ellyHello Fedi, @weirdtreething @9lore and I are looking for project name suggestions.
We've been discussing it on-and-off for the past year, but I think it's finally time to split #chrultrabook into two projects.
Just to be clear: There's no drama, not even close. It's just that project grew exponentially, and 95% of support requests come from Windows users, which we couldn't care less about.
Main reasons behind this decision are:
1. Inability to get funding.
Chrultrabook being Windows-centric meant people assumed that paid Windows drivers funded the project.
That is absolutely not the case, as the only person working on Windows support is CoolStar and money goes directly to her.
2. Increasing complexity.
Machines are getting more expensive, and a lot of stuff is changing under the hood. For instance:
- IPU cameras are now standard on higher-end models
- Tablet mode and other various sensors were handled by EC (with exception of Dell machines that used ISH). It seems like noving forward, ISH (Integrated Sensor Hub) will become standard.
- Intel also introduced THC (Touch Host Controller) and various other "co-processors" as they're cosplaying as ARM.
- Disappearing support from Google and CPU vendors (Intel, AMD). Previously, Google's engineers were unofficially helping us behind the scenes. As nearly entire ChromeOS team was fired, we no longer have connections at Google to help us with stuff like publishing firmwares in upstream. I raised this issue this year, and been told that no firmware from ChromeOS will be published in linux-firmware moving forward (at least when it comes to Intel-specific platforms).
Google cancelled AMD Chromebooks (resulting in layoffs at AMD and re-structuring), while Intel is just generally struggling. Many people who helped us with stuff like audio on Intel's side were fired as well.
- Lack of documentation of any kind. Schematics are under NDAs, ARM vendors like Qualcomm or MediaTek refused to share any kind of documentation (even if we offered to sign an NDA), which means that we need to reverse-engineer ARM64 platforms (hence why progress is slow on that end).
3. Documentation.
Because of Windows users, we had to over-simplify documentation to the point where it's not really useful... to anyone. Not to advanced users, and certainly not to developers who'd be interested in helping.
Splitting the project means new documentation, where we hope to document every machine (or at least platform) with as much detail as possible (see postmarketOS wiki pages or my github profile as examples of that).
We're also hoping to create wiki-like website to host pictures of mainboards listing chips and various measurements, which should help people trying to fix their systems and reduce amount of e-waste.
Small taste of how it could look like: https://cros.sakamoto.pl/w/Main_Page
4. Mental health.
Community became unmoderable. Choosing Discord as communication platform was a mistake, people straight up refuse to use forums or github issues.
On top of that, 95% support requests come from Windows users who ignore any rules whatsoever and ask repeated quesitons that are answered in FAQ on CoolStar's website.
Because of that, we lost motivation to work on project (despite loving tinkering with those machines) and burned ourselves out.
We want to leave that baggage behind and focus on what we want to do, without dealing with support requests from Windows users.
5. Re-focusing.
We want to work on ARM64 firmware. Current state of coreboot on ARM is frankly abominable as a whole, full of hacks and regressions.
We want to improve it, despite platform-specific quirks.
Ideally you should be able to just flash the firmware, plug USB with Linux on it, boot installer via EFI and have standard booting method.
So, that's why we decided to move forward with this decision. We really want to get back on track, but it won't be under "chrultrabook" umbrella anymore.
We've been discussing it on-and-off for the past year, but I think it's finally time to split #chrultrabook into two projects.
Just to be clear: There's no drama, not even close. It's just that project grew exponentially, and 95% of support requests come from Windows users, which we couldn't care less about.
Main reasons behind this decision are:
1. Inability to get funding.
Chrultrabook being Windows-centric meant people assumed that paid Windows drivers funded the project.
That is absolutely not the case, as the only person working on Windows support is CoolStar and money goes directly to her.
2. Increasing complexity.
Machines are getting more expensive, and a lot of stuff is changing under the hood. For instance:
- IPU cameras are now standard on higher-end models
- Tablet mode and other various sensors were handled by EC (with exception of Dell machines that used ISH). It seems like noving forward, ISH (Integrated Sensor Hub) will become standard.
- Intel also introduced THC (Touch Host Controller) and various other "co-processors" as they're cosplaying as ARM.
- Disappearing support from Google and CPU vendors (Intel, AMD). Previously, Google's engineers were unofficially helping us behind the scenes. As nearly entire ChromeOS team was fired, we no longer have connections at Google to help us with stuff like publishing firmwares in upstream. I raised this issue this year, and been told that no firmware from ChromeOS will be published in linux-firmware moving forward (at least when it comes to Intel-specific platforms).
Google cancelled AMD Chromebooks (resulting in layoffs at AMD and re-structuring), while Intel is just generally struggling. Many people who helped us with stuff like audio on Intel's side were fired as well.
- Lack of documentation of any kind. Schematics are under NDAs, ARM vendors like Qualcomm or MediaTek refused to share any kind of documentation (even if we offered to sign an NDA), which means that we need to reverse-engineer ARM64 platforms (hence why progress is slow on that end).
3. Documentation.
Because of Windows users, we had to over-simplify documentation to the point where it's not really useful... to anyone. Not to advanced users, and certainly not to developers who'd be interested in helping.
Splitting the project means new documentation, where we hope to document every machine (or at least platform) with as much detail as possible (see postmarketOS wiki pages or my github profile as examples of that).
We're also hoping to create wiki-like website to host pictures of mainboards listing chips and various measurements, which should help people trying to fix their systems and reduce amount of e-waste.
Small taste of how it could look like: https://cros.sakamoto.pl/w/Main_Page
4. Mental health.
Community became unmoderable. Choosing Discord as communication platform was a mistake, people straight up refuse to use forums or github issues.
On top of that, 95% support requests come from Windows users who ignore any rules whatsoever and ask repeated quesitons that are answered in FAQ on CoolStar's website.
Because of that, we lost motivation to work on project (despite loving tinkering with those machines) and burned ourselves out.
We want to leave that baggage behind and focus on what we want to do, without dealing with support requests from Windows users.
5. Re-focusing.
We want to work on ARM64 firmware. Current state of coreboot on ARM is frankly abominable as a whole, full of hacks and regressions.
We want to improve it, despite platform-specific quirks.
Ideally you should be able to just flash the firmware, plug USB with Linux on it, boot installer via EFI and have standard booting method.
So, that's why we decided to move forward with this decision. We really want to get back on track, but it won't be under "chrultrabook" umbrella anymore.
the esoteric programmer
WeirdTreeThing