You also see things such as this bug in liquibase (https://liquibase.jira.com/browse/CORE-3222 ) which broke the CLI around 7 months ago. And even though the fix is very simple, and the community submitted a pull request with the fix, the repo owner still has not merged it and the CLI remains broken.
Also, the mvn build of the last 2 releases is actually broken. (Tests no longer pass)
v3.6.2 : released around 6 months ago
v3.6.1: released around 9 months ago
Should we be worried about the future of Liquibase as an open source project? Is this basically what prompted someone to start the DB-Manul fork a few months ago? Can someone share some details on this?
Hi, sorry for the apparent lack of progress on Liquibase but I want to assure you it’s not due to a Datical vs. Liquibase priority problem and Liquibase as a project is definitely not abandoned.
Most of the problem comes down to a poor 3.7 vs. 3.6 release management on my part. After the 3.6 release I started on changes for 3.7 with a hope of being able to release that soon. However, for various reasons that release has dragged on far longer than I had hoped while always feeling like it is “nearly there”. My time has been spent more on 3.7 but, in the meantime the 3.6 release has suffered as I’ve not had the time to also inspect and test the pull requests.
Now that the holidays have finished up, I was already planning on shifting my focus back to a 3.6 release based on the open pull requests. I’m planning on having a new release out by Jan 15th. As far as the commits that go into the datical repository, I’ve tended to merge those changes back into Liquibase near release time for each major release, so the changes that are in there will be in 3.7 as well.
It is definitely apparent that the project has grown enough that it’s past my ability to manage both incoming contributions and future work. Pull requests are great because most of the bug/feature work is done, but it does still take time for me to review and test the changes as well as ensure they do not break anything else.
That has made the bottleneck and I would like to improve that. Is there anyone in the community that would be interested in taking a more active role in helping me with the pull request process? Someone that could merge sets of pull request into their own fork and be able to give me the feedback on “these sets of pull requests solve the problems they claim and do not break anything else” ? If anyone is interested, let me know.
