iOS 26 has been officially released, but we recommend holding off on updating until some Safari issues are addressed. These are visual issues that cause fixed position elements on iPhones and iPads to be misplaced after opening the on-screen keyboard — in Discourse this includes the header, mobile footer, composer, and chat.
We’re keeping a close eye on this bug and are attempting to build some workarounds in the meantime, but solutions are fairly limited due to the nature of the problem being tied to how Safari calculates the window height (which websites have no direct control over).
These issues appear to impact all websites with fixed elements and inputs.
If you’re interested in more details there’s a bug report filed with the webkit team here:
Yeah I’ve noticed this when trying to type a response or a new post on my iPhone. It’s been getting really annoying. I mainly use my PC now for Discourse forums as a result.
I have iOS 26 installed and have been using the Public Betas. It’s wonderful, but trying to respond to a new post or existing reply in Discourse can get a bit mucky. Will wait until a fix comes out.
Das stimmt nicht so genau das Problem ist erst seit IOS Beta 7 aufgetreten. Das macht es nicht besser aber ich hoffe auf eine schnelle Behebung des Problems.
There are often problems with betas that are resolved by the time the public release happens, and issues change even between beta versions. In this case it’s a very basic browser height calculation that’s broken and avoiding the problem is not at all trivial.
I don’t think it’s odd for us to put the onus on Apple to fix this, it’s a very sloppy issue that will impact thousands of sites beyond Discourse, and even occurs in Apple’s own support forums.
I think the link is tagged with Beta 7 because that’s when it was reported. I had the dev beta installed from day 1 in June and experienced the issue from the start.
It seems like it will be fixed in the next developer beta, according to the linked thread. Still, communication about this from Discourse was neither proactive nor precise.
Then the story goes you monitored this for Beta 1-9, held out hope it’s going to be fixed with the final release and then quickly tell people to not update (like they won’t) once it’s been pushed to the general public without the fix?
I get that addressing the problem is out of your hands but one would expect more attention to a show stopping problem on the most important platform.
I understand you’re upset about the bug, but you’re criticizing us for not notifying everyone about an issue in beta releases while also telling us the notification wouldn’t stop anyone from updating anyway… so it seems you’re holding us conditions that are impossible to satisfy.
In more positive news, someone from Apple’s side has responded to the report on their bug tracker, and it sounds like they already have a fix and it just needs to make its way through their software release process. https://bugs.webkit.org/show_bug.cgi?id=297779#c10
I was only made aware of this platform by this announcement. I had raised the issue in my community before realizing it’s not being addressed.
@awesomerobot You are responding to something I didn’t say. I was trying to figure out how your testing of the beta went to have led to this announcement. I suppose I don’t want to believe that you only became aware of the problem when the public update was pushed.
Glad to hear it is getting addressed. Ironically in a dev beta first. Might be time that someone in the team gets their device on them.
I am quite happy with how it went. It is not on Discourse to test on Apple’s behalf. If there is a webkit bug in iOS beta, I would expect it to be solved in production. I don’t want to be even informed about it here on meta. There might be hundreds of such bugs. Of course, it’s iOS beta. They should be all gone by the time it gets released. And if not – we will be notified here. It’s perfect.