Re: [MINOR]

2024-01-04 Thread Julian Hyde
I have merged https://github.com/apache/calcite/commit/3c19347cc45349a21a8c97d6f6e8d3e9f596070f with a new lint rule. I had to force-push to amended the previous two commits because the linter checks the N most recent commits. Julian > On Jan 4, 2024, at 11:28 AM, Julian Hyde wrote: > >

Re: [MINOR]

2024-01-04 Thread Julian Hyde
For the record, I think that ‘Lint’ and ’Typo’ are perfectly good commit messages. Simple changes should look simple. We have already documented the standards: https://calcite.apache.org/develop/#contributing. > On Jan 3, 2024, at 7:07 PM, Forward Xu wrote: > > If the fix is very small and

Re: [MINOR]

2024-01-03 Thread Forward Xu
If the fix is very small and there is no related work order or issue, we can briefly describe the problem and explain the reason for the fix when submitting the PR. If we need to standardize this type of PR submission, we can form some PR templates or documentation. Best, ForwardXu Julian Hyde

Re: [MINOR]

2024-01-03 Thread Julian Hyde
It's so difficult to agree on a vocabulary that everyone will agree on, I don't know whether we should even try. I follow the universal rule: "When in Rome, do as the Romans do". Which is to say, follow the existing standard (even if it's undocumented, which it probably is) and don't invent your

Re: [MINOR]

2024-01-03 Thread Benchao Li
The word [MINOR] is very coarse-grained to me, as Stamatis mentions above, it may contains small improvements to javadoc, site, test, error message, method/variable name, etc. And it is subjective to the author/committer whether a PR should be minor. Looking at the git history, I found it useful

Re: [MINOR]

2024-01-03 Thread Stamatis Zampetakis
The presence of the "minor" keyword in the commit summary is a bit redundant. I would argue that if the message is precise enough the person reading it can infer it is minor or not. Moreover, the minor classification is subjective. Some people consider minor things that do not change code at all.

Re: [MINOR]

2024-01-03 Thread Ran Tao
This format most likely comes from other open source projects. If calcite has its own specifications, such as how to set the title for PRs that do not require a jira name, IMHO, it can be introduced in the contribution doc. Commiters can also review PRs according to this specification. Best

Re: [MINOR]

2024-01-03 Thread Istvan Toth
Perhaps the square bracket convention ? If the ticket starts with CALICITE-\d+ , then make sure that the JIRA ticket id is between brackets. Also check for Gerrit Change IDs which are often added automatically, and a paint to remove. Istvan On Tue, Jan 2, 2024 at 10:50 PM Tanner Clary wrote:

Re: [MINOR]

2024-01-02 Thread Tanner Clary
I like the [MINOR] prefix because it makes it easy to identify simple commits (via grep or ctrl+f), the same way [CALCITE-1234] makes it easy to find commits related to [CALCITE-1234]. I also like that it maintains the "[...]" styling at the beginning of the commit message. Neither of these