You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
While there are currently no automated regression tests for language coders, the generated Haskell and Java code serves as a good regression test because it is so comprehensive. When a change is made to any code which may affect Haskell and Java code generation, we only have to generate all of Hydra's sources again to verify that nothing has changed. Not so with the other coders (TinkerPop, Avro, PDL, GraphQL, SHACL, JSON, YAML). These coders can easily become corrupted by upstream changes without anyone immediately noticing.
Find a happy medium between no tests at all for these languages, and checking in a huge amount of generated code which bloats the repository and serves no purpose other than as a regression test.
The text was updated successfully, but these errors were encountered:
While there are currently no automated regression tests for language coders, the generated Haskell and Java code serves as a good regression test because it is so comprehensive. When a change is made to any code which may affect Haskell and Java code generation, we only have to generate all of Hydra's sources again to verify that nothing has changed. Not so with the other coders (TinkerPop, Avro, PDL, GraphQL, SHACL, JSON, YAML). These coders can easily become corrupted by upstream changes without anyone immediately noticing.
Find a happy medium between no tests at all for these languages, and checking in a huge amount of generated code which bloats the repository and serves no purpose other than as a regression test.
The text was updated successfully, but these errors were encountered: