Has Discourse ever considered adopting RWD instead of AWD? I would like to hear the reasoning behind why AWD was chosen. 
Are there any technical difficulties that would prevent the implementation of RWD? I see this topic, but that was from more than four years ago. Personally, Iâm not a fan of having to toggle between âmobile viewâ and âdesktop viewâ on any website.
Jeffâs answer from the topic you linked to still applies today:
Iâm talking about one view (UI) to rule them all though. The mobile view on a desktop understandably doesnât look right (and vice versa), so why not have one view or user interface to adapt to all screen sizes? Screen sizes on mobile devices have changed a lot in the past four years or so. 4-inch screens are becoming a thing of the past, let alone the 3.5-inch screen for the iPhone 4 days.
Responsive web design should be a standard across the entire internet. I agree with everything you say. It would be convenient on all my devices. 
From my understanding, youâre asking for a âunifiedâ UI standard â if this is the case, Discourse doesnât currently have this on our roadmap.
I understand the desire to have a single simple stylesheet to cover all cases, as it feels good from a dev perspective, and is simple to reason about, but we err to give the best experience for each device - for instance we provide different sets of data to targeted mobile and desktop devices separately based on UI needs+constraints, which allows us to minimize traffic loads. This might not seem like such a big deal if youâre in a highly available, high bandwidth environment, but in other environments, being able to selectively slim down payloads depending on form factor need is a big win.
I suspect I can give you part of the answer:
Android.
Itâs not just a case of different screen sizes. Bandwidth and device performance vary radically.
The mobile view has had to be kept technically simpler with less moving parts to cope with the rather poor javascript performance of a lot of Android devices, especially earlier models.
See The State of JavaScript on Android in 2015 is... poor
If you look at the code source, there are simplifications across the board on the mobile view to cut down the complexity of the assets pushed down to the mobile devices.
As @ Falco points out here, you also want to be able to push down simpler assets not just send the whole desktop down and hide later because thatâs a waste of a lot of resources on the client (whole page in memory etc.)
This is especially important due to the fact that Discourse is a web app that uses a lot of javascript to work automagically and not a normal static web page.
Not everyone has the latest iPhone, which admittedly can crunch through most desktop websites with aplomb.
Perhaps one day it will be possible to deliver a single UI that auto-scales and responds? Iâm sure any dev team would love to deliver a one-size-fits-all approach if they could ⌠?
In a word, âcompromiseâ?
I think weâll be pretty close to fully responsive at some point, especially now that weâve dropped IE11 support. It would be nice to have one set of templates. As others have said though, we do optimize for mobile so weâll likely always have some adaptive features either way.
Oh also, are you finding cases where mobile isnât being detected properly? That might be something we can fix. Outside of development work I never have to manually toggle, and ideally that should be the case for everyone!
The state of JavaScript on Android over a half a decade later is still poor?
Of course, but donât you think that any modern iPhone nowadays could handle it?
That is a blessing. I will pay close attention to the roadmap. 
Yes, that is what I would like to see.
If I recall correctly the latest and greatest Android phones are now performing almost as well as 2017 iPhones.
That should be a good thing then, right? Shouldnât that allow for true RWD in Discourse?
Designing anything for Android right now is such a headache.
Yeah, thatâs what Iâm wondering.
Sure, any recent iPhone is 
(Un)fortunately a lot of people donât buy Apple.
Well the proof is in the pudding, and I think the answer is, yes, it is still relatively poor, e.g.:
Thatâs almost an order of magnitude slower, which, is ⌠wow, pretty slow! 
Now obviously there are reasons why Android works like that to provide a lot of flexibility with regards to what hardware it can run on 
but yeah, downsides.
I think it makes sense to have a desktop and mobile view.
One is landscape the other usually portrait.
One can be huge the other really small.
Squeezing buttons and menus onto a small screen is an art form.
It would be an amazing template that looked good on a large monitor and shrank down to be pixel perfect on a mobile.
I really like the way it works ⌠and the way Discourse looks has improved significantly lately kudos @ Kris
iPhone has majority market share in the United States and a couple of other major countries I believe. Am I wrong?
Haha. If only the US mattered. :).
Yes, thatâs global market share. Take a look at this site.
I think the point here, fundamentally, is Android must be taken into account due to its huge reach. In doing so you have greater technical constraints to take account of and Discourse have navigated that as well as anyone could expect? (or weâd be seeing lots of complaints from users about performance).
In general this is true of any web app approach.
It was brave and cutting edge to go the PWA route. Itâs done wonders for the industry having a major platform proving that route out instead of just falling back to using App Store apps which inhibit competition and destroy standardisation. I donât think itâs hyperbole to state that Discourse having stuck to its guns here has had profoundly positive effect on the industry: it might seem small but there was even more at stake here imho.
To get a good deal as a consumer we must avoid proprietary solutions and narrow support.
PS iPhones are amazing too 
