This is fun, but it feels suspiciously close to my old presentation about the exact same thing:

https://www.youtube.com/watch?v=xxX81WmXjPg

Maybe it's just a coincidence, but a shoutout would have been nice if not.

Coincidence. I hadn’t seen this.

I got my examples mostly from looking through https://github.com/jackbearheart/email-addresses/blob/master... and reading the EBNF in RFC 5322.

Will watch your video!

Your talk is a lot better than my quiz.

Interestingly you say comments aren’t valid in the middle of the local part but I thought they are, and the parser I used agreed. Maybe obsolete behaviour?

I didn’t realise quotes could be dot separated but it makes sense now I look back over the EBNF! That’s really cool.

Sorry I made you think I’d stolen your content, I can totally see why you’d think that. I promise you that was the first time I watched that video.

Same guy made https://jsdate.wtf, also posted to HN, and also containing material from someone else's talk I'm pretty sure.

Edit: I was wrong!

Similar to this I got my examples for jsdate.wtf through reading the v8 date parsing code and experimentation. Wasn’t aware there was a talk with similar examples for JS dates, would be interested to see it.

My bad, I was thinking about a much older talk[1], maybe something about the theme activated the same neurons. Sorry for making an accusation like that for what turns out was no reason.

[1]https://www.destroyallsoftware.com/talks/wat

It’s okay <3 it doesn’t surprise me that both of these topics have been covered before, and both are fairly small input spaces so there’s bound to be similarities or overlap.

I do love Gary’s “wat” talk, it was certainly part of the inspiration to make these.

There should have been questions without "@" as well...

Are there valid emails without an @? I’m very vaguely aware of bang paths but didn’t want to include them because they’re a bit too esoteric.

Canonically postmaster is valid without a domain in SMTP. Traditionally mail messages can be composed to local users without writing domains in their addressses, and the message submission service will add domains where necessary.

Well, of course, in a multi-user system you can send an email to "user". Typically used for system processes to email "root" back in the day, or "postmaster" or other system accounts, as well.

Bang paths host!user or even host1!host2!user were used in the past, and I think they are still supported by most MTAs. They are definitely valid syntax for email addresses. I mean, if you're talking about quotes and parentheses, ...

As far as I know, that emoji question due to the same reason IDN is allowed in URLs, and they even went as far as their own RFC for it https://datatracker.ietf.org/doc/html/rfc6530 (very similar to the 6532 cited repeatedly, but more important(?) because every system in between the sender and recipient needs to know about hops, whereas the recipient is the terminal node's problem

The LHS is, as far as I know, because the LHS of _all_ email addresses is "if it's deliverable", modulo the rest of the call-outs from the rest of the quiz

I scored 17/21 on https://e-mail.wtf and all I got was this lousy text to share on social media.

I think using example.com is a bit of a trick question. It is known as an example domain which is invalid for all purposes, except of course for serving as an example. :)

https://www.iana.org/help/example-domains

Yes. If I'd known that other questions would also use example.com, I would have answered the first question correctly.

I've actually been wondering for a while if example.com is a valid example email address.

Well, "valid," yes, but "deliverable" thankfully, no

  $ dig MX example.com.

  ;; ANSWER SECTION:
  example.com.  3600 IN MX 0 .

Almost certainly the same for its .org .net and .example friends. There were some other TLDs that were designed for use in documentation, but I don't have the RFC in front of me right now to cite them

Some of the other examples which are "technically valid but obsolete" I would consider safe to call invalid for practical purposes.

Yeah, I almost put a disclaimer at the end saying that while these examples are valid according to the spec, that has no bearing on whether they’re usable in practice. Email is the Wild West.

Exactly. Not valid because it can't be routed.

An explanation of the use of example.com would have informed users that addresses should be evaluated as such, otherwise, AFAICT, all of the examples provided are invalid.

Love it. Minor but question 8 should maybe be 'trailing' instead of 'tailing'?

Good eyes! Code is at https://github.com/samwho/e-mail.wtf if you’d like to open a PR

would be nice if what valid means was clearly defined before the test. what the rfc's regex allows is probably the least useful way to think about email validation.

I do have a short sentence on the homepage saying I used https://github.com/jackbearheart/email-addresses as the validator for the examples. It could be more clearly defined, though.

I see some of the sibling comments seem to imply that "valid" means that it's deliverable at that moment in time, which would make producing a quiz like this some nonsense because then they'd have to actually register xn--tp8h as a domain, and create an MX record for it

Funny. I definitely knew there were some unique characteristics, but this took it to a new level.

Thank you <3 which one got the biggest laugh?

I chortled audibly at the goatse one.

Some notes:

* Single-label domains are problematic in several ways. Traditionally they are used as local abbreviations on the assumption that TLDs can’t be mail domains – tho that’s false, because several ccTLDs have had MX records at one time or another. Still, mail software can vary in whether it might treat single-label domains as abbreviations or TLDs or both. And there’s the historical anomaly that RFC 2821’s syntax disallows single-label domains; this was a drafting mistake not an intentional change.

* Spaces around local parts are valid or not depending on which spec you are following, so question 7 is badly framed. You need to be clear whether you are parsing a mail address as in the SMTP envelope, or an address and display name as in a message header.

* Similarly, comments are not valid in SMTP so the questions about comments are also poorly framed.

* And the syntax of domain literals / address literals is specified by SMTP, so question 16 and 18 are based on not reading enough of the RFCs.

15/21.

I was expecting a question about simple%[email protected].

which I remember was a useful trick when I worked as a sendmail admin in the early 2000s.

Last week I created a script for finding which TLD's have MX records. Result:

    cf: ["mail.intnet.cf."]
    gp: ["ns1.nic.gp."]
    gt: ["aspmx.l.google.com.", "alt1.aspmx.l.google.com.", "alt2.aspmx.l.google.com.", "aspmx2.googlemail.com.", "aspmx4.googlemail.com.", "aspmx5.googlemail.com."]
    hr: ["alpha.carnet.hr."]
    km: ["mail1.comorestelecom.km."]
    mq: ["mx1-mq.mediaserv.net."]
    mr: ["mail.nic.mr."]
    tt: ["ASPMX.L.GOOGLE.COM.", "ALT1.ASPMX.L.GOOGLE.COM."]
    ua: ["mr.kolo.net."]
    ws: ["mail.worldsite.ws."]
    موريتانيا: ["mail.nic.mr."]
    عرب: ["your-dns-needs-immediate-attention.عرب."]

PS: Only http://uz/ seems to have a "working" HTTP server, returning the 500 status code.

I'm not sure if I'd count examples with spaces and comments towards valid emails, because one could argue these are not part of the address, they're email header syntax details. Like, sure, FWS from RFC can be used somewhere in the address, but it can also be in any other header. Also, CFWS and whatnot are going to get stripped anyway by the time you'd have to deal with SMTP (see Forward-path in RFC 5321, which doesn't allow `(nonsense in parentheses)`), so I don't think they should be considered part of the valid address anyway.