Topic: ckyf -- a tool for checking for non-free software

i was inspired by the poor job a tool with similar goals does:


this will not be a fork, this will be an entirely new tool. i think i can do a lot better than the original-- to the point where it might even be worth keeping, rather than blacklisting. it will have limitations, but it will be thorough enough that i think many of you would benefit from it.

i will keep you posted.

your suggestions are welcome. indeed, the blacklist here helped inspire it.


Re: ckyf -- a tool for checking for non-free software

Very nice idea

3 (edited by freemedia 2018-11-04 04:47:01)

Re: ckyf -- a tool for checking for non-free software

thanks. completely unoriginal, but if im the only person that uses this completely new edition then it will significantly advance and assist the development of something i care about, which is distro-libre.

im watching this video about bunnie huang today: https://www.media.mit.edu/videos/mltalks-2015-11-24/

and it says so many things i want people to know-- it talks about where we were with truly libre hardware design in 2015, whats taking so long, why are things expensive-- things that have real answers that we tend to gloss over with the geek/activist version of "freedom isnt free" (as in beer.)

which is true, but sort of glib compared to the details-- especially when the details in this video will help you understand, the cost of the novena hasnt got as much to do with the cost of freedom, its about design decisions and supply chains and where and how you manufacture, and yield vs scrap ratios. things i know little about, which those of us who want hardware that ryf more need more introduction to if we want to understand it beyond "freedom isnt free."

it gives you a little hope that people are working on this, that there is more progress to expect, beyond keeping your fingers crossed.

but my favourite part (more to do with this thread) is that nadya peek talks about how a lot of what huang is doing is sort of working on projects from the standpoint of an investigative journalist. the point is certainly not to make badges that light up in cool colour patterns.

the point is to demonstrate and investigate supply chain connections to quickly and reliably produce prototypes in quantities of 100, rather than 2 or 10,000-- which we had less data/process/streamlining on.

in other words-- this isnt just about the product and immediate applications. this isnt just about academic research.

what its really about is taking important questions, and building things that give us real answers to those questions. all the things ive worked on that im proudest of, from remixing multiple distros (which i had no experience with, but wanted to automate because im too lazy to do it any other way) to teaching language design as a beginner-friendly exercise, started with just questions i wanted the answer to.

if you really want to have fun, dont start from building products. everyone is doing that, youll get bored really quickly. ask someone who is tired of it, it can get really tedious and unrewarding to work on products all the time.

ive been working on ckyf and written very little code (but what ive written works, and a lot of the other parts ive actually written before for other things.)

the main reason i havent released it yet is that im half-incorporating (and trying to avoid) a really stupidly-licensed (non-software-licensed) blacklist from the fsf into a program that is gpl 3.

so im doing it transparently and i can deeply thank this very project-- hyperbola-- for their gpl3-licensed blacklist (even though i think a permissive gpl-compatible license would be better.)

thats better than a non-gpl-compatible license like the asinine gnu fdl used for the (otherwise extremely useful) libreplanet blacklist.

that being said, a lot of this is inspired by guidelines hosted on the libreplanet wiki. guidelines which are useful without copying them in whole or in part.

but here you go: if you do a blacklist, please:

a. use a permissive license (gpl3 is alright, but completely inappropriate. theres simply no need to copyleft a blacklist, which makes this excesssive)

b. use a gpl compatible free software license for your blacklist, not a documentation license.

otherwise, youre just wasting peoples time. copyleft was designed to prevent certain problems. those problems do not exist in the world of blacklists, so using copyleft for those is limiting without any real purpose. using the fdl is also limiting. by-sa 4.0 at least, is gpl3 compatible.

the fdl may have inspired better licenses, but its just awful that it is still used so often when it is such a lousy license for its purpose.

the only thing taking so long right now is the fdl, and the only thing saving me any trouble right now is the blacklist created for hyperbola. so thank you-- thank you for that thing!

as for people leaning on documentation licenses for data that is just as useful to a program as a human reader:

its all code! all of it is code. stop licensing it in a way that code cant use it because you insist its special-- it isnt. never use the fdl, use a free license that is gpl compatible, even if it is a license that is more appropriate for cultural/oer works.

(i lean on cc0 personally, but i dont expect people to do that. gpl compatibility is very important to me.)