Coming back to this a couple of weeks later, they appear to have changed their site, and https://citibank.com.sg now redirects to https://www.citibank.com.sg/

No, users expect to type what they see in the address bar, and get the same result every time.

Can we even call them address bars anymore?

And blaming website owners for working within the bounds of published standards is ridiculous and you know it.

If I asked any non-technical person under the age of 20 the difference between www.google.com and google.com they probably couldn't tell me. Users expect www.example.com to equal example.com. If a website isn't redirecting one to the other, they, are doing something that is incredibly anti-user and it is a good thing that google/apple are forcing them to do it different.

This is not about users knowing the difference between www and no www on a site that redirects one to the other, but about users who don't know the difference on a site that doesn't. See my comment https://news.ycombinator.com/item?id=17930243 for an illustration of one of the issues that may occur.

> Users expect www.example.com to equal example.com.

No, they don't. The expect that the address they write down works when they type it back in the address bar though.

> If a website isn't redirecting one to the other, they, are doing something that is incredibly anti-user and it is a good thing that google/apple are forcing them to do it different.

You keep claiming this with any sort of evidence or backing. This may be anecdotal, but I worked as tech support for a large org (300+ users) most of my career and have dealt with most types of users. I've only seen them interact with the address bar in one of two ways: explorer shortcuts on the desktop/browser bookmarks (few) or stick-it notes on the monitor or keyboard (many). I'd be willing to bet my next paycheck that not 5% of them are aware that google redirects to www.

At least for those users, none of them will benefit from chrome's mishandling of www. Many of them will suffer from it. I'd be willing to stand corrected with a proper study though.

Also, subdomains aren't only used to make http urls pretty, they are intended to be used to refer to actual physical hardware hosts that belong to a given domain. All published standards I know of are written with this in mind. None of them mandate that www be synonymous to the parent domain, not even those responsible for web technologies. Organizations that follow standards are not at fault for following standards.

If having two different hosts on www and the parent domain was truly "incredibly anti-user" (a dubious claim), let google introduce a rfc at relevant standards bodies and have it go through proper scrutiny first.

> If I asked any non-technical person under the age of 20 ...

Which is why technical specifications aren't written or maintained by non-technical people.

I think this is a fair comment but Chrome has been doing this a lot lately. They'll make a change and developers must scramble to fix their websites.

It's the same with autocomplete earlier this year. One day Google decides to ignore autocomplete="off" and all hell breaks lose.

Interesting to note, they have reverted this change. Google now respects autocomplete="off" in some scenarios (i.e. when autofill is not triggered via name attribute).

Note: autocomplete !== autofill