The hex 1337 mode certainly faces this limitation, but the emoji modes are actually working with more character space (albeit taking more space to store).
The all alphanumeric characters based 1337 mode should be able to approach / overtake standard UUID character space given its inclusion of the characters G-Z, though might need to add to the words list and tweak minimum word size to do so.
Thank you for the feedback! I'll be adding documentation for each mode's count of potential variations to clarify.
Yeah but there are subtle differences in emojis between platforms. I can't remember the exact details, but I remember running into it as a problem when I was buying emoji domains.[0] Essentially Emoji A would render identically to Emoji B on Android but on iOS Emoji A would render identically to Emoji C, so if you were ever entering them into a URL bar you'd have to have all of the possible combinations 301 to the canonical domain name.
[0] That was such a poorly preforming investment. At least I hit early on Bitcoin to make up for it.
For something a little more practical, I recently came across this postgres id generator[1], which Ive been playing around with as a solution for ordered, scalable bigint ids (that take up less space than standard uuids).
> This project is intended for entertainment purposes only - it is not recommended for use in your production or intended as a replacement to existing UUID generation mechanisms.
T-20 days until "Sir or madame, I have found your codes and used. We are experiencing highly critical outage of production system as result of bug. Kindly do the needful to resolve."
GitHub bug reports on personal projects make me substantially revise upward my estimate of the proportion of programmers who are just greedy (in the regex sense) copy / pasters.
> "Sir or madame, I have found your codes and used. We are experiencing highly critical outage of production system as result of bug. Kindly do the needful to resolve"
... "and revert back with the same"
my first thought was also "this is just another conspiracy to recruit humans to fuzz UID based systems with unicode" :D
Oh, is there a way to test the collisions for this? Like it generates a uuid, prints it to screen, then does a progress bar showing how many tries attempted and a timer for duration.
ENS allows to register names with emoticons, I think that regular URL allows that as well. The problem with those things is not that they're not memorable, but that there's no good UI to type then when you want to write the name.
That said, the project is great! I love such things:-)
25 comments
[ 3.3 ms ] story [ 57.1 ms ] threadThe all alphanumeric characters based 1337 mode should be able to approach / overtake standard UUID character space given its inclusion of the characters G-Z, though might need to add to the words list and tweak minimum word size to do so.
Thank you for the feedback! I'll be adding documentation for each mode's count of potential variations to clarify.
[0] That was such a poorly preforming investment. At least I hit early on Bitcoin to make up for it.
[1] https://rob.conery.io/2014/05/28/a-better-id-generator-for-p...
Ha! You can't tell me what to do!
GitHub bug reports on personal projects make me substantially revise upward my estimate of the proportion of programmers who are just greedy (in the regex sense) copy / pasters.
... "and revert back with the same"
my first thought was also "this is just another conspiracy to recruit humans to fuzz UID based systems with unicode" :D
... I appreciate this whole exchange, complete with emoji:)
I've always thought it was underused in decentralized authentication spaces.
That said, the project is great! I love such things:-)
> Generate cute UIDs, i.e. unique(ish) identifiers that are similar in appearance to UUIDs.
Ok - but what's the difference between them and UUIDs... What makes them "cute"?