Yes, I recently looked at the version management of Flink and SkyWalking, which is the same as this one.
Lucifer Tyrant <[email protected]> 于2023年3月25日周六 03:08写道： > We have only one branch dev，through bug fix or big feature should be merged > in dev branch，like hadoop or spark or flink they does not distinguish > different branch. If you want to and big feature in next version，release a > small version before it. > > JUN GAO <[email protected]>于2023年3月24日 周五16:07写道： > > > Yes, I agree with you about the version number defined. > > > > But, how can we manage the code branch? For example, we have the version > > 2.3.1 now, If we want to fix a bug, which branch should we use? dev > branch > > or the 2.3.0-release branch? > > > > If we use the dev branch to fix bugs, it means the new features can not > use > > the dev branch. Each feature needs to create an Individual branch? > > > > If we use the 2.3.0-release branch to fix bugs, it means the pr(to fix > the > > bug in 2.3.x) needs to merge into all of the 2.3.0-release branch and dev > > branch. > > > > Guangdong Liu <[email protected]> 于2023年3月24日周五 14:17写道： > > > > > I agree with Leonard and I would like to add an example. > > > > > > For the Apache Spark project, the version number is defined as follows: > > > > > > 1.Major version: indicates major changes that may require users to > > rewrite > > > their code and are not backwards compatible. The major version number > is > > > typically increased when there are incompatible changes to the API. For > > > example, upgrading from 1.x.x to 2.x.x would require code to be > rewritten > > > and would not be backwards compatible. > > > 2.Minor version: indicates backwards-compatible improvements to > > > functionality that may require users to modify their code. The minor > > > version number is typically increased when new functionality is added. > > For > > > example, upgrading from 2.3.x to 2.4.x may require code modifications > to > > > use new functionality, but does not require code to be rewritten. > > > 3.Patch version: indicates backwards-compatible bug fixes and > > > optimizations. The patch version number is typically increased when > bugs > > > are fixed or optimizations are made. For example, upgrading from 2.3.1 > to > > > 2.3.2 may only include bug fixes or minor optimizations and would not > > > require code modifications. > > > > > > Therefore, when Apache Spark is upgraded to version 2.0.0, there may be > > > incompatible changes to the API, requiring users to rewrite their code. > > > When Apache Spark is upgraded to version 2.4.0, there may be new > > > functionality added, requiring users to modify their code to use it. > When > > > Apache Spark is upgraded to version 2.3.2, it may only include bug > fixes > > or > > > minor optimizations. > > > > > > > > > -- > > > Best Regards > > > > > > ------------ > > > > > > Liugddx > > > [email protected] > > > > > > > > > -- > > > > Best Regards > > > > ------------ > > > > EricJoy2048 > > [email protected] > > > -- Best Regards ------------ EricJoy2048 [email protected]
