30 years of GNU [LWN.net]

30 years of GNU

Posted Sep 27, 2013 12:47 UTC (Fri) by hummassa (subscriber, #307) [Link]

Filename completion isn't hard per se, but "completion" is... have you read the files under /etc/bash_completion.d ??

30 years of GNU

Posted Sep 28, 2013 23:37 UTC (Sat) by marcH (subscriber, #57642) [Link]

bash_completion.d is always one of first thing I disable, reverting to the good old, pure "directory" completion.

This idea that the shell can predict every single verb+object combination I type is completely ridiculous and infuriating because always missing corner cases.

Also note nothing too complex ever works really reliably.

30 years of GNU

Posted Sep 29, 2013 0:00 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]

> Also note nothing too complex ever works really reliably.

Indeed, bash "completion" plugins are terrible and have caused me a lot of techsupport woe. It fails just often enough to make life miserable... e.g. people thinking that a file doesn't exist because it just won't complete.

Ideally computers should do the right thing. But when they can't do the right thing they should do the explicable thing. In our quest to make computers right more often we've added complexity that increases inexplicable exponentially when they don't get it right.

It's not always a good tradeoff.

30 years of GNU

Posted Sep 29, 2013 17:24 UTC (Sun) by jzbiciak (✭ supporter ✭, #5246) [Link]

e.g. people thinking that a file doesn't exist because it just won't complete.

That one drives me nuts, personally. For example, suppose I download a compressed tarball, not noticing if it's foo.tar.gz, foo.tar.bz2, or foo.tar.xz. If I type tar zxvf foo[TAB], the over-fancy completion rules won't complete the filename unless it's actually a tar.gz. I'd rather it expand the filename, so I can then go back and edit the flag I gave to tar based on what I see in the completion, but the completion scripts take the flags I've given tar already a bit too seriously.

30 years of GNU

Posted Sep 29, 2013 17:52 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]

The use of specific flags based on the format is entirely redundant for a long time and I am not sure why you bother to do that anymore. Tar does do the right thing automatically here.

30 years of GNU

Posted Sep 29, 2013 18:20 UTC (Sun) by jzbiciak (✭ supporter ✭, #5246) [Link]

I missed the memo on that. It appears a my webserver is old enough to not support autodetect, though. Likewise for the 32-bit Linux boxes I still have to build software on from time to time at work. (I just checked.)

So, still too soon to delete that from my muscle memory.

30 years of GNU

Posted Sep 29, 2013 18:27 UTC (Sun) by jzbiciak (✭ supporter ✭, #5246) [Link]

In case you're wondering, the machines at work are running RHEL WS4 (with a 2.6.9 kernel, even!), and my webserver is older still, running RH7 and a 2.4.x kernel. :-)

30 years of GNU

Posted Sep 29, 2013 18:32 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]

Sweet. But that doesn't really remove the fact that bash completion makes the behavior confusing.

Similar things happen for mplayer where mplayer identifies files by magic but bash completion has a predefined list of possible media files which will forever be incomplete (since magic based detection is how the tool actually works).

30 years of GNU

Posted Sep 29, 2013 19:51 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]

Individually programs that can extend or disable parts of bash completion as needed.

30 years of GNU

Posted Sep 29, 2013 20:07 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]

Sure. There is an infinite number of ways you can improve the completions.

But the fundamental complaint I was making there is that it increases complexity and makes it more inexplicable when it does fail.

Programs extending or disabling parts of it increase the complexity further and potentially just exacerbate the problem, because it makes the behavior even harder to mentally model.

30 years of GNU

Posted Sep 29, 2013 20:12 UTC (Sun) by rahulsundaram (subscriber, #21946) [Link]

Providing more functionality always increases complexity. I think it is manageable in this case and it is optional anyway.

30 years of GNU

Posted Sep 29, 2013 23:08 UTC (Sun) by marcH (subscriber, #57642) [Link]

It does not matter: life is too short.

30 years of GNU

Posted Sep 29, 2013 23:07 UTC (Sun) by marcH (subscriber, #57642) [Link]

> If I type tar zxvf foo[TAB], the over-fancy completion rules won't complete the filename unless it's actually a tar.gz.

"over-fancy"? You are too kind. Whoever spend their time implementing this type of over-crazy stuff should go and find themselves something useful to do.

The only (but significant) superiority of a command line interface over a graphical one is the lack of boundaries: I can enter the most surprising and unusual command and the computer will just do it - no questions asked. bash_completion.d breaks this. bash_completion.d brings the "Cannot find a program to open your file, wanna search the Internet?" type of frustration from the graphical interface world into the command line. A very impressive feat.

The even crazier thing is: how come many distributions enable this by default?

30 years of GNU

Posted Sep 30, 2013 9:21 UTC (Mon) by gowen (guest, #23914) [Link]

M-/ is your friend.

You can, of course rebind TAB to complete-filename and skip all the auto-completion. Use "help bind".

It's unfortunate you couldn't find it in the documentation, but its more unfortunate that your response to this was ranting about it on the internet, as if your opinion were the only sane/valid one.

30 years of GNU

Posted Sep 30, 2013 11:36 UTC (Mon) by jzbiciak (✭ supporter ✭, #5246) [Link]

[I]ts more unfortunate that your response to this was ranting about it on the internet, as if your opinion were the only sane/valid one.

Yeah, well, that's just, like, your opinion, man.

;-)

30 years of GNU

Posted Sep 30, 2013 12:19 UTC (Mon) by gowen (guest, #23914) [Link]

*chuckles*

Do you think there's a market for t-shirts showing Richard Stallman's face with the phrase "The Dude Abides" underneath?

30 years of GNU

Posted Sep 30, 2013 17:41 UTC (Mon) by Kluge (guest, #2881) [Link]

Absolutely.

30 years of GNU

Posted Oct 1, 2013 16:50 UTC (Tue) by egoforth (subscriber, #2351) [Link]

"The GNUde Abides"?

30 years of GNU

Posted Sep 30, 2013 17:58 UTC (Mon) by marcH (subscriber, #57642) [Link]

> It's unfortunate you couldn't find it in the documentation,

New, "smart" feature; so smart that now you must read documentation before using [TAB]...

You must be having a laugh, right?

30 years of GNU

Posted Sep 30, 2013 15:32 UTC (Mon) by rgmoore (subscriber, #75) [Link]

I can enter the most surprising and unusual command and the computer will just do it - no questions asked.

Of course that's still true when you use command line completion. All that completion will do is make it easier to enter simple, predictable commands by hitting the TAB button. You can still enter the most surprising and unusual command and the shell will still execute it. You just have to type the whole thing out yourself.

30 years of GNU

Posted Oct 3, 2013 5:01 UTC (Thu) by k8to (subscriber, #15413) [Link]

I really like the bash completion fu, and I think it is probably better (more convenient) for almost everyone. But it does have awkward corners, and makes using the shell more complicated, so you might be right that it should be off by default.

I certainly didn't mind manually enabling it on Debian.

30 years of GNU

Posted Oct 3, 2013 5:16 UTC (Thu) by gmaxwell (subscriber, #30048) [Link]

It probably would have been really awesome if it were just on another key instead of replacing the nice deterministic one. I've personally never found it useful, but I could imagine that someone might. My frustration was that I found that it greatly degraded something that was useful to everyone, and resulted in some unwelcome techsupport burden when it hit other people that haven't turned it off yet.

The responses here about using alt-/ to get the explicable behavior back are great. I'm not sure how I ever would have discovered that, since that difference in behavior is not pointed out in the bash man-page.

30 years of GNU

Posted Oct 3, 2013 6:45 UTC (Thu) by k8to (subscriber, #15413) [Link]

I don't remember how I learned it. I think it might have been in the README.Debian files for the bash completion package that I had to manually install.

bash's command line behavior is a complicated interplay between libreadline and bash and your configfiles. It's pretty hard to learn about in general.

30 years of GNU

Posted Oct 3, 2013 11:15 UTC (Thu) by khim (subscriber, #9252) [Link]

The responses here about using alt-/ to get the explicable behavior back are great. I'm not sure how I ever would have discovered that, since that difference in behavior is not pointed out in the bash man-page.

Really? That's strange: it's explained here:
$ man bash
…
complete-filename (M-/) Attempt filename completion on the text before point. …

One of the best things WRT GNU project is it's documentation: it's usually pretty well-written. At least user-facing one. Bash is not an exception.

P.S. There are many different way to complete your text in bash: as a filename, as a variable, as a username and so on. And of course TAB always meant “smart completion”. Only usually it was less “smart” than it is today.

30 years of GNU

Posted Oct 4, 2013 22:47 UTC (Fri) by marcH (subscriber, #57642) [Link]

> And of course TAB always meant “smart completion”. Only usually it was less “smart” than it is today.

Which means it Just Worked.

I did not realize this was called "smart completion". Now I completely understand the connection with Microsoft Word where the very first required is to immediately disable all the so-called "smart" features ("smart" quotes, "smart" initials, "smart" selection,...) in order to be allowed to type anything unusual that Microsoft did not expect.

