modify things. That will be determined by what kind of DRM laws are passed, and what kind of hacked hardware consumers are able to get through the black or grey market.

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 21:39 UTC (Wed) by cventers (subscriber, #31465) [Link]

> None of this, of course, has anything to do with whether hobbyists will
> be able to modify things. That will be determined by what kind of DRM
> laws are passed, and what kind of hacked hardware consumers are able to
> get through the black or grey market.

You claim to be a kernel developer in another one of your posts, and I
think this is one place where many kernel developers appear confused.

The anti-Tivoization clause in GPLv3 is a clarification intended to
prohibit manufacturers from implementing technical restrictions
preventing users from modifying the covered software work; something
embodied certainly in the spirit of GPLv2 if not strongly in plain legal
terms. In that sense, it has everything to do with whether hobbyists will
be able to modify things.

In fact, if the ability of hobbyists to modify things were something
kernel people considered important, and they still didn't think GPLv3 was
the right way to go about it, they might be better served by merely
acting on their own opinion rather than trying to stop people that have
other opinions from acting differently.

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 14:25 UTC (Wed) by cventers (subscriber, #31465) [Link]

> Most comments coming from the Kernel Camp are Trollish and FUDish.

Absolutely, and I think it is offending a lot of Linux users, developers
and evangelists. I have so much respect for the kernel community, having
sat with my toe in that swimming pool now for a few years, and some of
the comments made by Linus and others (especially to the press) have
forced me to ask a lot of tough questions - not about the comments but
about the people making them.

Disagreeing about a software license is one thing, but what I've stood
witness to is something entirely different. There's this great fear in
circulation (largely started by the kernel community) that GPLv3 is going
to split us down the middle and ruin us, yet they seem to totally miss
the irony of their choice of language and tactic.

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 21:15 UTC (Wed) by RareCactus (guest, #41198) [Link]

Linus chose the GPLv2 because he liked the license, not because he necessarily agreed with RMS in any other way.

I am a Linux developer, and I agree with Linus' choice, then and now. If you want politics, go over to the projects which do politics. We do code.

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 18, 2006 21:25 UTC (Wed) by cventers (subscriber, #31465) [Link]

Color me a little confused with your response. I'm not calling the kernel
guys out because they dislike GPLv3. I think they have as much right as
anyone else to have an opinion on what license best suits their code. I
think it is likely (given discussions on LKML, the views of the most
active kernel developers, and practical matters) that Linux will remain
GPLv2, and I'm perfectly okay with that. I'm not offended.

What I'm not okay with is the ad hominem attacks aimed at RMS and the
FSF, and the blatant lies about the GPL drafting process. That offends
me. If kernel people aren't going to use GPLv3, fine, but why not leave
the rest of us alone?

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 19, 2006 0:19 UTC (Thu) by walken (subscriber, #7089) [Link]

> I am a Linux developer, and I agree with Linus' choice, then and now. If you want politics, go over to the projects which do politics. We do code.

What about the document titled 'The Dangers and Problems of the GPLv3' - was that code or politics ?

Please get off your high horse. You can't dismiss people arguments as 'doing politics' when they're only replying to your own inflamatory statements.

FSF should separate GPLv3 changes (Linux.com)

Posted Oct 19, 2006 15:24 UTC (Thu) by GreyWizard (guest, #1026) [Link]

Well said!