Slashdot Mirror
RMS Objects To Support For LLVM's Debugger In GNU Emacs's Gud.el
An anonymous reader writes with the news that Richard Stallman is upset over the prospect of GNU Emacs's Grand Unified Debugger (Gud.el) supporting LLVM's LLDB debugger.
Stallman says it looks like there is a systematic effort to attack GNU packages and calls for the GNU Project to respond strategically. He wrote his concerns to the mailing list after a patch emerged that would optionally support LLDB alongside GDB as an alternative debugger for Emacs. Other Emacs developers discounted RMS' claims by saying Emacs supports Windows and OS X, so why not support a BSD-licensed compiler/debugger? The Emacs maintainer has called the statements irrelevant and won't affect their decision to merge the LLDB support.
... especially when someone acts freely and in a way you object to.
It also allows programmers to make money off their work. You know, "evil deceptive" poor people and middle class people.
It is more resentment as BSD is "actually" open as opposed to the handcuffed license he wants to impose on people. Will take BSD style licensing any day of the week over proprietary or GPL
There's a little more than is being reported. Here's some other RMS lines in the same thread:
First we have:
"More precisely, Apple intends LLVM and Clang to make GCC cease to be a
signal success and a reason for all sorts of companies to work on a
compiler that always gives users freedom. That would be a victory for
Apple and a defeat for freedom.
I don't know what LLDB is, or what it might do. I am going to find
out."
That's a little bit paranoid, but it is still a cautious statement.
Then:
"This question is a small part of a big issue which is more or less bad.
I want to find out what it is, and think about it. Please do not ask
me to rush to a conclusion without finding out what is happening."
Again, in all of his posts he mentions wanting to discuss it a bit more. RMS is pretty incendiary, eccentric, and often does or says crazy shit but... in this case it sounds like he said something alarmist to get attention and try to get some discussion, without stamping his foot down or flipping his shit. That he's being selectively quoted to make news is bad juju.
Some of us prefer others to voluntarily give back rather than be forced to.
Actually "users" don't touch source code, they might hire a programmer to do that, but then that's another developer.
GNU makes the right of "information" higher than that of people, ever.
Take your time. But whatever you find out is irrelevant to whether or not the Emacs maintainer will accept LLVM support into gud.el, at least as long as I'm the maintainer.
I think I like him.
#DeleteChrome
Say what? I disagree, but at least your rant made sense, right up unti you said:
>So yes, people who need to work for a living will prefer a BSD license over a GNU one.
BSD is only a hairs-breadth removed from public domain - it gives away pretty much all the rights that can be given, unlike GPL which retains many rights in order to impose reciprical giving on downstream developers.
I can only assume that by "people who need to work for a living" you are refering not to the people that did the actual work to create the BSD code, but rather to the exploitative sorts who happily harvest their code to incorporate into proprietary software without giving anything back.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
The most important tenet of GNU General Public License is that anyone who distributes a derived program is obligated to reciprocate by sharing the modified source. This is the "freedom" when RMS talks about "free software." Many other open source licenses such as BSD, MIT, and Apache concern more about attribution and no reciprocation, which more and more people seem to embrace instead. Many companies have a policy to use GPL code only in very specific cases and strongly forbids Affero GPL. If you are the author of some open source project and you want more people to use your code and make you famous, you'd care more about attribution and less about reciprocation. That's where GPL is losing ground.
I think RMS underestimates that many people are more than willing to exchange someone else's freedom for one's own fame. And famous projects tend to attract more contributors. I think RMS also overestimates that the proprietary code written by some company are worth contributing back to open source while most of them are garbage. Once he realizes his misunderstanding of people's motivation, he'd become less coercive.
I once had a signature.
Well, it is, but it does sod all to protect that openness, so BSDed software often ends up less open by the time you actually get a copy of it.
The only stuff the GPL doesn't let you do is remove other people's freedom. That should never be a problem unless you were planning to do that in the first place.
Ah, Open Source infighting: "We're not the People's Judean Front! We're the Popular Front for Judea. The People's Front is over there".
the issue is, can people make money selling software? You know, contributing to their own survival and success. Both for individuals and companies. RMS doesn't care about that.
Lets see what my grandmother wants while we're at. Her opinion is just as valid.
If your grandmother is a developer producing useful software available at no cost, used by millions all around the world, and both articulates a philosophy and draws up a license facilitating useful systems such as every Linux distribution, starts a foundation known around the world to advocate said philosophy and host said software, and encouages many people (even those who do not agree) to think about and discuss such matters ... then yes at that point I will begin to care about her opinion on this subject.
If you don't like RMS that's fine, if you think he's completely wrong that's fine too, but to dismiss his views the way you are doing is weak, cheap, and shows that you lack the emotional maturity to separate your personal feelings from the actual subject at hand. I hope that pointing this out will be useful to someone else, because as for you, I doubt I could reason with you in an adult manner. The really annoying part is: so many people are like this that they think it's normal.
It is more resentment as BSD is "actually" open as opposed to the handcuffed license he wants to impose on people.
This is an argument that can go on forever.
You could say GPL is more free, because it leaves you free to dual-license the product. People who want it free can get it free, people who want to restrict other people's freedoms can get a proprietary license. As far as freedom goes, it is just as free as BSD (but not as free in money).
You can't have a dual license in any meaningful way with BSD.
Ultimately the difference between the two licenses is this:
If you want to have your software used as much as possible, even though you might not see the uses, then use BSD.
If you want to have people (and corporations) give something back in exchange for their use of the software, then go with GPL (or GPL + proprietary option).
It's pointless to argue about 'freedom' and 'which is more free' because the arguments don't relate to why people use those licenses.
"First they came for the slanderers and i said nothing."
If you apply handcuffs to someone you can't then claim they are free. you say freedom is a matter of self control, this is true, UNLESS you are under the GPL where freedom is dictated to you under strict terms.
Well, it is, but it does sod all to protect that openness, so BSDed software often ends up less open by the time you actually get a copy of it.
Are you guys still spewing this BS meme? Letting someone use open code in a closed project does "sod all" to close off the original code.
that says you must share *all* of your code simply because it uses some LGPL library somewhere.
If it's full GPL you do, if it's LGPL you don't.
That's why everything useful is LGPL instead of GPL.
You could argue that LGPL isn't compatible with GPL and shouldn't be included in Linux >.>
There is a move toward BSD for a number of reasons. It is difficult to comply with the
new GPL on some modern platforms where bits of this and that are closed. Many of
the new SOC devices have NDA or binary blob support for this and that.
The new compiler has already kicked the GCC folk in the logjam of bugs and features.
As long as the bits packaged with GCC are GPL there is no big issue.
The Apple example is interesting. They have used LLVM and kit on some projects
but recently have given their internal changes back to the LLVM community. Not because
they had to but because keeping up with a wider and wider collection of diffs is harder
and harder to do.
RMS is correct to be concerned and the selection of hardware and software
needs to consider licence as well as long term maintenance of their product.
With the internet of things coming there will be litigation should a company
fail to maintain a bug fix stream during the reasonable life expectancy of the device.
Some cell phones distribute the blame... Google, Samsung, AT&T where each
is a congested pile of congesting making me feel samstung. I now only buy
half price unlocked refurbished hardware with removable batteries when a device
get too long in the tooth. My current phone was purchased because of tower investment
and no more. They are no longer updating it so I am looking hard at rooting it
for a list of missing features.
Truth is stranger than fiction, but it is because Fiction is obliged to stick to possibilities; Truth isn't. Mark Twain.
Well, of course you can't re-license GPL code. Who cares? You didn't write it. The only requirement is that you provide the source for the GPL code you use. Any modifications to GPL code result in more GPL code; interfacing with GPL code does not, unless you want it to, in which case it's a moot point anyway. You are free to write an application using GLIBC, the GNU C Library, licensed under the GPL, and not release the source for your application. You do have to provide the source for GLIBC, as per the terms of the GPL, but nothing requires you to share *your* code. In any version of the GPL.