Exactly the same kind of "smartness".

30 years of GNU

Posted Oct 3, 2013 6:14 UTC (Thu) by jzbiciak (✭ supporter ✭, #5246) [Link]

I'll be honest. When I first noticed the completion-fu, I was impressed. And most of the time, it does not bother me. But, when it gets in the way of something basic that I think should 'just work,' it incites my 'nerd rage.'

As an architect (both chip and software), I understand the desire to make the on-ramp to working with any particular platform shallower. I /also/ understand the frustration when the 'straightforward, dumb thing' gets usurped by the 'smart, but misapplied thing' (such as my tar example earlier).

I am officially on the fence, truth be told. I grew up with simpler behavior, and find less to complain about with it. The more complex completion.d behavior is definitely very clever, but is also more frustrating at times.

30 years of GNU

Posted Sep 30, 2013 0:02 UTC (Mon) by Richard_J_Neill (subscriber, #23093) [Link]

> That one drives me nuts, personally. For example, suppose I download a
> compressed tarball, not noticing if it's foo.tar.gz, foo.tar.bz2, or
> foo.tar.xz. If I type tar zxvf foo[TAB], the over-fancy completion rules
> won't complete the filename unless it's actually a tar.gz.

Use Alt+/ to force completion, if "smart" completion fails.

The other trick is to add to /etc/inputrc:
set show-all-if-ambiguous on

Two really useful cases for tab completion are with the package-manager:
apt-get install libfoo[TAB]
and with SCP (if you have ssh keys):
scp remote_host:filen..[TAB]

30 years of GNU

Posted Sep 30, 2013 11:44 UTC (Mon) by cortana (subscriber, #24596) [Link]

> Use Alt+/ to force completion, if "smart" completion fails.

Sorry for the low-content post but... THANK YOU SO MUCH FOR THIS!

completion

Posted Sep 30, 2013 12:53 UTC (Mon) by tialaramex (subscriber, #21167) [Link]

scp completion is rather dangerous, here's how:

1. Set a remote machine to disable SSH connections from hosts which pester with invalid credentials because maybe they're "script kiddies"

2. Change the credentials somewhere

3. Try to use SCP completion. Whack tab a few times wondering why it's not working

4. Congratulations you are now locked out of the remote machine with no clue as to what happened.

A systems administrator we hired locked himself out of his home systems this way, very amusing.

completion

Posted Oct 3, 2013 5:07 UTC (Thu) by k8to (subscriber, #15413) [Link]

I'm not a fan of log-into-remote-systems-on-tab-complete for sure. If nothing else, it introduces *crazy* latency into tab completion, which isn't really very pleasant.

As for the script-kiddy lockout, the sane thing to do is expire that kind of thing after some time interval, possibly progressive, so that once you realized you screwed up you just wait 30 minutes and can get back in.

If one is so dense as to keep spamming it over and over for over 30 minutes, well... maybe the lockout could work to teach one lessons in patience and care. :-D I know a younger me could probably have used them.

30 years of GNU

Posted Sep 29, 2013 10:38 UTC (Sun) by nix (subscriber, #2304) [Link]

bash_completion.d is for pikers. I use zsh completion, because no completion system is complete unless it's a double-layer system with the top layer and consisting of bytecompiled shell. :) (Also, I like the fuzzy and directory completions. I typo a lot.)

Where are the RMSs of the world?

Posted Sep 27, 2013 14:39 UTC (Fri) by SEJeff (subscriber, #51588) [Link]

The closest I would think is perhaps Lawrence Lessig, but he is certainly cut of a different cloth than RMS.

Where are the RMSs of the world?

Posted Sep 29, 2013 0:02 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]

Lessig is very self-promotion oriented. Its really not the same thing at all.

Who has a complete vision of something important but too obscure for other people to care about, who will carry that vision forward by their own sheer will even when everyone else considers it impossible or pointless?

Where are the RMSs of the world?

Posted Sep 27, 2013 14:43 UTC (Fri) by el_presidente (subscriber, #87621) [Link]

Maybe we don't need these things (as much) thanks to his self-sacrifice

Where are the RMSs of the world?

Posted Sep 27, 2013 14:56 UTC (Fri) by khim (subscriber, #9252) [Link]

yes, I said "free"

Do you understand the difference between “free” and “open source”?

But sometimes it pains me to think that there the number of Richard Stallmans hasn't exploded exponentially, while the popularity of Free Software has.

That's because free software have not exploded exponentially. That's why we've had GPLv3 fiasco and now are witnessing slow decay of GNU tower. Projects are either replaced (GCC with LLVM, readline with libedit, etc) or leaving (like GnuTLS did).

I don't like to talk about that when we are celebrating anniversary but free software is losing it's relevance. Till 2007 it looked like free software grows rapidly, too, but in reality the growth have come from “open source” side. When GPLv3 was invented to bring everyone back in “free software” camp it forced people's hand and most developers and users have chosen “open source” camp. Some have picked “free software” size (Samba is probably the most well-known example), but the overwhelming majority have rejected it.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 15:22 UTC (Fri) by coriordan (guest, #7544) [Link]

Neither GnuTLS nor its developers actually left. All major distros still use GCC. (Never heard of libedit, but I wouldn't call a Readline-killer the end of GNU.)

Much more importantly, if there is any decline in the GPL or copyleft licensing (I've seen people argue for and against this claim, with numbers), then I'm pretty sure it will be temporary.

The GPL did such a good job of fostering adoption and keeping free software out of legal problems, that some developers might have gotten the impression that changing the world is a risk-free endeavour. The current multiplication of mega corporation free riders, particularly in the smartphone market is likely to cause a backlash and a renewed interest in copyleft.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 16:28 UTC (Fri) by khim (subscriber, #9252) [Link]

Neither GnuTLS nor its developers actually left.

Really? Have you actually read the article? When main developers of your project (the ones who wrote over 90% of all the code in it) say gnutls is moving outside the infrastructure of the GNU project, actually move it to different place and and relicense the code it's pretty safe to say that yes, GnuTLS have left the GNU project. Well, FSF can insist that GnuTLS stays with GNU but that's will just mean that they'll keep some form on zombie around and real project (under some other name) will live in different place.

All major distros still use GCC.

Well, all GNU/Linux distributions still use GCC, sure. But that's kind of “by definition”. Non-GNU ones right now are exploring LLVM, but yes, most still use GCC (well, except for MacOS and iOS which are not Linux distributions, obviously). But for them it's question of “when”, not “if”. And non-GNU Linux distributions outnumber GNU Linux distributions today to such a degree that it's not even funny.

Much more importantly, if there is any decline in the GPL or copyleft licensing (I've seen people argue for and against this claim, with numbers), then I'm pretty sure it will be temporary.

Only time will tell. Most interesting projects as of late don't use the GPL (although some of them do), but as RMS correctly pointed out to be “truly free project” you must embrace the GPLv3+ and most projects are still on GPLv2+ (even some FSFs projects are on GPLv2+ because even FSF knows it can not relicense to GPLv2 and still keep it under their control: GLibC is prime example).

As I've pointed out GPLv2 is not “true free software license” — it sits on the border between “free software license” and “open source license” (and RMS himself explained) thus we need to look on GPLv3+ projects.

The GPL did such a good job of fostering adoption and keeping free software out of legal problems, that some developers might have gotten the impression that changing the world is a risk-free endeavour.

Hardly. GPL was broken by FSFs attempt to improve it. Note: no quotes! FSF truly believed that GPLv3 will improve the situation WRT free software… and back in 2006 this was my hope, too. Unfortunately this attempt backfired, free software and open source camps got a divorce and free software part is slowly withering away. As Landley explains it:

Copyleft is dying.

GPLv2 was category killer, synonymous with copyleft.
• terminal node of a directed graph of license convertability
• universal receiver
• A license was either GPL-compatible or it wasn't.
GPLv3 broke “the” GPL into incompatible forks that can't share code.
• Linux and Samba can't share code, implement 2 ends of same protocol.
• QEMU caught between GPLv2 Linux drivers and GPLv3 binutils/gdb processor instruction set descriptions. Can't take code from both.
• “GPLv2 or later” give to both but can't take code from either one.

Sad, but true.

The current multiplication of mega corporation free riders, particularly in the smartphone market is likely to cause a backlash and a renewed interest in copyleft.

Probably but unlikely. Since “an universal receiver” option is no longer available it makes it much less attractive. And since GPLv3 basically means that your creation will be rejected by most “serious” players… no, copyleft is not coming back any time soon. In five or ten years it may come back but it'll mean that we've faced a setback of such a cosmic proportions that we basically need to start from the beginning. I hope that we'll never reach that point.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 16:45 UTC (Fri) by drag (subscriber, #31333) [Link]

> Probably but unlikely.

