It’s time to get rid of the part of user-agent strings that identifies which browser you’re using. It should only include things like mobile/desktop, version of html supported, and JavaScript version supported.
It’s not that simple. A lot of browser “standards” are standards in that they achieve the same end result, but for whatever reason they take a different approach to getting to that result, so you often end up needing browser specific code. This is especially the case with CSS, which is why so many “standard” CSS properties still need a “-moz” or a “-webkit” version as well, decades in. The only way the website can know if they’re running the correct code for that browser is if they know what browser is being used, hence user agents. This is the reason that pop ups like this exist at all; sure they were lazy as fuck to not properly support Firefox, absolutely, but they wouldn’t have needed to support Firefox specifically at all if browsers could just get their shit together and fix the “standards”.
I would fucking cry tears of joy if browsers could standardize enough that writing browser specific code and needing the user agent was a thing of the past, but I really don’t see it happening any time soon.
I have to version check to workaround Chrome, FireFox and Safari bugs. Some things they fix and I can flag around version (eg: FF113 has buggy focus detection with Web Components), but some just have never been fixed (eg: Firefox does not support animated styles with CSS variables in Web Components).
That’s not to pick on FireFox. Chrome doesn’t support scrolling two elements simultaneously which breaks any type of fancy horizontal scrolling in horizontal tabs. Safari has some buggy implementation with ARIA tags for Web Components and [type=range] doesn’t follow spec for min.
If we were going to just not support new features because browsers are buggy, we’d never get any new features. It’s better to feature detect and that includes knowing what versions need workarounds.
It’s time to get rid of the part of user-agent strings that identifies which browser you’re using. It should only include things like mobile/desktop, version of html supported, and JavaScript version supported.
It’s not that simple. A lot of browser “standards” are standards in that they achieve the same end result, but for whatever reason they take a different approach to getting to that result, so you often end up needing browser specific code. This is especially the case with CSS, which is why so many “standard” CSS properties still need a “-moz” or a “-webkit” version as well, decades in. The only way the website can know if they’re running the correct code for that browser is if they know what browser is being used, hence user agents. This is the reason that pop ups like this exist at all; sure they were lazy as fuck to not properly support Firefox, absolutely, but they wouldn’t have needed to support Firefox specifically at all if browsers could just get their shit together and fix the “standards”.
I would fucking cry tears of joy if browsers could standardize enough that writing browser specific code and needing the user agent was a thing of the past, but I really don’t see it happening any time soon.
Definitely not.
I have to version check to workaround Chrome, FireFox and Safari bugs. Some things they fix and I can flag around version (eg: FF113 has buggy focus detection with Web Components), but some just have never been fixed (eg: Firefox does not support animated styles with CSS variables in Web Components).
That’s not to pick on FireFox. Chrome doesn’t support scrolling two elements simultaneously which breaks any type of fancy horizontal scrolling in horizontal tabs. Safari has some buggy implementation with ARIA tags for Web Components and
[type=range]doesn’t follow spec formin.If we were going to just not support new features because browsers are buggy, we’d never get any new features. It’s better to feature detect and that includes knowing what versions need workarounds.