APK quotes people (including myself) without context and should not be trusted. Just thought you should know.
RMS' philosophy is that the code should be free so you can do whatever you want with it. Unless you do something he doesn't like, in which case, he throws a tantrum. This is about making free software compatible with more free software, but since he doesn't like the licence on the other free software, he wants this restricted.
in a separate process that you don't link to, in a separate computer, and send it into space so it's not on the same planet.
No wonder you've never heard of that, you're too stupid to understand abstraction.
RMS hasn't been an active developer in years by his own admission. His role is largely advocacy and philosophy, and that appears to be the sole issue here. However, he doesn't seem, based on a reading of the thread, to have any formal ability to block the patch.
You can never go home again... but I guess you can shop there.
He has the freedom to throw a tantrum. You, and everyone else, also have the freedom to distribute a version of Emacs with LLVM support.
Read the GNU Manifesto sometime, would you?
That is what Copyleft is. That is what GPL is supposed to be: using the copyright laws that were designed to protect proprietary interests in away that instead protects Software Freedom, that enforces Software Freedom.
Just because you refuse to understand the terms and arguments doesn't mean you've uncovered some hidden truth or something. You don't like Software Freedom. You find enforced Freedom too restrictive. You want to choose to be free, or not to be free. That is fine.
People probably mod you down because you pretend that people with a different view must just be stupid, or something. These are different choices based on different values, there is no utility in complaining about other people's license choices.
You know best what license to use for software you write, I know best what license to use for software I write, and RMS knows best what license to use for software that the FSF writes. This is all as it should be.
If you can't link GPL code at work, that is because of choices your boss made, not because of choices that RMS made or some implied deficiency in the GPL. Remember, people who choose the GPL want to be protected from your boss. People who don't share the values of the GPL are excluded for real reasons. You don't have to agree with those reasons or share their values to recognize that they have reasons that are based on their values, and they have every right to license their software in the way that they do. And you should be aware most of them are getting paid to write their code, most GPL code is written by paid programmers. Paid by companies. For-profit companies. With bosses who choose GPL. For business reasons. That doesn't make them less Free.
It's a discussion you retard. He's arguing there are negative effects of supporting LLDB, other people are agreeing with him and others are arguing that isn't the case. He isn't censoring any of them, he isn't "throwing a tantrum", they're just having a discussion.
People like you are the reason these discussions should be closed instead of public. You make it public and every idiot with little grasp on the arguments will move a mountain to prevent the project from advancing an inch.
You're funny. You really are. Did you mean to be?
EMACS is the whole reason the GPL exists to begin with. Some jackass hijacked it into a commercial project and RMS's contributors were pissed. They went to HIM with the torches and the pitchforks demanding an explanation.
The copyright holder of EMACS will most certaily sue your sorry ass for such a stunt.
As much as some people (even RMS) might want to portrary the GPL as some sort of communist plot, it really isn't. It's just a way to keep contributors happy. It turns out that most people don't want their charity abused by some crass corporation.
It also has the nice side effect that it allows for those same crass corporations to collaborate with some level of trust.
A Pirate and a Puritan look the same on a balance sheet.
No, it's not throwing tantrum. RMS has a philosophy that users of software should have certain freedoms / rights (use, study & modify, redistribute, distribute). That's the gist of GPL and why he founded GNU. BSD-style license does not guarantee these freedoms, and Stallman sees wider adoption of projects using those licenses as a threat to free software. I do work on BSD-licensed projects, but I certainly do share his fear that this poses serious threat to free software in the long run.
I don't think it's a conspiracy or somehow widely orchestrated effort - more likely it's simply easier not to guarantee those rights and thus more attractive for commercial companies (participating in those projects), but I believe the threat to the freedoms is real.
Pay the licenses, or out-compete them with freely available code. There's considerable experience by now that says the latter will win every time.
Disinfect the GNU General Public Virus!
RMS has a philosophy that users of software should have certain freedoms / rights (use, study & modify, redistribute, distribute). That's the gist of GPL and why he founded GNU. BSD-style license does not guarantee these freedoms more likely it's simply easier not to guarantee those rights ...
