Yes but numbers also minimalize human errors. Typos, big/small letters etc.
Discussion
We're putting these IDs in Json, tho, to make them machine-readable. If we wanted humans to type kinds into notes, we'd have to use yml.
A computer doesn't care if a kind is "68325" or "tweet4". Neither does a developer, as they can just copy-paste.
And the json is full of gobbledygook strings, anyway. No one is typing that stuff in by hand.
And if a human wanted to know what is the kind for "event deletion" you could tell them that it's "delete_event", instead of "Kind 09".
Kind 5
But still, it's harder to ensure devs all call it delete_event and not deleteEvent or delete_evnt or Delete-event, be it on purpose or by accident. Kind numbers help ensuring that. But yeah, it's not the most human friendly in terms of readability.
I am not a programmer, but when I needed to code something back in the days, we just used ctrl-n (iirc) in vim to recall already written variable names. So I ended up using always the same name, once it was defined.
I think every IDE has some similar auto-complemtation nowadays?
x1 to xn in code just looks like straight out of the dis-assembler if you ask me.
What I mean is that different Devs might use different names, not necessarily the same one.
Are there not headers or libraries for everyone to include?
Maybe a stupid question, it's really a looong time ago I was doing such things.
People use different sdks or no sdks. Nostr is just a protocol, it's not one implementation everyone uses.