Didn’t notice this post before in the huge pile of other ones. Anyway, since this was posted:
Is there literally anything that we as individual non-engineers with a mix of Linux distros and hardware can do to help Roblox’s effort? My guess is no, but still figured I should ask in case there is something that will get me onto Linux faster.
The Linux community is one of our most dedicated and technically savvy communities, and their input is extremely valuable to us. We do not want to lose this community.
Rolling out a new product, such as the 64-bit Windows client on the scale we are, is rather challenging. As this is a staggered rollout, it will take some time. We need this time to address issues that arise with the introduction of new technology.
At this point, the biggest help would be to give us the benefit of the doubt and not assume malicious intent on our part.
I wrote a question about this a while ago (see here), but are you able to provide a rough estimate on approximately when A/B testing for the 64-bit update will finish before it starts being pushed out to everyone?
I just want to drop in to say thanks. Seeing this kind of two-way communication is a real breath of fresh air. Normally I just lurk in these sorts of conversations, but I felt like I should say something since I normally serve a lot of criticism.
That’s not to insinuate that I believe dropping Linux support is philosophically a good thing, of course, but I understand that there are likely roadblocks, especially where deep-reaching systems are involved, to preserving compatibility. It is good to hear that at least this is a point of consideration.
I would like to think I also speak for a lot of other people who are likely lurking on this thread right now, but who you likely won’t hear because they’re not replying. Genuinely - thanks. Your transparency is not going unnoticed
I know this is off topic, but I feel like it has to be said.
The fact that Bitdancer is STILL ANSWERING QUESTIONS DAYS AFTER THE FACT is the kind of reassurance needed for me to truly believe Roblox will not block WINE forever considering how communication has gone in the past.
I may not have a whole lot of interaction with Developer Relations, and I am very well aware each and every engineer does not have the time to invest in open communications like this (I work in the MSP field, it’s get the details then get off the phone), but I feel like this thread needs to be used by Roblox to establish better procedures for communicating with its platform. As previously stated, it’s been a long time since I’ve seen this level of engagement with any aspect of the community, and the fact it can be done at the shear size Roblox is now compared to 2008/2009 means it’s still possible if done right.
Has anyone here considered the android client??
If you can emulate that, then it’s not the end for Roblox on linux. I’m pretty sure it even supports mouse and keyboard.
Unless this blocks emulation of the android client? that would suck lol
No, you can’t run VMs such as “android emulators” in Wine/Proton/any compatibility layers in mind.
However, there are better solutions for Linux since Android basically uses the Linux kernel for most of it’s operations.
If you have AMD or Intel GPUs, you can try out Waydroid, which is a compatibility layer for running Android apps, with similar approach as Wine and Proton. Install the ARM64 compatibility libraries in using waydroid_scripts and install Roblox in it. You might need to use a compatible kernel such as Zen though.
If you have an NVIDIA GPU, bad luck though, you do need a VM-based emulator to run the android client. However if your CPU is from 2012 or later, you can use QEMU with KVM enabled + virtio-gpu, and install x86 Android ISOs such as BlissOS in it. Might not be as fast as compatibility layers, but still better than Windows VMs (which requires a GPU passthrough, and you might have to install either a Byfron’d 64-bit client or a Microsoft Store client that’s more buggy (such as mouse locking and acceleration issues) and will probably be Byfron’d very soon) and performance should be closer to the native method.