I think so also. If Free software is to succeed it has to do because of merit, not because it forces companies to comply to a world view. Even if you get people participating that seek to distribute proprietary code based on your open source code then largely it is still a net win because they are still participating. Your project wins when giving you patches causes a net win for the company or group giving you patches.

Due to the caustic legal environment we have to exist in the 'copyleft' licenses themselves can serve a important role based on your goals and the type of software you are using, but it's still probably better to pick a license with less restrictions simply to reduce the legal overhead and increase chances of participation when you can.

For GCC, in particular, the GPLv3 plays a role in popularizing LLVM, but I think a much larger and more important factor is GCC's developers introducing hurdles and anti-features into their software to prevent third party plugins. I think the license was much more of a 'huge straw that broke the camel's back' type thing.

I think this was extraordinarily destructive because while you were making life hard for third party proprietary developers from 'stealing' GCC features by working around copyright restrictions, but it was also entirely destructive to any third party open source developer wanting to use GCC features.

It's like shooting yourself in spite of your foot. Extremely destructive and limiting.

People need to realize that when you introduce anti-features into your project to make life harder for proprietary developers you are also making life harder for all developers, free software included.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 17:01 UTC (Fri) by khim (subscriber, #9252) [Link]

For GCC, in particular, the GPLv3 plays a role in popularizing LLVM, but I think a much larger and more important factor is GCC's developers introducing hurdles and anti-features into their software to prevent third party plugins.

GPLv3 had nothing to do with creation of LLVM. That is crystal clear: Apple hired Chris Lattner before start of GPLv3 process. But GCC's switch to GPLv3 in 2008 certainly changed things: after that point it was not about “if gcc will be replaced with clang” but about “when gcc will be replaced with clang”. Apple never used any GPLv3 code (which means that all GNU programs in MacOS are quite old by now) and there are other companies like that.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 17:57 UTC (Fri) by hummassa (subscriber, #307) [Link]

> Apple never used any GPLv3 code (which means that all GNU programs in MacOS are quite old by now) and there are other companies like that.

That is because Apple LOVES embrace and extend, as does other companies; it's a "make easy money" scam. At any point in the future the many contributors to the "liberally-licensed", non-copylefted software ecosystems will probably get bitten in their collective arses by this fact, and when (not if) it happens, they will come back running to copylet.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 21:07 UTC (Fri) by drag (subscriber, #31333) [Link]

> collective arses by this fact, and when (not if) it happens, they will come back running to copylet.

There are a huge and growing number of non-copyleft free software projects out there and have been for years. Nobody has had their 'arses bitten' that I can tell.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 22:23 UTC (Fri) by hummassa (subscriber, #307) [Link]

He, he. Lots of BSD'd projects just died (or almost just died) in the 1990s because their code was being legally used under proprietary -- better developed -- software. Microsoft Windows being one of the offenders (which BSD socket library was used, anyway?)

But you don't have to believe me, just wait half a dozen years and tell me I was wrong. Or the other way around.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 23:31 UTC (Fri) by drag (subscriber, #31333) [Link]

> He, he. Lots of BSD'd projects just died (or almost just died) in the 1990s because their code was being legally used under proprietary

BSDi products were sued because they were violating copyright and ended up settling out of court with Novel.

> better developed -- software. Microsoft Windows being one

Excuse me for the selective edit, but frankly that is BS, because; Microsoft Windows is not better developed, Microsoft didn't really have a operating system that was competitive with Unix (proprietary or free) until 2000, and the whole 'using BSD code' is entirely overblown and was the entire point of having TCP/IP stack cheaply licensed.

BSDi getting sued in the early 1990's doesn't support your contention at all, IMO. The only thing that it does illustrate is the destructive and toxic nature that copyright restrictions (and other IP) have on technological advancements.

GNU's healthy now, and will soon be moreso

Posted Sep 28, 2013 1:55 UTC (Sat) by khim (subscriber, #9252) [Link]

Microsoft Windows is not better developed, Microsoft didn't really have a operating system that was competitive with Unix (proprietary or free) until 2000

What are you talking about? Windows NT was released back in 1993. Sure, version suitable for desktop (with accelerated 3D and other goodies) was not released till 2000, but Windows NT was used on quite a few workstation in that period.

BSDi getting sued in the early 1990's doesn't support your contention at all, IMO. The only thing that it does illustrate is the destructive and toxic nature that copyright restrictions (and other IP) have on technological advancements.

Sure, but what this has to do with anything? Yes, BSDs suffered from BSDi processing and lost couple of years as a result but when USL v. BSDi case was resolved Linux was still pretty weak OS which was clearly inferior to BSD. But managed to take over (as even some BSD guys admit). The question is still open if it was because of copyleft of if it was because of something else (Linus personality?), but it looks that copyleft actually helped.

But that was different era: era when people (and not corporations) were main actors in FOSS projects, era when ASF tended to it's server and had no interest in office suites and, of course, era when copyleft meant tit-for-tat and nothing else. We still don't know how new and improved GPLv3-based copyleft will affect the balance. So far we are observing both positive and negative effects from it. E.g. Samba. Samba 4 will probably open some new opportunities but at the same time we can't use Samba with our Android devices (at least not on stock roms) because it's GPLv3-licensed now. Only time will tell which way this whole thing will go. Ditto for other GPLv3-based projects.

GNU's healthy now, and will soon be moreso

Posted Sep 28, 2013 10:44 UTC (Sat) by jengelh (subscriber, #33263) [Link]

>Linux [...] managed to take over[...]. The question is still open if it was because of copyleft of if it was because of something else (Linus personality?), but it looks that copyleft actually helped.

With respect to Linux (both in the sense of kernel and as operating system bundles called distributions), "quality of implementation" is the reason for many a user. Lately with all the security agency crap going on, I would prospect that open source (just open, not necessarily free) would garner an increase in followers.

GNU's healthy now, and will soon be moreso

Posted Sep 30, 2013 0:45 UTC (Mon) by jra (subscriber, #55261) [Link]

khim wrote:

> E.g. Samba. Samba 4 will probably open some new opportunities but at the
> same time we can't use Samba with our Android devices (at least not on
> stock roms) because it's GPLv3-licensed now.

You certainly can use Samba on your Android devices. You chose not to as you don't want to obey the license.

Don't use the word "can't" when you mean "won't". It's misleading.

GNU's healthy now, and will soon be moreso

Posted Sep 30, 2013 8:53 UTC (Mon) by khim (subscriber, #9252) [Link]

Don't use the word "can't" when you mean "won't". It's misleading.

That's why I use “can't” when I mean “can't”. Samba is not something you can just add to your stock Android because it requires root access to be actually usable (you can install it on non-standard port, but then Windows will not able to see it which makes it less then useful for most users). You either need to root the device or install special rom. In both cases you, most likely, will lose the warranty. Instead we are forced to use MTP, PTP, or something like Dropbox/Google Drive.

You certainly can use Samba on your Android devices. You chose not to as you don't want to obey the license.

Me? I've chosen nothing (well, except maybe for the phone model, but they all are identical WRT Samba support). Android developers have chosen not to use Samba—and most likely because it's license is GPLv3 now. Android tends to avoid even GPLv2 components and GPLv3 is big no-no.

GNU's healthy now, and will soon be moreso

Posted Oct 3, 2013 8:06 UTC (Thu) by madhatter (subscriber, #4665) [Link]

I'm sorry, but I think it's disingenuous to suggest that GPLv3 software can't run on Android phones. I open up my F-Droid client, and the screenful of freshly-released software, all ready to be installed on my Android phone, lists along with the packages ready to go, their licences: four GPLv3, three GPLv3+, and one MIT.

Much of the software I'm already running from f-droid is GPLv3-licenced, including Angulo, the APV PDF viewer, the Dasher text input client, the FasterGPS NTP client, and Vector Pinball.

It might be fair to say that you haven't chosen a handset that makes this choice easy, or that you haven't chosen to use your handset in a way that makes this choice easy, but I don't think it's fair to say that it's in some inherent way impossible.

For me, jra's point about "can't" vs "won't" stands.

GNU's healthy now, and will soon be moreso

Posted Oct 3, 2013 11:24 UTC (Thu) by khim (subscriber, #9252) [Link]

I'm sorry, but I think it's disingenuous to suggest that GPLv3 software can't run on Android phones.

Who said anything about random pieces of software? Of course you can do that! But if you are creating software which is supposed to be used by OS (samba or libstdc++) and it's distributed under GPLv3 then no, you can not do that. Either it's distributed separately (libstdc++) or it can not be used at all (samba).

It might be fair to say that you haven't chosen a handset that makes this choice easy, or that you haven't chosen to use your handset in a way that makes this choice easy, but I don't think it's fair to say that it's in some inherent way impossible.

Really? What kind of handset I should have used instead? Note: I've clarified in the very beginning: at least not on stock roms. Of course if you root your phone or install custom firmware then you can use anything you want including samba, but most users don't know how to do that (and don't want to know).

GPLv3 is incompatible with OS distribution?

