LLM script for error message improvement
Dear contributors, I hope you are doing well! I see there are contributors who are interested in working on error message improvements and persistent contribution, so I want to share an llm-based error message improvement script for helping your contribution. You can find a detail for the script at https://github.com/apache/spark/pull/41711. I believe this can help your error message improvement work, so I encourage you to take a look at the pull request and leverage the script. Please let me know if you have any questions or concerns. Thanks all for your time and contributions! Best regards, Haejoon
Re: LLM script for error message improvement
I think adding that dev tool script to improve the error message is fine. On Thu, 3 Aug 2023 at 10:24, Haejoon Lee wrote: > Dear contributors, I hope you are doing well! > > I see there are contributors who are interested in working on error > message improvements and persistent contribution, so I want to share an > llm-based error message improvement script for helping your contribution. > > You can find a detail for the script at > https://github.com/apache/spark/pull/41711. I believe this can help your > error message improvement work, so I encourage you to take a look at the > pull request and leverage the script. > > Please let me know if you have any questions or concerns. > > Thanks all for your time and contributions! > > Best regards, > > Haejoon >
Re: LLM script for error message improvement
+1 from my side, I'm fine to have it as a helper script On Thu, Aug 3, 2023 at 10:53 AM Hyukjin Kwon wrote: > I think adding that dev tool script to improve the error message is fine. > > On Thu, 3 Aug 2023 at 10:24, Haejoon Lee > wrote: > >> Dear contributors, I hope you are doing well! >> >> I see there are contributors who are interested in working on error >> message improvements and persistent contribution, so I want to share an >> llm-based error message improvement script for helping your contribution. >> >> You can find a detail for the script at >> https://github.com/apache/spark/pull/41711. I believe this can help your >> error message improvement work, so I encourage you to take a look at the >> pull request and leverage the script. >> >> Please let me know if you have any questions or concerns. >> >> Thanks all for your time and contributions! >> >> Best regards, >> >> Haejoon >> >
Re: LLM script for error message improvement
Additional information: Please check https://issues.apache.org/jira/browse/SPARK-37935 if you want to start contributing to improving error messages. You can create sub-tasks if you believe there are error messages that need improvement, in addition to the tasks listed in the umbrella JIRA. You can also refer to https://github.com/apache/spark/pull/41504, https://github.com/apache/spark/pull/41455 as an example PR. On Thu, Aug 3, 2023 at 1:10 PM Ruifeng Zheng wrote: > +1 from my side, I'm fine to have it as a helper script > > On Thu, Aug 3, 2023 at 10:53 AM Hyukjin Kwon wrote: > >> I think adding that dev tool script to improve the error message is fine. >> >> On Thu, 3 Aug 2023 at 10:24, Haejoon Lee >> wrote: >> >>> Dear contributors, I hope you are doing well! >>> >>> I see there are contributors who are interested in working on error >>> message improvements and persistent contribution, so I want to share an >>> llm-based error message improvement script for helping your contribution. >>> >>> You can find a detail for the script at >>> https://github.com/apache/spark/pull/41711. I believe this can help >>> your error message improvement work, so I encourage you to take a look at >>> the pull request and leverage the script. >>> >>> Please let me know if you have any questions or concerns. >>> >>> Thanks all for your time and contributions! >>> >>> Best regards, >>> >>> Haejoon >>> >>
Re: LLM script for error message improvement
I am sitting on the fence about that. In the linked PR Xiao wrote the following We published the error guideline a few years ago, but not all contributors adhered to it, resulting in variable quality in error messages. If a policy exists but is not enforced (if that's indeed the case, I didn't go through the source to confirm that) it might be useful to learn the reasons why it happens. Normally, I'd expect -Policy is too complex to enforce. In such case, additional tooling can be useful. -Policy is not well known, and the people responsible for introducing it are not committed to enforcing it. -Policy or some of its components don't really reflect community values and expectations. If the problem of suspected violations was never raised on our standard communication channel, and as far as I can tell, it has not, then introducing a new tool to enforce the policy seems a bit premature. If these were the only considerations, I'd say that improving the overall consistency of the project outweighs possible risks, even if the case for such might be poorly supported. However, there is an elephant in the room. It is another attempt, after SPARK-44546, to embed generative tools directly within the Spark dev workflow. By principle, I am not against such tools. In fact, it is pretty clear that they are already used by Spark committers, and even if we wanted to, there is little we can do to prevent that. In such cases, decisions which tools, if any, to use, to what extent and how to treat their output are the sole responsibility of contributors. In contrast, these proposals try to push a proprietary tool burdened with serious privacy and ethical issues and likely to introduce unclear liabilities as a standard or even required developer tool. I can't speak for others, but personally, I'm quite uneasy about it. If we go this way, I strongly believe that it should be preceded by a serious discussion, if not the development of a formal policy, about what categories of tools, to what capacity, to what extent are acceptable within the project. Ideally, with an official opinion from the ASF as the copyright owner. WDYT All? Shall we start a separate discussion? Best