You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@impala.apache.org by Zoltán Borók-Nagy <bo...@apache.org> on 2025/12/15 12:38:54 UTC
AI Tooling Guidance
Dear Contributors,
With AI tools getting more popular these days, please take some time to
read the ASF Generative Tooling Guidance:
https://www.apache.org/legal/generative-tooling.html
Kind regards,
Zoltan Borok-Nagy
Re: AI Tooling Guidance
Posted by Jason Fehr <ja...@apache.org>.
Thanks Michael!
On 2026/06/03 20:37:10 Michael Smith wrote:
> I'll update
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> to request that commit messages include "Assisted-by: <model> (<tool>)" if
> used.
>
> On Mon, Jun 1, 2026 at 2:44 AM Zoltán Borók-Nagy <bo...@apache.org>
> wrote:
>
> > Some tools (e.g. Claude) adds "Co-Authored-By" by default. However, that
> > tag should probably be reserved for humans.
> > "Assisted-by:" or "AI-assistant:" is more explanatory.
> >
> > On Fri, May 29, 2026 at 7:51 PM Jason Fehr <ja...@apache.org> wrote:
> >
> > > +1 to using "Assisted-by".
> > >
> > > The Apache guidelines at
> > >
> > https://www.apache.org/legal/generative-tooling.html#:~:text=When%20providing%20contributions,Tooling%2DProvenance%20file
> > > do not dictate any particular indicator, but I still think we should have
> > > an indicator that AI tooling was used to better communicate who/what
> > > exactly generated the change.
> > >
> > > On 2026/05/04 23:23:56 Quanlong Huang wrote:
> > > > I think "Assisted-by" used in the Linux communities is more
> > > > suitable/accurate. It means "AI is considered a tool (similar to a
> > > compiler
> > > > or static analysis tool) and not an author. The human contributor must
> > > > review, understand, and vouch for every line of code, regardless of
> > > whether
> > > > it was written by them or an AI".
> > > >
> > > > https://docs.kernel.org/process/coding-assistants.html
> > > >
> > https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
> > > >
> > > > Quanlong
> > > >
> > > > On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <
> > > boroknagyz@apache.org>
> > > > wrote:
> > > >
> > > > > Thanks, Michael. Yeah, I agree that we should have clear guidelines
> > > and I
> > > > > also don't think we should overuse "Generated-by" if it was only used
> > > for
> > > > > fairly trivial stuff.
> > > > >
> > > > > I also find the scope of "Generated-by" in the commit message too
> > > large as
> > > > > it doesn't provide any hint which parts were AI generated.
> > > > >
> > > > > I think it would make sense to add "Generated-by" to class/function
> > > > > comments that were mostly generated by AI.
> > > > >
> > > > > - Zoltan
> > > > >
> > > > > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <
> > > michael.smith@cloudera.com>
> > > > > wrote:
> > > > >
> > > > > > These guidelines seem to leave some of the specifics up to
> > individual
> > > > > > projects. I think we should agree on some clearer guidelines for
> > > > > > contributors, and post it on our wiki
> > > > > > <
> > > > >
> > >
> > https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > > > > >.
> > > > > >
> > > > > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do
> > > more
> > > > > than
> > > > > > accept small auto-complete suggestions (not much different to what
> > > > > > pre-generative tooling produced) that I can immediately understand,
> > > I add
> > > > > > "Generated-by: GitHub Copilot (<model>)" to the commit message,
> > like
> > > in
> > > > > > apache/impala#ac1c11dd
> > > > > > <
> > > > > >
> > > > >
> > >
> > https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > > > > >
> > > > > > or apache/impala#c0b3580
> > > > > > <
> > > > > >
> > > > >
> > >
> > https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > > > > >
> > > > > > .
> > > > > >
> > > > > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <
> > > boroknagyz@apache.org
> > > > > >
> > > > > > wrote:
> > > > > >
> > > > > > > Dear Contributors,
> > > > > > >
> > > > > > > With AI tools getting more popular these days, please take some
> > > time to
> > > > > > > read the ASF Generative Tooling Guidance:
> > > > > > > https://www.apache.org/legal/generative-tooling.html
> > > > > > >
> > > > > > > Kind regards,
> > > > > > > Zoltan Borok-Nagy
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Michael Smith <mi...@cloudera.com>.
I'll update
https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
to request that commit messages include "Assisted-by: <model> (<tool>)" if
used.
On Mon, Jun 1, 2026 at 2:44 AM Zoltán Borók-Nagy <bo...@apache.org>
wrote:
> Some tools (e.g. Claude) adds "Co-Authored-By" by default. However, that
> tag should probably be reserved for humans.
> "Assisted-by:" or "AI-assistant:" is more explanatory.
>
> On Fri, May 29, 2026 at 7:51 PM Jason Fehr <ja...@apache.org> wrote:
>
> > +1 to using "Assisted-by".
> >
> > The Apache guidelines at
> >
> https://www.apache.org/legal/generative-tooling.html#:~:text=When%20providing%20contributions,Tooling%2DProvenance%20file
> > do not dictate any particular indicator, but I still think we should have
> > an indicator that AI tooling was used to better communicate who/what
> > exactly generated the change.
> >
> > On 2026/05/04 23:23:56 Quanlong Huang wrote:
> > > I think "Assisted-by" used in the Linux communities is more
> > > suitable/accurate. It means "AI is considered a tool (similar to a
> > compiler
> > > or static analysis tool) and not an author. The human contributor must
> > > review, understand, and vouch for every line of code, regardless of
> > whether
> > > it was written by them or an AI".
> > >
> > > https://docs.kernel.org/process/coding-assistants.html
> > >
> https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
> > >
> > > Quanlong
> > >
> > > On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <
> > boroknagyz@apache.org>
> > > wrote:
> > >
> > > > Thanks, Michael. Yeah, I agree that we should have clear guidelines
> > and I
> > > > also don't think we should overuse "Generated-by" if it was only used
> > for
> > > > fairly trivial stuff.
> > > >
> > > > I also find the scope of "Generated-by" in the commit message too
> > large as
> > > > it doesn't provide any hint which parts were AI generated.
> > > >
> > > > I think it would make sense to add "Generated-by" to class/function
> > > > comments that were mostly generated by AI.
> > > >
> > > > - Zoltan
> > > >
> > > > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <
> > michael.smith@cloudera.com>
> > > > wrote:
> > > >
> > > > > These guidelines seem to leave some of the specifics up to
> individual
> > > > > projects. I think we should agree on some clearer guidelines for
> > > > > contributors, and post it on our wiki
> > > > > <
> > > >
> >
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > > > >.
> > > > >
> > > > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do
> > more
> > > > than
> > > > > accept small auto-complete suggestions (not much different to what
> > > > > pre-generative tooling produced) that I can immediately understand,
> > I add
> > > > > "Generated-by: GitHub Copilot (<model>)" to the commit message,
> like
> > in
> > > > > apache/impala#ac1c11dd
> > > > > <
> > > > >
> > > >
> >
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > > > >
> > > > > or apache/impala#c0b3580
> > > > > <
> > > > >
> > > >
> >
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > > > >
> > > > > .
> > > > >
> > > > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <
> > boroknagyz@apache.org
> > > > >
> > > > > wrote:
> > > > >
> > > > > > Dear Contributors,
> > > > > >
> > > > > > With AI tools getting more popular these days, please take some
> > time to
> > > > > > read the ASF Generative Tooling Guidance:
> > > > > > https://www.apache.org/legal/generative-tooling.html
> > > > > >
> > > > > > Kind regards,
> > > > > > Zoltan Borok-Nagy
> > > > > >
> > > > >
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Zoltán Borók-Nagy <bo...@apache.org>.
Some tools (e.g. Claude) adds "Co-Authored-By" by default. However, that
tag should probably be reserved for humans.
"Assisted-by:" or "AI-assistant:" is more explanatory.
On Fri, May 29, 2026 at 7:51 PM Jason Fehr <ja...@apache.org> wrote:
> +1 to using "Assisted-by".
>
> The Apache guidelines at
> https://www.apache.org/legal/generative-tooling.html#:~:text=When%20providing%20contributions,Tooling%2DProvenance%20file
> do not dictate any particular indicator, but I still think we should have
> an indicator that AI tooling was used to better communicate who/what
> exactly generated the change.
>
> On 2026/05/04 23:23:56 Quanlong Huang wrote:
> > I think "Assisted-by" used in the Linux communities is more
> > suitable/accurate. It means "AI is considered a tool (similar to a
> compiler
> > or static analysis tool) and not an author. The human contributor must
> > review, understand, and vouch for every line of code, regardless of
> whether
> > it was written by them or an AI".
> >
> > https://docs.kernel.org/process/coding-assistants.html
> > https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
> >
> > Quanlong
> >
> > On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <
> boroknagyz@apache.org>
> > wrote:
> >
> > > Thanks, Michael. Yeah, I agree that we should have clear guidelines
> and I
> > > also don't think we should overuse "Generated-by" if it was only used
> for
> > > fairly trivial stuff.
> > >
> > > I also find the scope of "Generated-by" in the commit message too
> large as
> > > it doesn't provide any hint which parts were AI generated.
> > >
> > > I think it would make sense to add "Generated-by" to class/function
> > > comments that were mostly generated by AI.
> > >
> > > - Zoltan
> > >
> > > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <
> michael.smith@cloudera.com>
> > > wrote:
> > >
> > > > These guidelines seem to leave some of the specifics up to individual
> > > > projects. I think we should agree on some clearer guidelines for
> > > > contributors, and post it on our wiki
> > > > <
> > >
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > > >.
> > > >
> > > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do
> more
> > > than
> > > > accept small auto-complete suggestions (not much different to what
> > > > pre-generative tooling produced) that I can immediately understand,
> I add
> > > > "Generated-by: GitHub Copilot (<model>)" to the commit message, like
> in
> > > > apache/impala#ac1c11dd
> > > > <
> > > >
> > >
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > > >
> > > > or apache/impala#c0b3580
> > > > <
> > > >
> > >
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > > >
> > > > .
> > > >
> > > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <
> boroknagyz@apache.org
> > > >
> > > > wrote:
> > > >
> > > > > Dear Contributors,
> > > > >
> > > > > With AI tools getting more popular these days, please take some
> time to
> > > > > read the ASF Generative Tooling Guidance:
> > > > > https://www.apache.org/legal/generative-tooling.html
> > > > >
> > > > > Kind regards,
> > > > > Zoltan Borok-Nagy
> > > > >
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Jason Fehr <ja...@apache.org>.
+1 to using "Assisted-by".
The Apache guidelines at https://www.apache.org/legal/generative-tooling.html#:~:text=When%20providing%20contributions,Tooling%2DProvenance%20file do not dictate any particular indicator, but I still think we should have an indicator that AI tooling was used to better communicate who/what exactly generated the change.
On 2026/05/04 23:23:56 Quanlong Huang wrote:
> I think "Assisted-by" used in the Linux communities is more
> suitable/accurate. It means "AI is considered a tool (similar to a compiler
> or static analysis tool) and not an author. The human contributor must
> review, understand, and vouch for every line of code, regardless of whether
> it was written by them or an AI".
>
> https://docs.kernel.org/process/coding-assistants.html
> https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
>
> Quanlong
>
> On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <bo...@apache.org>
> wrote:
>
> > Thanks, Michael. Yeah, I agree that we should have clear guidelines and I
> > also don't think we should overuse "Generated-by" if it was only used for
> > fairly trivial stuff.
> >
> > I also find the scope of "Generated-by" in the commit message too large as
> > it doesn't provide any hint which parts were AI generated.
