Hacker News Reader

Comment by fhd2 | Hacker News Reader

Sounds like the open source vs free software debate all over again. I don't have a strong position in that one, but I find GCC's position rather obvious: They want to support free tools, and they explicitly don't want to support proprietary tools.

I cannot believe that this isn't obvious to esr, of all people. Is this just him trying to start a flame war?

edit: I'm now aware that esr is not talking about license restrictions but technical restrictions here. I have yet to find any evidence of technical restrictions for political reasons though, and it looks like the folks responding to him on the mailing list are not sure what he means either.

View on HN

GCC's position [is] rather obvious: They want to support free tools

I don't have a strong position either, but unless you work for the FSF in their PR department, software developers should really be stating the FSF's position accurately, which is:

The Free Software Foundation only want to promote free^H^H^H^HFree Software Foundation-licensed tools, primarily tools that use its GPL license.

To the FSF, "free" is just a shorthand way of saying "Free Software Foundation-licensed". It doesn't mean "free" as normally understood. Just because the FSF wants to conflate the two to obscure what's going on doesn't mean the rest of the world should go along with the ruse.

(And it is clearly non-standard, which is easily demonstrated by how often FSF people have to explain what "free" means. As they say in politics, if you're explaining, you're losing. If there was no difference from the standard usage, no explanation would be needed. Therefore, the FSF usage is non-standard. QED)

FWIW I doubt anyone has a problem with the FSF's actual mission, since people are free (normal usage) to do what they want (and even encouraged to do so). It's the rhetorical duplicity of their PR that we shouldn't be supporting. Let the FSF's mission stand on it's own merits, rather than by trying to gain credibility/respectability by association with something else (in this case, our pre-existing affinity for freedom).

> To the FSF, "free" is just a shorthand way of saying "Free Software Foundation-licensed".

Not entirely true; the FSF recognizes lots of other licenses as "free".

Its true that the FSF thinks that its licenses are the most appropriate for promoting software freedom.

I think they're wrong: if you can convince people that Free Software has value, you don't need a copyleft license forcing them to give back, and if you can't convince them of that, a copyleft license doesn't help you get them to create free software, it just prevents them from engaging with it at all, gets them to commit to an alternative, and makes them less likely to to commit back even if they later realize a value in Free (since if they commit to a non-Free third-party alternative, the cost of switching it out is higher.)

"It doesn't mean "free" as normally understood ... have to explain what "free" means. As they say in politics, if you're explaining, you're losing"

And so you are losing. You are trying to explain what "free" is, and QED, your usage is non-standard.

Yes, the result is likely just going to become a reignite the flame war for a post and here on HN.

GCC authors clearly want that users can modify and share the compiler, regardless which version of the program the user got. The idea that someone would then try to use their work as a springboard to sue individuals who attempt modify or share an specific version is clearly seen as abhorrent behavior.

Yet, again and again, people disrespect the wishes of the authors and keep try to pressure them to change their views. "Allow us to sue some users by using your work" they say. No they get back, and yet they keep asking again and again and again.

No. It's not that at all. The GCC/LLVM debate has nothing whatsoever to do with licensing, but crippling the design of gcc for political aims. Educate yourself on the subject.

I can't claim to know anything about GCC's internals, but from what I know/read, it does have a plugin system and I don't see any indication of that being held back for political reasons (there are many other potential reasons for technical inferiority).

And I don't see how the FSF could ever have furthered their political goals by making it technically hard to use GCC in free software tools, quite the contrary. Would be quite interested in evidence of that.

I think I get what you mean now, you said above that GCC doesn't want to give external tools access to intermediate formats. But is that really the "anti plugin policy" esr is talking about?

https://lists.gnu.org/archive/html/emacs-devel/2012-12/msg00...

RMS says "freedom" is worth sacrificing functionality for. GCC is a monolithic mess of shit that is virtually impossible to debug, and produces incorrect code very frequently. Being forced into a terrible architecture where everything is smushed into everything else just to prevent people from being able to use gcc as a front end and something else as a backend is absolutely a case of making the code worse for the sake of trying to further a political agenda.

   GCC is a monolithic mess of shit that is virtually impossible to debug

This is the “RMS loophole” in the GPL. In principle you have the right to modify GCC to suit yourself; in practice, the barrier to entry is too high.

Please take your FUD and go elsewhere.

Educate yourself before questioning others level of understanding.

The GCC has an plugin system, but not one that allows proprietary modules. The people that care about that distinction only do so because of the licensing difference between GPL and proprietary licensing.

Reply on news.ycombinator.com