Proposal: Extend object IDs to support m...
j
Jake
03/31/2023, 2:03 PMt
Toi
03/31/2023, 2:06 PMYeah, I saw this! This will certainly help!
I have two questions about this proposal though:
1- Will ids be able to start with these special characters?
2- How long until release? 😆
j
Jake
03/31/2023, 2:07 PM2- it's early days
t
Toi
03/31/2023, 2:09 PM2- Do you mean it's too early to tell? (english is my second language, sorry)
j
Jake
03/31/2023, 2:10 PMyeah, it's too early to tell what other things we will find as this goes through implementation
t
Toi
03/31/2023, 2:10 PMGot it!
Any recomendation on how to workaround it?
j
Jake
03/31/2023, 2:13 PMyou could do base64 URL encoding and add a single letter prefix, that's not great
t
Toi
03/31/2023, 2:14 PMHm, it would work I think.
I'm also trying to figure out what combination of characters would generate a "+" at the begining of a base64 encoded string.
Maybe it would never happen for us
j
Jake
03/31/2023, 2:14 PMyeah, i've honestly never seen it
but that doesn't mean it couldn't happen
i've tried encoding emojis and other unicode things
t
Toi
03/31/2023, 2:16 PMIt seems to be only some combination of very weird characters hahaha
I guess encoding with base64 (url safe) works
Thank you very much!
j
Jake
03/31/2023, 2:19 PMyou're welcome!
btw, what backing datastore are you using?
t
Toi
03/31/2023, 2:31 PMPostgres
j
Jake
03/31/2023, 4:07 PMthanks!
make sure you check out 1.19 when it lands, it has some pretty big perf wins for postgres
t
Toi
04/24/2023, 8:27 PMHi @Jake , I saw that you already merged this id validation change. Any guesses on when will it land? 😁
Can't wait to try it out
Oh, just realized that it was released on v1.20. My bad