News for hackers. Stuff that matters.
Re:60x speed-up of Linux “perf” by dvdkhlng on Thursday September 09, 2021 @09:56AM (#28469727)
IMO it's not "GPL virality" that is at the root of the issue here. The authors' decision to license theis works this way causes the problem. You can still ask both libraries' authors for relicensing their work in a less restricted / more compatible way.

If they don't want to do that, then being authors and owning their copyrights, that's their right to do. That's nothing I would blame on the GPL.

I mean we should be grateful that these authors made their works available free of cost in the first place.

Re:60x speed-up of Linux “perf” by nickitolas on Thursday September 09, 2021 @11:55AM (#28471085)
As I understand it relicensing in many open source projects can be a very complicated matter, since in some cases it requires getting permission from all past contributors (Which could number in the hundreds)
Re:60x speed-up of Linux “perf” by guerrilla on Thursday September 09, 2021 @02:00PM (#28472643)
Can someone ELI5 to me where the problem is because both perf and bfd seem to be GPL. Why would the former not be allowed to link to the latter?
Re:60x speed-up of Linux “perf” by loeg on Thursday September 09, 2021 @02:12PM (#28472791)
GPLv2 and GPLv3 are difference licences, despite sharing a name. The GNU people (FSF) intentionally made the GPL3 incompatible with GPL2, to enforce the virality of the GPL3 license (which they perceive to be superior to the older GPL2).
Re:60x speed-up of Linux “perf” by Macha on Thursday September 09, 2021 @03:02PM (#28473330)
The purpose of the GPLv3 was to address the "tivo" clause, where a vendor sells you a device with some modified GPL preloaded software included. You can request the software under GPL, but if you can't load it on the device then the FSF feels that's not very good.

So GPLv3 bans that. That, by itself, is inherently an additional restriction not allowed by GPLv2. They don't need to go out of their way or have a conspiracy to deliberately make it incompatible. Especially since the GPL2, as distributed and suggested, includes the "or later" text which resolves this incompatibility to allow GPL3 software to use GPL2 code. If anything, this encourages GPL2+ as the default GPL license if maximum compatibility is your goal.

Others (e.g. Linus), are more focused on getting the code changes so if they want to use them in the original project they can, and don't feel being able to install it on the device it was built for is as high up on the priority list.

Re:60x speed-up of Linux “perf” by pabs3 on Thursday September 09, 2021 @09:12PM (#28476666)
Its also interesting to note that none of the GPL licenses require giving code back, so what Linus wants from the license doesn't happen via the license, only via cultural standards. Indeed, there are many companies and persons violating those cultural standards by not sending patches back to Linus, as well as many companies violating the GPL licenses by not distributing source to their customers.
Re:60x speed-up of Linux “perf” by pabs3 on Thursday September 09, 2021 @09:09PM (#28476651)
A couple of interesting posts from Software Freedom Conservancy make it clear that what Tivo did (breaking proprietary software when reinstalling Linux) is allowed, even by the GPLv3 and that allowing reinstall of libre software is a requirement of the GPLv2 license.

https://sfconservancy.org/blog/2021/jul/23/tivoization-and-t... https://sfconservancy.org/blog/2021/mar/25/install-gplv2/ https://events19.linuxfoundation.org/wp-content/uploads/2017...

Re:60x speed-up of Linux “perf” by boomboomsubban on Thursday September 09, 2021 @03:12PM (#28473466)
Most GPL licenses permits their code to be redistributed as the same GPL version or any later version of the GPL, Linux releases as GPLv2 or later for example.

Perf does not permit the license to be redistributed as anything except GPLv2, so it conflicts with GPLv3 code.

Re:60x speed-up of Linux “perf” by Someone on Thursday September 09, 2021 @03:38PM (#28473758)
“Linux releases as GPLv2 or later for example”

Linux doesn’t use “or later”. https://github.com/torvalds/linux/blob/master/COPYING:

The Linux Kernel is provided under:

  SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note
Being under the terms of the GNU General Public License version 2 only, according with:

  LICENSES/preferred/GPL-2.0
With an explicit syscall exception, as stated at:

  LICENSES/exceptions/Linux-syscall-note
In addition, other licenses may also apply. Please see:

  Documentation/process/license-rules.rst
for more details.

All contributions to the Linux Kernel are subject to this COPYING file.

Re:60x speed-up of Linux “perf” by Matthias1 on Thursday September 09, 2021 @04:27PM (#28474262)
Sure, it's not the license itself. But this does read like a problem. It's difficult to imagine that the authors of libbfd licensed it under GPLv3 because they wanted to ensure that no one would link it against perf. Maybe they do have strong feelings about supporting that license.

But this is an example of the ugly side of open source licensing. A lot of people don't have strong feelings about ensuring the distribution terms of their code, and just don't care how it's used. And in those cases, it can be annoying to have e.g. opensource.org insist that GPLv3 is the best option.

Edit: To be more specific, I mean that this problem could be avoided if both softwares were released under a public-domain-equivelent license. But of course, that will never happen.

Re:60x speed-up of Linux “perf” by Dylan16807 on Thursday September 09, 2021 @06:42PM (#28475544)
The difference in license only matters because of a broad interpretation of virality. The easy way to show this is to mentally replace GPL with LGPL and notice that the problem disappears.
Re:60x speed-up of Linux “perf” by for1nner on Thursday September 09, 2021 @11:44AM (#28470940)
> The authors' decision to license theis works this way causes the problem.

I'm not sure how you can rationally make an argument for a framework that literally limits the use of improved or more performant code/technology/understanding simply because it does not share a philosophical view of IP/ownership/sharing.

This is the core of the problem with IP in the real world as well, and I find it mind-boggling that we don't look at the SYSTEM and see it as problematic.

Re:60x speed-up of Linux “perf” by aseipp on Thursday September 09, 2021 @01:25PM (#28472209)
Actual IP law and how it conducts itself in the real world has plenty of issues to be sorted out, but there are tons of potential "inefficiencies" in the real world that could be exploited and rectified across a vast array of things, but aren't, due to combinations of social politics, law, philosophy, etc preventing it. That some potential efficiency is on the table in theory, but can't be reaped due to those things, is nowhere remotely close to "mind boggling"
Re:60x speed-up of Linux “perf” by pessimizer on Thursday September 09, 2021 @12:49PM (#28471767)
It doesn't limit it any more than if it were proprietary and you couldn't afford it.