> >
> > I think it would make sense to add "Generated-by" to class/function
> > comments that were mostly generated by AI.
> >
> > - Zoltan
> >
> > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <mi...@cloudera.com>
> > wrote:
> >
> > > These guidelines seem to leave some of the specifics up to individual
> > > projects. I think we should agree on some clearer guidelines for
> > > contributors, and post it on our wiki
> > > <
> > https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > >.
> > >
> > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do more
> > than
> > > accept small auto-complete suggestions (not much different to what
> > > pre-generative tooling produced) that I can immediately understand, I add
> > > "Generated-by: GitHub Copilot (<model>)" to the commit message, like in
> > > apache/impala#ac1c11dd
> > > <
> > >
> > https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > >
> > > or apache/impala#c0b3580
> > > <
> > >
> > https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > >
> > > .
> > >
> > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <boroknagyz@apache.org
> > >
> > > wrote:
> > >
> > > > Dear Contributors,
> > > >
> > > > With AI tools getting more popular these days, please take some time to
> > > > read the ASF Generative Tooling Guidance:
> > > > https://www.apache.org/legal/generative-tooling.html
> > > >
> > > > Kind regards,
> > > > Zoltan Borok-Nagy
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Quanlong Huang <hu...@gmail.com>.
In the future if we use AI in code review, the label also helps us to use a
different model which reduces the risk of shared biases and ensures that
the reviewer catches flaws the generative model might have overlooked.
On Wed, May 6, 2026 at 2:16 AM Michael Smith <mi...@cloudera.com>
wrote:
> I like that framing.
>
> I've been procrastinating because I'm not sure how the label would change
> how we review PRs. It mostly becomes a record of potential copyright
> violation if, in the future, that becomes a thing for a particular
> tool/model.
>
> Code review in open-source has to be thorough no matter what, and relies on
> trust between people established over time to scale. It's not worth the
> review time if the contributor hasn't clearly put in effort to understand
> and validate their contribution.
>
> On Mon, May 4, 2026 at 4:25 PM Quanlong Huang <hu...@gmail.com>
> wrote:
>
> > I think "Assisted-by" used in the Linux communities is more
> > suitable/accurate. It means "AI is considered a tool (similar to a
> compiler
> > or static analysis tool) and not an author. The human contributor must
> > review, understand, and vouch for every line of code, regardless of
> whether
> > it was written by them or an AI".
> >
> > https://docs.kernel.org/process/coding-assistants.html
> > https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
> >
> > Quanlong
> >
> > On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <
> boroknagyz@apache.org>
> > wrote:
> >
> > > Thanks, Michael. Yeah, I agree that we should have clear guidelines
> and I
> > > also don't think we should overuse "Generated-by" if it was only used
> for
> > > fairly trivial stuff.
> > >
> > > I also find the scope of "Generated-by" in the commit message too large
> > as
> > > it doesn't provide any hint which parts were AI generated.
> > >
> > > I think it would make sense to add "Generated-by" to class/function
> > > comments that were mostly generated by AI.
> > >
> > > - Zoltan
> > >
> > > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <
> > michael.smith@cloudera.com>
> > > wrote:
> > >
> > > > These guidelines seem to leave some of the specifics up to individual
> > > > projects. I think we should agree on some clearer guidelines for
> > > > contributors, and post it on our wiki
> > > > <
> > >
> >
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > > >.
> > > >
> > > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do
> more
> > > than
> > > > accept small auto-complete suggestions (not much different to what
> > > > pre-generative tooling produced) that I can immediately understand, I
> > add
> > > > "Generated-by: GitHub Copilot (<model>)" to the commit message, like
> in
> > > > apache/impala#ac1c11dd
> > > > <
> > > >
> > >
> >
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > > >
> > > > or apache/impala#c0b3580
> > > > <
> > > >
> > >
> >
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > > >
> > > > .
