Hey everyone,
We have seen a deluge of posts (on Reddit etc.) about x86 emulation on Asahi Linux in the past few weeks. While it is of course OK to be excited about recent developments, we want to remind everyone of a few things:<li>This stuff is bleeding edge. ABIs are changing all the time. Things may break with any random package upgrade. There is a reason we are not shipping this for end users yet. You are entirely on your own if you choose to try it.</li><li>The state of things today should in no way be taken to be representative of what x86 gaming on Asahi Linux can achieve. Put another way: It's okay to be excited about the things that <em>do</em> work, but the bugs, limitations, and brokenness you might experience are not where we expect to be.</li><li>In particular, if you experience issues or bugs, chances are that as a user you have absolutely no idea what the root cause is or how easy it is to fix. We've been through issues as dumb as "all GPU memory is leaked forever" and "microVMs only have a hardcoded 4GB of RAM", never mind the obvious "without TSO everything is slow". Obviously these issues are not acceptable, nor are they hard to fix, but if you experience the effects you might wrongly conclude that stuff is very broken in much deeper or hard to fix ways than it actually is, and therefore leave disappointed and very misled thinking it's going to take months or years to fix these dumb issues.</li><li>We don't talk about timelines for a reason. Anyone claiming "X will be here next quarter" or "X won't be here for a year" is making things up, either way. Stuff will be done when it's done. Until then, any speculation about when things will be ready from random people is pure speculation, and not based on any objective reality.</li>
We're excited about what there is to come, and we do intend to package and make this all available to users - once it's ready. In the meantime, our recommendation is always to wait until then. You are free to experiment of course, but please be mindful that you don't imply anyone or everyone should try this, and avoid writing "easy-to-follow" guides or scripts that present themselves as being suitable for end users. We don't want news coverage to happen before things are ready, as that will only hurt the project. If you are a journalist watching the progress, we would appreciate it if you wait until things are officially released before publishing any articles about this.
Asahi Linux /
npub15z…3t7sv
2024-06-02 05:57:14
Author Public Key
npub15zxvdefgmppkljpvr6m8cu7pgkv76c7a2dxxlph0z738qx2lq6hql3t7svPublished at
2024-06-02 05:57:14Event JSON
{
"id": "9667103a599f7f165c491f36991f31a75fc20fc1e71476c0735ac1004e276e58",
"pubkey": "a08cc6e528d8436fc82c1eb67c73c14599ed63dd534c6f86ef17a270195f06ae",
"created_at": 1717307834,
"kind": 1,
"tags": [
[
"proxy",
"https://social.treehouse.systems/@AsahiLinux/112545486239767733",
"web"
],
[
"proxy",
"https://social.treehouse.systems/users/AsahiLinux/statuses/112545486239767733",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://social.treehouse.systems/users/AsahiLinux/statuses/112545486239767733",
"pink.momostr"
]
],
"content": "Hey everyone,\n\nWe have seen a deluge of posts (on Reddit etc.) about x86 emulation on Asahi Linux in the past few weeks. While it is of course OK to be excited about recent developments, we want to remind everyone of a few things:\u003cli\u003eThis stuff is bleeding edge. ABIs are changing all the time. Things may break with any random package upgrade. There is a reason we are not shipping this for end users yet. You are entirely on your own if you choose to try it.\u003c/li\u003e\u003cli\u003eThe state of things today should in no way be taken to be representative of what x86 gaming on Asahi Linux can achieve. Put another way: It's okay to be excited about the things that \u003cem\u003edo\u003c/em\u003e work, but the bugs, limitations, and brokenness you might experience are not where we expect to be.\u003c/li\u003e\u003cli\u003eIn particular, if you experience issues or bugs, chances are that as a user you have absolutely no idea what the root cause is or how easy it is to fix. We've been through issues as dumb as \"all GPU memory is leaked forever\" and \"microVMs only have a hardcoded 4GB of RAM\", never mind the obvious \"without TSO everything is slow\". Obviously these issues are not acceptable, nor are they hard to fix, but if you experience the effects you might wrongly conclude that stuff is very broken in much deeper or hard to fix ways than it actually is, and therefore leave disappointed and very misled thinking it's going to take months or years to fix these dumb issues.\u003c/li\u003e\u003cli\u003eWe don't talk about timelines for a reason. Anyone claiming \"X will be here next quarter\" or \"X won't be here for a year\" is making things up, either way. Stuff will be done when it's done. Until then, any speculation about when things will be ready from random people is pure speculation, and not based on any objective reality.\u003c/li\u003e\n\nWe're excited about what there is to come, and we do intend to package and make this all available to users - once it's ready. In the meantime, our recommendation is always to wait until then. You are free to experiment of course, but please be mindful that you don't imply anyone or everyone should try this, and avoid writing \"easy-to-follow\" guides or scripts that present themselves as being suitable for end users. We don't want news coverage to happen before things are ready, as that will only hurt the project. If you are a journalist watching the progress, we would appreciate it if you wait until things are officially released before publishing any articles about this.",
"sig": "5323d92b5e8042b765798e8095c025d6c995ab0bdd2a2e8e652ac2194fde177891339782eb64782ec6907d666b2daf792f73b87cb66cd7f3c87c85e1219e93d2"
}