PDA

View Full Version : source and license of ShaderEffect example



bvanevery
04-05-2013, 01:31 PM
ClanLib-3.0/Examples/Display_Shaders/ShaderEffect is pretty spiffy, in that it gives a simple example of procedural content generation. This is the sort of approach I imagine doing in my game; that is, writing a lot of GLSL 3.30 code. I noticed however that the concept of this example is not original. ClanLib-3.0/Examples/Display_Shaders/ShaderEffect/Resources/fragment_shader.glsl attributes it as:

// Shader code by Frequency (http://www.frequency.fr/)

I went to that site and found the original work, "To the Road of Ribbon" (http://www.frequency.fr/index.php?option=com_content&task=view&id=69&Itemid=70). It has source code, but it's not GLSL source code as it appears in ClanLib.

rombust, did you obtain a GLSL 3.30 version of this demo from somewhere? If so, I'd like to know where, so that I can find more of the same. Or, did you look at the "To the Road of Ribbon" .asm source, extract the GLSL code that's embedded there, and bump it to #version 330? Or did you do something else?

What was the license on the original code? Both for purposes of programming, and for attribution? There is no license in the sources at the Frequency website. I think such things fall into a legal grey area and aren't "public domain." That is to say, it can make ClanLib and users of this example legally exposed, if the original author up and decides to make a fuss about it. Or authors: "Frequency" is not a person, but a group (http://www.frequency.fr/index.php?option=com_content&task=blogcategory&id=19&Itemid=72), whose members change over time. The .nfo file attributes various people; legally speaking all of them are Joint Authors. That means any of them could say, "Hey, pay us money."

I think it would be appropriate to name the work itself, "To the Road of Ribbon," and also briefly state that this example is a derivative work. Unless you feel that's not appropriate and have made a new work... in which case, you shouldn't be attributing Frequency as having written the shader code. I think an IP lawyer would say this is a derivative work though, as it looks and behaves just like the original. You might consider putting the original .nfo in ClanLib and then a README saying how you derived the work.

sphair
04-05-2013, 01:52 PM
The shader was found online at https://www.shadertoy.com/view/MsfGzr, where the original author (XT95) of the demo uploaded it. From their terms:

"All the shaders you create in Shadertoy are owned by you. You decide which license applies to every shader you create. We recommend you paste your preferred license on top of your code, if you don't place a license on a shader, it will be protected by our default license:

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License."

I basically upgraded it to use a uniform block, otherwise used it as it was. Most likely, the current shader in the example is a placeholder until we have developed something on our own. For now, we should add the above license information in the example.



I went to that site and found the original work, "To the Road of Ribbon" (http://www.frequency.fr/index.php?option=com_content&task=view&id=69&Itemid=70). It has source code, but it's not GLSL source code as it appears in ClanLib.

rombust, did you obtain a GLSL 3.30 version of this demo from somewhere? If so, I'd like to know where, so that I can find more of the same. Or, did you look at the "To the Road of Ribbon" .asm source, extract the GLSL code that's embedded there, and bump it to #version 330? Or did you do something else?

What was the license on the original code? Both for purposes of programming, and for attribution? There is no license in the sources at the Frequency website. I think such things fall into a legal grey area and aren't "public domain." That is to say, it can make ClanLib and users of ClanLib legally exposed, if the original author up and decides to make a fuss about it.

The .nfo file does indeed say it's by Frequency, but I think it would be appropriate to name the work itself, "To the Road of Ribbon," and also briefly state that this example is a derivative work. Unless you feel that's not appropriate and have made a new work... in which case, you shouldn't be attributing Frequency as having written the shader code. I think an IP lawyer would say this is a derivative work though, as it looks and behaves just like the original.

bvanevery
04-05-2013, 02:16 PM
The shader was found online at https://www.shadertoy.com/view/MsfGzr, where the original author (XT95) of the demo uploaded it. From their terms:

"All the shaders you create in Shadertoy are owned by you. You decide which license applies to every shader you create. We recommend you paste your preferred license on top of your code, if you don't place a license on a shader, it will be protected by our default license:

IANAL, but I don't buy that Shadertoy has any legal right to impose any kind of default license at all. The shader has probably got the license the Joint Authors gave it: none / ambiguous / not public domain. If Shadertoy really had a right that would hold up in court, they would patch any submissions to include their default license. Instead, Shadertoy says in their TOS:


Shadertoy is not responsible for the content created by its users, the users are responsible for that content. [...]

