Menu
▾
▴
opensc-devel
|
From: Ludovic R. <lud...@gm...> - 2025-07-17 09:10:29
|
Hello, I received the request bellow from GitHub. I had a look at the OpenSC organisation configuration on github. For now the ChatGPT Connector (installed one month ago, not by me) has access to the OpenSC/libp11 project only. I don't know who installed the ChatGPT Connector. The ChatGPT Connector app by openai is requesting updated permissions Read-only access to Checks Read-only access to Members Read-only access to Commit statuses The ChatGPT Connector already has access to: *Read and write* access to *Actions* *Read and write* access to *Contents* *Read and write* access to *Issues* *Read-only* access to *Metadata* *Read and write* access to *Pull requests* *Read and write* access to *Workflows* What should I do with the new request? Bye ---------- Forwarded message --------- De : GitHub <no...@gi...> Date: jeu. 17 juil. 2025 à 04:38 Subject: [GitHub] ChatGPT Connector is requesting updated permissions To: Ludovic Rousseau <lud...@gm...> Updated Permissions Request The GitHub App ChatGPT Connector is requesting additional access to your organization. Hello there! You’re receiving this email because the GitHub App *ChatGPT Connector*, which is currently installed on your *OpenSC* organization, has updated its permissions and is requesting additional access. Review permission request to accept or reject this change <https://github.com/organizations/OpenSC/settings/installations/72073247/permissions/update> You may choose to ignore this request, in which case ChatGPT Connector will retain its current permissions. You can view pending requests directly by visiting your organization’s settings page and clicking on Applications. If you run into problems, please contact support. <https://support.github.com/contact?tags=dotcom-integrations> Link not working? Paste the following link into your browser: https://github.com/organizations/OpenSC/settings/installations/72073247/permissions/update Thanks! -- Dr. Ludovic Rousseau |
|
From: Michał T. <Mic...@st...> - 2025-07-18 15:25:16
|
Hi Ludovic, I have requested access for the ChatGPT Connector and have also handled the request for additional read-only access permissions. No action is required from you. Please let me know if you are not comfortable with installing the ChatGPT Connector for libp11. The connector is simply a matter of convenience. It is not essential for the project, and I can remove it if necessary. Best regards, Mike On 7/17/25 11:10 AM, Ludovic Rousseau wrote: > Hello, > > I received the request bellow from GitHub. > I had a look at the OpenSC organisation configuration on github. For > now the ChatGPT Connector (installed one month ago, not by me) has > access to the OpenSC/libp11 project only. > > I don't know who installed the ChatGPT Connector. > > The ChatGPT Connector app by openai is requesting updated permissions > Read-only access to Checks > Read-only access to Members > Read-only access to Commit statuses > > > The ChatGPT Connector already has access to: > *Read and write* access to *Actions* > *Read and write* access to *Contents* > *Read and write* access to *Issues* > *Read-only* access to *Metadata* > *Read and write* access to *Pull requests* > *Read and write* access to *Workflows* > > What should I do with the new request? > > Bye > > ---------- Forwarded message --------- > De : *GitHub* <no...@gi...> > Date: jeu. 17 juil. 2025 à 04:38 > Subject: [GitHub] ChatGPT Connector is requesting updated permissions > To: Ludovic Rousseau <lud...@gm... > <mailto:ludovic.rousseau%2Bg...@gm...>> > > > Updated Permissions Request > > The GitHub App ChatGPT Connector is requesting additional access to > your organization. > > Hello there! > > You’re receiving this email because the GitHub App *ChatGPT > Connector*, which is currently installed on your *OpenSC* > organization, has updated its permissions and is requesting additional > access. > > Review permission request to accept or reject this change > <https://github.com/organizations/OpenSC/settings/installations/72073247/permissions/update> > > > You may choose to ignore this request, in which case ChatGPT Connector > will retain its current permissions. > > You can view pending requests directly by visiting your organization’s > settings page and clicking on Applications. If you run into problems, > please contact support. > <https://support.github.com/contact?tags=dotcom-integrations> > > Link not working? Paste the following link into your browser: > https://github.com/organizations/OpenSC/settings/installations/72073247/permissions/update > > > Thanks! > > > > -- > Dr. Ludovic Rousseau > > > _______________________________________________ > Opensc-devel mailing list > Ope...@li... > https://lists.sourceforge.net/lists/listinfo/opensc-devel |
|
From: Alexander B. <al...@al...> - 2025-07-18 20:19:44
|
Hi Mike, Why was it given read-write access? Are we gonna start vibe coding OpenSC? Cheers, Alex |
|
From: Michał T. <Mic...@st...> - 2025-07-18 20:55:29
|
Hi Alex, Alexander Burke wrote: > Why was it given read-write access? It can submit pull requests when ordered to do so. Also, it was only given access to the libp11 subproject, and not the entire OpenSC. > Are we gonna start vibe coding Open SC? Definitely not, although I've seen worse pull requests submitted by humans... You may think of it as yet another potential (sic!) contributor. Best regards, Mike |
|
From: Ondřej S. <on...@su...> - 2025-07-19 06:56:21
|
