LLVM's Clang Begins Better Supporting Musl Libc

[caligula]

Why would anyone want a lightweight libc? Libc is supposed to be heavy and a beast to work with. Glibc was so good when Drepper was the lead.

[paulpach]

MIT license. No thanks

The MIT License gives you too much freedom? You only want software that places more restrictions on you like the GPL?

Or perhaps you mean that you won't contribute to it because you want to put a little more restrictions on the software you make? If that's the case, it's simple: get any MIT code you want, add your own modifications and redistribute the whole thing under the GPL or whatever license you deem appropriate.

[cb88]

The MIT License gives you too much freedom? You only want software that places more restrictions on you like the GPL?

Or perhaps you mean that you won't contribute to it because you want to put a little more restrictions on the software you make? If that's the case, it's simple: get any MIT code you want, add your own modifications and redistribute the whole thing under the GPL or whatever license you deem appropriate.

Burn... In all seriousness though I've ran Alpine Linux and Void Linux both which run in quite low ammounts of ram even with heavy desktop environments... So another point might be that sometimes a license that "gets out of the way" leads to better productivity. It has been the case in the past were GPL code has become such a job to police the license for the developers quit developing and just sue all the time... and if you do that you suck just as much as the person violating the license IMO.

[Michael_S]

Burn... In all seriousness though I've ran Alpine Linux and Void Linux both which run in quite low ammounts of ram even with heavy desktop environments... So another point might be that sometimes a license that "gets out of the way" leads to better productivity. It has been the case in the past were GPL code has become such a job to police the license for the developers quit developing and just sue all the time... and if you do that you suck just as much as the person violating the license IMO.

I have nothing against the MIT license. Open is open. But I'd like a link to someone who wrote GPL code that switched to spending all of their time in lawsuits. That sounds like a wild story someone invented.

Further, the whole point of the GPL is to keep the code under the GPL. So releasing code under GPL and then suing violators is kind of the point. If you didn't plan to do that, you would have gone MIT/BSD/Apache in the first place.

[starshipeleven]

The MIT License gives you too much freedom?

The issue isn't my freedom, is everyone else's. There is a large movement that thinks that without some rules and enforcement stuff will just get used without contributing back.

If that's the case, it's simple: get any MIT code you want, add your own modifications and redistribute the whole thing under the GPL or whatever license you deem appropriate.

Does not solve the above issue.

Burn... In all seriousness though I've ran Alpine Linux and Void Linux both which run in quite low ammounts of ram even with heavy desktop environments... So another point might be that sometimes a license that "gets out of the way" leads to better productivity.

musl is a library designed for embedded devices, of course it uses less ram.

musl lacks various things though (systemd for example cannot run with musl), and glibc runs better performance-wise. http://www.etalabs.net/compare_libcs.html

It has been the case in the past were GPL code has become such a job to police the license for the developers quit developing and just sue all the time...

This is bullshit.

[cb88]

This is bullshit.

No it isn't.

https://en.wikipedia.org/wiki/BusyBox#GPL_lawsuits

Also, Rob Landley goes into detail about these issues and he's pretty sharp, he maintained busybox for awhile and also maintains the next generation of core utilities for android (toybox)

The problem with the GPL is that it expects someone to give you something for free... (development time). MIT on the other hand gives you something for free in hopes that you will give back for free as well seeing the clear mutual benefit. It also avoids spending time in court to the largest extent possible.

While you can build closed systems with MIT code... quite frankly it doesn't matter, any anyone that isn't a GPL zealot and says otherwise is a hypocrite.

[starshipeleven]

No it isn't.

That's just lawsuits over license violations. Only 50% of the assignment I gave you. Now show me where the developers quit developing as a result. busybox is still pretty much commonplace everywhere

and also maintains the next generation of core utilities for android (toybox)

which is practically unknown apart from that bunch of asses at Cyanogenmod that ship it instead of busybox and NO FUCKING APPLICATION RUNS WITH IT SO WHY GOD WHY!? so I need to install busybox manually anyway.

yes I read his articles, good read and nice analysis I agree with (android needs an integrated system to compile itself so it can become self-sustaining) but you cannot just u-turn and break compatibility with applications for lulz just because, while 99.999% of firmwares don't even ship with busybox anyway.

The problem with the GPL is that it expects someone to give you something for free... (development time).

Uh? GPL expects you to give back any modifications that has been made to the original code.

You take big-huge amount of development time from the community, you give back some little amount of development time to the community.

As long as you are using stuff as-is you don't need to payback in any form.

MIT on the other hand gives you something for free in hopes

... it does not end up taken and used by corporations that make tons of cash on it without contributing back a damn. It's already hard to get corporations to not steal GPL code, go figure getting corporations to contribute back to permissive-licensed code out of good will of their hearts.

While you can build closed systems with MIT code

Which means? closed source are better? worse? both? none?
You can sell opensource stuff too, most consumers will never know.

[cb88]

That's just lawsuits over license violations. Only 50% of the assignment I gave you. Now show me where the developers quit developing as a result. busybox is still pretty much commonplace everywhere

which is practically unknown apart from that bunch of asses at Cyanogenmod that ship it instead of busybox and NO FUCKING APPLICATION RUNS WITH IT SO WHY GOD WHY!? so I need to install busybox manually anyway.

yes I read his articles, good read and nice analysis I agree with (android needs an integrated system to compile itself so it can become self-sustaining) but you cannot just u-turn and break compatibility with applications for lulz just because, while 99.999% of firmwares don't even ship with busybox anyway.

Uh? GPL expects you to give back any modifications that has been made to the original code.

You take big-huge amount of development time from the community, you give back some little amount of development time to the community.

As long as you are using stuff as-is you don't need to payback in any form.

... it does not end up taken and used by corporations that make tons of cash on it without contributing back a damn. It's already hard to get corporations to not steal GPL code, go figure getting corporations to contribute back to permissive-licensed code out of good will of their hearts.

Which means? closed source are better? worse? both? none?
You can sell opensource stuff too, most consumers will never know.

Dude... take a chill pill.

[paulpach]

The issue isn't my freedom, is everyone else's. There is a large movement that thinks that without some rules and enforcement stuff will just get used without contributing back.

Ahhhh, so the issue is that you don't want competition from code that has less restrictions than your own.

[starshipeleven]

Dude... take a chill pill.

Pills bounce off my beard.