There's a similar plot in the Silicon Valley TV show (Piped Piper v.s. Hooli).
This is particularly saddening for me because I remember back in 2011/12 (around that time) I desperately wanted an Iphone so that I could do one thing - download Snapchat. I wanted a totally new/expensive phone to do one single thing - use Snapchat. Sure I was 19, and it was slightly known as the sexting app but man oh man, there was this sort of aura of mystery surrounding it. It was literally the "coolest" app in my peer group (hell, in my age group at the time). It was soooo difficult to search for people, so you had to know their user before you could add them. It was as if the app was so public (sharing pictures that could be screenshot), yet so private (you tended to only add people you knew, and the porn accounts).
It's slightly funny to watch Evan Speigel still think that he knows whats best for Snapchat (e.g., the UI changes, monetization). He makes his changes now and tells the users to shutup and deal with it (bad idea). Now that I'm older, everyone I know still uses the app, but not nearly as much, and it's not nearly as "cool" or fun. It seems to be just another app on our phones now - not "the" app.
But in some ways I understand Snapchat's changes. Those servers, engineers, creatives, and yellow bikes cost money. You can't pay for an office lease in Venice Beach with "cool" points. However, with that said, I do think Snapchat went overboard with the monetization efforts. They followed in the footsteps of Google, Facebook, etc - not knowing that they were so much cooler and so much more fun than anyone else. And in trying to monetize, they ruined the experience, and thus the app. The "scale" thing [gives Reid Hoffman a side eye] doesn't always have to apply to everything. Sometimes just having an amazing app that does what it does is ok - the partnerships, click bait, and "publishers" be damned.
RIP Snapchat. The only app I've ever bought a brand new to device for.
I definitely think we will look back on my generation (Generation Y) as the guinea pigs for this whole "connected social world" thing. So far it's been too much, too fast, and we have done an abysmal job handling it.
There are two major parts to a spreadsheet app, in my view:
1. A UI to change sheet layout and formulas, and to view computed values.
2. A "programming language"-like engine to derive computed values from the formulas and sheet layout.
In (2) you also have to do a bit of topological sorting to make sure you compute things in the right order, but it's not terribly tricky.
A spreadsheet is kinda like a REPL, where the "R" and "P" are the spreadsheet UI and the "E" is the programming language part.
In Excel and Google Sheets (but not in my program) there are some fun data structures you might use to make formulas like `SUM(A:A)` (i.e., sum the whole column) run in a reasonable amount of time. Essentially, data structures that take advantage of the sparsity of the grid. I don't know off-hand the best way to go about something like that, but if you didn't want to invent anything yourself I guess you could just throw everything into an off-the-shelf spatial acceleration structure (an r-tree or a quadtree or something.)
There's quite a big surface area to the problem, if there's something more specific you want to know, please ask! I can't promise to respond quickly, but I should get back sooner or later.
After that, then came the VC dump with the price now at $59 - $60 for weeks with the bagholders entering at >$200 or >$400 on listing day.
When I see charts like that, that is what you call a VC pump and dump.
[0] https://www.coinbase.com/price/internet-computer