That’s pedantic and not even that right. The browser is an agent, not a bot. A bot takes semi-autonomous actions after being instructed by a human, for an undetermined amount of time; A browser just does something it’s told for a few seconds (or as long as the website’s open)

By your definition would a screen reader turn the browser into a bot? If not, if a I program my own “bot” isn’t it just working as my own agent?

Maybe depends on whether you are "driving" the agent personally?

If it’s my bot, it’s literally doing what I told it to do (how I bought a PS5 in my sleep, actually).

I'll just use Selenium to direct the browser then.

I suspect this how a large part of web UI spam occurs. Selenium combined with the latest browsers provide almost perfect control over a real web browser. It makes me wonder if this is how Google tests its "I am not a bot" test. Basically, generate billions of actions -- surfing a web UI -- with slightly different mouse movements (Selenium). Then, use machine learning to generate a network to differentiate between robots and humans. (The humans training data would come from a highly quality population... like all Googlers using internal Google assets/websites.)

Using a headless browser would muddy the waters then.

Yes, it seems everybody has forgotten the meaning of User Agent.

Probably the nuance that is not spelled out (and certainly should be) is that AirBNB expects only interactive user agents, and not automated/batch user agents.

All those lines of code in a browser are automation.

Not to mention screen readers.

Do you always go for the technicality and the gotcha? Isn't that best applied elsewhere rather than here? Because sure, lines of code in a browser are technically automation, but they don't automate your actions so your whole point is moot. And you know this, I'm sure, so you're wasting your (and everybody else's) time.

They do exactly in fact automate your actions.

You decided you wanted to kick off a rather large process by clicking a link, and the browser performed countless individual and dynamic actions as a result.

Just to take one macroscopic crude simple example since we apparently don't have much imagination to work with here, did you evaluate every one of the individual http fetches in that page to decide if you wanted to fetch or render tgem, or did your ublock origin plugin do that for you?

You may not like the inconvenience implied by the fact that a statement like "no automation" is invalid and therefor unusable, but that doesn't change the fact. It's inconvenient not incorrect.

> since we apparently don't have much imagination to work with here

Oh, I'm much more likely to engage in a great conversation after an ad hominem. Sigh.

I am well aware of how browsers work and what they automate for you. But you called them bots. You lumped them with several other types of software, which doesn't make sense. I never said you were incorrect — I said you were going for the technically correct and the gotcha.

Do browsers include automation? Yes. Do things get automated when you use them? Yes. Is software development automation at heart? Yes. The fact that you are correct about these things doesn't make the deception any less obvious: you are using these technicalities to misrepresent what a browser is — which doesn't make sense in the context of the thread -, and that's all I pointed out.

Pedantic was not an insult, but unimaginative was, got it.

Eh, arguably you can do it without bots by opening a telnet (or openssl s_client) connection and typing the commands yourself. Though ironically some websites will block your attempts because it doesn't look enough like a browser.

You also can't really pass Captcha's this way, as far as I know.