[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Hacking License

Il December 7, 2018 7:03:47 PM UTC, Thorsten Alteholz <debian@alteholz.de> ha scritto:
>Hi Giacomo,

Hi Thorsten, thanks for pointing out these issues.

>On Fri, 7 Dec 2018, Giacomo Tesio wrote:
>> If you can help me understand the problems you see, we could try to
>> design a new test that make them evident together.
>some terms are ambiguous and need to be defined. For example how do 
>you want to use "shall"?
>What is an organization? Is a one man company an organization or a user?

GNU GPL doesn't define "organisations" or "shall" either.
I can add organisation if it's ambiguous here but I'd like to understand what makes it clear in GPL instead.

>What are activities?

A synonym of actions.

>According to your definitions:
>  "Hacker" refers to any Copyright holder of the Hack.
>"Copyright" means copyright-like laws that apply to other literary works.
>As a conclusion: "Hacker" refers to any copyright-like law holder of the Hack. This does not make sense.

Copyright definition comes from GPLv3.
I will fix Hacker definition to clearly means any Copyright holder as law intend it.

The idea here is to bind the copyright to the modifications (the license only grants right to modify the Hack to Users) so that all hackers are really Copyright holders (unless they exclusively assign the Copyright to a third party and renounce to it).

This empower hackers, even as employees within a company.
Since the company has no right to modify the source, (by default) it has to convince the hackers to do it and to share their copyright.

>Further: "Human" is every live being with humans among its genetic ancestors.
>Is a trout a human? Yes, because a trout is a live being that has other trouts among its genetic ancestors.

The Judge is free to choose the base point of such recursion.
However, it would be hard to sustain that the judge, the parties and the developers are not humans so the definition must be wide enough to encompass all of them.
This in turn will likely encompass all the humanity (and if the judge is free to discard some other group, we have a totally different kind of problems)

So basically that definition is there to prevent discrimination against any group or minority or even against  people affected by genetic issues and so on.

>What is the trout readable form of the software?

This is an interesting question. I will reflect about it.
But I don't think this is relevant here.

>"know-how required to perform any of the activities" is not specific 
>enough. What exactly do I have to give the user?

The relevant original documentation you use yourself during development, deploy and so on.

You cannot use security through obscurity, for example.
You cannot make impractical to modify the software by taking code comments private and so on...

I will find a way to improve the wording.
By know-how I meant all of these things though.
Maybe I should simply use "documentation" insteas of know-how?

>These are only the verbal shortcomings that came to my mind after
>reading your text.

Well... I really thank you for sharing.
I'll try to address them as well as possible.

>Organizations get less grants, this is clearly against DFSG #6. Even 
>Organizations need to modify the code.

It doesn't limit what members of the organisation can do with the software.

It doesn't restrict what people IN the business can do with it.

OTOH businesses doesn't have minds that can study or hands that can code, there are always people that do these things and businesses can copy and sell copies of the Hack or services based on it on their behalf.

>Currently I can sell Debian on CDs. According to 3.4, if software under
>the HACK license is in Debian I am not allowed to do this anymore. This is against DFSG #9.

This is a really good point! Thanks!

The goal is not to force the distribution of the tools with the Hack but to impose their availability.
The goal is to avoid that a big company limit the freedom of users through say a custom compiler they don't distribute.
GPL imposes similar conditions through the definition of "Corresponding Source".

I could add "or by other reasonable means" so that the distribution can be left upstream.

OTOH consider that the tools in question are simply a C compiler and a standard libc and they are already in the distribution.

I will reflect about this. I might remove or completely reformulate the clause if it turns out that it is a problem. Or I could make it more similar to GPLv3 if that wording looks more clear.

>This license text is so vague that its consequences for Debian users are 
>not clear. So if it were up to me, I would not accept software under this 
>license to be part of Debian. I would even struggle with me to accept it for the non-free part.

I really thank you for sharing your concens in a way I can understand and address.

Given clause 3.4 implications (that to be honest I missed), I agree with you: without addressing it, a software under this License couldn't be distributed with Debian.

I'll do my best to fix this.

As for being vague, maybe there are more terms you think should be defined?

>Sorry, Debian is open for new licenses all the time, but I am afraid
>the HACK license needs a fair amount of rework.

I'm not in a rush.
I will fix these issues as soon as possible and try again.

Meanwhile, if you see other issues you want to share I would really appreciate them.

>   Thorsten


Reply to: