Fixed typo in nta examples in years (+ CI fix for forks)

[Micovec]

Fixed typo '20209-12-17T09:30:46.0Z' with '2019-12-17T09:30:46.0Z' in nta examples.

In the linked issue I mentioned it should be 2020 but given that the /PublicationDelivery/PublicationRequest/RequestTimestamp is 2019-12-17T09:30:46.0Z and not 2020-12-17T09:30:46.0Z I believe I was originally mistaken.

Fixes: #1048

EDIT @thbar: also in this PR, a fix to ensure the test workflow (run) will run when we get PR from forks.

[skinkie]

@TuThoThai likely needs backports to all releases.

[skinkie]

@thbar we are again in the situation were the ci process does not work. When will it be fixed?

[thbar]

The required check on run is mine; announced in #1036 last month, explicitly flagged as "to be tested & verified".

Current PR does exactly that, and it surfaced something older than the check: ci.yml is on: push, so a PR from a fork produces no push here and the workflow has never once been triggered by one; true of all 338 runs still on record.

Until June that stayed invisible, because fork contributions were only ever validated after merge if I understand well, when the merge commit hit the target branch.

So the debt is the missing trigger, not the enforcement:

-on: push
+on:
+  push:
+  pull_request:

       - name: Commit changes
+        if: github.event_name == 'push'

The gate is needed because a fork's pull_request run gets no write token. Target branches stay normalized exactly as today.

@skinkie you can take it from here if you want (a patch directly on this PR will be fine I think) - I won't have much time before mid of august.

[thbar]

And thanks @Micovec as you are helping us uncover a bit of pipeline technical debt that is worth solving 😄

[skinkie]

@skinkie you can take it from here if you want (a patch directly on this PR will be fine I think) - I won't have much time before mid of august.

I have no clue what to do.

[thbar]

@skinkie issue is fixed with the proposed patch in 8a2ee38.

[thbar]

Thanks @Micovec ; re-approving since my CI fix (which we'll have to backport to v2.0 if you wanna target PR there too) dismissed the previous review.

[thbar]

@skinkie you'll have to re-approve, since I'm the last pusher.

[thbar]

Thanks @Micovec @skinkie