This is the second time I’ve hit things about example.com
When I am going to test an email sender soon, what domain should I friggin use for the test addresses? I want the mails I use for fake test data to not link back to me personally, unless I am specifically looking to receive and look at said mail.
I use something like
user@invalidOr
user@domain.invalidinvalida reserved TLD meant for basically this exact use-case.See RFC 2606
Time to reset the dB again
I used user@example.com … so should that be fine or not?
testwould work too
In my opinion example.com should just return HTTP
418 I'm a teapotstatus response code.That’d confuse the hell out of me if i missed a single config line when updating defaults. Maybe we could have a “you didn’t update all the defaults ya dummy” status code that only example.com uses
I’m very curious about how much bandwidth it consumes. Even with the small amount of content, I wouldn’t be surprised if it’s over 1Gbps.
I wouldn’t be surprised if it’s over 1Gbps.
The site is about 2kB. In order to consume 1Gbps it would need about 62 000 visitors every second, assuming no one uses any kind of cache or anything. Or 5,4 billion visitors daily. Not impossible, but I’d guess the real number isn’t quite that much.
I was thinking more about automated usage. There’s probably a lot of devices that hit it to check for internet connectivity. The linked article says that most of the traffic is automated and they saved bandwidth by putting some content in a JS file (since most of the automated hits don’t download or run the JS).
probably a hell of a lot. I do a lot of travelling for work and thus I’m always connecting to open wifi spots (with a vpn naturally) and many times my browser won’t pick up the capture portal automatically so I use example.com or neverssl.com to trigger the capture portal. It’s also the site I use to a very quick/dirty internet test.
I imagine most people use it the same way.
deleted by creator
I was going to be rude, but I thought I’d just do you the courtesy and paste the relevant section from the article:
As it tends to be heavily trafficked, the overall bandwidth is a key consideration, and to that end one of the changes we made this week was to split the page content into a more basic page, augmented by a separate Javascript file with additional content beyond that. Since most of the traffic to the site is automated it tends not to fetch the Javascript file, reducing overall data requirements.
They’re using JS to reduce the load.
deleted by creator
Yeah specifically the JS holds translated versions of the sentence “This domain is for use in documentation examples without needing permission. This is not a service; avoid relying on it for testing and monitoring purposes.” The raw HTML returns just that + CSS. The JS injects translations of that sentence in multiple languages
deleted by creator
So basically what is happening now. The JS allows it to happen on the client side, taking a HUGE load off the server trying to parse the ua string, find the right translation, bring up the file, and respond. They probably have a single binary that has the HTML, CSS, and JS all compiled in and update it once every year.
Thanks, when I go back to work at the IANA on Monday I’ll take care of it