Posted Oct 3, 2013 20:11 UTC (Thu) by kmacleod (guest, #88058) [Link]

But if you are creating software which is supposed to be used by OS (samba or libstdc++) and it's distributed under GPLv3 then no, you can not do that. Either it's distributed separately (libstdc++) or it can not be used at all (samba).

Can you provide a link to an in-depth article or previous thread that gives background and details on "GPLv3 is incompatible with OS distribution"? An FSF statement along those lines would also work.

GPLv3 is incompatible with OS distribution?

Posted Oct 3, 2013 22:26 UTC (Thu) by khim (subscriber, #9252) [Link]

Can you provide a link to an in-depth article or previous thread that gives background and details on "GPLv3 is incompatible with OS distribution"?

But of course!

And before you'll say “hey, but they could have just opened up their system” you need to read another article.

Actually you need just three lines from said article: Mobile phone sales is not like cameras or cars or clothing. In all those other industries, if you make a good product and offer it at competitive prices, you will have success in the market. In mobile, even if you have the best phone ever made, it is utterly IRRELEVANT if the carriers decide not to support it.

Combine these two articles—and here is your answer. It's as simple as 1-2-3:
1. Mobile phone without carriers support is DOA.
2. Carriers demand control over devices.
2. GPLv3 guarantees that you (and not carrier!) can control the device you've bought.
As you can see items number two and number three are very clearly incompatible and thus GPLv3 have to go.

You can notice that Apple does not cede control over iPhone to carriers (there are no carrier logos, etc). That's true, Apple certainly managed to pressure carriers more then any other handset manufacturer to date. But carriers still have some measure of control: even if technically Apple (and not you!) decides how your phone behaves carriers still decide many things. For example carriers determine if your iPhone will support tethering or not.

P.S. Situation is slowly changing but I think it'll be few more years till anyone will be able to create truly successful phone which will be successful despite carriers opposition.

GPLv3 is incompatible with OS distribution?

Posted Oct 4, 2013 1:33 UTC (Fri) by kmacleod (guest, #88058) [Link]

Ah, I misunderstood your post. I thought you meant a technical/legal incompatibility.

GPLv3 is incompatible with OS distribution?

Posted Oct 4, 2013 10:28 UTC (Fri) by nix (subscriber, #2304) [Link]

In mobile, even if you have the best phone ever made, it is utterly IRRELEVANT if the carriers decide not to support it.

This is, of course, not true in most of Europe. It's a US thing, I think.

The thought of not being able to choose my mobile provider and phone independently is baffling. Why on earth would anyone want to cut themselves off from choice like that? Oh yes, because that way they get a 'free' phone (and an increase in contract cost that means they pay much more in the next year or two than they would have done to get the phone). i.e., it's a phone mortgage.

I'm not really sure that phones cost enough to be worth taking out a mortgage for them -- particularly not when a side-effect is to greatly restrict your choice of phone.

GPLv3 is incompatible with OS distribution?

Posted Oct 4, 2013 11:14 UTC (Fri) by dlang (✭ supporter ✭, #313) [Link]

In the US, if you try to get phone service without the contract and 'free phone', you really don't save very much. I'm going through this process right now, and you really have to hunt around and _not_ go with the main carriers to save anything. A few years ago I purchased a tablet that could take a SIM card and when I went to get service, most of the phone stores I went to didn't understand what I was talking about. That's getting better now, but still only AT&T and T-Mobile can get you a SIM, The other two major carriers (Sprint and Verizon) require different models of the devices that only work with their networks.

And like it or not, the US is a huge block of customers for phone companies. the EU may or may not have more total phone users, but that market is far more fragmented in terms of the advertizing needed, the different language versions needed etc.

in the US, most phone manufacturers spend exactly zero money on advertizing, it's only in recent years that you would find any phone advertizing from Samsung or HTC (Motorola and Apple have been advertizing for a lot longer). Since the vast majority of users get their phone from their carrier, based strictly on what the carrier has on display in their store, there hasn't been much need.

One good thing about the 'smartphone wars' is that people are seeing that there are a bunch of different options, and so people are getting interested in getting a specific phone rather than just picking whatever is in stock.

as for developing fully open phones, if you were a phone manufacturer and knew that investing a large chunk of money into your own system would anger the carriers that control access to a very large number of customers, would you be eager to do so?

GPLv3 is incompatible with OS distribution?

Posted Oct 5, 2013 14:56 UTC (Sat) by kleptog (subscriber, #1183) [Link]

Actually, in the EU there is a huge market for SIM only plans, I think largely *because* of the fragmentation. Given that your mobile phone plan really only cheap in the country you bought it, it's not uncommon to buy an extra SIM when you go on holiday.

There are even providers that specialise in this: you select a country, a time period and a number of MB/minutes and they'll mail you a SIM card which you pop in when you reach your destination and pop out again when you go home. They don't even really need to advertise either, people who need it will go looking for them. But it does make locked phones somewhat worthless for normal use.

Ofcourse, for europeans going to the US, they can buy one SIM for one large area, which is reasonably attractive. I imagine this market will slowly disappear when the cost of roaming becomes more reasonable, either by competition or government regulation.

GNU's healthy now, and will soon be moreso

Posted Oct 4, 2013 11:05 UTC (Fri) by paulj (subscriber, #341) [Link]

Again, as others have already have pointed out, you are using "can't" when you really mean "won't".

Nothing stops someone putting together a phone and using GPLv3 licensed software, if they are willing to abide by the GPLv3.

GNU's healthy now, and will soon be moreso

Posted Sep 28, 2013 6:39 UTC (Sat) by did447 (guest, #49454) [Link]

> I think this was extraordinarily destructive because while you were making life hard for third party proprietary developers from 'stealing' GCC features by working around copyright restrictions, but it was also entirely destructive to any third party open source developer wanting to use GCC features.

Absolutely, if you read compiler conferences papers most (all?) use LLVM now, even when they release their own code as GPL3 software!

GNU's healthy now, and will soon be moreso

Posted Sep 29, 2013 17:54 UTC (Sun) by nix (subscriber, #2304) [Link]

a much larger and more important factor is GCC's developers introducing hurdles and anti-features into their software to prevent third party plugins.

Since this never happened, I doubt it.

No 'hurdles' or 'anti-features' were ever introduced (the developers would be unlikely to do that, given that they were almost universally in *favour* of plugins!). Rather, the major improvements that *could* have happened earlier to improve the modularity of the huge ball of hair that was GCC didn't quite happen as early as they might have. That's all. It's still a ball of hair, certainly compared to LLVM, with a much higher learning curve.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 17:39 UTC (Fri) by coriordan (guest, #7544) [Link]

> Have you actually read the article?

Yes. But unlike you, I also checked to see to see what happened afterwards. The article was about Nikos taking GnuTLS out of GNU, but all the GNU releases since the article have all been made by... Nikos.

> GPLv3 basically means that your creation will be rejected by
> most “serious” players

GPLv3 means your software will be rejected by corporations that never wanted users to get the freedoms anyway. If your goal is to do unpaid work for corporations, then GPLv3 is not for you. If your goal is for users to be free to use, modify, run and redistribute, then GPLv3+ is your best bet.

GPLv3 didn't drive anyone away. It just unmasked a few hangabouts that weren't our friends to begin with.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 18:10 UTC (Fri) by hummassa (subscriber, #307) [Link]

> GPLv3 didn't drive anyone away. It just unmasked a few hangabouts that weren't our friends to begin with.

YES. Thank you. I agree 100%.

Sure, does Apple (or Google) produce great quality software? Yes, it does! Does it produce software that is guaranteed to stay free forever? Uh-oh. Wait and see. I will support copylefted software as long as I am alive, for this reason alone.

GNU's healthy now, and will soon be moreso

Posted Sep 28, 2013 3:39 UTC (Sat) by wahern (subscriber, #37304) [Link]

It's free as long as you have the source code with the permissive license. And a company shipping GPL code could stop updating it at any time--e.g. MySQL.

Also, unfortunately, no GPL code authored by a real person is guaranteed free forever, because US copyright law gives individuals an inalienable right to terminate a license at 35 years after publication, regardless of the terms of the license. (In practice this is nothing to worry about, but corporate lawyers like to bring this up and it's caused no end of bickering between Stallman, the FSF, and proprietary software advocates.)

Don't get me wrong, I'm not anti-GPL. But the differences between GPL and BSD, et al licenses are more subtle. Does the GPL nudge companies to contribute more? In some cases, yes; in others, no, because they simply avoid it.

For example, my company encourages me to work on and release Free Software (and I have, and Apple and Google and Hulu use my FOSS software), but they ask I use a BSD or MIT license. GPL licenses give them headaches and they don't want to be hypocrites about it. GPL licenses have real costs when distributing products, because even if compliance isn't a roadblock, it still takes time and effort. This is why everybody appreciated the U. of California Regents rescinding the advertising clause from the original BSD license.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 20:41 UTC (Fri) by khim (subscriber, #9252) [Link]

I also checked to see to see what happened afterwards.

Hmm… really? Have you checked the GNU ftp site?

