As someone handling dozens of OpenBSD servers and VMs at work, I dont care about copyright and licenses anymore.
Its 2026, just shut up and give us at least one modern filesystem already!
Vibe-Coded Ext4 for OpenBSD
61–70 of 84 posts
Re: Vibe-Coded Ext4 for OpenBSD
#62> So as of today, the Copyright system does not have a way for the output of a non-human produced set of files to contain the grant of permissions which the OpenBSD project needs to perform combination and redistribution. This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright! You can distribute it at will, not due to any…
> This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright!
This opinion is simplistic. LLMs are trained with pre-existing content, and their output directly reflects their training corpus. This means LLMs can generate output that matches verbatim existing work. And that work can very well be subjected to copyright.
Re: Vibe-Coded Ext4 for OpenBSD
#63> So as of today, the Copyright system does not have a way for the output of a non-human produced set of files to contain the grant of permissions which the OpenBSD project needs to perform combination and redistribution. This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright! You can distribute it at will, not due to any…
> This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright! This opinion is simplistic. LLMs are trained with pre-existing content, and their output directly reflects their training corpus. This means LLMs can generate output that matches verbatim existing work. And that work can very well be subjected to copyright.
Language models are good at translation and retrieval. This also extends to computer languages. LLMs translate from GPL to other licenses the same way Google translate turns French to English, except that the source material is implicitly stored in the LLM.
Re: Vibe-Coded Ext4 for OpenBSD
#64Earlier quoted context omitted.
Under the premise advanced in the quote, copyright is not being violated because there is none. Thus, the quote makes no sense as stated. It may be that, additionally, copyright is in fact being violated (I don't believe it myself), but if so that's a separate argument.
The premise of the quote does not contain the assumption that there is no copyright to the code. In fact the various contributors do not advance an opinion about whether code written by an AI can be granted copyright. Rather they are saying that it is obviously derivative of code that is under copyright, that is only distributed under terms which, however many dry cleaners process it, will still conflict with the lic…
> Rather they are saying that it is obviously derivative of code that is under copyright
Derivatives are not subject to copyright, unless they are close to, and contain substantial verbatim copies from, the original. It's a virtual certainty that a vibe-coded Ext4 FS is none of the above.
Redefining copyright as some weird patenting of similar ideas is absurd.
Re: Vibe-Coded Ext4 for OpenBSD
#65Regardless of license status, I'd be very hesitant to trust a vibe-coded filesystem implementation with my data.
Re: Vibe-Coded Ext4 for OpenBSD
#66> So as of today, the Copyright system does not have a way for the output of a non-human produced set of files to contain the grant of permissions which the OpenBSD project needs to perform combination and redistribution. This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright! You can distribute it at will, not due to any…
This all relies, as the article points out, on everyone looking directly at code that both looks like and works like the only extant codebase for EXT4 and nonetheless concluding that in fact the computer conjured it from the aether. If I wrote a program that zipped up the Linux kernel source, unzipped it, and grepped -v for comments it would not then be magically transformed into unattributable public domain software…
> If I wrote a program that zipped up the Linux kernel source, unzipped it, and grepped -v for comments it would not then be magically transformed into unattributable public domain software.
That's not the case here. A re-implemented piece of software that does not contain meaningful verbatim excerpts from the original is not subject to the copyright of the original.
Re: Vibe-Coded Ext4 for OpenBSD
#67Earlier quoted context omitted.
This all relies, as the article points out, on everyone looking directly at code that both looks like and works like the only extant codebase for EXT4 and nonetheless concluding that in fact the computer conjured it from the aether. If I wrote a program that zipped up the Linux kernel source, unzipped it, and grepped -v for comments it would not then be magically transformed into unattributable public domain software…
> If I wrote a program that zipped up the Linux kernel source, unzipped it, and grepped -v for comments it would not then be magically transformed into unattributable public domain software. That's not the case here. A re-implemented piece of software that does not contain meaningful verbatim excerpts from the original is not subject to the copyright of the original.
that is not certain. if you read code and then reimplement it using the original code as reference, the claim has been made that this falls under the copyright of the original because the new code is derived from the old code. unfortunately this particular situation has not yet been tested in court. but clean room implementations are done specifically to avoid the risk reading the original code poses. if this was clear cut then clean room development would not be needed.
this is similar to creating an extension to some program, because the extension could not be written without the original even if the interface the extension is using is a public API. the claim has been made that the copyright of the original program applies. i think the linux kernel is an example here.
see also these questions on stackexchange:
https://softwareengineering.stackexchange.com/questions/2087...
https://softwareengineering.stackexchange.com/questions/8675...
Re: Vibe-Coded Ext4 for OpenBSD
#68Earlier quoted context omitted.
The premise of the quote does not contain the assumption that there is no copyright to the code. In fact the various contributors do not advance an opinion about whether code written by an AI can be granted copyright. Rather they are saying that it is obviously derivative of code that is under copyright, that is only distributed under terms which, however many dry cleaners process it, will still conflict with the lic…
> Rather they are saying that it is obviously derivative of code that is under copyright Derivatives are not subject to copyright, unless they are close to, and contain substantial verbatim copies from, the original. It's a virtual certainty that a vibe-coded Ext4 FS is none of the above. Redefining copyright as some weird patenting of similar ideas is absurd.
see my response here: https://news.ycombinator.com/item?id=47557250
Re: Vibe-Coded Ext4 for OpenBSD
#69Earlier quoted context omitted.
> This seems extremely confused. The copyright system does not have a way to grant these permissions because the material is not covered under copyright! This opinion is simplistic. LLMs are trained with pre-existing content, and their output directly reflects their training corpus. This means LLMs can generate output that matches verbatim existing work. And that work can very well be subjected to copyright.
Language models are good at translation and retrieval. This also extends to computer languages. LLMs translate from GPL to other licenses the same way Google translate turns French to English, except that the source material is implicitly stored in the LLM.
this is disputed. see my comment here, especially the stackexchange links: https://news.ycombinator.com/edit?id=47557250
Re: Vibe-Coded Ext4 for OpenBSD
#70Earlier quoted context omitted.
> If I wrote a program that zipped up the Linux kernel source, unzipped it, and grepped -v for comments it would not then be magically transformed into unattributable public domain software. That's not the case here. A re-implemented piece of software that does not contain meaningful verbatim excerpts from the original is not subject to the copyright of the original.
that is not certain. if you read code and then reimplement it using the original code as reference, the claim has been made that this falls under the copyright of the original because the new code is derived from the old code. unfortunately this particular situation has not yet been tested in court. but clean room implementations are done specifically to avoid the risk reading the original code poses. if this was cle…
> this is similar to creating an extension to some program
There's no such thing as "an extension to some program". A derivative work is a work that contains the original. Using the privileges provided by copyright law, the creator may impose licensing restrictions on how the original work is used - but that's contract law, not copyright.
For example the GPL and the AGPL define different sets of use restrictions, none of that matters in this case because the original work is not being reproduced or used per se.
As I already said in my other, down-voted comment - copyright is only about verbatim, or near verbatim copies, in whole or in part - it's the spirit that both judgment and the letter of the law are supposed to follow. Copying of functionality is not subject to copyright.
For example, one can use the same topic for a work of poetry for a similar aesthetic effect and that doesn't infringe other poems.
The GPL used a hack to stretch copyright law into a near opposite but stretching it further goes into absurd territory, achieving the opposite of what the GPL claims to protect.