News for hackers. Stuff that matters.
Re:Richard Stallman: How I do my Computing by Aethaeryn on Saturday April 07, 2012 @03:27PM (#3811866)
Personally, rms doesn't bother me and he's free to do as he pleases, but what annoys me is the people in rms related threads online who act as if the development of FOSS[0] is entirely dependent upon him. There were a lot of parallel developments in software, which is not a zero sum game.

I find a lot of rms's essays interesting and I regularly use a lot of GNU software, especially emacs. Still, rms is often divisive and counterproductive. We must call it "GNU/Linux"? We must use one of a handful of "totally free distros" instead of even distros like Fedora and Debian? Stallman played a vital early role, but no role was necessary.

If GNU never existed, we wouldn't have emacs/gcc/etc., but there are alternatives and those alternatives would have become more developed over time.[1] Linux would have used the BSD programs instead of the GNU programs or they would have written their own.

(In fact, if Linus never invented Linux, there would still be FreeBSD, NetBSD, and OpenBSD. These are all, as far as I know, less dependent on GNU applications (if at all) and many people in the BSD communities see the GPL as restrictive.[2])

If the FSF never was founded, we wouldn't have the GPL and we might not have gotten copyleft at all, but we would still have BSD, MIT, Apache, etc., licenses. "Free software" is an ideology that is largely dependent on Stallman, but "open source" software is inevitable because of the benefits it provides society.

-----

While one particular philosophy of freedom and computing were largely driven by rms's writings, the practical benefits of sharing source under a free license are not at all dependent on some broad philosophical goal. It's just more profitable for businesses to not have to pay for software (while at the same time not being "pirates"), to have access to the source code (for patches, even if the upstream refuses them), and so on. I find it hard to imagine a company with lots of servers like Google existing if they had to pay Windows or UNIX license fees (especially if the competitive pressure from Linux and BSD never existed). Software like Apache's httpd would have existed even if rms never wrote anything, in my opinion.

Companies don't care about the freedoms that rms cares about. They care about making money. It just so happens that you make more money in the long run[3] if you don't get forced into overly restrictive terms by a vendor who then locks you into their proprietary technology, continually increasing fees and restrictions over time. Open source software gets major contributions by people who are paid to do this by big corporations who would rather do this than pay for closed source software. It just so happens that the terms allow end users, startups, and so on to benefit from the big businesses who are ultimately acting in their own best interests. If rms had his way, this virtuous circle might not have come about because rms is very much opposed to any sort of compromise at all.[4]

Stallman is the opposite of this sort of pragmatism that is common in the "open source" world. To rms, one ought to act ethically, and anyone who does not do so is essentially evil. At one point, he said that you didn't have to make software for money, you could just be a waiter. (I can't find the source, but it stuck with me.) The implicit assumption behind this statement is that free software is inherently not profitable in the world where Internet distribution makes charging for distribution of free software impractical. In rms's mind, something like GNU would exist even if one could not make money off of it, had to get a day job, and development had to be sponsored by big not-for-profit organizations like the FSF.

It's noble to be this committed to a philosophy, but fortunately the world does not exist as the "worst case scenario" that rms prepared for. The rest of the FOSS world has largely moved beyond this mentality. Companies like IBM and Red Hat sponsor development for, and make money off of, "Linux and open source". They use those terms instead of "GNU/Linux and free software" because the latter are ideologically loaded, and hence have the potential to hurt their businesses by scaring away potential customers.[5] The people who wanted FOSS in the 1990s got creative and found ways[6] to get massive amounts of cash to sponsor their development so that they could get paid to do what they love. Apparently, you can make software without charging directly for copies.[7]

So I can see other reasons why people active in FOSS would want to not be associated with rms: his complete hostility toward "open source", major Linux distributions, the common use of the term "Linux" instead of "GNU/Linux", the complete lack of the ability to compromise, focus on philosophy and politics over code, etc. If people associate all of the open source movement like rms[8], it could scare away some who would otherwise contribute and help make the world a better place. The useful contributions rms made was his source code (such as GNU emacs) and he did make many useful contributions in the past, but if his ideologies scare off someone from making major source code contributions, then now he's being counterproductive. It doesn't matter how notable his name/initials are if he's making a very public mistake. In the end, all that matters in the FOSS world is the source code.

-----

[0] FOSS means "Free and Open Source Software" and is a compromise term between "free software" fanatics and those who prefer the term "open source". All three terms are largely equivalent in meaning, but have different ideologies attached.

[1] I think one of the appeals of FOSS is that there is no one leader, no one ideology, no one organization or company, and so on. There is not even just one program for any given niche, there tends to be at least two. Often the alternatives to a given program are inferior in some way, but if the major program they are alternative to never existed, they probably would have been made more capable or forked at some point. (Or another program would have been written from scratch to fill the niche.)

[2] Even the so-called "permissive" licenses are restrictive when compared to the WTFPL. :-P

[3] It's not surprising that contemporary big businesses care about profit. It is surprising that some pay attention to their long run interests! :-P

[4] If you don't believe me (and this very article that's being linked to implies the same thing), see the GPLv3 controversy and also note that the Linux kernel is still under the GPLv2.

[5] If you're making money, the term "free software" is misleading. Also, the term is not unambiguous in English: Google it and you'll find a lot of freeware. It would also be confusing for a company like Red Hat to "charge for free software". "Freedom Software" probably would have made more sense than "free software".

[6] One of the compromises they made was dropping "free software" for "open source" and rms never forgave them.

[7] The ease of distribution and copying that broadband Internet provides should encourage other industries to find ways to make money without charging for copies, as well. Music, movies, books, and so on, probably have ways to make money without the traditional "charging for copies" model. Also note that if you do not charge for individual copies of your software, making all or most of your source code available under an open source license is probably a great cost-reducing technique that just about everyone should use. Sites like reddit have source code available on Github, and they benefit from it because the value for reddit is the domain "reddit.com" and the posts located there, not the software that runs it.

[8] Very few people would want to live in as pure of a free software environment as rms does. I personally don't know how he can get any work done on a tiny netbook. I use two 1080p 24 inch monitors.

Re:Richard Stallman: How I do my Computing by zach95 on Saturday April 07, 2012 @04:43PM (#3812076)
This general argument pioneered by ESR against RMS is something that you hear a lot: the idea that open source software is just better. (And it is that too!)

#1: it's too hard to say whether the free software movement would have evolved without RMS. We do know that Linux was able to bootstrap from GNU which was founded some 7 years earlier and without that, it probably wouldn't have taken off the way it did. By the time Linus came around, the only thing missing from GNU was its kernel. The Berkeley people got bogged down with the BSDi lawsuit, and you might ask yourself how that happened if their license was so freedom preserving. If it wasn't for that, I imagine that BSD would be in the place linux is now.

#2: People say it's divisive not to have one way of thinking about open source, but why this need to do it one way only? Surely there is a place for the GPL for some projects, and RMS isn't forcing anyone to GPL their software. The FSF counts lots of other licences as GNU-compatible.

RMS just wants people to understand that if they license their software restrictively they're not doing the hacker community any great service, even if they claim otherwise.

#3: "To rms, one ought to act ethically, and anyone who does not do so is essentially evil." That's not just RMS who thinks that. You're going to have a hard time justifying why unethical people who harm others are essentially good. You make it seem like it's radical to stand on the side of ethical behavior. Maybe it is.

Re:Richard Stallman: How I do my Computing by Aethaeryn on Saturday April 07, 2012 @05:27PM (#3812176)
> it's too hard to say whether the free software movement would have evolved without RMS.

Hypothetical speculation of how history would have turned out if certain details were changed a little bit are hard to argue because they're hypothetical and thus a realm of opinion. There does not exist an actual alternate universe in which a given event never happened that we can just observe to see how things worked out differently.

I think that where we're heading in software is inevitable. Yes, RMS contributed to it early on and yes, things might have taken off more slowly without the FSF and GNU, but in 100 years, I think we'll have won no matter what. This is an opinion. There's no empirical way to evaluate it. I think the nature of easy copying via the Internet has created a sort of inevitable force, like gravity, and that fighting it is ultimately futile. Am I right? We'll see how things are in 2112.

-----

> Surely there is a place for the GPL for some projects, and RMS isn't forcing anyone to GPL their software.

If you think that proprietary software is evil instead of just bad, then yes, the GPL is good for some things because it is "freedom-preserving". I could easily imagine an alternate reality in which Mac OS X used a BSD-licensed Linux kernel.

It could also be argued, for the sake of argument, that the GPL is ultimately a bad thing. It requires copyleft, sharing with the same license, and about half of FOSS has been "infected".[1] Very few projects do the copyright assignment that the FSF does. If there is a currently unseen issue with the GPL, it would be very hard to get every single contributor who owns copyright to switch every major project from the GPL. If you say "GPL v2 or later" or "GPL v3 or later" in your software's license, you are dependent on one organization for your "upgrade path" of licenses. An advantage that the permissive licenses have is that there isn't one major permissive license: there are lots of them. It's a more diverse ecosystem, without one point of possible future failure and without any dependency on one organization.

> RMS just wants people to understand that if they license their software restrictively they're not doing the hacker community any great service, even if they claim otherwise.

Those companies are not doing themselves a great service, either. The big proprietary software companies will realize this eventually, or like the big media companies who are so resistant to change, they'll go out of business in the long run. Reinventing the wheel is expensive, time consuming, and there's no guarantee you'll get the best programmers on board. You'll also get stuck into other vendors' lock-ins unless you're a very big company like Microsoft, so you'll lose more than you'll gain by going with closed source, non-free software.

-----

> You make it seem like it's radical to stand on the side of ethical behavior. Maybe it is.

It's not radical to focus on doing ethical behavior (or maybe it is, but it's not the radical thing I'm talking about). I do admire him for his courageous ability to take an ethical stand against all odds. I just don't think it's the most productive way to achieve to his goals.

If you want all software to be free, then you need to convince companies making software to make it free. If you can't do that, then you need to make your own company and put those older companies out of business. There's only so many things that an organization or individual not making any money can do. In the world as it is, money matters. Maybe that's not how things should work, and maybe all software should be made by the FSF, but very radical structural changes are probably never going to happen.

People went with "open source" not because they hate freedom or rms, but because they needed to "sell" the concept to companies like IBM.

-----

[1] http://osrc.blackducksoftware.com/data/licenses/

Re:Richard Stallman: How I do my Computing by zach95 on Saturday April 07, 2012 @06:28PM (#3812346)
Instead of responding to things that might be factually inaccurate, let me just save time and reframe the debate between ESR and RMS.

ESR thinks that open source makes good business sense since open source is better than closed source. Computing is something that extends our ability to get things done, but isn't an end in itself. We need to be pragmatic about what businesses want, and how the business world works in order to perform our role of helping people do business.

RMS doesn't care so much about what the business world wants or the fact that open-source is indeed better.

He sees hacking and computing as one of the greatest achievments of mankind. Because it's a great achievement of mankind (like writing, or alphabets, or language itself) he thinks that it should be made available to everyone, and never closed off. In other words, for the people who live in it, spending their lives working on it and in it, software is culture. It's not just an application that you use like a word processor. And in the hacker culture, there needs to be a free flow of ideas like there is in the scientific community.

This is reflected in the different licenses. The GPL "restricts" people from trying to close off one of the greatest achievements of human kind for the sake of some people who want to earn some more money off of it. The BSD-style licenses consider themselves more practical in that they allow for better commercial exploitation by eventually imposing restrictions. This is why Apple is so heavily involved in converting FreeBSD away from GPL code Currently I think it's about 50% GPL and OpenBSD is more like 60%. They plan to replace the GPL code with BSD licensed code so they don't have to contribute to the project eventually.

Clearly some code isn't the equivalent of Shakespearean drama, and it won't matter that humanity is deprived of it. Microsoft word comes to mind. But for the really amazing software, the software that involves the cutting edge free flowing of ideas, and I would argue that the linux kernel falls into this category, GPL gives it the ability to serve as a lasting monument to human ingenuity, as well as a way of transmitting culture. It's no coincidence that the linux kernel powers nearly all web servers, smartphones and supercomputers.

Re:Richard Stallman: How I do my Computing by zach95 on Saturday April 07, 2012 @04:13PM (#3811981)
"I personally don't know how he can get any work done on a tiny netbook. I use two 1080p 24 inch monitors."

