appending a non-existent locale to standard wrapper does not load

[pixelzoom]

https://github.com/phetsims/phet-io/issues/1901 is relevant to phetsims/qa#872 and phetsims/qa#873

[zepumph]

Cherry pick: https://github.com/phetsims/phet-io-wrappers/commit/84da4632b21a7163b3e7922604d2ced29b5d4133

Test: follow steps in the first comment of https://github.com/phetsims/phet-io/issues/1901 and make sure a bad locale will work.

[samreid]

@zepumph and I cherry-picked the fixes and this is ready for QA verification, following the steps described in https://github.com/phetsims/phet-io/issues/1901.

[zepumph]

I confirmed in ph-scale and basics 1.6.0-rc.2. QA please feel free to close.

[Nancy-Salpepi]

Confirming that the standard wrapper opens (bad locale ignored) in ph Scale and ph Scale basics.
However, https://github.com/phetsims/phet-io/issues/1901#issuecomment-1401423861 says that with phetioDebug=true, I should see an assertion when a bad locale is used and I do not. Neither does @KatieWoe.

[KatieWoe]

I attempted this on the standalone wrapper, but phetioDebug=true didn't seem to work. Is this expected?

After discussing with @samreid this aspect is expected. Noting that what @Nancy-Salpepi saw in #274 (comment) also happens on my end.

[arouinfar]

@samreid is it possible that https://github.com/phetsims/phet-io-wrappers/commit/84da4632b21a7163b3e7922604d2ced29b5d4133 didn't make it into the 1.6 branch?

[samreid]

@KatieWoe and I discussed this in a zoom call. The "standalone wrapper" isn't actually a wrapper and hence doesn't support wrapper query parameters like phetioDebug.

[arouinfar]

The "standalone wrapper" isn't actually a wrapper and hence doesn't support wrapper query parameters like phetioDebug.

Oh, I missed that part of @KatieWoe's comment. However, @Nancy-Salpepi is reporting that there isn't an assertion in the Standard PhET-iO Wrapper for a bad locale when phetioDebug=true. I tested rc.3 and I'm not seeing an assertion, either.

[samreid]

Thanks, I reproduced the same problem. I saw that in phetsims/beers-law-lab#320 (comment) (5 days ago) @zepumph and I collaborated to improve the locale initialization sequence. It seems it would be better for fallbacks to work in the usual way even if assertions are enabled. So I think we decided we no longer want the behavior listed in https://github.com/phetsims/phet-io/issues/1901#issuecomment-1401423861 (2 weeks ago). Sorry this wasn't clearly communicated. Therefore, this issue can be closed.

[arouinfar]

@samreid good to know. We should update phet-io-guide.md because it currently tells clients to expect an assertion.

Relevant section: https://github.com/phetsims/phet-io-sim-specific/blob/master/client-guide-common/client-guide/phet-io-guide.md#use-of-query-parameters-with-the-standard-phet-io-wrapper

If you specify a locale for which a translation does not exist, two things may happen. If phetioDebug=false, the locale query parameter is ignored and the simulation will retain the localeProperty used when creating the Standard PhET-iO Wrapper. If phetioDebug=true, an assertion will be thrown.

[arouinfar]

@samreid I updated master in the commit referenced above. The section now reads:

If you specify a locale for which a translation does not exist, the locale query parameter is ignored and the simulation will retain the localeProperty used when creating the Standard PhET-iO Wrapper.

[matthew-blackman]

Fixed and confirmed in rc.4. @arouinfar said that we can test this ourselves, so we are ready to go to production on that one.

[samreid]

@matthew-blackman and I cherry-picked the change from @arouinfar into ph-scale and ph-scale basics, checked in the RCs and double checked in the production version. Closing.