The article was about Nikos taking GnuTLS out of GNU, but all the GNU releases since the article have all been made by... Nikos.

You mean non-GNU releases on non-GNU sites?

Last GNU versions are 3.0.26/3.1.5. All versions after that (3.0.32/3.1.14/3.2.4 currently) are released not as part of GNU project and they are not released via usual GNU channels. Sure, they were made by Nikos. And FSF was even provided redirects from old addresses @ http://www.gnu.org/software/gnutls/ to new addresses @ http://www.gnutls.org/. That's cool: this means that they kept enough common sense to understand that even if legally they can try to keep GnuTLS as GNU “zombie project” it'll just make them a laughing stock.

But the fact remains: GnuTLS have left the GNU project and it's not coming back.

GPLv3 means your software will be rejected by corporations that never wanted users to get the freedoms anyway.

Sure. Few corporations do.

If your goal is to do unpaid work for corporations, then GPLv3 is not for you. If your goal is for users to be free to use, modify, run and redistribute, then GPLv3+ is your best bet.

Well, that's one way of saying it. Another way is to say that if you do care about users then GPLv3 is not for you. If you do care only about software freedom (and it does not matter for you if users will ever see or care about your software) then yes, GPLv3 is your best bet.

As I've said before: GPLv3 separated people who think about software as means of doing something and people who think about software (and software's freedom) as goal. And most people care more about people then they care about unfeeling bytes.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 23:16 UTC (Fri) by pizza (subscriber, #46) [Link]

>Well, that's one way of saying it. Another way is to say that if you do care about users then GPLv3 is not for you. If you do care only about software freedom (and it does not matter for you if users will ever see or care about your software) then yes, GPLv3 is your best bet.

Come now, "user's freedom" is precisely what the GPL was written to ensure. Users, by definition, are those that use the code. Those who don't use the code, aren't users, and are irrelevant in this equation.

Now, encouraging folks to use that code is another matter entirely, but that is precisely their choice to make.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 23:27 UTC (Fri) by coriordan (guest, #7544) [Link]

> really? Have you checked the GNU ftp site?

Why yes, I have. It contains a README saying it's a GNU project, and an AUTHORS file saying FSF is the copyright holder. Got any other evidence that proves I'm right?

> Another way is to say that if you do care about users (...)

First, that makes no sense because if they only care about users, to the point that they give up on software freedom, then why use a free software licence in the first place? (And "because we can trick people into unpaid work" isn't valid.)

Second, this mass exodus that you base your claims on never happened. Some companies walked away. That's a pity. But if the only way to keep them was to promise they could circumvent the licence, then we just have to accept that their presence was just a fad and there was no way to keep them.

I don't have links or stats about GPLv3, but John Sullivan of FSF makes a solid case for there being a big increase in the use of GNU licences:

http://faif.us/cast-media/FaiF_0x23_Is-Copyleft-Framed_sl...

Slides 14 to 17 show that between 2005 and 2011, the percentage of Debian packages under GNU licences went from 71 to 93. That's looking pretty healthy.

GNU's healthy now, and will soon be moreso

Posted Sep 28, 2013 1:26 UTC (Sat) by khim (subscriber, #9252) [Link]

It contains a README saying it's a GNU project, and an AUTHORS file saying FSF is the copyright holder.

Really? For GnuTLS 3.2.4? Perhaps you are living in some other universe then me. Transdimensional communication is probably bigger discovery then mere squabbles between “free software” and “open source” proponents! We should investigate this phenomenon further…

Got any other evidence that proves I'm right?

Not really. It's kinda pointless to offer evidence to someone who refuses to admit that black is black and white is white.

First, that makes no sense because if they only care about users, to the point that they give up on software freedom, then why use a free software licence in the first place?

Not free software license. Open source license. ESR explained why quite well in his well known book. It has nothing to do with “free software” jihad, it's just a better software development model. Well, it's better in certain circumstances and worse in some other cases, but I digress.

I don't have links or stats about GPLv3, but John Sullivan of FSF makes a solid case for there being a big increase in the use of GNU licences:

http://faif.us/cast-media/FaiF_0x23_Is-Copyleft-Framed_slides.pdf

One phrase from said “solid case” says it all: When it comes to looking at a different, well-vetted frame of software — like what's in Debian GNU/Linux — GPL family use is very high. If you ignore the fact that most software novadays are developed not for inclusion in traditional Linux distributions but for the Web, Android and iOS then yes, you can conclude that GPL use is growing. But it's not because copyleft is winning. That's because developers are moving in other directions and develop things for other platforms. Also note that even in said “well-vetted frame of software” GPLv3 only managed to grab 10-20% of all GPL packages. And we know that if you value freedom you should use GPLv3, not GPLv2, right?

GNU TLS copyright

Posted Sep 28, 2013 7:15 UTC (Sat) by oldtomas (guest, #72579) [Link]

coriordan It contains a README saying it's a GNU project, and an AUTHORS file saying FSF is the copyright holder.
khim Really? For GnuTLS 3.2.4? Perhaps you are living in some other universe then me. Transdimensional communication is probably bigger discovery then mere squabbles between “free software” and “open source” proponents! We should investigate this phenomenon further...

OK. Let's see. Just downloaded from gnutls.org The AUTHORS file (at top level) says, among other things:

GnuTLS AUTHORS -- Information about the authors.
Copyright (C) 2000-2012 Free Software Foundation, Inc.
See the end for copying conditions.

The copyright holder for GnuTLS is Free Software Foundation, Inc., 51
Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.

So I seem to be living on coriordan's universe (I guess I like it more there, btw.)

GNU TLS copyright

Posted Sep 28, 2013 14:16 UTC (Sat) by khim (subscriber, #9252) [Link]

Close, but no cigar.

khim: Hmm… really? Have you checked the GNU ftp site?

coriordan: Why yes, I have. It contains a README saying it's a GNU project, and an AUTHORS file saying FSF is the copyright holder. Got any other evidence that proves I'm right?

khim: Really? For GnuTLS 3.2.4?

oldtomas: OK. Let's see. Just downloaded from gnutls.org The AUTHORS file (at top level) says, among other things:

$ whois gnutls.org
…
Registrant Name:Nikos Mavrogiannopoulos
Registrant Organization:Nikos Mavrogiannopoulos
…

Does it look like GNU ftp to you?

It's impossible to call out copyright assignments thus obviously copyright for old code will remain with FSF forever (well, forever less one day, but who's counting). But note that “(C) Free Software Foundation, Inc.” notices mostly cover 2009-2012. New code (in gnutls_dtls.c, heartbeat.c, rsa_psk.c and dozen other files) are “Copyright (C) 2013 Nikos Mavrogiannopoulos”, “Copyright (C) 2013 Frank Morgner”, “Copyright (C) 2013 Adam Sampson <ats@offog.org>”, etc. Which means that copyright is no longer assigned to FSF. Old versions include copyright from other people (so much for FSF being the “sole copyright owner”), but these were mostly supplementary files not written by Nikos. This is far as you can pull former GNU project from under the FSF's umbrella.

GNU TLS copyright

Posted Sep 28, 2013 19:04 UTC (Sat) by oldtomas (guest, #72579) [Link]

    $ whois gnutls.org
    …
    Registrant Name:Nikos Mavrogiannopoulos
    Registrant Organization:Nikos Mavrogiannopoulos
    …

Does it look like GNU ftp to you?

Oh, for $DEITY's sake, khim!. We were discussing the license of the GNUTLS code, not the owner of the ftp site whence the sources come from. Let's stay focused, pretty please.

New code (in gnutls_dtls.c, heartbeat.c, rsa_psk.c and dozen other files) are “Copyright (C) 2013 Nikos Mavrogiannopoulos”, “Copyright (C) 2013 Frank Morgner”, “Copyright (C) 2013 Adam Sampson <ats@offog.org>”, etc. Which means that copyright is no longer assigned to FSF.

There you might have a point. Now, if the current maintainer doesn't change the AUTHORS file (which he would be free to do: not removing the copyright by FSF, mind you, but augmenting it, like he very well does in the file mentioned by you, e.g.:

 * Copyright (C) 2009-2012 Free Software Foundation, Inc.
 * Copyright (C) 2013 Nikos Mavrogiannopoulos

then I'd tend to assume that he's fine seeing the overall copyright in the hands of the FSF).

Still it's a departure of the assignment model preferred by the FSF, so I'll concede you a partial one on that. A far cry from "leaving", as you posted in your original article, though).

GNU TLS copyright

Posted Sep 28, 2013 20:15 UTC (Sat) by khim (subscriber, #9252) [Link]

Oh, for $DEITY's sake, khim!. We were discussing the license of the GNUTLS code,

Really? Sorry, to disasppoint you, but we were talking about an article which clearly says the development of gnutls is moving outside the infrastructure of the GNU project and that Nikos (one of two principal developers of GnuTLS) no longer consider GnuTLS a GNU project. Nothing more, nothing less.

coriordan made ridiculous claim that neither GnuTLS nor its developers actually left and another even more ridiculous one that all the GNU releases since the article have all been made by… Nikos. This is what we discussing.

not the owner of the ftp site whence the sources come from.

