Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Do What the Fuck You Want to Public License (wikipedia.org)
38 points by frostnovazzz on May 19, 2013 | hide | past | favorite | 63 comments



It's a bad license for two reasons: One, it does not cover the case where it's not possible to place stuff in the public domain (CC0 does). Two, it's alienating some users unnecessarily. Licenses are practical, there's no reason to invent a new term for concepts that already have well-known identifiers. Put your jokes outside the license section.


(disclaimer: I wrote WTFPL v2)

I fart on your recommendation of the CC0.

Your first “reason” is totally bogus. There is no need to cover cases where it’s not possible to place stuff in the public domain, because there is no need to place stuff in the public domain. That’s a fabrication of the CC0 license and one of the reasons why the OSI does not recommend it. A license shouldn’t even mention public domain since it’s not a widely accepted concept.

Your second “reason” sounds like a matter of taste to me, especially since the CC0 goes to great lengths defining new terms such as “Commons” and “Affirmer”. If there’s a joke here, it’s the CC0, in needing more than a thousand words to basically say “do what the fuck you want”.

Moreover, here is a quote from the OSI about the CC0: “the Committee felt that approving such a license would set a dangerous precedent, and possibly even weaken patent infringement defenses available to users of software released under CC0”.

It’s also worth noting that the WTFPL predates the CC0 by nine years. Nine years to fill a non-existing void, and they couldn’t even do it properly.


Another reason not to use it would be pre-approved licences in corporations. I can use software released under many popular free licenses without issues. But something like this one forces a roundtrip through the legal department which takes days / weeks.

If you want your software to be business/corporate friendly, make sure you're using something popular (and preferably not mentioning patents or special linking clauses).


There are many other licenses that are similarly permissive, but that lawyers actually have vetted.

If you ever want interesting commercial adoption of software, you need to make sure the associated ip / software licenses are solid


Well, the FSF has approved the WTFPL as compatible with the GPL. [1] Certainly that doesn't mean it's been as vetted as the other licenses, but it counts for something IMO.

[1] - http://www.gnu.org/licenses/license-list.html#WTFPL


You could always just dual license. ;)


Exactly. I think of legalese as a computer language. Publishing your software globally means that your license document will be run through many sepatate jurisdictions. Each jurisdiction is akin to a unique runtime interpreter.

I prefer to stick with idiomatic legalese to avoid running into interpreter specific bugs :-)


I use MIT for this. Already well known, stablished, and has the same effect.

It is also less offensive to people who get offended by words.

That can be an advantage or a detriment; it depends on you.


Exactly, the MIT license is my favorite out of all of theme. Apache, BSD, CC, etc - needlessly complicated and full of lawyer-lingo. MIT speaks to me and my priorities.


This may be nitpicking, but it seems rather restrictive to force renaming of the license in the case of its modification. That is not very "do what the fuck you want to" ...


This sounds like a more care-free way to do the same thing as http://unlicense.org

Personally I'd choose unlicense instead of WTFPL.


First, unlicense is a bad name for a license that is so different from no license.

Second, the first sentence and the third paragraphs do not work in most of the world: you cannot put a work in the public domain, even in countries that "recognize copyright laws".


How is this different from an impolite MIT license?

Edit: also, for the interested, there is a free e-book from O'Reilly explaining the major free software licenses http://oreilly.com/openbook/osfreesoft/book/index.html


Thanks for the link. I remember how much trouble I had trying to pick a license when I first released code (even though it wasn't a big project or likely to be used by many people). I think I finally made my decision through research on StackOverflow and Wikipedia. This book would have been very useful.


the main difference is that this essentially places the work in the public domain, where the MIT license means it's freely available, but the copyright is still the creator's. Also, thou have to retain the MIT license and distribute it with all subsequent works.


I don't think this is true. MIT explicitly allows sublicensing, meaning you can essentially just change the license on any derivative. The copyright notice needs to remain, though.


OMWFL

Obey my will forever license. Copying or creating derivative works shall irrevocably enslave you to my will forever.


BeerPL

I'd like a beer.


So, if I can just do what the Fuck I want to, then can I download your code and then re-license it as proprietary and charge royalties to everyone else who is just doing what the fuck they want to with it? I mean - I'm doing what the fuck I want to soooooo......

This is probably a stretch or something, but really, my point is that this license is so incredibly open to interpretation that I get worried about using code licensed with it every time.


Of course you can. Companies use MIT/BSD/Apache/etc licensed work in proprietary systems all the time, that's what they are for. You're better off using one of those since they leave no room for confusion.


With anyone who uses your re-licensed version, yes. Not the people who aren't.


I tend to do things with less legalism, with the copyheart (which is not a licence)

http://copyheart.org/


So no one can safely use your code, and you can be sued because you don't disclaim liability properly. Great idea!


why should I have to 'disclaim liability'? Because our litigious society would like to take advantage me and sue my pants off for no real reason? Then the flaw is not in my code, is it?


Because those few lines of boilerplate you would have to add exist in order to make explicit your stance on issues you probably don't even know exist, and their inclusion would come at basically zero cost.

Programmers of all people should appreciate that the "plain English" explanation of something is often sorely lacking in comparison to a full and complete implementation.


well, I guess my answer is that copyheart is not for the faint of heart, but if you feel like standing up to our lawyerocracy, then you should do it. And you shouldn't have to worry about 'issues you didn't know exist'. There is this concept of 'mens rea', you can't be criminally guilty if you shouldn't have known it was a crime. As for civilly, anyone who sues you for something silly is a jerkface. Don't negotiate with terrorists.


Am I allowed to modify copyhearted material? Distribute modified versions? Perform them? Incorporate them into works that are copyrighted?


why are you asking for permission? You've already lost.

http://mimiandeunice.com/2011/06/23/allowed/


See also it romantic counterpart: https://github.com/pygy/The-Romantic-WTF-Public-License

/shameless plug


wow didn't know this was fsf approved.

the in-your-face-ness of the name makes it a great tool for talking about copyright with someone and making a point.


Hackers care too much about licenses...


Not really. I mean, yes, more than the average person, but that's because code licensing helps make the world go 'round, and means we don't have to reinvent the wheel every time we want to do something.

It's mostly because modern copyright law really doesn't want knowledge to be shared easily, or until the author has had a good, long profit run. Or until Disney has a good long profit run with the author's work.


My gamedev bros use this for our libraries, and I'd love to see wider adoption of it--mostly due to the spirit of the thing.

However, the number of people who come out and bawww about not picking a more widely recognized license (generally, LGPL or BSD) is disappointing; folks, if you want good licenses, you have to set an example.

I implore my fellow developers to use this wherever feasible--show some guts.


Sorry, but please don't. The license has serious problems for developers. To start simple, first, while yeah, a judge is probably going to find you can do whatever you want, because it's not very explicit, there will be arguments over what you were allowed to do. It is entirely possible a judge will decide what is really meant is that you can do whatever reasonable thing you want to.

Worse, it does not effectively disclaim liability for the software developer. He claims he doesn't want "obnoxious things", such as "reproducing a huge disclaimer that is written in all caps".

The reason that all caps language is used is because of a history of court decisions and laws around what is required to effectively disclaim implied warranties. Right now, the license does not properly disclaim any implied warranties, which puts the developer at risk.

Worse than this, the text block he suggests putting in text if you want to disclaim warranty (http://www.wtfpl.net/faq/) would likely be ineffective in a lot of places.

I know everyone likes to think they won't get sued. But it does happen. It will wipe you out. It isn't worth taking a moral stand against a few lines of text in a source file.

To answer your other point, if you want good licenses, you have to actually know what the hell you are doing.

I really do understand that people want to not have to give a shit about this stuff, and have simple and clear licensing. Really. But this is not the path to accomplishing that.

The legal and political world takes a long time to adapt. Things like like free and open source software will eventually be sanely handled in the law. WTFPL is not going to be anywhere in that history.


> The reason that all caps language is used is because of a history of court decisions and laws around what is required to effectively disclaim implied warranties. Right now, the license does not properly disclaim any implied warranties, which puts the developer at risk.

Just trying to learn a bit about US law here. This part intrigued me, and from what I've Googled so far, it's because of a requirement for the warranty to be "conspicuous". [1][2] Is all caps the most commonly accepted minimum as advised by lawyers (especially when they're advising business types who couldn't care less about putting it in red or bold or something else)?

[1] http://www.mslater.com/post/104847-why-are-warranty-disclaim...

[2] http://www.adamsdrafting.com/all-capitals/


The all-caps is because of conspicuousness, yes, but there are other requirements.

