From: Trans Date: 2007-10-16T02:59:05+09:00 Subject: Re: MIT vs. Ruby/GPL License On Oct 15, 9:20 am, Chad Perrin wrote: > On Tue, Oct 16, 2007 at 12:37:04AM +0900, Trans wrote: > > > Uh oh. We were doing so well! A potentially flaming subject kept to a > > congenial ember. > > > Tell you what. Lets now take a different bent: Here is what I think I > > learned from this thread. > > > 1) Licensing can be complicated ;) > > 2) Unless you have specific reasons to do otherwise pick a very simple > > one like MIT. > > Specific reasons are either: > > 1) I demand all derivative works also be OSS. > > 2) This software is mine mine mine (non-OSS or partial-OSS) > > 3) Another lib, that I must use, is forcing my hand. > > 3) Ruby license is not a good license to use, but unfortunately so > > many Rubyists have used it we are kind of stuck with it --see 2.3 > > above. Thankfully, clause 4 of the license gets us out of point 2.1 > > (barely?). > I'm not quite as well-versed in the MIT license as I am in some others. > It is often compared with the BSD license as having similar terms, > however, so I'll speak about the BSD license here (with which I'm more > familiar). > > The BSD license actually provides for the code covered by the license to > be forevermore "open source". There's a common misconception that the > BSD license and similar licenses somehow allow you to make the code in > them proprietary. That's not legally possible, as the BSD license > contains an inheritance clause that mandates further copies of the code > be distributed under the same license. The difference between that and > the voracious nature of the GPL is that the BSD license allows covered > code to be "wrapped" in code of another license, such that a complete > project can be distributed under the licensing terms of the copyright > holder's choice while the specific parts of it that are licensed BSD > remain under the BSD license's terms. The GPL demands that everything > that is connected to code it covers must also be distributed under its > terms. That's good to understand. > So . . . the MIT license may very well be quite suitable to situations > where you "demand that all derivative works also be OSS". Does it? I thought it was just the opposite. > That really > depends on how you mean that statement to be interpreted. Well, I was thinking "GPL" in this case. > As for the flame war prone nature of license discussions, I guess I > probably should have left well enough alone rather than asking someone to > keep his/her personal opinions down to a dull roar. The obvious pro-GPL > implications of a preceding message in this thread seemed to require some > correction, and I made the mistake of not realizing attempting to cut the > rhetoric away from the facts would be something like throwing fuel on a > fire. I apologize for that misjudgment. Well, we are all entitled (even to argue if we want ;) But classically I think the "war" of license is between GPL and BSD isn't it? That's probably a huge over simplification, but that tends to be my shallow assessment of the field. And that was really the crux of my dividing line for point 2.1. T.