Re: [eclipselink-dev] [External] : EclipseLink 3.0 branch

[Lukas Jungmann <lukas.jungmann@xxxxxxxxxx>]

> On Feb 9, 2021, at 7:21 AM, Piotr Żygieło <piotr@xxxxxxxxxx> wrote:
> 
>> 
>> @Piotr Żygieło
>> 
>>> The PR 985 reveals the commit that 3.0.0-RELEASE branch was pointing at (dd5cf6aad9). Perhaps this could be considered, as it has versions for development updated already
>> 
>> I don't see a branch for "3.0.0-RELEASE"
>> (https://urldefense.com/v3/__https://github.com/eclipse-ee4j/eclipselink/tree/3.0.0-RELEASE__;!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8uMsyPgAM$ ).
> 
> Because of this event:
> https://urldefense.com/v3/__https://github.com/eclipse-ee4j/eclipselink/pull/985*event-4084747964__;Iw!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8u16KpEds$ 
> 
>> Based on PR 985 (https://urldefense.com/v3/__https://github.com/eclipse-ee4j/eclipselink/pull/985__;!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8uhYAPAno$ ,
>> was "3.0.0-RELEASE" from Radek Felcman's fork?
> 
> No. This PR was not from a fork.

Piotrek is right here. eclipselink-bot@xxxxxxxxxxx is an artificial user who works against the main repo. Since direct push to master is forbidden (for everyone - I hope), the “bot" creates a branch, does his work and waits for someone to merge his work back to master when appropriate.

What was wrong here is that his PR should have been “rebased” only for all his commits to be presented in the master’s history, instead of being squashed and rebased


> 
>> Also, I don't think "dd5cf6aad9e06cd2ad4a140d380abb976be5964b" is the
>> correct commit hash. Commit "3986bdbeae8e0e04e5be4a7076f2bda2ee1a09a5"
>> matches the version.properties file in
>> https://urldefense.com/v3/__https://mvnrepository.com/artifact/org.eclipse.persistence/eclipselink/3.0.0__;!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8uLmvtZWY$ 
> 
> dd5cf6aad9 is NEXT commit, just after the release.
> 
>> It is interesting that the tag "3.0.0"
>> (https://urldefense.com/v3/__https://github.com/eclipse-ee4j/eclipselink/releases/tag/3.0.0__;!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8uLTmantE$ )
>> appears to have its own commit
>> "3986bdbeae8e0e04e5be4a7076f2bda2ee1a09a5". However, if you view the
>> commit history for "master", you don't see this commit listed.
> 
> Because it was not merged to master.

It was “squashed & merged” instead of “rebased"

> 
>> I had
>> always thought tags were associated with OTHER commits and were not
>> their own commit...
>> but that may not be important.
> 
> Understanding this part of git might be important.
> 
>> What is important,
>> in my opinion, is that the "version.properties" file for the
>> EclipseLink 3.0.0 release lists tag 3.0.0's commit "3986bdb". That is
>> what we should fork from, in my opinion.
> 
> And then you will have to update versions, from fixed, released 3.0.0
> to something "next", like 3.0.1-SNAPSHOT. This work is already done in
> dd5cf6aad9. That's why I propose, if you choose to branch off 3986bdb,
> do it from dd5cf6aad9 instead.


I can just hit the button to restore the `3.0.0-RELEASE` branch if that is what we want. But keep in mind that this branch has never been intended to evolve further, it’s purpose was just to record the change of the version number and nothing else. Also note that that branch name does not follow the branch naming convention used by the project. The tag, on the other hand, follows existing convention and points to correct commit.

thanks,
—lukas



> 
> Piotrek
> _______________________________________________
> eclipselink-dev mailing list
> eclipselink-dev@xxxxxxxxxxx
> To unsubscribe from this list, visit https://urldefense.com/v3/__https://www.eclipse.org/mailman/listinfo/eclipselink-dev__;!!GqivPVa7Brio!LszX0lLmfn5JmOk-ASF9kP4OwgCUS46jd3OC712CpNYGSBSEtdkWJ8s_EP8uzb1Gbpo$