Preserve node marks when cloning a node#1337
Conversation
|
@jbeder, could you review this, please? |
|
@jbeder, ping? |
|
@vvd170501 : could you rebase this PR? I believe this is a nice addition! |
|
Under which circumstance do I not want to preserve the markings? |
377a93d to
e8d92ca
Compare
I don't really have any use cases for not preserving the markings, but this was the original behavior, and someone might have relied on it. I don't know if yaml-cpp uses semantic versioning, and even if it does, versions are still below 1.0, so breaking changes might be acceptable, but I chose to keep the default behavior just to be safe. |
|
If I would decide, I would leave the @jbeder : Should |
|
Let's go with simpler: no preserveMarks arg. Let's make sure to make a note about this (@SGSSGene I've vaguely tried and horribly failed to keep a running list of "breaking changes since last release" or even "notable changes since last release". I wonder if maybe it would be possible to use hashtags on commits, e.g., #breaking: "what changed" or #change: "what's changed" or something, which could then be rolled up at release time. What do you think?) |
I was already heavily impressed and wondered how you kept track of the changes for 0.9.0 😅 |
@jbeder : |
|
I'm a big fan of pure linear history, even if it involves rewriting PR commits. So I would just squash and edit the commit message to add the tag. |
|
In fact, feel free to edit commit messages to make them clearer in general. |
By linear commit history, you mean no hit merges, but squash and rebase, so each PR is most of the times a single commit? I have to check later if I can edit the commits of PRs for this project. |
|
Yep |
\#log: behaviour change: cloning a node preserves markings
1dab263 to
47f0ae0
Compare
Solves #812