let git do the cqc versioning, add optional version suffix in cqc folder structure #98
Conversation
| ``` | ||
|
|
||
| Commits to the `cqc` branch MUST contain the commit hash of the commit that was validated in the commit message. |
There was a problem hiding this comment.
You mean here or in the document? 😆
What i mean with this is that the system implementing the commits to the cqc branch must include the hash of the commit that was validated in the commit message that adds the validation results. not sure how to make that clearer.
There was a problem hiding this comment.
OTOH this might lead to huge commit messages that contain the commit hashes when many branches are validated, maybe leave this out or change to SHOULD?
There was a problem hiding this comment.
What i mean with this is that the system implementing the commits to the cqc branch must include the hash of the commit that was validated in the commit message that adds the validation results. not sure how to make that clearer.
Ahh you mean the commit hash of the validated commit must be included in the commit message in the cqc branch?
If yes, maybe include the word message in the specs.
There was a problem hiding this comment.
OTOH this might lead to huge commit messages that contain the commit hashes when many branches are validated, maybe leave this out or change to SHOULD?
Is it possible that multiple branches get validated in parallel?
There was a problem hiding this comment.
Ahh you mean the commit hash of the validated commit must be included in the commit message in the cqc branch?
If yes, maybe include the word message in the specs.
i mean it is right there:
Commits to the
cqcbranch MUST contain the commit hash of the commit that was validated in the commit message.
Is it possible that multiple branches get validated in parallel?
You mean from a technical standpoint or according to spec? It is possible, hence the $branch in the folder path {$branch}/{$package} in the cqc branch specs. However, i guess from a technical standpoint, validating each branch and then committing the results per branch to cqc is easier. This would mean that commit messages to cqc would have the format {$branch}/{$commithash}, where $branch is the validated branch and $commithash is the hash of HEAD commit on $branch. would that be clearer?
There was a problem hiding this comment.
Whoops, somehow didn't parse the message at the end. But I really struggled to understand that sentence without the extra discussion here. So maybe would still be beneficial to split that sentence into two, decreasing information density in exchange for clarity.
There was a problem hiding this comment.
You mean from a technical standpoint or according to spec? It is possible, hence the $branch in the folder path {$branch}/{$package} in the cqc branch specs. However, i guess from a technical standpoint, validating each branch and then committing the results per branch to cqc is easier. This would mean that commit messages to cqc would have the format {$branch}/{$commithash}, where $branch is the validated branch and $commithash is the hash of HEAD commit on $branch. would that be clearer?
Not sure what would be the best solution. I'm just wondering whether there is any benefit of having cqc on different branches on the (potentially) same commit hash, as these branches are structurally equal. Going a step further, the branch name doesn't really matter at all besides human readability. The only important bit is the hash.
The rationale is that we do not need a per-commit folder to version cqc, because git will track the history of these files anyways. For more linking goodness, we now only require the commit hash in the commit message to the cqc branch.