Why?

Let's stay focused, pretty please.

Well, let's. One property of GNU project is that it's source is distributed via ftp.gnu.org: we strongly recommend using ftp.gnu.org to distribute official releases. Another one is that GNU package maintainers are supposed request either to assign the copyright to the Free Software Foundation or to sign a copyright disclaimer to put this change in the public domain. And yet another one is that GPL license upgrade is recommended.

A far cry from "leaving", as you posted in your original article, though.

Really? License itself is important sign of GNU project, but far from the only sign. It's not even the most important one: there are plenty of non-GNU projects under GNU license. But if author of GnuTLS says that he no longer consider GnuTLS a GNU project, stops assigning copyright to FSF, no longer uses GNU-provided facilities, switches from GPLv3 back to GPLv2 (note that zombie project which proudly claims that this project is part of the GNU Project still as proudly claims that it's license is GNU General Public License v3 or later, too) then what exactly ties it to GNU project? One measly file?

For all practical purposes GnuTLS have left the GNU project, if they are keeping few attribution files to keep the name I would not blame them.

GNU TLS copyright

Posted Sep 29, 2013 18:04 UTC (Sun) by nix (subscriber, #2304) [Link]

One property of GNU project is that it's source is distributed via ftp.gnu.org

GNU is more vaporous and less dictatorial than you seem to believe. Note that both GNOME and Bazaar are 'officially' GNU projects (in the sense that 'RMS says they are'), and neither is distributed from ftp.gnu.org. Heck, RMS stamps TeX and X with the GNU imprimatur (they are 'part of the GNU operating system'), but in that case not even the projects' maintainers would agree.

Even 'developed by GNU' is a woolly thing. A few core things were originally developed by funding from the FSF, but even those were mostly not developed by RMS, and many of them are still maintained by their original developers. Some of those have got pissed off with RMS and said 'we are no longer GNU' -- but this is just semantics. It makes almost *no* difference, because GNU is no longer the sole, nor even primary, umbrella under which free software sits.

The world is bigger than GNU now -- GNU has transmitted its message effectively enough that it is no longer as important as it was. And this is a *good* thing, not a thing to be afraid of.

GNU TLS

Posted Sep 30, 2013 11:55 UTC (Mon) by coriordan (guest, #7544) [Link]

This discussion started with your claim of a "slow decay of GNU tower" and you've proved nothing to support that claim.

All you've shown is that there have been some arguments within one GNU project and the main developer of that project has decided to start keeping his copyrights (possibly because he wants to enforce the licence against someone).

From there, you dramatise, make assumptions, and exaggerate a few details to make a claim that GnuTLS is no longer part of GNU, and from there you jump to the claim that GNU is decaying.

Do you really expect a 30 year project, with a hundred sub-projects and thousands of developers, to produce as much software as GNU has, without any internal arguments? This is par for the course in such a massive undertaking and the continued existence and growth of GNU is a tribute to the rarely-praised management skills of RMS!

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 18:26 UTC (Fri) by hummassa (subscriber, #307) [Link]

> • “GPLv2 or later” give to both but can't take code from either one.

Sure it can. It only has to choose when it does.

If I combine a program that is v2+ with a v2 patch, I can redistribute the result under v2-only license.
If I combine a program that is v2+ with a v3 patch, I can redistribute the result under v3-only license.
If I combine a program that is v2+ with a v3+ patch, I can redistribute the result under v3+ license.

And so on.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 19:26 UTC (Fri) by southey (subscriber, #9466) [Link]

All of your examples are illegal! These all require changes in the license especially as each change further restricts the original license. You cannot relicense any code unless you are the only copyright holder.

While you could redistribute the package under a different but compatible license, it is would be wrong include all included code under that compatible license. Rather the original program will always remain GPL v2 or later and only your patch will be covered under the compatible license you choose.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 22:19 UTC (Fri) by hummassa (subscriber, #307) [Link]

You are not relicensing. You are distributing under the terms the copyright holders allowed you to distribute. Go read all the terms of the GPLv2 and v3 a couple of times and you'll get some illumination after some time.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 22:33 UTC (Fri) by hummassa (subscriber, #307) [Link]

I will clarify this a little bit:

When you say:

"You can redistribute this work (and derivatives) under the terms of the GPL, version 2 or later", then you are saying:

A. you can redistribute this work (and derivatives) under the terms of the GPLv2
B. OR you can redistribute this work (and derivatives) under the terms of the GPLv3
C. OR you can redistribute this work (and derivatives) under the terms of the GPLv3.1

and so on.

If you need to redistribute some work that is a derivative of both a GPLv2 work (terms under letter A, above) and a GPLv2+ work (A|B|C|...), you still have the option of distributing it under the terms of A. Likewise, if you need to redistribute some work that is a derivative of both a GPLv3+ work and a GPLv2+ work (B|C|... and A|B|C|... respectively) then you still have the option of distributing it under the terms of B|C|... etc etc.

GNU's healthy now, and will soon be moreso

Posted Sep 27, 2013 22:26 UTC (Fri) by coriordan (guest, #7544) [Link]

There's no legal problem.

The only way to distribute a combination of v2+ and v3+ is to abide by v3 (or a later version when one is published). So it's fair to tell people that the project "is under v3". It's simplifying slightly, but not to the extent of being misleading.

To be precise, the project doesn't have a licence. The works (authors' works) have licences.

(If I unilaterally claimed that every part of the project, including the v2+ parts written by others, was v3+, or if I edit all the other authors' copyright notices to say v3+, then yes, either of those acts would be illegal, but that's not the case here.)

GNU's healthy now, and will soon be moreso

Posted Sep 29, 2013 18:03 UTC (Sun) by jzbiciak (✭ supporter ✭, #5246) [Link]

The main issue is that the original project is unlikely to want to change its license terms in order to accept a contribution into its mainline. So, while you can legally combine code in the ways you list and redistribute the result in the resulting intersection between licenses, the original project would probably reject the submission because it changes the licensing for the whole project.

I'm pretty sure that's what was meant by "give to both but can't take code from either one." If my project starts as a GPLv2+ project and wants to remain a GPLv2+ project, it cannot accept a V2 patch, V3 patch or V3+ patch into mainline.

Where are the RMSs of the world?

Posted Sep 27, 2013 15:29 UTC (Fri) by thumperward (guest, #34368) [Link]

I don't like to talk about that when we are celebrating anniversary

Given your longstanding tendency to derail almost any discussion in which you participate with a forcefully-worded contrarian position, I rather fancy that was exactly your intention.

Where are the RMSs of the world?

Posted Sep 27, 2013 16:42 UTC (Fri) by oldtomas (guest, #72579) [Link]

Now, now. I think you are being unjust on khim.

I don't agree on all (s)he says (actually I tend to disagree more than to agree, especially on this topic), but his/her posts are articulate and polite, and represent a respectable position. Sometimes strongly opinionated, yes, but always preferable to a bland non-opinion.

Where are the RMSs of the world?

Posted Sep 28, 2013 11:58 UTC (Sat) by pboddie (guest, #50784) [Link]

I find that comments generally come across as being a lot more polite if they don't all feature "Really?" as the very first sentence, optionally misspelt for dramatic effect or preceded by some other exclamation-enhancing padding. I'm sure we all make similar mistakes, but toning down the drama makes it more likely that people might keep reading, potentially increasing the benefit to all concerned.

Where are the RMSs of the world?

Posted Sep 28, 2013 23:44 UTC (Sat) by marcH (subscriber, #57642) [Link]

> but toning down the drama makes it more likely that people might keep reading, potentially increasing the benefit to all concerned.

I hope you never considered any kind of job as a journalist. You could even fail to be a "LWN journalist"

Where are the RMSs of the world?

Posted Sep 29, 2013 0:06 UTC (Sun) by gmaxwell (subscriber, #30048) [Link]

> Projects are either replaced (GCC with LLVM

Which is being stuffed full of patents by qualcomm and apple by their employees (some of whom were major GCC contributors while they worked with other employers…), which _no one_ receives a patent license to...

If LLVM/Clang were under the Apache license or had a contributor agreement that released access to patents under free software compatible terms you might have something to say there, but as it's currently licensed I think that we're just seeing a latently proprietary grow in strength... which isn't good news, but it's not something new.

Where are the RMSs of the world?

Posted Sep 29, 2013 14:32 UTC (Sun) by Del- (guest, #72641) [Link]

When a thread evolves I often find it interesting to go back to the starting point. In this case khim, it seems like you have been moving the goal post as you move along. Your starting point was:

> That's because free software have not exploded exponentially. That's why we've had GPLv3 fiasco and now are witnessing slow decay of GNU tower. Projects are either replaced (GCC with LLVM, readline with libedit, etc) or leaving (like GnuTLS did).

