Address–search unification
In the early days of web browsers implementing built in
Internet search, e.g. via Google, this was done by having
secondary text field to the right of the address bar.
Usually (always?) this one was smaller by default but
could be resized. Today however the address bar and the
search bar has been unified into one single bar. And in
the worst way imaginable. When something is entered into
the this bar, the web browser will first classify the
entered text into one of three categories: 1) URL, 2)
maybe URL, or 3) not URL. In the second case, it till
treat the text as an URL and if it's doesn't resolve,
treat it as not a URL. Text that is not an URL is sent
to a search engine. This means that URL's that are
mistyped and text that is pasted by mistake (pasted
from the wrong clipboard, or previously copied text
that was not successfully replaced with a new text) is
sent to a search engine. This means that any web browser
that for financial reasons have a default search engine
that analyses the user's searches is in practice spyware.
There are three ways to resolve this. The first alternative
is to go back to the good old split, which I prefer. The
second alternative of course is to not even over this
feature. And the third option is require the search to be
prefixed with a particular search keypress or text, e.g.
how Brave lets (unfortunately it does not force) the user
type a colon followed by a letter for selecting the search
engine.