Although Shadertoy owns the data storage, databases and the Shadertoy site, the users retain all rights to their creations and can decide a specific license for it (section "What License Will my Shaders Have?").

IANAL but my reading of that is, the Author can revoke the license at any time, and can say Shadertoy had no right to impose a default license. The Author has signed away nothing, not didly squat, except he's given Shadertoy a right to reproduce the work worldwide "for the limited purpose of operating, promoting, and improving Shadertoy." Shadertoy has covered it's own butt, not yours.

Also, XT95 is not the only author with rights to the demo. All the Joint Authors have the same rights, no matter how big or small their contribution was. You'd need to get a permission or license from at least one of the Joint Authors, to be on a firm legal footing. Any of them can give it to you. At least under US law, it doesn't matter squat what the other Joint Authors want to have happen. All Joint Authors can trade equally in the work, can license to other parties in any way they want, and merely have a duty to share profits with other Joint Authors. (I learned all this after working on a Wesnoth campaign "To Lands Unknown" for 4 months, then was ousted by the primary author.)


For now, we should add the above license information in the example.

No you shouldn't. Not sure exactly what you should do, but a claim by Shadertoy doesn't mean squat.

sphair
04-05-2013, 02:31 PM
I love these kind of discussions, so I have updated the example accordingly. Hope you like the new version!

bvanevery
04-05-2013, 03:09 PM
I love these kind of discussions, so I have updated the example accordingly. Hope you like the new version!

No I don't. I wonder if rombust likes it either, but if he's good with having his worked nuked, who am I to say. The new "shader" (which produces an entirely red screen and doesn't show off procedural anything) does solve any possible legal issue, but I just got done telling the Ogre crowd you "had something" for this in your SVN repository. Maybe nobody cares and won't go looking, but if they do, they're going to wonder why I mentioned it. (EDIT: well now they're not, because I deleted it.)

In your shoes, I would try bugging the author who posted on Shadertoy to put a proper license statement in his code. "Hi, we're from ClanLib and we really like your work. We'd like to use it as our 1st demo for our GLSL 3.30 capabilities. We already have it working, we're just wondering about the license. Can you...?"

In my shoes, I will see if Shadertoy or other sites have interesting GLSL 3.30 shaders that are properly licensed already. And I may have a discussion with Shadertoy about getting their license pollution issue on a proper footing.

sphair
04-05-2013, 04:02 PM
Although rombust does most of the work in clanlib these days, this shader was added by me, so i nuked my own work.

Secondly, feel free to do what you suggest. I can't be bothered to follow up licensing issues myself, so until someone else figures this out, the example stays red (or whatever someone else decides to replace it with).


No I don't. I wonder if rombust likes it either, but if he's good with having his worked nuked, who am I to say. The new "shader" (which produces an entirely red screen and doesn't show off procedural anything) does solve any possible legal issue, but I just got done telling the Ogre crowd you "had something" for this in your SVN repository. Maybe nobody cares and won't go looking, but if they do, they're going to wonder why I mentioned it.

In your shoes, I would try bugging the author who posted on Shadertoy to put a proper license statement in his code. "Hi, we're from ClanLib and we really like your work. We'd like to use it as our 1st demo for our GLSL 3.30 capabilities. We already have it working, we're just wondering about the license. Can you...?"

In my shoes, I will see if Shadertoy or other sites have interesting GLSL 3.30 shaders that are properly licensed already. And I may have a discussion with Shadertoy about getting their license pollution issue on a proper footing.

bvanevery
04-05-2013, 04:32 PM
Although rombust does most of the work in clanlib these days, this shader was added by me, so i nuked my own work.

My bad. rombust wrote a Makefile for it; I thought he wrote the shader too. It seems I didn't pay good attention when you answered my question about how it got written.



Secondly, feel free to do what you suggest. I can't be bothered to follow up licensing issues myself, so until someone else figures this out, the example stays red (or whatever someone else decides to replace it with).

I will find appropriate fishing holes, or else berate the ***holes I do find for inappropriateness. (Or not. Sounded like I better pun when I wrote it.)

Hm, ShaderToy is using WebGL (https://www.khronos.org/webgl/), essentially OpenGL ES 2.0 for the web. I imagine I could find a "better" GLSL 3.30 shader somewhere. I've looked at a lot of ShaderToy demos and making license statements doesn't seem to be the culture. Only 1 guy's doing it, possibly a guy who runs the site. He's doing "// License Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License." I think one could do better than a non-commercial chunk of code for ClanLib. Anyways, having a licensing discussion over there sounds non-productive.