> > > >
> > > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <
> > boroknagyz@apache.org
> > > >
> > > > wrote:
> > > >
> > > > > Dear Contributors,
> > > > >
> > > > > With AI tools getting more popular these days, please take some
> time
> > to
> > > > > read the ASF Generative Tooling Guidance:
> > > > > https://www.apache.org/legal/generative-tooling.html
> > > > >
> > > > > Kind regards,
> > > > > Zoltan Borok-Nagy
> > > > >
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Michael Smith <mi...@cloudera.com>.
I like that framing.
I've been procrastinating because I'm not sure how the label would change
how we review PRs. It mostly becomes a record of potential copyright
violation if, in the future, that becomes a thing for a particular
tool/model.
Code review in open-source has to be thorough no matter what, and relies on
trust between people established over time to scale. It's not worth the
review time if the contributor hasn't clearly put in effort to understand
and validate their contribution.
On Mon, May 4, 2026 at 4:25 PM Quanlong Huang <hu...@gmail.com>
wrote:
> I think "Assisted-by" used in the Linux communities is more
> suitable/accurate. It means "AI is considered a tool (similar to a compiler
> or static analysis tool) and not an author. The human contributor must
> review, understand, and vouch for every line of code, regardless of whether
> it was written by them or an AI".
>
> https://docs.kernel.org/process/coding-assistants.html
> https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
>
> Quanlong
>
> On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <bo...@apache.org>
> wrote:
>
> > Thanks, Michael. Yeah, I agree that we should have clear guidelines and I
> > also don't think we should overuse "Generated-by" if it was only used for
> > fairly trivial stuff.
> >
> > I also find the scope of "Generated-by" in the commit message too large
> as
> > it doesn't provide any hint which parts were AI generated.
> >
> > I think it would make sense to add "Generated-by" to class/function
> > comments that were mostly generated by AI.
> >
> > - Zoltan
> >
> > On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <
> michael.smith@cloudera.com>
> > wrote:
> >
> > > These guidelines seem to leave some of the specifics up to individual
> > > projects. I think we should agree on some clearer guidelines for
> > > contributors, and post it on our wiki
> > > <
> >
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > > >.
> > >
> > > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do more
> > than
> > > accept small auto-complete suggestions (not much different to what
> > > pre-generative tooling produced) that I can immediately understand, I
> add
> > > "Generated-by: GitHub Copilot (<model>)" to the commit message, like in
> > > apache/impala#ac1c11dd
> > > <
> > >
> >
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > > >
> > > or apache/impala#c0b3580
> > > <
> > >
> >
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > > >
> > > .
> > >
> > > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <
> boroknagyz@apache.org
> > >
> > > wrote:
> > >
> > > > Dear Contributors,
> > > >
> > > > With AI tools getting more popular these days, please take some time
> to
> > > > read the ASF Generative Tooling Guidance:
> > > > https://www.apache.org/legal/generative-tooling.html
> > > >
> > > > Kind regards,
> > > > Zoltan Borok-Nagy
> > > >
> > >
> >
>
Re: AI Tooling Guidance
Posted by Quanlong Huang <hu...@gmail.com>.
I think "Assisted-by" used in the Linux communities is more
suitable/accurate. It means "AI is considered a tool (similar to a compiler
or static analysis tool) and not an author. The human contributor must
review, understand, and vouch for every line of code, regardless of whether
it was written by them or an AI".
https://docs.kernel.org/process/coding-assistants.html
https://docs.rockylinux.org/10/guides/contribute/ai-contribution-policy/
Quanlong
On Thu, Jan 29, 2026 at 10:07 PM Zoltán Borók-Nagy <bo...@apache.org>
wrote:
> Thanks, Michael. Yeah, I agree that we should have clear guidelines and I
> also don't think we should overuse "Generated-by" if it was only used for
> fairly trivial stuff.
>
> I also find the scope of "Generated-by" in the commit message too large as
> it doesn't provide any hint which parts were AI generated.
