Harsh loading for non cached centers

[kevinherdez]

If a user hasn't loaded a center's forecasts/warnings yet, the experience from going from the No Center Experience into the center experience is quite harsh

[rustynwac]

Is this dependent on API updates?

[kevinherdez]

We can do some work on the frontend to make this better. This could go hand-in-hand with #1154 where we center on the zone polygons immediately with the information we have from the map-layer call, maybe show a loading spinner on the cards if the forecast isn't fetched yet at all, and then fall back in the case that that fails with a toast saying "Failed to fetch the full forecast. Please try again", or, if we know it's old data, say something like "The forecast information expired X hours ago" like we do in the full forecast

This could also be helped by API updates where we fetch all the forecasts in one go. In the short discussion we had in the Map Project API updates, it didn't seem like that was a solution that the NAC was super okay with as it could be very costly on the backend

[rustynwac]

They're doing a bunch of API work now so we might be able to get some changes made end of summer

[stevekuznetsov]

It should be exactly the same amount of "cost" on the backend as us making N calls to fetch the data - it's the same load, after all. Once V3 API is ready to go, we should look at adding the endpoints we need to make the app's life easier.

[kevinherdez]

Their worry more is, is it necessary to make all those calls up front if the users mights never use the data. I have a document with the changes that we're proposing and some discussion with them that I can share with you guys

[stevekuznetsov]

Sure - if and only if some smaller fetch is workable and dynamically loading it in during panning can work. I am highly suspect of the load claim, and it costs nothing to implement it (even locally) and verify the actual compute cost.