And unix was originally written on a machine with less than 144 kilobytes of RAM using a teletypewriter...

What amazing contributions to the world's software have these enormous screens allowed you to make?

Re:Richard Stallman: How I do my Computing by Aethaeryn on Saturday April 07, 2012 @04:49PM (#3812088)
>And unix was originally written on a machine with less than 144 kilobytes of RAM using a teletypewriter...

> What amazing contributions to the world's software have these enormous screens allowed you to make?

A tool is a tool, and is different than a person. When it comes to programming, the person is more important than the tool, but the tool does affect the person. You can theoretically do roughly the same work on any computer with any equivalent application, but certain computers and editors allow you to do such work more quickly.

Just because you can do something using the "wrong tools" doesn't mean that you should. Most of the original Unix was written using ed, too.[1] That doesn't mean that ed ought to be used today unless it is the tool that happens to make you the most productive. A real hacker using ed can program better than a newbie with a modern editor, obviously. That doesn't mean that real hackers should use ed.

ed was used out of necessity. So were teletypes and 144 KB of RAM. Good hackers could still do work with such tools. However, they themselves think that they're more productive with modern tools or else the hackers who used ed back in the day would still be using ed.

I actually do use ed for some quick/light editing because it doesn't get in the way. It's the perfect editor if you never make any mistakes. When you do make mistakes, something like emacs is immensely more useful because of all the ways you can correct your program from within the editor. If I used ed for everything, I would probably have to write dozens of helper scripts (in sh, of course!) to help me edit my source code. In something like emacs, those scripts have already been written and integrated into the editor and this is a good thing. After all, one should not reinvent the wheel unnecessarily.

-----

I hope you understand what I was getting at in my analogy to ed. You can make amazing contributions using any weak tool, and you can make no contribution at all using a strong tool. This doesn't mean that one should willingly use weak tools unless they have a pretty good reason to do so. People did great things on weak tools in the old days because they had no other choice, not because those tools somehow make someone great. Using ed isn't going to allow me to create the next great operating system.

You can have a powerful computer, or lots of screen space and have those tools be wasted, obviously. The specs of your hardware do not make you more capable of a programmer. On the other hand, to willingly suffer weak tools for ideology shows that rms puts his ideology ahead of his hacking. Otherwise, he would have a setup that is designed to optimize hacking productivity, so that he could do more things in less time. Sure, he could hack great things using just about any device with a Unix terminal and a keyboard, but he could hack in a more productive manner using proper tools.

Using roughly equivalent tools is a matter of taste (e.g. vi and emacs), but he's using a rather weak machine simply out of a desire to have a totally 'free' machine. Ironically, he's limiting his own freedom of choice in hardware for what he believes is total software freedom.

rms is a great hacker. In the domain of programming, he's better, smarter, and more experienced than me. That being said, his uncompromising commitment to his ideology is not getting in the way of me, it's getting in the way of him. There are other groups that inconvenience themselves greatly due to their philosophical beliefs. For instance, the Amish. They don't harm me or even affect me, but they do miss out on a lot of cool stuff by shunning technology. I probably can't persuade someone who is already Amish to go get an awesome smartphone, but if the Amish were actively evangelizing, I would try to discourage other people from giving up the conveniences of modern technology.

Perhaps I should ask what amazing contributions to the world's software rms has done in the last decade as opposed to, say, the 1980s?

-----

[1] http://catb.org/~esr/writings/taoup/html/ch13s02.html

Re:Richard Stallman: How I do my Computing by zach95 on Saturday April 07, 2012 @05:05PM (#3812131)
I have to grant you a few points. #1: I've heard RMS complain about his lemote netbook, about it being slow and with a small screen, although I supposed he could attach an external monitor. As far as the lemote netbook, I wouldn't buy one because lemote is an arm of the Chinese government and the loongson processor already has design flaws, and is licensed from MIPS. It beats me why he doesn't use an OpenRISC platform if he wants something just for cli editing with emacs.

#2: RMS seems to have stopped hacking software, so his computing requirements seem to be limited to what's necessary to produce his writing. He also seems to be getting support and service from other people's computers/servers and other people. I'm not sure that any average person would want or be able to do things exactly like him. Since he is trying to be a role model, he should model something that everyone could do and would want to do.

#3: Instinctively I want to say you're probably right that hackers should use the best tools available, and most of them do. However, I've known too many Russian/Finnish hackers who grew up with computers like clones of the sinclair zx80 who would surely not have been as awesome had they started off on supercomputers. Something about the easy availability of resources and lack of simplicity seems to spoil some of the fun. This is why I think so many hackers are drawn to simpler devices like ardunio or embedded microcontrollers.

Re:Richard Stallman: How I do my Computing by Aethaeryn on Saturday April 07, 2012 @06:00PM (#3812263)
> I supposed he could attach an external monitor.

I'm not sure it's powerful enough to handle one 1080p external screen. It's certainly not powerful enough to handle two. (My desktop can't handle three with its current graphics card and driver.)

I also think that you need X for multi-monitor setups. When I Ctrl+Alt+F2 (C-M-F2 if you are used to emacs) for a console, it just mirrors the screens. Even if you could get it to not mirror, I'm not sure what meaningful advantage you would get over just tiling lots of terminals in a very basic window manager.

-----

> RMS seems to have stopped hacking software, so his computing requirements seem to be limited to what's necessary to produce his writing. He also seems to be getting support and service from other people's computers/servers and other people. I'm not sure that any average person would want or be able to do things exactly like him.

Yes. Maybe I'm subconsciously being overly critical because he chose to stop hacking and start writing philosophy. Different people obviously have different needs, and those needs are best served by different devices.

That being said, he seems to use his netbook because he has no other choice, not because he wants to. Perhaps the Free Software Foundation (FSF) should spin off a Free Hardware Foundation (FHF) if he wants devices that can run GNU/Linux (as he puts it) without needing any sort of proprietary firmware at all. It seems unhackerish (if that's a word) to just accept circumstances rather than to try to build your way out of them. I'd gladly use a FHF motherboard/CPU for the next box I build if they were decent -- and if they existed!

-----

> However, I've known too many Russian/Finnish hackers who grew up with computers like clones of the sinclair zx80 who would surely not have been as awesome had they started off on supercomputers. Something about the easy availability of resources and lack of simplicity seems to spoil some of the fun.

Yes. I hate it when classes "teach" programming with IDEs that think (and code) for you. I took a class that required the use of Eclipse, so I went ahead and coded in emacs anyway. It's very vital to build up things like muscle memory (from basic editors that don't get in your way) and resourcefulness (from poor hardware) if you want to get good at programming. It's easier to learn a good concept out of necessity than just because you've been told that it's good.

That being said, once you've learned how to do something hard, if you don't want to do it again you shouldn't have to. The best environments for learning might not be the same as the best environments for use once you know what you're doing. (Although I would, in my biased opinion, think that something like emacs comes close. It doesn't get in your way or try to think for you like a "modern" IDE, and yet it has lots of depth if you know the right commands or download the right extensions. It seems to encourage learning to obtain more functionality, rather than just having menus upon menus. I'm not opposed in principle to IDEs, but none of them seem to have sane, minimalist defaults. It's a sad state of modern software where an example of bloat in the 1980s is now an example of minimalism today.)

-----

> This is why I think so many hackers are drawn to simpler devices like ardunio or embedded microcontrollers.

It's also just fun to build things. I'm looking forward to the Raspberry Pi. I don't know what I'm going to do with it when I get it, but I'm sure I'll find some use for hardware that inexpensive.

-----

P.S. I hope I'm not giving the impression that I hate rms. I wouldn't be so picky here if I thought that his end goal wasn't worth achieving.

Re:Richard Stallman: How I do my Computing by zach95 on Saturday April 07, 2012 @06:47PM (#3812380)
>>It seems unhackerish (if that's a word) to just accept circumstances rather than to try to build your way out of them. I'd gladly use a FHF motherboard/CPU for the next box I build if they were decent -- and if they existed!

As I mentioned above, they do exist: http://orsoc.se/127/langswitch_lang/en/

Even the bus architecture is open and the processor itself can be debugged or modified. As other people have mentioned, there is coreboot (coreboot.org) for a number of other platforms.