Query strings are awesome. Especially for one-page applications.

I build a lot of internal applications, and one of my golden UI rules is that a user should be able to share their URL and other users should be able to see exactly what the sender did.

So if you have a dashboard or visualization where the user can add filters or configurations, I have all of their settings saved automatically in the URL. It's visible, it's obvious, it's easy, it's convenient.

>There is also a moral question here about whether it is okay to modify a given URL on behalf of the user in order to insert a referral query string into it. I think it isn't.

These dogmatic technical screeds are all so weird to me. They usually reveal more about the authors lack of experience or imagination than provide a useful truism.

▲

keane 7 hours ago | parent | next [-]

Yes, query strings often enable useful features! But Chris's post, "no unauthorised query strings", is only regarding third parties adding them.

▲

legitster 6 hours ago | parent [-]

But... like... that's a weird hill to die on.

> If I wanted to know I’d look at the Referer header; and if it isn’t there, it’s probably for a good reason. You abuse your users by adding that to the link.

The reason is that the referrer headers are a usability and privacy nightmare. It's weird for the author to jump to such a conclusion.

This referral information is being done purely as a courtesy to the webhost. If we imagined a world in which ChatGPT or Wikipedia launched massive hugs of death on referral links without attributing themselves, that is a much, much worse outcome.

	▲	kyralis 4 hours ago \| parent [-]
		There's a referrer header, if the client wishes to send it. If they don't, the "courtesy to the web host" is done at the expense of the client. This particular web host takes umbrage at other sites taking advantage of their clients that way, which seems reasonable to me.

▲

jimmaswell 6 hours ago | parent | prev [-]

A relatively minor impact concern is that query strings create a new cache entry both in the browser and typically on server-side caches unless configured otherwise, so you might want to use URL fragment parameters if the parameters are only used by clientside JavaScript but the server response is the same.