UCC, for example (most contracts/licenses for software would fall under UCC, rather than common law), says basically (I'm pasting a capsule summary from a case book, i can cite the underlying provisions in the actual code, but this summary is accurate):

First, the seller must clearly and explicitly do so at the time of contracting by using words that communicate the risks that fall on the buyer.

To exclude the implied warranty of merchantability, the language of the disclaimer must mention “merchantability” and, if written, must be conspicuous.

To exclude the implied warranty of fitness, the exclusion must be in writing and be conspicuous. Thus, fine print disclaimer may be inoperative.

If an express warranty is also given, words or conduct relevant to its creation and words or conduct tending to negate or limit a warranty must be construed whenever reasonable as consistent with each other.

Any words of disclaimer or modification of warranty should be strictly construed against the seller. In general, words or conduct tending to negate or limit a warranty are inoperative to the extent that construction is unreasonable.


I really do understand that people want to not have to give a shit about this stuff, and have simple and clear licensing. Really. But this is not the path to accomplishing that.

You've hit it on the head--unfortunately, what else are we to do?

I don't trust the general legal intuitions of ourselves and our peers, since we seem to be so rubbish at creating a useful legal and political climate for our industry (based on ~30 years of case law).

So, in the meantime, let's instead live in our ideal world, pull no punches, and cross the legal bridges when (and if) they come up. Perhaps a proper rallying will be tenable then, instead of us all being chickens.

As for licenses, I think Apache currently is best for "serious grown-up usage", no, what with the patent indemnification and all, or am I misinformed?


"I don't trust the general legal intuitions of ourselves and our peers, since we seem to be so rubbish at creating a useful legal and political climate for our industry (based on ~30 years of case law)."

This is entirely because of the penchant for the average software/etc engineer to stick their head as far in the sand as it can go, and then complain that this is ineffective. It is only in the past few years that this has changed, and headway has already started to be made behind the scenes.

"As for licenses, I think Apache currently is best for "serious grown-up usage", no, what with the patent indemnification and all, or am I misinformed?"

This is, IMHO, correct. The only downside to apache is that the defensive patent termination clause only terminates patent rights. If there were no patents on the software, and they sue you over patents, they can go right on using it. Sadly, it's not possible to make a license that does terminate both copyrights and patents on being sued, and still hae such a license be GPL compatible.


> The only downside to apache is that the defensive patent termination clause only terminates patent rights. If there were no patents on the software, and they sue you over patents, they can go right on using it.

IANAL - could you explain what you mean by this?


Sure: Apache grants a patent license and a copyright license. The patent license (but not the copyright one) terminates if i sue you. So right now the situation is:

You license your stuff under apache.

It has no patents on it.

I sue you for patent infringement.

I can still use your stuff, since the only thing that got terminated was the patent license I had received. There are no patent rights, so nothing was really terminated. So in effect, I can sue you without worrying about how it impacts my ability to use your software.

This is true of most current open source licenses, btw. Suing does not terminate anything.

A better situation would be: Apache grants a patent license and a copyright license. Both patent and copyright license grants terminate if i sue you for patent infringement.

Then it would be

You license your stuff under apache.

It has no patents on it.

I sue you for patent infringement.

I can't use your software anymore.

The problem is such termination clauses are at the very least GPLv2 incompatible (They have been tried). Because everyone wants license compatibility, and people like GPLv2, nobody can really make this situation. In fact, it makes the situation worse.

Worse than this, what ends up happening, often, is that Apache 2 projects (which provide some protection at least if you own patents, since you can countersue) get pressured by GPL projects to move to a less protective licenses. See, e.g., https://github.com/twitter/bootstrap/issues/2054

This was, IMHO, a bad move for the Bootstrap project (full disclosure - the company I work for refused to let Bootstrap relicense the code we contributed because we viewed this as harmful to the open source ecosystem. This is/was drupal's problem, not Bootstrap's, and you shouldn't make a bad licensing decision to make up for another)


Ah, I see the distinction that you're making, though I would disagree with the statement that the latter situation is "better".

If I'm release software with that clause, I'm in a position to take advantage of anybody who uses the software, because they would have to cease using it before they could sue me, even on legitimate grounds.

That said, I'm not surprised that you (as a lawyer) would prefer the latter; it exposes your client to less liability, and lawyers tend to be very risk-averse, for obvious reasons.


You gave them software for free, it's not clear what legitimate grounds you want to accept liability for, but that's up to you :)

Given the "get sued by people over patents for releasing popular open source software" occurs very often for companies that deal with open source software and have money, ...

As for risk aversion, I am probably one of the least risk averse lawyers you will meet.

My goal in this case is not to protect my clients, my goal is to see the open source ecosystem stop being destroyed for dumb reasons.


Nevermind, I just re-read the text of v2, and it seems more limited in scope that what I previously thought.

> If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed.

I was also trying to thing of some pathalogical edge case in which I release something under an Apache license in order to preempt a lawsuit that I've infringed on someone else's patent... but I guess that doesn't make sense, since that would require that they accept the terms of the license to begin with.

So, I guess that's not really as much of a problem as I thought.


I don't trust the general legal intuitions of ourselves and our peers, since we seem to be so rubbish at creating a useful legal and political climate for our industry (based on ~30 years of case law).

The DWTFYWT license is better in this respect than MIT/BSD/Apache?

Edit: Or is worse enough that it'll help bring about a revolution? I'm … skeptical.


If a warranty disclaimer from another well-vetted license were used in the WTFPL, would that solve most of the problem? (Sans the judge ruling that it's only limited to reasonable things.)


Mostly, but at that point, why use it? Why not use MIT without the copyright notice reproduction clause?

It would be

1. Better written

2. As easy to understand

3. Well tested and effective.


>there will be arguments over what you were allowed to do

much like the GPL then


In my experience, most of the people arguing over it haven't read it (or don't understand basic copyright concepts). It's a pretty clear license...but, it gets talked about a lot by folks who misrepresent it, either willfully or out of ignorance.


To be blunt, the name is horrible and LGPL, BSD, MIT, and CC have been around long enough for folks (and not unimportantly in a lot of situations: corporate lawyers) to understand what they mean. It also doesn't look like a particularly good license if a bad actor gets me in front of a judge because they used my code and failed miserably.


Eh, I may get dragged up in front of a judge and sued one day, but I'd rather know that I did it as somebody with some chutzpah.


That will be cold comfort when you waste years of your life and probably your entire life savings because of it.

"Welp, we can't buy that house we were saving up for 15 years for, because I decided i wanted to take a stand and use the WTFPL". No offense, really, but that's not chutzpah, it's stupidity.


In what kind of a scenario would this happen?


Any one in which you got sued? The average lawsuit of this type, win or lose, is going to cost you 75-100k if it goes to court, minimum. That assumes it doesn't drag on, etc.


I suspect none, honestly, but it sounds scary, doesn't it?

Everybody who does software would prefer a better, less legally climate, and yet few seem willing to put skin in the game.

A better question is this: since we seem hellbent on not having any sort of accredited professional status, why are we surprised that this is an issue?


You suspect none?

Tell this to the java model railroad guy.

A single person who never expected to be sued over open source model railroad software.

He in fact, spent over 100k on the defense, AFAIK, and it did in fact cause him serious financial hardship. This is despite the fact that his lawyer (a wonderful woman) was doing the best she could to keep his costs down.


This: Jacobsen v. Katzer ?

You know, I was expecting to be really really sad reading that.

Like, I pictured a really nice old dude with a big white bushy beard and conductor's overalls and a blue hat crying as he had to sell off his pike to pay legal fees against some overzealous corporate lawyer, only to pass away a few weeks later from diabetes (his now-grown children laying in his coffin the one train car that escaped the lawsuit-caused liquidation of assets, a reminder of better times).

Instead, I found a tale of people being assholes to each other over copyright and patents and cybersquatting--further evidence that we should re-examine all those things.

Again, so happy to be wrong.


You must have read a different case than I did.


How is your case identified?


It is the same case, but I know the details personally for various reasons, and it is not as he describes.


I was going off of the wikipedia article--I'm sorry if I am misinformed. I make no claims that a wiki is anything more than a useful first-step onto the path of knowledge.


Chutzpah is nice, but I like to pick causes that have a chance of being worth the hit to my family. If I really cared about the purpose of this license, I would spend time on the political side to get clear laws about placing works into the public domain.


Is BSD not a good license, then? Legally it seems hardly different than this license (not necessarily zero, but hardly worth getting excited about), just more legalistically phrased.


The MIT license is essentialy "do whatever the fuck you want" in legal terms: http://mit-license.org/




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

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

Search: