Report hash of branch revisions
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
codetree |
Fix Released
|
Medium
|
Unassigned |
Bug Description
This is a feature request. If the codetree maintainers agree to the feature, I'm happy to write the code and submit a MP.
As part of our service deployment strategy, we're deploying new services with every new code revision, rather than upgrading charms / service code. Currently our mojo spec includes code that reads the 'collect' file and produces a unique hash based on the branch names and revisions. That hash is then incorporated into the juju service name.
To do this, we have ended up re-writing much of the codetree config file parsing routines. I'd like to include that functionality inside codetree itself. My idea is to add a new option to codetree: '--print-hash'. If set, instead of codetree actually building the branches on disk, it would print something like:
~~~
branch_hash: 8b1a9953c461129
~~~
This hash is built from the branch locations and their current revisions (or their pinned revisions, in the case of 'revno=X' in the config file).
What do the codetree maintainers think? Is this something you'd consider adding to codetree itself?
Related branches
Changed in codetree: | |
importance: | Undecided → Medium |
status: | Incomplete → Confirmed |
Changed in codetree: | |
status: | Confirmed → Fix Released |
Sure, this sounds like a good idea. Happy to review when if/when you have a branch.