Judas
04-05-2013, 06:40 PM
The example was never about showing OpenGL 3.3. It was about showing how you can use the ShaderEffect class to quickly set up the input/output for a shader program, without having to create each lower level object yourself. Sphair simply chose to use a very fancy WebGL as the demonstration shader.

Rombust wanted to verify that it was working under Linux. The Debian non-free Nvidia driver was too old (3.2) to be able to run the example. You could probably have gotten it to work with OpenGL 3.2 by changing the #version to 150 (3.2) instead of 330 in the shader file.

For almost all purposes you can see OpenGL ES 2 as a subset of OpenGL 3, which is also why a WebGL shader was chosen in the first place. It took less than 5 minutes to port.

bvanevery
04-05-2013, 07:57 PM
All good and true, but I personally am looking at ClanLib for OpenGL 3.x capabilities, which is why I'm interested in finding appropriately licensed GLSL 3.30 shaders somewhere. Running the open source NVIDIA driver is a non-priority to me, as I want to ship 3D intensive games on Linux. I expect anyone I'd eventually sell a game to will have a proprietary driver, or the open source driver will have sufficiently improved by the time I ship anyways. Or multiple #versions or directories could be offered or whatever. My main interest is "find something that looks cool and actually is cool under the hood." For sake of comparison, Ogre doesn't seem to have anything like what I'm interested in. All of their .glsl examples are #version 120 or #version 150, with the exception of some tesselation code that is #version 400, but I'm not sure it works and I don't have 4.x HW anyways.

Hm, downloading Linux demos from pouet.net (http://pouet.net/) hasn't been productive so far. Need sites with GLSL as a search parameter.

Hm, demoscene is proving to be quite a learning curve. They don't seem to have an open source mentality, but rather, are driven more by personal artistic products. I haven't finished surveying yet though.

The versioning differences finally sunk in, that GLSL #version 150 = OpenGL 3.2.

bvanevery
04-08-2013, 01:43 AM
I finished my survey of the demoscene, and especially of source code available at Pouet.net (http://www.pouet.net/sourceprod.php). Nothing ever compiled on Linux, and license statements are rare. Somehow I completely missed the new wave of GLSL demos that are supposed to be out there. Maybe they just don't have source. I'd probably have to ask on their BBS to make any progress on this. Judging by their archives, which I've extensively perused on the subject of open source and MIT / BSD / zlib licensing, I don't expect much. So for now, I consider that investigation "concluded" and will move on to some other approach. It may be that nothing particularly worth having exists off the shelf, but if it is out there, I don't think it's in the demoscene.

Judas
04-08-2013, 03:24 PM
Different communities tend to cling to different licenses for no good reasons.

For example, in the Apple open source community everything is always MIT style licensed. Even the simplest silliest small 500 lines libraries demand that you credit them. I'm sure that if you emailed such an author and asked why this license, he'd probably reply that he picked it as it was tradition in their community to use this particular license. Likewise in Linux communities its GPL. And in demo scene communities its usually just something random if any license at all, since quite frankly nobody gave a damn about licensing when they made Amiga 500 and C64 demos. After all, when people weren't writing demos they'd be busy copying tapes and writing trainers for games. For fun.

If you truly want a GLSL replacement for the ShaderEffect example, the easiest and most effective way is to write to someone and ask if you can reuse his shader under ZLib terms instead. He's probably reply back something like "sure, whatever" while being just as bored about licensing stuff as the rest of us. :)

bvanevery
04-08-2013, 07:18 PM
Different communities tend to cling to different licenses for no good reasons.

For example, in the Apple open source community everything is always MIT style licensed. Even the simplest silliest small 500 lines libraries demand that you credit them.


If the usual point of a community is to have such things be reused in commercial software, rather than just for funzies or hobby or student purposes, then license clarity is essential. I agree that not all communities value this, hence the differences. People who don't care about commercial reuse don't fit well in MIT / BSD / zlib driven communities. Actually your comment is interesting given that ClanLib is zlib licensed. Perhaps you don't see why 500 lines of code would be of any commercial use to someone else, that everyone should just write those 500 lines for themselves over and over again.

Or maybe you are objecting to the crediting nature of a zlib license? I will admit, I don't care for it, and won't license my own work that way. But it is commercially useful, hence lumpable as MIT / BSD / zlib.



