Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I want to hear more about that. I think the Apache license has been incredibly successful. Sure certain companies have taken httpd or other products and rebranded and tried to sell them but they never approach the popularity of the community supported branch - at least not for any product I can think of. I also agree the GPL has been very successful and neither provide a counter-example that disproves the benefits offered by both models.


It's nearly impossible to imagine there would be any reprogrammable SoC-based consumer devices had the Linux kernel not been GPL. It's not a pretty world as-is, but at least vendors have an imperative to dump "source" out there somewhere There's a built-in "shame" mechanism there where vendors want to at least look like they're doing the right thing by upstreaming their drivers.

An Apache-licensed kernel would appear in phones as a useless blob.


It is not nearly as pervasive but there are BSDs used in embedded devices (or more commonly, IT "appliances"). Neither of you have offered an actual counter example - such as a BSD/Apache/MIT licensed software that was ruthlessly exploited or that died off because no one would share their code. In fact my experience has been that communities using those licenses are also very open to sharing and they are thriving in that environment.


Good grief. Which have not been exploited? All unix derivatives came from the same set of (originally open) base code, they shared nothing. They all died (to first approximation). All TCP implementations in the early days were copies of the BSD code. They shared nothing. Does Apple contribute back to Mach or BSD the stuff they do with Darwin for iOS? Does Microsoft publish their modifications to their BSD-derived networking stack?

Can I get a buildable kernel tree for my Android phone? Yup.


The simple answer is that software has social value, and that it ethically ought be modifiable by its users.

Two key things have driven my desires for a GPL world. (1) Reading about historical software now unavailable and (2) working in the corporate world as a tools developer.

The question is pretty much, are you weighing the distributor's freedom more than the user's freedom? When you want someone to be able to use your program and not publish their fixes, MIT-style licenses become more your style. This is preferred by corporations, AFAICT. However, when you want the end user to be able to modify software and to be able to trust the software, you start wanting the GPL world. Of course, the idea of a typical non-technical user doing technical work is illogical. But with the GPL, they have the freedom to hire people to do technical work. That can be really important for things like crypto!

Another way to consider the situation is that the GPL enables viral transmission, and more broadly, forcibly enables a survival of the fittest. If software is really good and GPL, it'll get used, and the users must publish to their users, and so forth. So good software can now survive longer, perhaps after its starting point's death. This in particular struck home with me when I was reading of the software systems from the '50s and '60s. They were doing things then that were not to be done for decades. As a social value, some of those old software systems would have been tremendous. I'd like to build on old & reliable stuff, not reinvent the wheel!

A concrete way to approach this idea is to consider infrastructure tooling. I write that, every day. It's my job. Sometimes I ask myself: Don't a hundred other companies already do this? What if software licenses were so written as to require these tools to also be published? We've been able to publish one Mercurial extension, and it has been a great experience. But what if we did that with everything? I wish all the companies that developed infrastructure opened their source and worked with each other to build a common world of amazing. The GPL style license is key, because it requires this republishing thing. Under the GPL, sharing becomes required. And then someone else can build on that, and they must also share their additions.

That's the kind of world I want to live in. Not a world where we guard our little towers replicating each other's work time after time. We can go do different things, safe in the knowledge that in the common areas, improvements will be published and worked on by all.

edit: With respect to RMS-

In particular, his short story Right to Read really hit home with me when I read it in 2008 or so. I have read one or two other things by him from the... 90s I think... that really indicated he understands the logical outworkings of certain processes, as his predictions came to pass in the late '00s. I figure if you're going to agree with someone, you had best agree with someone who seems to be proven right. :-)


I'm familiar with the ideology, I was hoping for some relevant counter-examples related to "less free" licenses and issues that has caused.


This doesn't qualify as issues caused by less free licenses, but because of strong copy-left (GPL in this case), I have access to the source code for devices such as my television. The point in a GPL heavy world would be that it is cheaper for companies to release their software under GPL because of the network of other GPL software they can then use. This likely won't help (much) with software companies, but for many companies the software is just a component to make their product work.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: