You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@calcite.apache.org by Mihai Budiu <mb...@gmail.com> on 2026/01/10 17:08:58 UTC
incorrect commit in main
Hello all,
By mistake I have merged a PR yesterday without squashing a commit.
Since there has been another PR merged following it, I propose to leave it as it is.
[<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
*
Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
*
[<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
I will try to be more careful in the future.
Thank you,
Mihai
Re: incorrect commit in main
Posted by Mihai Budiu <mb...@gmail.com>.
I have updated the main branch.
Thank you everyone for your patience.
Mihai
________________________________
From: Julian Hyde <jh...@gmail.com>
Sent: Saturday, January 10, 2026 6:11 PM
To: dev@calcite.apache.org <de...@calcite.apache.org>
Subject: Re: incorrect commit in main
I guess the question is “should downstream tools assume that the project never rewrites history?”. Whichever answer we choose, there is a cost.
Your script could use “ git merge-base --is-ancestor <commit_hash> main “ to detect whether history has been rewritten.
> On Jan 10, 2026, at 11:19 AM, Mihai Budiu <mb...@gmail.com> wrote:
>
> The build of our compiler includes the hash of a commit in Calcite's main; we track main periodically, since we want to incorporate bug fixes more frequently than Calcite releases.
>
> I can change the commit in our build, but maybe there are other projects which do this as well?
>
> Mihai
>
> ________________________________
> From: Julian Hyde <jh...@gmail.com>
> Sent: Saturday, January 10, 2026 10:40 AM
> To: dev@calcite.apache.org <de...@calcite.apache.org>
> Subject: Re: incorrect commit in main
>
> Personally, I would rebase main (including the following commit). It’s unlikely that you will break someone, because the contents of main will be unchanged by your rebase.
>
>> On Jan 10, 2026, at 9:09 AM, Mihai Budiu <mb...@gmail.com> wrote:
>>
>> Hello all,
>>
>> By mistake I have merged a PR yesterday without squashing a commit.
>> Since there has been another PR merged following it, I propose to leave it as it is.
>>
>> [<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
>>
>> *
>> Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
>>
>> *
>> [<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
>>
>> I will try to be more careful in the future.
>>
>> Thank you,
>> Mihai
>>
>>
Re: incorrect commit in main
Posted by Mihai Budiu <mb...@gmail.com>.
I will squash the commit in main soon. Please refrain from merging more stuff for a few hours. Thank you
________________________________
From: Julian Hyde <jh...@gmail.com>
Sent: Saturday, January 10, 2026 6:11:28 PM
To: dev@calcite.apache.org <de...@calcite.apache.org>
Subject: Re: incorrect commit in main
I guess the question is “should downstream tools assume that the project never rewrites history?”. Whichever answer we choose, there is a cost.
Your script could use “ git merge-base --is-ancestor <commit_hash> main “ to detect whether history has been rewritten.
> On Jan 10, 2026, at 11:19 AM, Mihai Budiu <mb...@gmail.com> wrote:
>
> The build of our compiler includes the hash of a commit in Calcite's main; we track main periodically, since we want to incorporate bug fixes more frequently than Calcite releases.
>
> I can change the commit in our build, but maybe there are other projects which do this as well?
>
> Mihai
>
> ________________________________
> From: Julian Hyde <jh...@gmail.com>
> Sent: Saturday, January 10, 2026 10:40 AM
> To: dev@calcite.apache.org <de...@calcite.apache.org>
> Subject: Re: incorrect commit in main
>
> Personally, I would rebase main (including the following commit). It’s unlikely that you will break someone, because the contents of main will be unchanged by your rebase.
>
>> On Jan 10, 2026, at 9:09 AM, Mihai Budiu <mb...@gmail.com> wrote:
>>
>> Hello all,
>>
>> By mistake I have merged a PR yesterday without squashing a commit.
>> Since there has been another PR merged following it, I propose to leave it as it is.
>>
>> [<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
>>
>> *
>> Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
>>
>> *
>> [<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
>>
>> I will try to be more careful in the future.
>>
>> Thank you,
>> Mihai
>>
>>
Re: incorrect commit in main
Posted by Julian Hyde <jh...@gmail.com>.
I guess the question is “should downstream tools assume that the project never rewrites history?”. Whichever answer we choose, there is a cost.
Your script could use “ git merge-base --is-ancestor <commit_hash> main “ to detect whether history has been rewritten.
> On Jan 10, 2026, at 11:19 AM, Mihai Budiu <mb...@gmail.com> wrote:
>
> The build of our compiler includes the hash of a commit in Calcite's main; we track main periodically, since we want to incorporate bug fixes more frequently than Calcite releases.
>
> I can change the commit in our build, but maybe there are other projects which do this as well?
>
> Mihai
>
> ________________________________
> From: Julian Hyde <jh...@gmail.com>
> Sent: Saturday, January 10, 2026 10:40 AM
> To: dev@calcite.apache.org <de...@calcite.apache.org>
> Subject: Re: incorrect commit in main
>
> Personally, I would rebase main (including the following commit). It’s unlikely that you will break someone, because the contents of main will be unchanged by your rebase.
>
>> On Jan 10, 2026, at 9:09 AM, Mihai Budiu <mb...@gmail.com> wrote:
>>
>> Hello all,
>>
>> By mistake I have merged a PR yesterday without squashing a commit.
>> Since there has been another PR merged following it, I propose to leave it as it is.
>>
>> [<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
>>
>> *
>> Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
>>
>> *
>> [<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
>>
>> I will try to be more careful in the future.
>>
>> Thank you,
>> Mihai
>>
>>
Re: incorrect commit in main
Posted by Mihai Budiu <mb...@gmail.com>.
The build of our compiler includes the hash of a commit in Calcite's main; we track main periodically, since we want to incorporate bug fixes more frequently than Calcite releases.
I can change the commit in our build, but maybe there are other projects which do this as well?
Mihai
________________________________
From: Julian Hyde <jh...@gmail.com>
Sent: Saturday, January 10, 2026 10:40 AM
To: dev@calcite.apache.org <de...@calcite.apache.org>
Subject: Re: incorrect commit in main
Personally, I would rebase main (including the following commit). It’s unlikely that you will break someone, because the contents of main will be unchanged by your rebase.
> On Jan 10, 2026, at 9:09 AM, Mihai Budiu <mb...@gmail.com> wrote:
>
> Hello all,
>
> By mistake I have merged a PR yesterday without squashing a commit.
> Since there has been another PR merged following it, I propose to leave it as it is.
>
> [<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
>
> *
> Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
>
> *
> [<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
>
> I will try to be more careful in the future.
>
> Thank you,
> Mihai
>
>
Re: incorrect commit in main
Posted by Julian Hyde <jh...@gmail.com>.
Personally, I would rebase main (including the following commit). It’s unlikely that you will break someone, because the contents of main will be unchanged by your rebase.
> On Jan 10, 2026, at 9:09 AM, Mihai Budiu <mb...@gmail.com> wrote:
>
> Hello all,
>
> By mistake I have merged a PR yesterday without squashing a commit.
> Since there has been another PR merged following it, I propose to leave it as it is.
>
> [<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>CALCITE-7367<https://issues.apache.org/jira/browse/CALCITE-7367>] NULLS FIRST throws ClassCastException when sorting arrays<https://github.com/apache/calcite/commit/073942c19377ef527892b0c2ecb2fa3f3a9f3db6>
>
> *
> Change Map literal implementation to Java Map<https://github.com/apache/calcite/commit/3e1091b73d74c9238fb09548c897c6a6350676bb>
>
> *
> [<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>CALCITE-7366<https://issues.apache.org/jira/browse/CALCITE-7366>] RexLiteral<https://github.com/apache/calcite/commit/a124dc7bee1195085d333515a2af57cbeb388870>
>
> I will try to be more careful in the future.
>
> Thank you,
> Mihai
>
>