I'm sure that if you emailed such an author and asked why this license, he'd probably reply that he picked it as it was tradition in their community to use this particular license. Likewise in Linux communities its GPL.

I did talk someone into GPL --> LGPL, or LGPL --> MIT once, can't quite remember, exactly because as you said they'd never done their homework on license differences. I did a little education for them and they were pretty easy to sell. I'm not going to bother unless it's actually worth having though.



And in demo scene communities its usually just something random if any license at all, since quite frankly nobody gave a damn about licensing when they made Amiga 500 and C64 demos. After all, when people weren't writing demos they'd be busy copying tapes and writing trainers for games. For fun.

So I gathered from reading their archives extensively. Their concerns are rarely mine.



If you truly want a GLSL replacement for the ShaderEffect example, the easiest and most effective way is to write to someone and ask if you can reuse his shader under ZLib terms instead. He's probably reply back something like "sure, whatever" while being just as bored about licensing stuff as the rest of us. :)

Except that, I'd have to keep looking through demoscene stuff to find something I actually wanted. In source code, with a GLSL shader, on Linux, it doesn't exist on pouet.net. Some demos I think are cool enough to want to spread far and wide, but many or not. Watching videos of them is more work than fun. So... let's say I watch a good video. Then I have to determine if it was written with GLSL or not. That's work. Unless I can find a site where everything is sorted on GLSL in the 1st place. Written in GLSL --> actually worth watching --> pester authors for MIT / BSD / zlib source code *might* be worth doing. But I'm feeling pretty burned out on demoscene scrounging after doing a solid 2 days of it. I learned a lot, but it's not looking like a way to get compelling visuals in the free and clear.

I could ask on their BBS, but I can already tell, just from the kind of answer you've given me here, that people will just jump all over it and get annoyed. Someone may be helpful anyways, but I just don't feel like getting drawn into a discussion on pouet.net about why I don't actually like most people's work and why I'm not just writing it myself.

Judas
04-08-2013, 09:14 PM
Actually your comment is interesting given that ClanLib is zlib licensed. Perhaps you don't see why 500 lines of code would be of any commercial use to someone else, that everyone should just write those 500 lines for themselves over and over again.

Or maybe you are objecting to the crediting nature of a zlib license? I will admit, I don't care for it, and won't license my own work that way. But it is commercially useful, hence lumpable as MIT / BSD / zlib.

The catch about the MIT license is that the crediting needs to be in the manual or the about dialog. Which means in commercial products you as a developer now need to get permission to use the code from management, and do extra work making sure this is written in a proper place. Usually that's more work for me than just writing my own 500 lines of code.

However my point here isn't which license is "best", but rather that almost all small open source projects just slam on a random license. Which license it is depends on which community they originate from - rarely what the license actually says. Go to a website like json.org (http://json.org) and notice how you can almost always predict what license it will be using based on what language/community the library originates from. I seriously doubt most of those making those JSON parsers cared much. They just liked their language and the principle of using JSON as data exchange. The reality is that most probably would have chosen Public Domain if they really knew what they were doing when slapping on those random licenses.


I could ask on their BBS, but I can already tell, just from the kind of answer you've given me here, that people will just jump all over it and get annoyed. Someone may be helpful anyways, but I just don't feel like getting drawn into a discussion on pouet.net about why I don't actually like most people's work and why I'm not just writing it myself.

Yes, licensing issues are only interesting to Debian's legal team and Slashdot troll stories. :)

Rest assured however, that all the ClanLib source code itself is licensed under a ZLib license with no legal problems. Well except when someone sues you for using the Patent Pending 2+2=4 function somewhere. In short: you're ****ed if a mega corp hates you no matter what you do, so why worry so much?

bvanevery
04-08-2013, 09:49 PM
The catch about the MIT license is that the crediting needs to be in the manual or the about dialog.

That's not true (http://opensource.org/licenses/MIT). Where did you get that idea? It says, "The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software." A license in the top directory of the library sources is generally fine. That's all Ogre does, for instance. When I was younger, I used to put my copyright and the license in every single one of my files, but in hindsight that's a tad excessive.

EDIT: Now I think I know where you got your idea from. The XFree86 project had (has?) an end user documentation requirement (https://en.wikipedia.org/wiki/MIT_License). This is a modified MIT license, not a straight one as generally understood in OSI terms today. Too bad XFree86 gave you this goofy idea. :p Well, the original BSD license also had an advertizing clause, that article says. Anyways that's all ancient history and not relevant today.