regards, Maciej Szymkiewicz Web:https://zero323.net PGP: A30CEF0C31A501EC On 8/3/23 18:33, Haejoon Lee wrote: Additional information: Please check https://issues.apache.org/jira/browse/SPARK-37935if you want to start contributing to improving error messages. You can create sub-tasks if you believe there are error messages that need improvement, in addition to the tasks listed in the umbrella JIRA. You can also refer to https://github.com/apache/spark/pull/41504, https://github.com/apache/spark/pull/41455as an example PR. On Thu, Aug 3, 2023 at 1:10 PM Ruifeng Zheng wrote: +1 from my side, I'm fine to have it as a helper script On Thu, Aug 3, 2023 at 10:53 AM Hyukjin Kwon wrote: I think adding that dev tool script to improve the error message is fine. On Thu, 3 Aug 2023 at 10:24, Haejoon Lee wrote: Dear contributors, I hope you are doing well! I see there are contributors who are interested in working on error message improvements and persistent contribution, so I want to share an llm-based error message improvement script for helping your contribution. You can find a detail for the script at https://github.com/apache/spark/pull/41711. I believe this can help your error message improvement work, so I encourage you to take a look at the pull request and leverage the script. Please let me know if you have any questions or concerns. Thanks all for your time and contributions! Best regards, Haejoon OpenPGP_signature Description: OpenPGP digital signature
Re: LLM script for error message improvement
Besides, in case a separate discussion doesn't happen, our core responsibility is to follow the ASF guidelines, including the ASF Generative Tooling Guidance (https://www.apache.org/legal/generative-tooling.html). As far as I understand it, both the first (which explicitly mentions ChatGPT) and the third acceptance conditions are not satisfied by this and the other mentioned PR. On a side note, we should probably take a closer look at the following 'When providing contributions authored using generative AI tooling, a recommended practice is for contributors to indicate the tooling used to create the contribution. This should be included as a token in the source control commit message, for example including the phrase “Generated-by: ”.' and consider adjusting PR template / merge tool accordingly. Best regards, Maciej Szymkiewicz Web:https://zero323.net PGP: A30CEF0C31A501EC On 8/3/23 22:14, Maciej wrote: I am sitting on the fence about that. In the linked PR Xiao wrote the following >We published the error guideline a few years ago, but not all contributors adhered to it, resulting in variable quality in error messages. If a policy exists but is not enforced (if that's indeed the case, I didn't go through the source to confirm that) it might be useful to learn the reasons why it happens. Normally, I'd expect -Policy is too complex to enforce. In such case, additional tooling can be useful. -Policy is not well known, and the people responsible for introducing it are not committed to enforcing it. -Policy or some of its components don't really reflect community values and expectations. If the problem of suspected violations was never raised on our standard communication channel, and as far as I can tell, it has not, then introducing a new tool to enforce the policy seems a bit premature. If these were the only considerations, I'd say that improving the overall consistency of the project outweighs possible risks, even if the case for such might be poorly supported. However, there is an elephant in the room. It is another attempt, after SPARK-44546, to embed generative tools directly within the Spark dev workflow. By principle, I am not against such tools. In fact, it is pretty clear that they are already used by Spark committers, and even if we wanted to, there is little we can do to prevent that. In such cases, decisions which tools, if any, to use, to what extent and how to treat their output are the sole responsibility of contributors. In contrast, these proposals try to push a proprietary tool burdened with serious privacy and ethical issues and likely to introduce unclear liabilities as a standard or even required developer tool. I can't speak for others, but personally, I'm quite uneasy about it. If we go this way, I strongly believe that it should be preceded by a serious discussion, if not the development of a formal policy, about what categories of tools, to what capacity, to what extent are acceptable within the project. Ideally, with an official opinion from the ASF as the copyright owner. WDYT All? Shall we start a separate discussion? Best regards, Maciej Szymkiewicz Web:https://zero323.net PGP: A30CEF0C31A501EC On 8/3/23 18:33, Haejoon Lee wrote: Additional information: Please check https://issues.apache.org/jira/browse/SPARK-37935if you want to start contributing to improving error messages. You can create sub-tasks if you believe there are error messages that need improvement, in addition to the tasks listed in the umbrella JIRA. You can also refer to https://github.com/apache/spark/pull/41504, https://github.com/apache/spark/pull/41455as an example PR. On Thu, Aug 3, 2023 at 1:10 PM Ruifeng Zheng wrote: +1 from my side, I'm fine to have it as a helper script On Thu, Aug 3, 2023 at 10:53 AM Hyukjin Kwon wrote: I think adding that dev tool script to improve the error message is fine. On Thu, 3 Aug 2023 at 10:24, Haejoon Lee wrote: Dear contributors, I hope you are doing well! I see there are contributors who are interested in working on error message improvements and persistent contribution, so I want to share an llm-based error message improvement script for helping your contribution. You can find a detail for the script at https://github.com/apache/spark/pull/41711. I believe this can help your error message improvement work, so I encourage you to take a look at the pull request and leverage the script. Please let me know if you have any questions or concerns. Thanks all for your time and contributions! Best regards, Haejoon OpenPGP_signature Description: OpenPGP digital signature