[Disclaimer: not a lawyer, this is not legal advice]<p>I understand they're trying to get rid of cascading attribution requirements for enterprise (which is part of the reason NIH syndrome became so prevalent!), BUT I really dislike this incarnation for two primary reasons:<p>1. It has conflicting branding. How is a newbie gonna know the difference? How many will misclick it because it sorts first, or because they don't know it's different than true MIT?<p>2. Some users may read 0 suffix as "original", "optimal", or "first" (0-index), further polluting the branding.<p>If they rereleased under another name these problems would be alleviated, but I don't think their intentions are entirely altruistic here.
The MIT-2 License:<p>Like the MIT license but you must put <i>2</i> instances somewhere in your attribution. It is per-license, so the whoever includes your work in their own project must put 4 instances, and so on.
Oh for the love of God, don’t introduce confusingly named variants of widely known licenses. If a new license is really needed — fine, if lawyers think that’s necessary, it probably is — but don’t name it <i>that</i>. No one will bother to check whether MIT and MIT-7BZ3/2 are slightly incompatible.
I prefer zlib over this. If you're shipping my code in a binary I don't care about attribution, but if you're reproducing my entire repo in source code form don't claim that you wrote it
One issue with this is that there's no obligation to keep the license on it-- which means that the entire disclaimer and lack of warranty can get sheared off the software.<p>Then in turn downstream users can legitimately obtain it without knowing that you're attempting to disclaim liability.
I predict a wave of articles and blog posts about how open source doesn't work in about 5 years, after the new generation of developers that Amazon convinces to use this licence complete the process of being out competed by Amazon using their own IP.
Is... Is Amazon allowed to just use the name "MIT" for its own unrelated license? Doesn't MIT have some say over the use of the name "MIT"?<p>Can I make my own license and call it MIT-2, to make my license sound much more established and prestigious than it really is? Or is this a case of "it's OK because Amazon does it and nobody will sue them for it"?
IANAL, but assuming a jurisdiction in which public domain dedications are likely to be legally valid (such as the US), is there any real difference between this license and a public domain dedication? Isn’t this functionally equivalent to the public domain, even if it is technically not?<p>Public domain code, I can do whatever I want with it, even publish it with all indications of its origins removed. Legally fine. But it seems like the same thing here-for a downstream consumer at least.<p>From the original developer’s viewpoint - the warranty disclaimer may give some additional legal protection, but you can put a warranty disclaimer in a public domain dedication too. People will debate whether warranty disclaimers in public domain dedications have the same legal force as those in copyright licenses-but maybe that too is a theoretical debate, because how big is the risk of someone being successfully sued by a code reuser who has zero commercial relationship (support or consulting contracts, etc) with the original developer-has that ever actually happened? Or are all these warranty disclaimers just lawyers saying “it’s never happened, but we can’t totally rule out the remote possibility that one day it might, so let’s just include some legalese to improve our position if anyone ever tries to make it happen - a disclaimer doesn’t cost anything significant anyway”?
So MIT is "do attribute me" and I guess MIT-0 is "you don't need to attribute me (but it's fine if you do)", I wonder if there are any "explicitly do NOT attribute me" licenses?
I use this license for a few of my NES libraries, since they're meant to be used in homebrew games where memory is a real problem. 32kB of PRG ROM (and the associated 8k of CHR ROM) doesn't leave much extra room for text rendering systems or fonts, and every byte counts. Sure if there's room, attribute away, but for simpler games there's <i>really</i> no need. It's fine :)
Didn't realize Mark decided to do this after leaving HP Enterprise, but crazy that it took 4 years to get published by Amazon.<p>For an open source evangelist, this is an odd way to engage the community. Really makes the FSF's principles look good, with the four freedoms and such.
For interested readers: you can find a quick comparison of licenses at <a href="https://choosealicense.com/appendix/" rel="nofollow">https://choosealicense.com/appendix/</a><p>MIT No Attribution is included in the comparison.
> i.e., a license is preferable to a public domain dedication<p>Are we that easy to fool by lawyers? The public domain is the public domain. Before someone comes out and says that the concept of dedicating IP to the public domain doesn't exist in some countries or is full of hidden perils, I'll ask you to provide a single case where code was attributed to the public domain and that became a problem for the author or for the users.
Isn’t this effectively the same as WTFPL? Of course I’m not a lawyer but they both seem to have total overlap in purpose and effect. The disclaiming aspects of MIT are basically restating defaults anywhere they’re legally enforceable.
Now I want a MIT--1 license that <i>prohibits</i> attributing the code to the author in any way, including in legal filings, and then you don't need any other part of the license.
At that point just go fully PD or CC0. Why have some random license nobody has heard of or wants to read about when you can just slap a CC0 sticker on it.
Isn't MIT trademarked [1] ? Can anyone take a trademarked name for a license?<p>[1] <a href="https://tsdr.uspto.gov/#caseNumber=73797219&caseType=SERIAL_NO&searchType=statusSearch" rel="nofollow">https://tsdr.uspto.gov/#caseNumber=73797219&caseType=SERIAL_...</a>
> a license is preferable to a public domain dedication<p>Who says that and why? Why should I care about the complicated terms of a license - especially when such a license expresses essentially the same as "public domain"?<p>Or in other words: is there any real rationale behind the above statement?
What happens once there are layers of execution of the lack of attribution, but then someone downstream comes into a battle over provenance? They have to backtrack now even more poorly documented provenance in discovery? Sounds awful.
I think NetBSD licensing is best :<p><a href="http://netbsd.org/about/redistribution.html" rel="nofollow">http://netbsd.org/about/redistribution.html</a>
GNU Affero General Public License<p><a href="https://www.gnu.org/licenses/agpl-3.0.en.html" rel="nofollow">https://www.gnu.org/licenses/agpl-3.0.en.html</a>