Which means in commercial products you as a developer now need to get permission to use the code from management, and do extra work making sure this is written in a proper place. Usually that's more work for me than just writing my own 500 lines of code.


I think your blase about licensing has made you unaware of what various licensing terms actually are. (EDIT: Either that or you are substantially older than I am, remembering some of these goofy advertizing clauses in their original use! I doubt that though.) Some time spent at OSI (http://opensource.org/) reviewing the major licenses would probably benefit you. MIT is the most permissive license out there. All the license says is, "It's licensed. The original author wrote this. No warranties." That's a minimalistic legal protection, but it could be important if some predatory company takes your code, tries to claim you didn't write it, and that you're infringing on them. I believe people can try to assert copyright over public domain stuff, but I've never looked into the details of such shenanigans. Very little of value ever ends up truly in the public domain.



However my point here isn't which license is "best", but rather that almost all small open source projects just slam on a random license.

I say it's not as random as you suppose, although I have run into people who could have rolled a d6.


The reality is that most probably would have chosen Public Domain if they really knew what they were doing when slapping on those random licenses.

If they had any idea what they were doing with licenses, or why they exist, I highly doubt that. They would have chosen MIT, as I always do, and Ogre does.



Yes, licensing issues are only interesting to Debian's legal team and Slashdot troll stories. :)


They are of interest to commercial game developers and I'm wondering what that implies about your own pursuits.



Rest assured however, that all the ClanLib source code itself is licensed under a ZLib license with no legal problems. Well except when someone sues you for using the Patent Pending 2+2=4 function somewhere.

Some licenses deal with patent reciprocation. I haven't gotten into that. If I thought it a potential problem, I would.


In short: you're ****ed if a mega corp hates you no matter what you do, so why worry so much?

That's unrealistically pessimistic. Look you don't have to go through any license drills for your own work, that's your prerogative. But it seems like you think about your own development on an awfully small scale if you don't take license issues seriously. Well, at least in the open source sense. You could of course short circuit all issues by writing totally proprietary software.

rombust
04-09-2013, 08:29 AM
Awesome thread :)

It's so funny, we hear the same conversation every couple of years.

There are lots of licenses out there. And a lot of confusion. There is no "best licence", it is just what the developer prefers at the time.
What muddies the waters is that some people change the license text, adding their own clauses.

For example, the MIT license, there are many variants ... https://fedoraproject.org/wiki/Licensing:MIT?rd=Licensing/MIT

I like the ZLib style license the best, because it's simple, short and easy to understand.

I use ClanLib in the commercial environment, and have to be very careful on the License. (I use ClanLib 2.3)

If ClanLib is not license clean, Debian avoids it. Hence that's why the Linux (.tgz) release is missing certain files:


echo "Removing configure.exe, SpriteSpeed1-CL09.exe, SpriteSpeed1-CL08.exe"
rm "$CLANLIB_FNAME/configure.exe"
rm "$CLANLIB_FNAME/Tests/Display/SpriteSpeed1/SpriteSpeed1-CL09.exe"
rm "$CLANLIB_FNAME/Tests/Display/SpriteSpeed1/SpriteSpeed1-CL08.exe"

echo "Removing microsoft window's GUI theme"
rm -r "$CLANLIB_FNAME/Resources/GUIThemeAero"
rm -r "$CLANLIB_FNAME/Resources/GUIThemeLuna"
rm -r "$CLANLIB_FNAME/Resources/GUIThemeAeroPacked"
rm -r "$CLANLIB_FNAME/Resources/GUIThemeLunaPacked"

(from http://clanlib.org/wiki/HowToRelease and http://clanlib.org/wiki/ReleaseScript )

When taking source code from the Internet, there is no way to know if the license at the top is genuine. We just have to hope that it is.

We try to be careful, but it's extremely difficult. To a point where the hobbiest looses motivation.