>
> I think it would make sense to add "Generated-by" to class/function
> comments that were mostly generated by AI.
>
> - Zoltan
>
> On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <mi...@cloudera.com>
> wrote:
>
> > These guidelines seem to leave some of the specifics up to individual
> > projects. I think we should agree on some clearer guidelines for
> > contributors, and post it on our wiki
> > <
> https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> > >.
> >
> > I do have GitHub Copilot enabled, and my rule-of-thumb is if I do more
> than
> > accept small auto-complete suggestions (not much different to what
> > pre-generative tooling produced) that I can immediately understand, I add
> > "Generated-by: GitHub Copilot (<model>)" to the commit message, like in
> > apache/impala#ac1c11dd
> > <
> >
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> > >
> > or apache/impala#c0b3580
> > <
> >
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> > >
> > .
> >
> > On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <boroknagyz@apache.org
> >
> > wrote:
> >
> > > Dear Contributors,
> > >
> > > With AI tools getting more popular these days, please take some time to
> > > read the ASF Generative Tooling Guidance:
> > > https://www.apache.org/legal/generative-tooling.html
> > >
> > > Kind regards,
> > > Zoltan Borok-Nagy
> > >
> >
>
Re: AI Tooling Guidance
Posted by Zoltán Borók-Nagy <bo...@apache.org>.
Thanks, Michael. Yeah, I agree that we should have clear guidelines and I
also don't think we should overuse "Generated-by" if it was only used for
fairly trivial stuff.
I also find the scope of "Generated-by" in the commit message too large as
it doesn't provide any hint which parts were AI generated.
I think it would make sense to add "Generated-by" to class/function
comments that were mostly generated by AI.
- Zoltan
On Tue, Jan 27, 2026 at 9:59 PM Michael Smith <mi...@cloudera.com>
wrote:
> These guidelines seem to leave some of the specifics up to individual
> projects. I think we should agree on some clearer guidelines for
> contributors, and post it on our wiki
> <https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala
> >.
>
> I do have GitHub Copilot enabled, and my rule-of-thumb is if I do more than
> accept small auto-complete suggestions (not much different to what
> pre-generative tooling produced) that I can immediately understand, I add
> "Generated-by: GitHub Copilot (<model>)" to the commit message, like in
> apache/impala#ac1c11dd
> <
> https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41
> >
> or apache/impala#c0b3580
> <
> https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6
> >
> .
>
> On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <bo...@apache.org>
> wrote:
>
> > Dear Contributors,
> >
> > With AI tools getting more popular these days, please take some time to
> > read the ASF Generative Tooling Guidance:
> > https://www.apache.org/legal/generative-tooling.html
> >
> > Kind regards,
> > Zoltan Borok-Nagy
> >
>
Re: AI Tooling Guidance
Posted by Michael Smith <mi...@cloudera.com>.
These guidelines seem to leave some of the specifics up to individual
projects. I think we should agree on some clearer guidelines for
contributors, and post it on our wiki
<https://cwiki.apache.org/confluence/display/IMPALA/Contributing+to+Impala>.
I do have GitHub Copilot enabled, and my rule-of-thumb is if I do more than
accept small auto-complete suggestions (not much different to what
pre-generative tooling produced) that I can immediately understand, I add
"Generated-by: GitHub Copilot (<model>)" to the commit message, like in
apache/impala#ac1c11dd
<https://github.com/apache/impala/commit/ac1c11dd8256e8a81e138f43663de06610441d41>
or apache/impala#c0b3580
<https://github.com/apache/impala/commit/c0b35807543f96cb382240717ab552d17f6f78a6>
.
On Mon, Dec 15, 2025 at 4:39 AM Zoltán Borók-Nagy <bo...@apache.org>
wrote:
> Dear Contributors,
>
> With AI tools getting more popular these days, please take some time to
> read the ASF Generative Tooling Guidance:
> https://www.apache.org/legal/generative-tooling.html
>
> Kind regards,
> Zoltan Borok-Nagy
>