fix: separate active and inactive streams visually

[razgraf]

Discussed in #544

^{Originally posted by **IaroslavMazur** June 8, 2023}

---

As suggested in #544, active streams should keep a lower amount of sand in the bottom area, while settled streams should only display a (larger) bottom sand area - no top.

[PaulRBerg]

Banger! I will push a commit to address the feedback below.

But, while reviewing this, I realized that if we want to be 100% accurate with the design, we should also introduce a dichotomy in the upper bulb:

• HOURGLASS_UPPER_BULB_LARGE: used for PENDING
• HOURGLASS_UPPER_BULB_SMALL: used for STREAMING

How perfectionists are we, @razgraf? I'm torn 😁

The Solidity part is easy so I will let you decide what to do here (I know it's a busy period) - either way would work for me.

[razgraf]

How perfectionists are we, @razgraf? I'm torn 😁

I think it actually depends on a couple of "points of view".

From the POV of being technically accurate - yes, we could either introduce another type of bulb or completely hide the sand while the stream is "pending".

From the POV of visual design - having an empty hourglass (no sand) would look quite ... empty 😅 We would eliminate a central piece of the design.

From the POV of being technically efficient - every new bulb introduces bytes of extra code which pushes us again towards that contract limit (not sure how big of a problem this is, but I remember it causing some headaches). Of course, without sand that wouldn't introduce a lot of data, but with yet another special sand-state/bulb, it would.

[PaulRBerg]

MSTM! I will merge this as is, then. Just one comment on this:

having an empty hourglass (no sand) would look quite ... empty 😅 We would eliminate a central piece of the design.

Well, hourglasses with no sand in the lower bulb are possible in the physical word - yes, they don't last for long, because gravity gets in the way - but they can exist for a few milliseconds.

[PaulRBerg]

Anyhow, the good news about the NFT descriptor is that we can always upgrade it later on, should the need arise for this.

[PaulRBerg]

Note: integration tests are failing because the hard-coded values in test_TokenURI_* have to be updated.

I'll work on this shortly.

[IaroslavMazur]

@PaulRBerg, just to confirm: in your original comment above, you were suggesting that the lower bulb of a PENDING Stream hourglass is empty, while the upper one is full, right?

Asking because that's the idea I had when analysing the hourglasses in the image posted by @razgraf above, and it'd be sad for the idea to be lost in case I've wrongly assumed this is what you've suggested, as well 😅

Aside from that, I like how the issue from #544 has been addressed. Well done, @razgraf! 👏

[PaulRBerg]

in your original comment above, you were suggesting that the lower bulb of a PENDING Stream hourglass is empty, while the upper one is full, right?

Yes!

I agree that it would be nice to have a comparatively larger upper bulb when the stream is PENDING, but this is an exceptionally busy period for all of us, and we gotta choose our priorities wisely.

I will open an issue about this to track it as a feature request!

[PaulRBerg]

On second thoughts, we might want to keep the design as is for PENDING (i.e. no empty lower bulb) because of the NFT metadata cache issue.

[PaulRBerg]

Thanks all for your help on this! Fantastic team work.

All CI checks passed - merging.