I do the same thing with a gmail address -- just do myemail+uniqueid@gmail.com, because gmail drops everything after the + for routing the email but still tells you which address it was sent to.
Most of the time this is not a problem because you can create the account again with the correct email, but I have lost a few good usernames in some popular websites because I pulled this trick on them, Gravatar is one of these websites and the simplest one to explain.
I wanted a specific username that was available; I signed up with an email that looks like username+gravatar@gmail.com and the operation went through, then when I tried to log in, the website kept saying that the email was incorrect. I tried to reset the account using the username instead, which would force them to send a reset link to the email associated to this account, but the emails never arrive. I gave up that and other good usernames in many websites because of the inappropriate handling of email addresses by some popular web libraries.
Also, the same way you learnt this “trick” other programmers discovered it too and implemented a stripping algorithm to remove the unique ID during the registration process, so it is not very useful nowadays unless the website is very obscure and their developers either are rookies or do not have time to prevent this.
Reputable websites don't have an incentive to strip out the ID. It is only websites like "sold out" dating websites or otherwise nefarious websites that have that incentive in the first place. So, name+id@domain.com is still useful for filtering and indexing, and in fact if a website does strip out the ID you should ask yourself why would they need to do that.
At first I believed it was some scam but it soon got clear that my account was accidentally siphoning mail from someone using this supposedly clever scheme...
So, instead of doing johndoe+hn@example.com, I can do hn@example.com.
Fastmail has an interesting variation of this that supports multiple users on the same domain: hn@johndoe.example.com. My wife and I both use that feature heavily on our domain.
Unfortunately, there is a mathematical theorem (greater idiot theorem) that makes it technically impossible to generate a new random string that becomes an inbox that goes to your inbox.
How do I know that it's technically impossible?
Because otherwise Gmail would have done it already, since this takes 20 minutes and fixes a glaring security issue that gives spammers access to your email that you didn't grant.
I can't have an IQ 50 points higher than all Gmail engineers, so it must stand to reason that I'm the greater idiot. Hence there must be something preventing this obvious solution. QED.
so for sure it isn't possible for some reason, even if someone else is doing it.
what other explanation can there be? why would they just have a glaring security hole like that, allowing anyone you try to give a trackable email to with + to just remove it with a regex? It doesn't make sense. there must be some reason it can't work.
The + need not have been intended as a spam filter; I use it for bona fide filtering of separate clients’ projects.
It's more effective to have an entire domain at your disposal and just invent $vendor-specific-thing@mydomain.com whenever an email address is needed. On the server end all you need is a catchall to handle it, and you can blackhole whatever addresses become spam magnets (while also knowing which vendor's responsible).
This rfc merely acknowledges that some systems do this. It doesn't say the must or even should.
I wrote about this here: https://idbloc.co/blog/background/2019/04/14/keep-your-addre... and it’s one of the main reasons https://idbloc.co exists