Hacker News Reader

Comment by rectang | Hacker News Reader

There's a long tradition of licenses with a rebellious spirit, either deliberately or incidentally drafted in such a way that makes them hard to interpret. Before this "GLWTPL" there was the WTFPL, the JSON license, etc.

People who choose such a license should be aware that they are making downstream adoption of their product more difficult. (For some authors that is surely the point.)

View on HN

Our company has, in its policy, the allowance of licenses that fall into the OSI spec. This license, along with the WTFPL, are patently denied due to their joke licenses.

Regardless of how amazing your code is, you look like a kid when you need to tag your code with one of these joke licenses. It loses a ton of credibility in the corporate world where open source is taken very seriously.

On the Internet, nobody knows you're a kid. The legality of the license and the quality of the code matters, not the childishness.

That's why I find the WTFPL a bit misleading, I tend to prefer the official french version of it (LPRAB -> http://sam.zoy.org/lprab/) ; the main point of this one is the "j’en ai RIEN À BRANLER." part, i.e, "I don't care at all".

It leaves little room to potential ambiguities regarding whether or not you should expect the author to care about concerns like, looking like a kid.

Oh no! Some people, serving a corporation, who would leverage someone's work in their business processes to a profit without compensating the author of the work, have an opinion, which they take very seriously, on the manner in which the author of the work should license the free work so that they can legally use that free work in the manner l that works best for them!?!?! :O

Ya'll get so lost in the process of exploitation that you forget you're actually exploiting people.

Perhaps they would do well to realize that if someone uses a "joke licence" they are implicitly stating that the code in question, according to their personal standards, probably isn't suitable for production environments. In some offices however, this same code might be viewed as worth trying to build a blockchain currency system with. Coughcough*

Programmers write things all the time that they wouldn't inflict on anyone else. They're written to get a one-time thing done and no more than that, however, they do sometimes realize that someone else might be facing that very same one-time problem, so there's no harm in at least letting other people play with said abomination.

The entire point of the license is to highlight how ridiculous it is that one should have to be/retain a lawyer in the process of generating sequences of 0's and 1's to run on a computer.

Programming is mathematics. You stop doing things like trying to 'license' 2+2 on x86, and we'll stop making goofy licenses to mess with you and give your lawyer a heart attack.

We all know if you see something you really like but don't like the licensing of, you'll just rewrite it differently, and slap All Rights Reserved on it anyway. Thus is the joy of the Turing Machine. What you do can be redone in an infinitude of different ways. You insisting on licensing schemes just holds back code from being utilized, and builds legal structures through which software writers and users can be abused.

I've recently thought of creating a TTCACOI (Take This Code And Choke On It) license as of late.

I mean, clearly if forced binding arbitration can force you to relinquish civil rights through Contract law, one should be able to compel auto-esophegeal blockage through licensing terms right?

These licences, if they have a point, fail big time on making them. There is a copyright law out there, and source code, being copyrighted product, needs to be licenced appropriately if you want to share it. If you want the copyright law changed or abandoned, which I personally do, well, inform people about it and do some activism maybe. These ridiculous "licences" do not inform people of anything, and also they obscure the reusability of the product.

In this case the point of the licence is that the original author knows they have done something terrible and/or inadvisable (probably out of desperation), but still... The code worked once and it might be useful to someone else in a bind, but the author absolutely does not want anything to do with whatever new problem it was leveraged against, or the fallout if this time it doesn't work.

A licence is not provided with free software in order to convince you that you should use that person's code. It's provided because some people are just shirty enough to steal other people's work, claim it for themselves, and then eventually sue the original author for copyright infringement.

How does do-whatever-you-want make adoption more difficult?

Serious answer: Most companies big enough to care about this sort of thing have a list of approved open source licenses, which their lawyers have read and decided they're ok with. Using a license not on that list means you need to talk to the lawyers. That usually takes a few days of back and forth emails.

As an example, WTFPL is on IBM's list of acceptable open-source licenses.

If the license isn't on a pre-approved list I have to send it to the in house legal team. Good Luck With That!!

Complying with the letter of an ambiguously drafted license is quite challenging, but it's important if you are an entity with a lot of assets and thus a plump target for a lawsuit.

The JSON license has made a lot of lawyers a lot of money.

The licenses aren't "do-whatever-you-want". They have jokes in them that are legally confounding.

Reply on news.ycombinator.com