What rights do BSD contributors lose? All the community code exists, the community can continue without the commercial changes, the community is not required to use some commercial fork. They lose nothing if some contributor chooses not to give back. Furthermore, users of GPL'd code decide not to give back at times too. They can use some a commercial fork internally and benefit from community work and not give back. Also, various commercial users of BSD code have a pretty good track record of contributing back.
What rights do BSD users lose? **IF** they care about "free software" or access to the source code they can just avoid commercial/closed forks and stick to the community based code.
The GPL does *not* offer greater freedom, it creates restrictions to force behaviors it believes benevolent. Forced benevolence may or may not be a good thing but it is not freedom.
... larger corporations will build non-free proprietary improvements on BSD licensed code without contributing back, and to continue to be productive as a programmer you will be forced to pay for the licenses on their proprietary tools ...
Really? How did they degrade your performance? You have access to and may enhance and contribute to the exact same source code they did. You lost *nothing*, not one line of code, not one opportunity to add a new line of code.
And if your code was GPL based a corporation may do the exact same thing. They may fork and enhance a GPL based tool for internal use only and not share. They can continue to benefit and merge all your work and the rest of the communities work as well.
BSD-style license does not guarantee these freedoms, and Stallman sees wider adoption of projects using those licenses as a threat to free software.
But why is that the case? I know there is the contrived case of a codebase being improved and re-packaged under a proprietary license but that just doesn't happen, part of the reason is the original codebase is still there and other people can still use and improve upon it. The most popular web server in the world is licensed under this model and it hasn't happened there.
For someone so convinced that his way is the best way, he's sure paranoid that some other way is going to take over.
RMS did the very same thing to GNUstep. GNUstep currently supports both GCC and LLVM/Clang. The project does this for good reason: because Objective-C is better supported in clang than it is in gcc. GCC doesn't even consider ObjC as a release critical compiler and LLVM/Clang looks on it as central. Additionally clang supports many modern features of ObjC that gcc lacks and shows no signs of ever attaining.
RMS specifically indicated that supporting LLVM/Clang by mentioning it on our wiki page (http://wiki.gnustep.org/index.php/ObjC2_FAQ) was harming the GNU project in an important place. Our response was swift and unanimous against remove it since all we are doing is providing user choice and, given that GCC is inferior to LLVM/Clang for ObjC, we MUST support LLVM/Clang. To date we have gotten no response from RMS.
I think it's grossly unfair of RMS to request this. By supporting Clang and LLVM and LLDB we are not impacting user freedom. All we are doing is offering users a choice which, last time I checked was completely okay. What we have here is a problem where RMS sees his role in the FLOSS community diminishing because someone has come up with a faster, more useful and better support compiler.
If anyone has damaged the FSF it is not the folks at Clang/LLVM it is RMS and the FSF itself. They have systematically impacted developer freedom by doing the following to GCC:
https://gcc.gnu.org/ml/gcc/2005-01/msg00008.html
"One of our main goals for GCC is to prevent any parts of it from being
used together with non-free software. Thus, we have deliberately
avoided many things that might possibly have the effect of
facilitating such usage, even if that consequence wasn't a certainty.
We're looking for new methods now to try to prevent this, and the outcome
of this search would be very important in our decision of what to do." -- RMS
This is terrible! Why would you do this?! RMS is trying to achieve through technical means what proprietary software companies try to do via copyright and IP law.
RMS is risking an all out rebellion of pretty much all of the FSF/GNU projects if he keeps this up. My advice to the FSF and to RMS is to allow developer freedom and stop viewing LLVM/Clang as a threat or a setback for it is neither.
GC
Gregory Casamento
## Chief Maintainer for GNUstep
BSD-style license does not guarantee these freedoms, and Stallman sees wider adoption of projects using those licenses as a threat to free software.
But why is that the case? I know there is the contrived case of a codebase being improved and re-packaged under a proprietary license but that just doesn't happen, part of the reason is the original codebase is still there and other people can still use and improve upon it. The most popular web server in the world is licensed under this model and it hasn't happened there.