> On 18. 7. 2025, at 22:56, Michał Trojnara via Opensc-devel <ope...@li...> wrote: > You may think of it as yet another potential (sic!) contributor. Who would be author then to give copyright assignment to the project and who owns the IP to the LLM generated code? These are important questions that need to be answered before merging anything into an open source project under free software license. > I've seen worse pull requests submitted by humans... Don’t use this as argument. Ever never in any context. Ondrej -- Ondřej Surý (He/Him) |
|
From: Michał T. <Mic...@st...> - 2025-07-19 21:44:31
|
Ondřej Surý <on...@su...> wrote: > Who would be author then to give copyright assignment to the project and who owns the IP to the LLM generated code? I will. Best regards, Mike |
|
From: Graham L. <mi...@sh...> - 2025-07-23 10:23:12
|
On 19 Jul 2025, at 22:44, Michał Trojnara via Opensc-devel <ope...@li...> wrote: > Ondřej Surý <on...@su...> wrote: > > Who would be author then to give copyright assignment to the project and who owns the IP to the LLM generated code? > > I will. You won't, and this has been settled in law. https://en.wikipedia.org/wiki/Monkey_selfie_copyright_dispute Regards, Graham -- |
|
From: Alexander B. <al...@al...> - 2025-07-21 08:45:40
|
Hi Mike, You can't legally do that, because the code comes from (1) a program you did not write, which (2) was trained on code written (and copyrighted) by others which it is regurgitating, which is also likely to be licensed differently. This is what the Wikimedia Foundation has to say about it: https://en.wikipedia.org/wiki/Wikipedia:Large_language_models_and_copyright This is a legal and technical minefield and I encourage you, and the project, to obtain legal advice before using this "tool". Best wishes, Alex |
|
From: Michał T. <Mic...@st...> - 2025-07-21 10:54:00
|
Hi Alex, On 7/21/25 10:45 AM, Alexander Burke via Opensc-devel wrote: > You can't legally do that, because the code comes from (1) a program > you did not write, which (2) was trained on code written (and > copyrighted) by others which it is regurgitating, which is also likely > to be licensed differently. Ad 1. You could argue the same about any output of non-AI code generators and/or compilers. I respectfully disagree. I came up with the prompt, and I retain intellectual property rights to the generated code. Ad 2. I myself was trained using code written (and copyrighted) by others. I tend to reuse good programming patterns I have seen before. This is how learning works. AI is no different. There is no such thing as 100% originality. You are always inspired by, and you always reuse, thoughts previously expressed by other people. > This is what the Wikimedia Foundation has to say about it: > https://en.wikipedia.org/wiki/Wikipedia:Large_language_models_and_copyright Isn't it better to use a more specific source, such as https://openai.com/policies/terms-of-use/ ? /As between you and OpenAI, and to the extent permitted by applicable law, you (a) retain your ownership rights in Input and (b) own the Output. We hereby assign to you all our right, title, and interest, if any, in and to Output. / > This is a legal and technical minefield and I encourage you, and the > project, to obtain legal advice before using this "tool". Well, life is a legal, technical, and emotional minefield. Common sense helps, though. I understand that the legal status of AI-generated code is still under debate, and interpretations may vary across jurisdictions. While I am confident in my approach, I remain open to further discussion and will reconsider my position if authoritative legal guidance becomes available. Regardless of how the code is generated, I ensure that all contributions are diligently reviewed for license compatibility and compliance with project policies. I believe that maintaining these standards is essential for the integrity and sustainability of the project. Have a nice day, Mike |
|
From: Juraj Š. <ju...@sa...> - 2025-07-21 12:59:37
|
On Sat, 2025-07-19 at 23:44 +0200, Michał Trojnara via Opensc-devel wrote: > > Who would be author then to give copyright assignment to the > > project and who owns the IP to the LLM generated code? > > I will. Dear Michał I understand you intend to mislead others about code authorship. This might put the project at risk. As if merging LLM-generated garbage were not risky enough... > You could argue the same about any output of non-AI code generators > and/or compilers. I respectfully disagree. I came up with the prompt, > and I retain intellectual property rights to the generated code. A LLM merely transforms code originally written (and therefore copyrighted) by others. To use the output, one needs permission from the original author(s). Which is indeed how output of non-AI code generators & compilers is normally treated. > I myself was trained using code written (and copyrighted) by others. > I tend to reuse good programming patterns I have seen before. We all do that. Programming patterns are not protected by copyright. Code is. Copy-pasting has always been risky. > the legal status of AI-generated code is still under debate The current (rather clear) status means that such code simply cannot be included in OpenSC or indeed in any free software. Best, Juraj |
|
From: Paul W. <ma...@al...> - 2025-07-23 09:44:39
|
В Mon, 21 Jul 2025 14:44:08 +0200 Juraj Šarinay <ju...@sa...> пишет: > > the legal status of AI-generated code is still under debate > > The current (rather clear) status means that such code simply cannot be > included in OpenSC or indeed in any free software. Very interesting, and very important! Don't you know, is there any deeper research about the subject? |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X