In my view the failure of GPLv3 to succeed universally is as much about knowledge and Linus Torvalds as anything else. As long as linux sticks to GPLv2, it will be difficult to drive momentum over to v3. Also most people really lack the knowledge, so it is just as much about our failure to educate the world on the benefits of GPLv3 over GPLv2. I fully agree that making v2 and v3 incompatible was a stupid mistake, which certainly is on the top three list of reasons too.

Now, you seem (willfully?) to want to pollute that discussion with the departure from FSF and GNU for some high profile developers. I believe that part is not about freedom. It just as well an be about copyright transfer aversion, the fact that important tools like git and cmake are not in the GNU stack, and some strong minded developers that like to go their own way rather than fit in in a larger community. That does *not* mean that those same developers are all about open source and not about freedom. It is a very different topic all together.

You have a fair argument about many choosing GPLv2+ over GPLv3+, but that has probably everything to to with the fact that v2 and v3 are incompatible. It dos *not* mean that those same developers are less freedom loving. For those who choose plain GPLv2 I believe their is a small minority that are even concious about it, and those do it just because they don't want RMS to decide future licensing terms for their software. It does *not* mean that they are all about open source and not about freedom.

Now you saw documentation that Debian sees a strong growth for copy-left (even exponentially?), and your answer is that Debian is not where the action is at. I can only assume that you are one of those who claim that the desktop is dead, and never knew servers existed. Even if I forgive you this misstep, and consider only smartphones and tablets, you do not seem to see the larger picture. Both Qt and Mono uses copy-left licensing. Applications like VLC is bound to surface once these tools are more easily deployable (which is just about now). In particular, I feel confident that applications from KDE will find it's way to the handhelds, and that it will make a change. On graphical drivers we already see clear signs (lack op copy-left there has little to do with lack of freedom love. I suggest you ask Luc Verhagen and Rob Clark yourself, I feel no need to.

Where are the RMSs of the world?

Posted Sep 29, 2013 17:27 UTC (Sun) by khim (subscriber, #9252) [Link]

If we “go back to the starting point” then we should probably go back to The GNU Manifesto and the infamous printer problem.

Manifesto says it quite clearly: So that I can continue to use computers without dishonor, I have decided to put together a sufficient body of free software so that I will be able to get along without any software that is not free.

Free software movement starts with the idea that proprietary software is not just something practically problematic, but something ethically so awful that it needs to be eliminated. He later debated the fact that he's using proprietary systems to develop GNU but concluded that it's acceptable as long as he only uses proprietary system to develop free replacement:

22. two_front_teeth: Suppose your doctor told you that you needed a medical procedure to survive but that the procedure would require inserting a device inside of your body which ran proprietary software. Would you be willing to have the procedure done to save your life?

RMS: The only way I could justify this is if I began developing a free replacement for that very program. It is ok to use a nonfree program for the purpose of developing its free replacement.

And this is what free software is all about. It's not about producing some software which anyone can take and tinker with. It's about complete elimination of proprietary software. That's goal is too large thus GNU project does not try to eliminate all the proprietary software from all areas of life but it aims to eliminate them from the realms of operation systems: GNU will remove operating system software from the realm of competition. You will not be able to get an edge in this area, but neither will your competitors be able to get an edge over you.

Now, for the “open source software”. If you'll read what Stallman writes about it you'll see that it accuses it's followers because they “miss the point of Free Software”. Note which words are used there, BTW: We in the free software movement don't think of the open source camp as an enemy; the enemy is proprietary (nonfree) software… Proprietary add-on software and partially nonfree GNU/Linux distributions find fertile ground because most of our community does not insist on freedom with its software.

So the left goalpost is as following: “free software camp” is at war (where else do you talk about enemies?) and it's goal is complete elimination of proprietary software. Remember that GNU does not endorse Debian GNU/Linux because people can readily learn about these nonfree packages by browsing Debian's online package database and Fedora because it makes an exception for certain kinds of nonfree firmware. Obviously the usefullness of the OS is not important if it means that freedom of saoftware is compromised.

Now for the right goalpost. “Open source camp” does not have a leader similar to Stallman, but if anything can serve as it's manifesto it's The Cathedral and the Bazaar which simply states that “given enough eyeballs, all bugs are shallow” which immediately puts to the rest an idea that proprietary software is somehow evil. No, it's not evil, it's just inferior. And it's inferior for the simple reason that not enough “eyeballs” are seeing it's source code. Which basically means that proprietary software makes perfect sense if the only way to get enough “eyeballs” is to pay for them (games are typical example). In a way his accusation about “missing the point” are true: “open source camp” is all about best software, not about free software.

In my view the failure of GPLv3 to succeed universally is as much about knowledge and Linus Torvalds as anything else.

Of course. But if you take a look on GPLv3 from “free software camp” POV then it's obviously better weapon than GPLv2 and that means everyone should upgrade. People from “open source camp” (and Linus is quite obviously one of them) note that it's trying to make software to be a “weapon of mass opinion” and usually refuse to participate in said war.

Now, you seem (willfully?) to want to pollute that discussion with the departure from FSF and GNU for some high profile developers.

No. I want to point out that even maintainers of GNU software (who must be more-or-less by definition in “free software camp”) are not all that ready to make their software to be a “weapon of mass opinion”.

It just as well an be about copyright transfer aversion, the fact that important tools like git and cmake are not in the GNU stack, and some strong minded developers that like to go their own way rather than fit in in a larger community.

No, it' not about git (GNU does not like git, but it certainly allows it's use—look here, for example) and not about CMake (gnutls still does not use CMake, after all). It's about strong-minded persons, yes, but not about developers:

(a) I felt particularly frustrated when FSF (when gnutls started around 2000) was insisting the transfer of the copyright to it, even though I had decided to transfer the copyright to FSFE (this is a very old issue but it had great influence on me as I realized that the transfer of rights was not simply for protection against copyright violations).
(b) The feeling of participation in the GNU project is very low, as even expressing a different opinion in the internal mailing lists is hard if not impossible.
(c) There is no process for decision making or transparency in GNU. The only existing process I saw is "Stallman said so" (this may not be bad, unless some threshold of disagreement has been reached - but then it is fair to be able to disassociate myself from the project).

But the practical trigger which driven the exodus was GPLv3 relicensing: one of the very first things GnuTLS project did after exodus was switch back to LGPL2.1+ for libraries (programs were kept under GPLv2 because they are not embedded in other programs). Which is obviously an anathema for “free software soldiers” who insist that you should use GPL for libraries, not LGPL (and as we've seen before GPLv3, not GPLv2).

You have a fair argument about many choosing GPLv2+ over GPLv3+, but that has probably everything to to with the fact that v2 and v3 are incompatible. It dos *not* mean that those same developers are less freedom loving.

Yes, it does. Or rather perhaps that they have different definition of freedom and said definition may not have complete elimination of proprietary software as a natural goal.

Now you saw documentation that Debian sees a strong growth for copy-left (even exponentially?), and your answer is that Debian is not where the action is at.

It's one explanation, but there are another one. Thanks for reminding me. As you've noted most people really lack the knowledge which means that if they want to pick GPL for a new project they most likely will pick GPLv3 simply because it's the “latest and greatest one”. What we should really look to see if people support “free software camp” or “open source camp” is number of projects which used GPLv2+ and switched to GPLv3+. Out of large, high-profile projects only GNU projects and Samba did such switch, but it'll be interesting to see how many smaller projects did that.

Both Qt and Mono uses copy-left licensing.

Which is important for them because they sell Qt and Mono under proprietary licenses, too. Which is far from “free software camp”'s idea that “proprietary software must die” as one can imagine.

On graphical drivers we already see clear signs (lack op copy-left there has little to do with lack of freedom love.

WTH does “freedom love” has to do with anything? “Free Software” is about quite particular freedom: “freedom from proprietary software”. One may argue that it's similar to infamous Apple's freedom from porn, but it does not change the fact that it's central to the idea of “Free Software”. Sometimes guys from “free software camp” are doing tactical concessions and even offer their software under quite liberal licenses (hey, RMS himself wrote It is wise to make some of the Ogg Vorbis code available for use in proprietary software, so that commercial companies doing proprietary software will use it, and help Vorbis succeed in competition with other formats that would be restricted against our use.) but all these things are justified as tactical movements with the end goal still the same: complete elimination of proprietary software. Somehow I don't see Luc Verhagen and Rob Clark as guys who will advocate this goal.

Where are the RMSs of the world?

Posted Sep 30, 2013 8:03 UTC (Mon) by pbonzini (subscriber, #60935) [Link]

> But the practical trigger which driven the exodus was GPLv3 relicensing: one of the very first things GnuTLS project did after exodus was switch back to LGPL2.1+ for libraries (programs were kept under GPLv2 because they are not embedded in other programs). Which is obviously an anathema for “free software soldiers” who insist that you should use GPL for libraries, not LGPL (and as we've seen before GPLv3, not GPLv2).