The standard example is Unix. Each computer manufacturer ported it to his computers, then improved it to make his product (hardware plus software) more appealing in the marketplace. The improvements weren't shared, so there was fragmentation: applications would run on some Unix systems but not others. POSIX and Single Unix System were attempts to fix the problem by standardizing certain parts of the user-mode API, but they weren't enough.
The problem was finally solved by a clean-room reimplementation of the utilities (GNU) and the kernel (Linux). Both are available under the GPL, which requires improvements to code to be released in source if the binary is distributed.
GPLv3 started because RMS saw that companies were using the GPL in a manner that was compliant to the letter but not to the spirit. Back then, the GNU haters laughed at him, as usual, because "who would want to run code on a set-top box". Nowadays, the vast majority of the end-user devices are tivoized (Android, Apple, Microsoft, ...), and users can't do anything with the code that runs on them, including fixing security bugs and auditing it to find out what it does with all their personal data, let alone (God forbid!) run their own programs on it. So the introduction of the GPLv3 wasn't a whim as you are implying, it was actually sensible and farsighted.
Also, I should point out that the LLVM/clang situation is a bit more complex. If I recall, LLVM came about because the gnu toolchain deliberately obfuscates it's output and interoperability interfaces with other tools even within the toolchain. This strategy was chosen because the outputs of the individual software tool in the toolchain were not, and could not be protected by the GPL (any version). It would have been possible for a proprietary product to be developed that didn't link to gcc (or another part of the toolchain) to take the useful output of gcc (e.g. a parsed abstract syntax tree) and use it to any number of cool things. The product could still be distributed with gcc (and the required accompanying notices) but the rest of the code would be locked up, because it doesn't link to gcc, only depends on it at runtime. This violates the spirit of the GPL which is not only to make software free, but to keep it free.
It's also not just proprietary software. For example, the Eclipse license is not GPL compatible, so even if GCC were cleanly structured you wouldn't be able to create a library incorporating GCC code and link it in to Eclipse. The same applies to anything Apache licensed. One of the benefits of LLVM is that you can use the code with projects of any open source license (except, apparently, emacs) and not have to worry about the incompatibility.
I am TheRaven on Soylent News
It didn't solve it, it just created yet another entry to the UNIX wars, it didn't supplant the major players BSD, Darwin, HP-UX, AIX, Solaris. The problem was that developers had many systems to target, Linux hasn't solved that, it is solved by having multi-platform frameworks and language standards.
And yet today GNU/Linux is the premier Unix-like system which everyone targets, after Microsoft and Apple. IBM actively supports it, even on their mainframes, though they haven't forgotten AIX. Dell advertises GNU/Linux on some of their offerings. Oracle offers a distribution of its own, though it hasn't forgotten Solaris. HP offers GNU/Linux on their Integrity servers along with HP-UX. Because of the GPL, GNU/Linux does not fragment, as Unix did.
He has the freedom to throw a tantrum. You, and everyone else, also have the freedom to distribute a version of Emacs with LLVM support.
... coming soon GPL v4.
Clause IV "...any code GPLv4 may not include, link, or run on any non GPLv4."
But serisouly GPLv3 started because of his tantrum with Tivio. It would not surprise me if he did a version 4 if clang takes over.
Tivo, not Tivio. And Tivoization is a real problem. It's frustrating enough that I'm locked out of modifying my own devices without risk of breaking them or the law. But it's even more frustrating when I have a device that purports to run free open source software that I can't modify. (I'm looking at you almost every Android device ever.) It's against the whole point of the GPL if I can't tweak the code to fix or improve my device because the manufacturer locks me out.
rms is often not the most mature orator, but he saw a problem, and he fixed it to the best of his abilities. Don't belittle GPLv3 by saying it started with a tantrum.
Freedom and power are not the same thing. When someone distributes a non-free derivative of a Free Software program, the distributors are exercising power over the users of that derivative.
Digital Citizen