Could be that the integrated the Gameface library incorrectly I guess? Still interested in more details from you.
Literally quoted from the article. Standard 2D UI like that can be done in as little a single draw call (or so I have read, never actually done it)
Great "DX" now when your building it, but good luck maintaining it over the long term.
"The last remaining draw calls are used to render all of the different UI elements, both the ones that are drawn into the world as well as the more traditional UI elements like the bottom bar and other controls. Quite a lot of draw calls are used for the Gameface-powered UI elements, though ultimately these calls are very fast compared to the rest of the rendering process. "
With this minimalistic, flat-style soulless UI, the correct number of draw calls spent on UI should be single digits....
Why on earth would they try to optimize how the UI renders when they're having big issues elsewhere?
It's absolutely not the most important, or even in the top 10 most important problems here, but it shows really illustratively how much they care about making a game which performs in an acceptable way. (which is: not that much)
Also, it's not even one or two high-poly models dragging the performance down, what I was aiming at is that the game suffers from death by a thousand cuts - the LoDs are only a part of the issue, almost every part of the game is done in a sub-optimal way. So while the UI is not a significant part of the frame time, if they fix the most glaring performance issues, they will find that there won't be a silver bullet, the game is just a pile of small performance problems all the way down.
All it shows is that optimizing something that was already fast enough was not a priority. But why would you want it to be?
I remember back when Apple removed the SD card slot from their phone. It wasn't a deal-breaking change, because their phones had lots of memory built in... but at the time, it seemed to me that the decision makers that did that, would continue to make similar decisions, each one making me unhappier and unhappier. So I bought accordingly. And I feel justified in retrospect.
Way back in the design stages of this game, somebody though that using web tech for their UI was a good idea, and it didn't get vetoed. That indicates bad judgement somewhere in the chain. Who knows whether it's developers, managers, or what... but the best predictor of bad decisions is previous bad decisions. It's at least a dysfunction smell.
Not "metaphorically", but "literally"? Or are you using "literally" in its non-literal sense? And "banned", not "discouraged", or simply "ridiculed" like you're trying to do?
That is literally (to use the term in its literal sense) an extremely brash statement, quite a lot to walk back in reverse into the shrubberies. Have you actually tried to develop a UI in Unity that approaches the quality you can easily (and cheaply and quickly and maintainably) implement in a web browser? And have you ever tried to find someone to hire who was qualified to do that (and then put them to work on the UI instead of the game itself), compared to trying to find someone to hire who can whip out a high quality performant web user interface in a snap, that you can also use on your web site?
Not to mention that you used web tech to call for the literal banning of web tech.
Why does it matter if it's 5 render calls or 500? The developers clearly have plenty of work to do optimizing the other 70ms, it doesn't make sense for them to spend any time working on this.