Actually you're wrong. The relicensing of GnuTLS was from LGPLv3+ to LGPLv2.1+, not from GPL. The reason is the incompatibility between LGPLv3+ and GPLv2. It would have been possible, alternatively, to use a dual-licensing LGPLv3+/GPLv2-only. But it's somewhat more complicated and gives little or no gain.

"Why your next library should be GPL" is an rms essay but FSF practice doesn't necessarily follow that.

There is plenty of "flagship" GNU software that hasn't moved to LGPLv3, for example GTK+. Note that even for non-FSF-assigned software, a single v3+ contribution could be enough to effectively prevent the software from being distributed under GPLv2.

Where are the RMSs of the world?

Posted Sep 30, 2013 15:48 UTC (Mon) by nix (subscriber, #2304) [Link]

Actually you're wrong. The relicensing of GnuTLS was from LGPLv3+ to LGPLv2.1+, not from GPL.

Quite. I can't help noticing that for all khim's very very long replies on this thread, he(?) was apparently unable to just download a pre-GNU-departure copy of GnuTLS and look at its COPYING files (in particular noting that there was still a COPYING.LESSER there).

Where are the RMSs of the world?

Posted Sep 30, 2013 15:10 UTC (Mon) by Del- (guest, #72641) [Link]

khim, I believe we have just about exhausted the capabilities of this forum with regards to long posts/threads. It will be too much to address all your points, even though there is a lot to address, maybe another time.

First up, I like to put some documentation behind statements. GPLv3 has it's problems, that we can both agree too. However, your comment that GNU and Samba are the only high profile projects using it is at best misleading. In total GPLv3 is just behind Apache in popularity:
http://www.blackducksoftware.com/resources/data/top-20-op...
which basically places it on a split second place behind GPLv2.

It does not make much sense for us to discuss other people's opinions. Still, I believe all three we have mentioned, Linus Torvalds, Luc Verhagen and Rob Clark may very well share my opinion that proprietary drivers are inherently a bad idea. Personally, I think proprietary development in general is a bad idea, so that puts me in the GNU camp I guess. Can't speak for others there. I am not sure exactly where to put you though, you seem to be more interested in blowing up the indifferences found among strong minded open developers.

Where are the RMSs of the world?

Posted Sep 29, 2013 17:50 UTC (Sun) by nix (subscriber, #2304) [Link]

I don't like to talk about that when we are celebrating anniversary

Why not, these days you talk about nothing else. (Seriously: every single article with anything even tangentially related to GNU gets this in it. It's getting a little tiring.)

Where are the RMSs of the world?

Posted Sep 30, 2013 0:22 UTC (Mon) by rsidd (subscriber, #2582) [Link]

As the article you linked clearly says, the difference between "free software" and "open source" is philosophical, not legal. Every OSI-approved licence that I know of is also FSF-approved (please point to any counterexamples). So it makes no sense to say a project 'left' free software for open source,

Where are the RMSs of the world?

Posted Sep 30, 2013 2:01 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link]

It is *primarily* philosophical but there are institutional differences between OSI and FSF as well that has resulted in atleast one license approved by OSI but declared as non-free by FSF

https://www.gnu.org/philosophy/historical-apsl.html

Where are the RMSs of the world?

Posted Oct 4, 2013 7:48 UTC (Fri) by rfontana (subscriber, #52677) [Link]

It is (at least today) a mistake IMO to equate 'free software vs. open source' to 'FSF vs OSI'. I am seeing that quite a bit in various contexts.

Regarding the differential between FSF-free and OSI-approved, the examples I'm aware of involved licenses that were approved by the OSI but considered nonfree by the FSF. It is possible that those OSI-approved licenses, if reviewed by the OSI today, would not be judged consistent with the Open Source Definition. I believe history may explain this difference: at a certain moment in the past, the OSI seemed eager to approve certain licenses then being manufactured by the business world, while the FSF was eager to demonstrate its willingess to reject such licenses on articulated principle. I suppose that may be what you mean by 'institutional differences', but I would just emphasize the historical element. The two organizations were, to some degree, in competition with one another in that time period (I would not say they are today).

Disclaimer: FSF is a former client and I am currently an OSI board member, but I am just speaking for myself here.

free software and open source

Posted Sep 30, 2013 7:53 UTC (Mon) by coriordan (guest, #7544) [Link]

The differences in licences is trivial. IIRC there was one licence approved by FSF and rejected by OSI, and two approved by OSI but not FSF, but none of those licences had widespread use, so that can generally be ignored.

The bigger difference is that software on a locked, tivoised, or DRM-using device cannot be considered free software, but it can still be open source. Free software is about users really having freedom. Open source is about the developers favouring one development model over another.

Where are the RMSs of the world?

Posted Sep 27, 2013 17:25 UTC (Fri) by coriordan (guest, #7544) [Link]

Where are the RMSs? Yeh, I ask myself this question too from time to time.

Clearly, there's no one that has all his attributes. Nobody has that same capacity to influence or inspire people to take a stand for a software cause. Others can give speeches that are equally interesting, but I don't think their speeches have the same long term impact.

OSS will never have a manifesto because it's only about making software development efficient. An improvement in efficiency can be great (or terrible; example: better DRM) but it's not part of what makes humans human, or what makes societies societies. Freedom, on the other hand, is a necessity of humans and societies, so it's not surprising that some people make big, long term sacrifices and struggles for freedom in the domain of computing.

With computing being increasingly interwoven into our daily lives, it's possible that the cause of computing freedom has grown too broad for one person. RMS is doing a good job of tackling new problems, but there are limits to what he can do. I mean, he calls for a boycott of Facebook, and I agree that that's a great idea, but it's not enough. The campaign against Facebook is still waiting for its full-time RMS. Same for privacy. Same for culture. (In the mean time, our RMS's comments on these issues are good stopgap advice.)

Where are the RMSs of the world?

Posted Sep 29, 2013 22:13 UTC (Sun) by jebba (✭ supporter ✭, #4439) [Link]

Jacob Appelbaum. See his LCA keynote.

Where are the RMSs of the world?

Posted Sep 30, 2013 15:27 UTC (Mon) by njwhite (subscriber, #51848) [Link]

I would agree with this. I'd also add Benjamin Mako Hill. They're both fantastic at explaining what software freedom means and is for, in a way that inspires action.

Eben Moglen, while he doesn't at all qualify as a newer generation, is a similarly great speaker.

Being an RMS

Posted Oct 5, 2013 15:10 UTC (Sat) by coriordan (guest, #7544) [Link]

But will Appelbaum or Mako Hill take this skill and dedicate their lives to creating a movement?

The problem is, it takes more than charismatic speeches to be an RMS.

These people can entertain, but will they do the thankless work of asking people to change their values?

(I've never heard either of their presentations, so these are general questions rather than being specifically about those two guys.)

Being an RMS

Posted Oct 5, 2013 16:52 UTC (Sat) by jebba (✭ supporter ✭, #4439) [Link]

Applebaum is the key developer of tor. That shows a lot more moxy/dedication than just being entertaining. Here's his his LCA presentation. Highly recommended (note this was in December of 2012, so it is before the avalanche):
http://www.youtube.com/watch?v=GMN2360LM_U

I don't know much about Mako Hill other than seeing his name a lot over a long number of years.

Being an RMS

Posted Oct 5, 2013 17:08 UTC (Sat) by coriordan (guest, #7544) [Link]

Thanks for the link.

I didn't mean to belittle anyone by calling them entertaining. Good public speaking is a real skill, and very necessary for a leader (it shouldn't be, but it is). Developing Tor is great. But RMS managed to (learn to) give good speeches, and lead the GNU software project, and then also created a movement. I don't know exactly what causes that latter part to happen, but that's the part I don't see yet in the rest of our community.

Lessig kinda made a movement, with Creative Commons. Unfortunately, he never set a goal of giving everyone rights or telling people to demand rights. He told copyright holders that they can grant rights if they wish.

I could make a long list of people who have made a success of important projects, and who work selflessly, etc. but if RMS retired tomorrow, I don't see who could step into his role - or even a significant sub-part of his role.

Being an RMS

Posted Oct 5, 2013 17:36 UTC (Sat) by jebba (✭ supporter ✭, #4439) [Link]

"The weekend culminated with an address by GNU founder Richard Stallman, in which he announced that privacy and security--especially against government surveillance--is now one of the GNU's project's primary purposes. Free software is one of the most important tools we have to protect ourselves against surveillance, and the address reflected an ongoing mobilization in the community to meet these challenges." [1]

Tor is reportedly one of the few truly surveillance resistant programs in the world. Imagine the daily grief Applebaum goes through just to write his free code (heavily watched, harassed, detained at borders, property seized, threats, threats to friends/family, etc.). He certainly has his fans. If it turns into a movement like RMS wants, Applebaum will likely play a large role--the movement doesn't need "one leader" either.

If freedom from surveillance doesn't become a movement...well, that would be a sad panda time for all.

[1] https://www.fsf.org/blogs/community/global-celebration-fo...