vehicle_types.json is not required

[nbdh]

Hi @isabelle-dr and contributors,

thanks for the good work!

I failed to raise this during yesterdays workshop as I only noticed after the slot was finished, but in my opinion, the validator should not declare potentially valid feeds as invalid. A feed rather should be considered invalid only if it clearly is not, while stating it would be valid still does not mean it could not potentially be invalid (e.g. because free_bike_status refers to vehicle types that are not declared).

Full disclosure: Our implementation of vehicle_types.json is still in progress and even though the spec says we should be publishing it, we are not required to do so and yet our feeds are considered invalid ;)

[CLAassistant]

All committers have signed the CLA.

[netlify]

✔️ Deploy Preview for wizardly-lichterman-770f54 canceled.

🔨 Explore the source changes: 414bd5a

🔍 Inspect the deploy log: https://app.netlify.com/sites/wizardly-lichterman-770f54/deploys/616ffdd41fc406000713a5e7

[netlify]

❌ Deploy Preview for competent-payne-690ca9 failed.

🔨 Explore the source changes: 414bd5a

🔍 Inspect the deploy log: https://app.netlify.com/sites/competent-payne-690ca9/deploys/616ffdd40fe2b100076fd970

[netlify]

❌ Deploy Preview for wizardly-engelbart-5c48ca failed.

🔨 Explore the source changes: 414bd5a

🔍 Inspect the deploy log: https://app.netlify.com/sites/wizardly-engelbart-5c48ca/deploys/616ffdd4b1b8e00007f70ddb

[netlify]

✔️ Deploy Preview for unruffled-hugle-914373 ready!

🔨 Explore the source changes: 414bd5a

🔍 Inspect the deploy log: https://app.netlify.com/sites/unruffled-hugle-914373/deploys/616ffdd47c5e5e000716d0ba

😎 Browse the preview: https://deploy-preview-41--unruffled-hugle-914373.netlify.app/

[netlify]

❌ Deploy Preview for kind-pike-a3f3f8 failed.

🔨 Explore the source changes: 414bd5a

🔍 Inspect the deploy log: https://app.netlify.com/sites/kind-pike-a3f3f8/deploys/616ffdd477246d00079f3f94

[PierrickP]

@nbdh Hello,
I would like to try implementing the right conditional rule (between files).
If i'm sure it's 100% impossible, we can be more permissive on this and merge this PR

[nbdh]

Hi @PierrickP,

thanks for your reply. Of course properly resolving the "to do" by implementing the conditional rule is the way to go.

Still I'd argue that in the meantime, no potentially valid feeds shall be declared invalid. Imho, even if the validator states that a feed is valid, it might later turn out it is not, while stating that it is not valid should mean that it definitely does not match the specification.

[isabelle-dr]

Hi @nbdh and @PierrickP.

I'm OK merging this PR and having that file as optional until we find a way to write this rule properly in the validator.
I opened issue #54 to keep track of it, merging this PR!