[License-review] [2nd Resubmission] ModelGo Attribution License, Version 2.0

[Moming Duan]

Hi Carlo,

Thanks for your review.

> while reviewing the modifications to decide on the approval (actually it should be better to withdraw and resubmit at the end of the discussion) I noticed that the modifications only applied to the MG-by and not to the Open Source version.
> 
> Comparing the two texts, moreover I see that here the liability disclaimer says "to the maximum extent permissible under applicable law, the Licensed Materials are provided on an “as is" and “as available” basis without any representation, warranty, *condition* or term of any kind (whether express, implied, statutory or otherwise)... 
> 
> "condition, or term" does not appear in the other license and I think it does not make sense adding it at all, since the license has terms and conditions.  
> 
> At least, however, I think you should port the improvements of the Attribution license to the Open Source one, I really don't see why, while sharing like 95% of the provisions,  this has not been done to the identical ones here: the rationale applies equally.

It seems you’re comparing the 2nd resubmission of MG-BY with the (1st) resubmission of MG-BY-OS (which was renamed to MG-BY-SA in the 2nd resubmission). For clarification, the 2nd resubmission of the MG license text is in the email with the subject:

[2nd Resubmission] ModelGo Zero License, Version 2.0
[2nd Resubmission] ModelGo Attribution-ShareAlike License, Version 2.0


> I also object using "Source Code Form" for "preferred form for making modifications, since most evidently the definition includes way more than source code and it is not good practice to define something with a name which is misleading.

This definition refers to GPL-3.0, which states:

The "source code" for a work means the preferred form of the work  for making modifications to it.

I am also not sure that MG-BY-SA can meet the approval requirements. And as Pam previously mentioned, there remains a question of whether copyleft can effectively function in the context of ML models. I look forward to OSI’s decision and hope to engage in a deeper discussion on this topic in the future.

> 
> I am still dubious that you can include the Output in the Derivative definition and that a license can purport to control it. This shall undergo more discussion, here and elsewhere, as it is a central point upon which I tried to make an impression every time I discussed porting Open Source concepts to the world of AI.
> 
> For the "zero", I think the title is misleading, because it has no "zero" conditions, and I have not made a final opinion yet.

Please refer to the 2nd resubmission version of MG, where I have already removed all the conditions as suggested by McCoy Smith. Thanks.


Best,
Moming

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20251030/1a2d873d/attachment.htm>