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

Re: GPT partlabel support for Mach



On Sat, May 30, 2020 at 10:32:36PM -0400, Joshua Branson wrote:
> Which webpage made you seethe? Was it this one?
> 
> https://www.gnu.org/licenses/why-assign.en.html
It was
http://git.savannah.gnu.org/cgit/gnulib.git/tree/doc/Copyright/request-assign.changes?id=1e972a8a37c153ddc15e604592f84f939eb3c2ad,
pointed to by
https://www.gnu.org/prep/maintain/html_node/Copyright-Papers.html,
which is what had come up on my cursory search, and demands I post my
postal address (to do what, exactly, with?) and "full legal name"
(a joke, but not funny) "(in ASCII characters)", which makes
not doing it easier, since I don't have one, even if I wanted to,
and birth year? Why the birth year of all things?

> To be fair the FSF is not a corporation.  It is a non-profit charity.
> The FSF will not make any money on your code contribution.
It's still a legal entity from the american empire with, inherently,
more power and money than me, that wants my personal data to further its
lobbying goals and questionable morals (cough rms). Colour me sceptical.

> I remember someone's code contribution making a Hurd
> debugger much more stable.  Samuel *really* wanted to add that patch to
> the Hurd, but since the developer was not willing to assign copyright,
> Samuel could not add the code properly.
Mayhap the lesson there is that this sort of predatory behaviour is
the reason the components of the GNU system not used in Linux
distributions are in such a sorry state, and not that Big Bad Businesses
are going to Eat Our Code.

Well, that or the broey no-documentation-as-a-feature attitude with the
signature GNU hospitality permeating the bug-tracker.

> The video talk that encouraged me to assign any assignments to the FSF
> (to be fair, I'm not really a developer...so I have never submitted
> copyright assignments), was this one:
> 
> https://www.youtube.com/watch?v=Fox1CuoP2QD
Poignantly, all that link shows is "Video unavailable".

> The basic idea behind that talk is this.  You have three ways to
> try to make "Free or open source software".
> 
> 1)  non-copyleft licenses (MIT, BSD, etc. or public domain)
> 2)  copyleft (GPL) + copyright assignments
> 3)  copyleft (GPL) w/o copyright assignments.  Each developer holds their
> individual copyright.
> 
> 1) Almost always results in a proprietary version of the software.  This
> means that some users are at the mercy of the developers, which is
> usually not good.
Well, if I had my way I would do carpentry, heavy machining,
oil refinery, power generation, and extract my own conflict minerals
from the global South, but I only do one and a half of those.
This puts me at mercy of developers who can do art at the Магнитогорский
Металлургический Комбинат and decidedly less art at Bełchatów.

And it puts them at ours. Specialisation; est la vie.

> 2) Is the tried and true route.  No company is dumb enough to try to
> make Emacs or GCC proprietary.  They would get sued, and they would
> quickly lose the lawsuit.  User freedom is protected.  Every two years
> or so, some University is upset because their "killer feature" for GCC
> will not be merged upstream until they assign copyright.  The FSF has
> never folded on this.  
Twice in so many paragraphs you have now provided examples of
The Software and therefore the user ending up worse off in the name of
"user freedom".

To you this is acceptable, since the FSF has (had) a near-monopoly
on some areas of development, and I doubt its choke-hold will release
any time soon. But, well, none of the software I regularly
contribute to slash maintain slash originally authored has a licence
this restrictive, and I haven't felt the need to sue anyone over it.

Maybe I prefer my software to be used and improve instead of
be either (a) not used or (b) used but in secret and not improve
because that'd be illegal or whatever. Maybe I'm just not enough
of a puritan.

> 3) […] It is up to each individual developer to enforce compliance[…]
Isn't that what happened here, since the code I added to was stolen from
Linux in the first place and that seemed to be good enough to merge?

This header, as found in most suckless' repositories' arg.h has
resonated with me more than any GPL shilling ever has:
1 /*
2  * Copy me if you can.
3  * by 20h
4  */
That sounds more succinct and representative of what I want
(my) software to be than the ever-growing 35k of the third GPL.

But hey, I'd rather things that work instead of lawyers;
наб

Attachment: signature.asc
Description: PGP signature


Reply to: