Decide on extent of OOP in metadata model #1133

Activity

1. lukpueh

   changed the title [-]Should we create classes for dictionary attributes of inner metadata classes (Targets/Snapshot/Timestamp/Root)[/-] [+]Decide on extent of OOP in metadata model[/+] on Sep 10, 2020
2. lukpueh

   added

   discussionDiscussions related to the design, implementation and operation of the projectDiscussions related to the design, implementation and operation of the project

   on Sep 10, 2020
3. lukpueh

   added this to the Foundations milestone on Sep 10, 2020
4. lukpueh

   added

   documentationDocumentation of the project as well as procedural documentationDocumentation of the project as well as procedural documentation

   decision recordOutcome of this discussion should be tracked in a decision recordOutcome of this discussion should be tracked in a decision record

   on Sep 10, 2020
5. sechkova commented on Nov 20, 2020

   

   sechkova

   on Nov 20, 2020

   Contributor

   More actions

   Seems like up to now there is no better suggestion than implementing classes for metadata fields #1139 which can "self-validate" in a similar fashion as in-toto does (with its ValidationMixin #1030) and I think this type of validation is the strongest argument "for" classes.

   Maybe wait until a suggestion of an implementation appears in #1139 so we can see it in practice? Meanwhile a brighter idea may come up 🤷
6. lukpueh commented on Nov 20, 2020

   

   lukpueh

   on Nov 20, 2020

   MemberAuthor

   More actions

   Thanks for reviving the issue, @sechkova. I also think that the metadata validation argument is very compelling. Because in order to validate the metadata schema, we need to define it in some way or another anyway, so we might as well use Python classes, which would also be useful for normal operations.

   I can draft an ADR next week.
7. lukpueh

   closed this as completedin #1229on Dec 1, 2020