For example, we used to have a very nice simple parallax scroller example, called "shadow of the beast", that uses graphics and music from an old Amiga 500 game. imho, realistically Sony would not care about the license issue, but technically we had to remove it.

http://clanlib.org/w/images/5/54/Example_shadowofthebeast.png

A side note:
When I created Core/Cypto code for ClanLib 3.0, I required a BigInt library (for the creation of the prime numbers for the RSA key pairs). I found the original MPI library. This library originally had a public domain license. Since then Netscape(Firefox) used it and change the license.
Hence the following comment on "big_int.cpp"


// This class is based on the original MPI library (not NSS, because of license restrictions) with some modifications.
// Some ideas and algorithms are from NSS (Netscape Security Suite). Where they have been used, the function contains a reference note
//
// Note, since September 2011, I believe the MPI homepage is now: http://spinning-yarns.org/michael/mpi/
// The license is as follows
// This software was written by Michael J. Fromberger,
// http://www.dartmouth.edu/~sting/
//
// See the MPI home page at
// http://www.dartmouth.edu/~sting/mpi/
//
// This software is in the public domain. It is entirely free, and you
// may use it and/or redistribute it for whatever purpose you choose;
// however, as free software, it is provided without warranty of any
// kind, not even the implied warranty of merchantability or fitness for
// a particular purpose.


That webpage not longer exists! (it may do in the wayback internet archive though, but i'm not that interested to check)

bvanevery
04-09-2013, 12:14 PM
There are lots of licenses out there. And a lot of confusion. There is no "best licence", it is just what the developer prefers at the time.
What muddies the waters is that some people change the license text, adding their own clauses.

For example, the MIT license, there are many variants ... https://fedoraproject.org/wiki/Licensing:MIT?rd=Licensing/MIT

The top text claims, "There are many MIT variants, all of which are functionally identical." But the ones with documentation clauses are not functionally identical to the nowadays canonical OSI "MIT license" (http://opensource.org/licenses/MIT). The whole XFree86 vs. XOrg split happened over that very issue. To people in the know, that's a rather charged, provocative statement, that all these "MIT" licenses are functionally identical. Did that webpage author just innocently summarize history, paying insufficient mind to issues that others consider important? Or did the author have a political axe to grind?

OSI exists in part to put a restraint on license proliferation, so that open source doesn't suffer these brand identity problems of terminology. Google Code is even more stringent, only allowing a small selection of license possibilities for one's project. This is all in the name of commercial license purity, so that when we hit the brass ring for one of our projects, we don't have some yokel getting sue-happy on us. You can't just take random stuff off the internet without a proper license statement, not if you care about your legal exposure.

Judas
04-09-2013, 01:54 PM
When I grew up as a boy, seeing my first game on the ZX Spectrum, I immediately started dreaming: "When I grow up I want to debate the importance of OSI, the difference between 1st generation MIT licenses and later iterations, and what options Google made available in their license drop-down box for their sourceforge clone and speculate about their motives".

Oh wait, I wanted to write games. Not become a lawyer.

rombust
04-09-2013, 02:07 PM
Yeah the system is a mess.

The silly patent system is worse.

I especially feel very sorry for hobby developers in Amerca. Always worried that their newly created game will have to be destroyed because a small segment of the code uses mega coorp xyz obsure patent from 1999.

bvanevery
04-09-2013, 02:11 PM
Magnus, I've had this discussion in order to press the importance of what commercial developers are looking for in a MIT / BSD / zlib licensed 3d engine. You don't seem much concerned about this, and it makes me wonder to what degree you're concerned about commercial development with ClanLib in general. Maybe it's not actually a warning sign and just a question of scale or preoccupation, your "right to be blase" about some aspect of what you work on. At least I hope so.

I cannot of course make you more focused on license issues than you wish to be, but I surely hope you'll stop thinking inaccurately about licenses, that people asking for "MIT" are showboating about attributions in documentation or splash screens etc. I will point out that Ogre developers "get it," that's why they changed from LGPL to MIT a few years back.

rombust
04-09-2013, 03:19 PM
I will point out that Ogre developers "get it," that's why they changed from LGPL to MIT a few years back.

ClanLib did the same from LGPL to ZLIB many years ago (around 2005 I guess).

I am not sure what the problem is here.

ClanLib only allows ZLib compatible license to be included. (I've been told "NO" in the past by trying to include LGPL code)

A "pure" MIT license to be included won't be a problem. As long as it hasn't been altered to include adverts.

All the developers here have used ClanLib in commercial environments.

If a developer wishes to fix problem areas, that's fine.

If a developer doesn't care, that's fine.

Developing a free SDK doesn't make money. Development is only for personal satisfaction.

sphair
04-09-2013, 03:20 PM
You keep making references to ogre. I think it is important that you realize that we could not care less about what they do.