News for hackers. Stuff that matters.
Re:Theo de Raadt on LLVM license change by Koshkin on Saturday October 01, 2016 @11:48AM (#12618337)
Everybody enjoys free (as in "freedom") software. Turns out, even quality-wise free software is often much better than non-free, closed-source software. On the other hand, it may be hard to really understand the rationale behind it. The idea goes against the principles of capitalism (based on property rights). Also, if it is completely OK for hardware design to be proprietary, why should it not be OK for software? What is so special about it? The arguments put forward by proponents of free software do not seem very convincing, and sometimes their personalities do not help attracting people to their cause. Often, this ideal becomes a nuisance even in the world of open source. If you want to use Debian, for instance, you are pretty much forced to use "non-free" repositories (or a "non-free" Debian re-distribution) for your setup to be useful.
Re:Theo de Raadt on LLVM license change by rubber_duck on Saturday October 01, 2016 @12:15PM (#12618473)
>Turns out, even quality-wise free software is often much better than non-free, closed-source software.

This is only true in the very narrow segment of 'tools for geeks by geeks', 'tools that big software companies need but don't provide a competitive edge/need to be OSS for adoption so they opensource' (Apple, Google, Microsoft) and companies that sell support for OSS (RedHat, Novell, Canonical, etc.)

As soon as you move from IT infrastructure/dev tools there is no contest between OSS and commercial packages or they only cover small a niche. Compare GIMP vs Photoshop, Blender vs Maya/3DSMax/ZBrush, Cycles/Yafaray vs VRay/Renderman/Lux, OSS CAD tools, LibreOffice vs Office, etc. They aren't in the same league most of the time in both quality and functionality supported.

Re:Theo de Raadt on LLVM license change by cyphar on Saturday October 01, 2016 @01:39PM (#12618949)
The point of the free software movement is not quality, it's user freedom. Would you prefer having an electronic heart implant that is proprietary and is "totally safe trust me", or is free software but isn't the best thing in the world to interface with. I know which I'd pick, and the same logic applies to everything else I rely on in my daily life.
Re:Theo de Raadt on LLVM license change by rubber_duck on Saturday October 01, 2016 @02:17PM (#12619148)
The parent comment was making a statement about quality to which I responded.

And if the proprietary software has been an industry standard and deployed in millions of cases before I would have no problem taking the closed source variant. It would be nice to have the source but in most cases I wouldn't know what to do with it anyway as it requires specialist knowledge, so I'd be relying on the community vetting it vs. legal requirements for approval by the regulatory agencies, I mean it would still be better to have the source but I could live without it.

Anyway my point isn't even related to preference, I'm just saying that the conditions for successful OSS software are limited to one of the models under which it makes sense commercially for the companies using it (most often IT giants), some sort of commercial support/free software deal or in rare cases it's a labor of love for geeks. Most of the software written out there doesn't fall nicely in to those categories and so OSS is not viable in those areas.

Re:Theo de Raadt on LLVM license change by NzNz on Saturday October 01, 2016 @01:13PM (#12618808)
Your point is valid, but also incredibly narrow, the type of tunnel vision that seems to affect all computer using professional artists, who somehow started thinking they were the majority of computer users. Not only is that not the case, but you're such a tiny minority that Apple didn't shy away from dropping you all, dropping software like Aperture, simplifying Final Cut and orienting it toward hobbyists and slashing its price down to a measly $300, and all those moves are coming from the company that used to be known as the friendliest toward artists. Your kind keeps harping about "ah, my firewire port" and so and so and Apple literally doesn't care. You're not the target audience anymore, you represent nothing. The quality of open source software is, by far, when you consider most needs, more than adequate and often superior to closed source.

Most of the examples you give aren't even close to being evidence of a superiority of closed source in a debate of closed vs open, they are examples of the superiority of a few specific graphic powerhouses like Adobe. Is there anything like Photoshop in the world of closed source software? It is the only one. Closed source doesn't magically make things better. There are plenty of closed source competitors to Photoshop that aren't even close to what GIMP achieves, even more so with that plugin : http://gmic.eu/ Which replicates almost all the most important functions of modern Photoshop. What GIMP misses, is not power, but productivity, by missing non destructive editing. On the other hand, with something like gmic, gimp is far, far more powerful than ANY other piece of software that isn't called Photoshop. Considering the extremely small amount of people who are actively working on GIMP and its plugins and that none of their are paid for it, I'd say they do an incredible job. Everything you can do in Photoshop, you absolutely can do with GIMP, for eg things like these : http://blog.patdavid.net/2014/02/getting-around-in-gimp-gmic... Work just as good. If you're not a professional working on deadline who truly needs a productivity boost through constant modification of something you're commissioned to produce (which is where Non Destructive Editing shines best), you don't really need anything other than the GIMP.

So I'll quote your first paragraph and throw it back at you:

>>This is only true in the very narrow segment of 'tools for geeks by geeks', 'tools that big software companies need but don't provide a competitive edge/need to be OSS for adoption so they opensource'

Your views are only true in the very narrow segment of 'tools for artists that cost big bucks which represent exactly 0.01% of the software used in the entire world'. If most people don't have a use for the things you mention, like code editors and compilers, they don't have a use for things like Photoshop either. And if we're going by % of software in use, there are far more web servers out there running open source programs that there will ever be photoshop installations. Much less packages like Maya which fill an even much smaller niche. Niche, a word you seem fond of using for talking about OSS. What is more niche between a web browser and maya?

Re:Theo de Raadt on LLVM license change by rubber_duck on Saturday October 01, 2016 @01:28PM (#12618885)
Browsers fall in to the 'things we need to ship but the source being closed isn't a competitive advantage' of huge software corps (Mozilla being an outlier).

I listed artist tools because that's the thing I have first-hand experience with and I'm sure others here do as well but it's faaar from being limited to that.

List superior OSS alternatives to things like SAP/Dynamics/Salesforce or CAD like SolidWorks/AutoCAD. Where are all the OSS banking systems ? Every industry I've seen has their own specialized tools that cost big $ and has stuff that took years to purpose build, the sort of thing that just isn't viable under OSS model. Industrial control applications, and let's not even go to heavily regulated stuff used in medical devices and such.

I'm not the one who's tunneling here, there's more software out there than dev tools, browsers and server OSes, very little of it is OSS, even less can compete on quality.

Re:Theo de Raadt on LLVM license change by icebraining on Saturday October 01, 2016 @02:33PM (#12619225)
List superior OSS alternatives to things like SAP/Dynamics/Salesforce

We've been making good money moving companies from SAP to Odoo.

Re:Theo de Raadt on LLVM license change by cyphar on Saturday October 01, 2016 @01:37PM (#12618934)
I'll try to answer your baiting questions.

> Also, if it is completely OK for hardware design to be proprietary, why should it not be OK for software?

Personally I also have an issue with hardware designs being proprietary, but because we still have issues with software I prefer to spend my energy on the issue we've had better traction on.

> What is so special about it?

The big difference between hardware and software is that software can be in principle changed, modified and updated by the end user without too much hassle. Hardware cannot be practically modified (have you ever tried to modify a chip?). And hardware designs are useless to most users, because they'd need to have someone fab the components for them. However, every user has someone they know who knows how to program, and thus they can practically take advantage of the freedoms that they ought to have a right to.

> If you want to use Debian, for instance, you are pretty much forced to use "non-free" repositories (or a "non-free" Debian re-distribution) for your setup to be useful.

This is simply untrue. I use openSUSE Tumbleweed and have no issues with the proprietary repositories removed. I cannot comment on Debian, but I know a few people who use Debian without the universe repositories enabled. What precisely is "forcing" you to use non-free software? Maybe you should reconsider who you're blaming for that, and maybe try removing that proprietary garbage from your life.

Re:Theo de Raadt on LLVM license change by Koshkin on Saturday October 01, 2016 @02:04PM (#12619081)
But the distinction between the hardware design and software is blurry. What about microcode? What about designs targeting FPGAs? A basic idea in modern systems design, in fact, is that there is no preconceived notion about where hardware should end and software should begin.
Re:Theo de Raadt on LLVM license change by icebraining on Saturday October 01, 2016 @02:43PM (#12619274)
The FSF draws a line that I find interesting (if not necessarily my own): whether it can be updated with a new version. If not, then it's equivalent to a circuit, and it's not as important that it be Free (though it must be non-malicious).
Re:Theo de Raadt on LLVM license change by gwright on Saturday October 01, 2016 @11:57AM (#12618387)
What principle of capitalism is being violated? If anything, the freedom to set a price, even $0 if you so choose, sounds like the free market at work.
Re:Theo de Raadt on LLVM license change by Koshkin on Saturday October 01, 2016 @01:13PM (#12618802)
The issue has never been about price.
Re:Theo de Raadt on LLVM license change by gwright on Saturday October 01, 2016 @01:23PM (#12618856)
OK, then what principle of capitalism are you referring to?
Re:Theo de Raadt on LLVM license change by Koshkin on Saturday October 01, 2016 @01:39PM (#12618942)
Property rights. The "free software" principle proclaims that software (possibly, among other things) should be "freed" from the possibility of being anyone's private property. (This also somehow gets mixed with this odd idea that software should be considered a form of "free speech".)
Re:Theo de Raadt on LLVM license change by cyphar on Saturday October 01, 2016 @01:48PM (#12618992)
Free software still maintains copyright, if that's what your referring to. However, you're mixing (as is common these days) property rights and other rights (such as copyright). Specifically, you probably intended to use the confusing term "intellectual property".

Software is not a physical thing, so it's not clear who is the "owner" of the work. If you buy a book, you are the owner of the book but are not the copyright holder. In a similar fashion, if you buy (or download) some software, you are the owner of the software but are not the copyright holder. So free software does not proclaim that ownership is evil, it just states that all users should have rights that people have had throughout history when it comes to similar intstructional works such as recipies.

Also the idea that software is a form of "free speech" does not come from the free software movement. I think you're getting confused, and would recommend reading the GNU website[1] to better understand precisely what free software refers to.

[1]: https://www.gnu.org/philosophy/free-sw.en.html

Re:Theo de Raadt on LLVM license change by Koshkin on Saturday October 01, 2016 @03:15PM (#12619425)
> Free software still maintains copyright

Otherwise known as "copyleft", which means that everyone has the right to copy; it, moreover, demands that this right to copy may not be infringed upon by alteration of the software. There is no trace of the originally intended meaning of the word "copyright" in there; if anything, it's just its opposite.

Re:Theo de Raadt on LLVM license change by cyphar on Saturday October 01, 2016 @09:47PM (#12620730)
Not all free software licences are copyleft licenses. While the GNU GPL is a very well known and widely used example, it is by far not the only free software licence. The Apache 2.0, MIT, BSD and other commonly used licenses are free software licenses because they provide the recipient the four essential freedoms. Copyleft is not mentioned in the free software definition[1]. It's mentioned as an optional way to license free software works.

In addition, even with copyleft every contributor has copyright over their changes. The crucial part of the copyleft hack is that it requires that distributing derived software must be done under the same license. You may choose to not distribute the software, and then the GNU GPL doesn't require you to do anything.

[1]: https://www.gnu.org/philosophy/free-sw.en.html

Re:Theo de Raadt on LLVM license change by gwright on Saturday October 01, 2016 @03:07PM (#12619391)
Private property in the traditional sense is certainly a critical component of capitalism, but "intellectual property" rights are something quite different.

I don't think there is a conflict between support of capitalism and criticism of intellectual "property" rights and the associated legal constructs.

Re:Theo de Raadt on LLVM license change by icebraining on Saturday October 01, 2016 @02:50PM (#12619317)
Property is a system for allocating rivalrous[1] resources. Applying property to freely copyable elements makes no sense. Supporting capitalism in no way requires defending the creation of government-granted monopolies under the cover of property rights.

[1] https://en.wikipedia.org/wiki/Rivalry_(economics)

Re:Theo de Raadt on LLVM license change by dbolgheroni on Saturday October 01, 2016 @02:30PM (#12619216)
I remember seeing arguments like this in the 90's.