Post

Both of those *should* have not been huge changes. But that assumes the feature was designed with sane ideas of change in mind.

it’s hard to model things perfectly! and fixing mismodeled data is also hard!

Oh absolutely. It would either be one of those changes that is pretty quick & mostly a pain for testing edge cases… OR one that ended up needing 5 new services, a db migration or two, and an act of god There never seems to be a middle ground for that type of thing.

Working closely with fintech, I have a few. We used to sell product combos with high discounts to incentivize purchases. We pivoted to offering the combos as single SKUs with one discount to make it technically easier... Major tech challenges around determining ownership due to mismatched SKUs.

Dropbox adding work+personal accounts is a good one too

you gotta assume that these were both enormous efforts that were probably years in the making. all to make it work the way it should’ve worked in the first place

Sometimes “time to market” is just more important than a feature working the intended way.

yes?

put another way: I'm interested in hearing about times when people modeled their data in the wrong way and why

twitch has “shared chats”??

well so here at Notion Labs Inc., we have the concept of a "block"

i have heard stories of a large mysql cluster, and a system wherein "the schema that we used is the schema for your account" pretty much forever the vast majority of the data is write once read maybe and so backfilling is a huge waste of resources, so it all gets adapted per storage unit

I also wonder if someone asked if that kind of functionality should have been there early on, but people decided it would take too long to build.

certainly! if I was building slack a dozen years ago no way would I have chosen to support this then

At Felt, we changed the way map layers are nested, which was invisible for the user with the current feature set, but unlocked a bunch of other future features. It was an ~8 person, multi-month effort touching *a lot* of code; the near-term aim being “no one notices anything”.

If I recall correctly, it went down without a hitch and indeed, no one noticed anything.

what’s an example of a future feature it enabled?