Gino Miceli

The last proposal sgtm, all things considered. On Fri, Sep 23, 2022, 11:35 AM Scott Olsen ***@***.***> wrote: > We could land somewhere in the middle: > > model.toLocalDataStoreObject() >...

We'll either hard code an offline basemap, or disable the feature altogether.

Similarly, we could add a helper called `logErrors` that reports them to logcat, but doesn't actually eat the error.

I think we generally wouldn't be adding features zoomed out this far. @sergeydolgov1, wdyt?

+1. Updated priority, release, pipeline. Please edit freely.

Should we really though? Posted [this question](https://groups.google.com/forum/#!topic/rxjava/sIFNQng5yg4) asking about an alternative approach, a là: ``` disposeOnClear( addFeatureClicks .switchMap( f -> dataRepository .saveFeature(f) .doOnComplete(() -> showFeatureSheet(f)) .doOnError(t -> onAddFeatureError(t)) .onErrorComplete() .toObservable())...

@scolsen, I discussed this with former teammates at [FAO](www.fao.org) today (@cdanielw, @lpaolini) who work extensively with RxJS in [SEPAL](https://github.com/openforis/sepal), and consensus is that it's better to wrap and unwrap in...

Yes, I was thinking the same thing. Afaict, the closest use of Optional is Java 9's "ifPresentOrElse()". What would the equivalent be here? ifSuccessOrElse? On Tue, May 21, 2019, 2:25...

I'm partial to `ValueOrError` because of existing precedent in some C++ libraries. On the other hand, I also like `Result` because it's less verbose. I'll need to see it used...

Sorry to vacillate on this, but after more experimentation, I'm back to thinking handling errors with side-effects rather than stream values will lead to more readable, self-documenting code. Consider this...