docs(conventions): define artefact-linking label convention - #146
Conversation
…ationships Signed-off-by: Bedirhan Yilmaz (bedirhan-yilmaz) <bedirhan.yilmaz@sap.com>
Signed-off-by: Bedirhan Yilmaz (bedirhan-yilmaz) <bedirhan.yilmaz@sap.com>
…e structure Signed-off-by: Bedirhan Yilmaz (bedirhan-yilmaz) <bedirhan.yilmaz@sap.com>
ee2a87e to
ef044c9
Compare
|
Since the K8S docs and ODG labels use kebab-case, I suggest we take the same approach for the label format |
| ### Label Name | ||
|
|
||
| ``` | ||
| ocm.software/artefact-references |
There was a problem hiding this comment.
nit: Do we think we will move towards a new spec version with artifacts instead of resources and sources soon? Currently, it seems rather like we will deprecate sources and keep the rest of the spec. So maybe ocm.software/resource-references would be more fitting?
There was a problem hiding this comment.
The idea here was to make this label agnostic of resources and sources and allow usage on both of those artifacts. For this use case, the same terminology as in the spec was used.
There was a problem hiding this comment.
For this use case, the same terminology as in the spec was used.
This actually diverges from the spec, as its spelled artefact here. So ironically the "conventions" document is now inconsistent with itself and does the opposite of its stated intent.
Going through this PR it does not appear this was a conscious decision? I believe the established consensus is to use the american spelling and this was overlooked
Signed-off-by: Bedirhan Yilmaz (bedirhan-yilmaz) <bedirhan.yilmaz@sap.com>
Signed-off-by: Bedirhan Yilmaz (bedirhan-yilmaz) <bedirhan.yilmaz@sap.com>
#### What this PR does / why we need it #146 introduced a regression in the [uniform naming of `artifact`](#146 (comment)). Looking at #39 and #40 this was an issue previously and `artifact` won. Presently the conventions document, whose purpose is to `ensure uniformity` uses two different spellings. #### Which issue(s) this PR is related to <!-- Usage: `Related to #<issue number>`, or `Related to (paste link of issue)`. --> Signed-off-by: Johannes Neisener <johannes@neisener.de>
7 participants
What this PR does / why we need it
Adds a convention to the OCM spec for expressing cross-artefact relationships via labels. This allows a derived resource (e.g. an SBoM) to declare which subject resource it relates to within the same component version, using a structured label placed on the derived artefact.
Which issue(s) this PR is related to
Fixes open-component-model/open-delivery-gear#126