…
You must log in or register to comment.
Jokes aside, I have been blocked many times by overzealous email validation. Yes, my email has a plus sign in it. This is allowed under RFC5322, so deal with it. It is better to have no validation at all than incorrect validation.
A plus sign? That’s nothing, LOL
Quote:
If you disagree, or have any other comments, feel free to email me at
'*+-/=?^_`{|}~#$@[IPv6:2602:f977:800:0:e276:63ff:fe72:3900]– if your mail client lets you, that is.
I like this issue in the form of a quiz
TIL:
🫱@🫲
is a valid e-mail address.
Because of one smartass customer who insisted on doing exact RFC 822 validation, I implemented exactly that. And yes,
zKcknV|NGv.lI66vR#@X`QcRK4K.R`?NpA.Gc2Kqzue9.%&nb1kGWp/./#Och$RQvis one of the test cases for a valid addr-spec. See (or generate) some others at https://github.com/mormegil-cz/rfcemailvalidator
I dont know if it’s just me, but this comment is breaking the rendering of Voyager
I needed to use a code block for that address as several apps had a problem when I tested escaping the back ticks in the address for the inline code. Not sure if you mean that as it renders in it’s own line or if anything else is broken
The best email validation is just sending an email to whatever provided by the user. If user receives an email and validates it, than its validated.
Email address spec is convoluted and this is indeed the best way. Noobs and ninja do it this way, normies try to validate before sending email
Email validation for a form should at most look for
- at least one character
- followed by @
- followed by at least one character
- followed by .
- followed by at least two characters
Sending an email can take a few minutes. Form validation is instant.
Which would still not be perfect because “foo@bar”, “foo@[123.123.123.123]” and “💩 @[IPv6 :::1]” are all technically valid email addresses.
It looks like the only validation that doesn’t block something valid pretty much would start and end at “It has at least one @ symbol, and something on both sides”.
So I can’t be directly
bezos@aws?
Even worse is when they strip the plus sign out after the fact and then you can’t log in anymore because you didn’t realize that’s what has happened.
This is criminal. You already send me a validation email, just check for an @ and leave me be
Yees this has happened to me before but with passwords. They have some length limit that they clamp to so you can’t login after registering and I have to do a password reset right after signing up. Happened multiple times to me.
I had a website not let me enter a proton.me email address, when I changed it to my custom.fyi address, it worked fine. They wanted a three letter TLD.
That was my best customer support interaction ever. Company did not let me register with a “new” TLD email address, as “this is not a valid email address”. I wrote them from that email address. They respondend to that email address with “this is not a valid address”. I wrote back “how are we writing, then?” and never heard back 😂
Same although for a totally different reason. There are some services that really don’t like gtlds and they will say your address is invalid if it doesn’t end in .com, .net, or .org…all my serious domains are gtld…so some services have emails on meme domains because the only domains I have with traditional tlds are memes
Not sure if you also do aliases as well but I’ve seen an increase in websites flagging providers like addy.io as well. Extremely annoying that so many websites think they are so important that they refuse an alias.
I had a site refuse my email address for my .net domain. Like wtf, if it’s not .com it’s not a real email address? Idk what that was about.
The issue this is referring to is because the user cannot paste into a text field. And the user was not rude about it either.
So instead of fixing the actual problem, the developer went nuclear and removed the validation. A dick move in my opinion given the developer’s attitude.
~It’s more sad than funny. 🤷♂️~
IMO as a developer this is a sane change. There’s no telling when the format of the first-party api key will change. They may switch from reference tokens to JWT tokens tomorrow. The validation should be using the token and seeing if it works.
If they had made the change for that reason, sure. But the actual stated cause was some pretty thing.
I don’t know what that repo does. But, chances are the dude was just fucking tired of dealing with curseforge. Total garbage scum software.
It’s prism. A multi-launcher for Minecraft Java edition.
I’m guessing removing the validation fixed the pasting, which means it did fix the actual problem?
So the users realized their mistakes and stopped complaining……and other jokes public project maintainers tell themselves while laughing in tears
Prismlauncher! I remember browsing through the changelog and spotting this, made me chuckle internally.
If only I had a penny each time a user told something doesn’t work when it shouldnt’ve.
I’ve done this before. it’s funny when the users are all, “why??!” and to respond with, “because you asked for it!”
