Legacy BIOS Support Remains Important For Some On Fedora, May Shift Responsibility To SIG

[phoronix]

Phoronix: Legacy BIOS Support Remains Important For Some On Fedora, May Shift Responsibility To SIG

Earlier this month the change proposal was laid out for Fedora 37 looking to deprecate legacy BIOS support. That kicked the hornets nest with many Fedora users expressing their desire to see Fedora legacy BIOS support continue whether it be for running the Linux distribution on dated hardware or even just for VMs without UEFI boot. It's looking more like that responsibility of legacy BIOS support may instead be shifted to a new special interest group (SIG) to take up the work of maintaining and testing that pre-UEFI boot support...

https://www.phoronix.com/scan.php?page=news_item&px=Fedora-Legacy-BIOS-Possible-SIG

[M@GOid]

Of course I support the removal of BIOS support. The less people using Fedora, the better.

[gamerk2]

Wouldn't it be something is Linux were architected in a way where BIOS support (and support for other legacy features) could be done in a sandboxed way that would "just work" years or decades down the line regardless of changes made to the rest of the kernel.

[skeevy420]

I'm going to parrot my talking points from the last article about this:

All they have to do is adopt the Clover Bootloader for x86_64 platforms. It has the ability to emulate a UEFI so Fedora can shift to an entirely EFI based setup while still supporting legacy BIOS systems. They can even use the same disk partition setups outside of the tables (MBR and GPT); especially so if they go with a method using volumes like BTRFS, ZFS, Reiser, or even LVM+(Ext4,XFS).

While that bootloader is normally used for Hackintosh setups, it has all the features that Fedora needs and would be their best choice going forward if they want to keep supporting older systems while advancing the boundaries.

If any of y'all have a more pragmatic solution, I'd like to hear it. "Fuck the BIOS" isn't a pragmatic solution. That only helps to increase the surplus stockpiles of e-waste.

[skeevy420]

Wouldn't it be something is Linux were architected in a way where BIOS support (and support for other legacy features) could be done in a sandboxed way that would "just work" years or decades down the line regardless of changes made to the rest of the kernel.

There's a bootloader that already exists and does just that.

[clockwork]

Of course I support the removal of BIOS support. The less people using Fedora, the better.

Ok... I'll bite. Why is that?

[willmore]

If their UEFI support was good, this would be a bit less worrying. I'm sitting here on a laptop that I have to manually select a boot partition and file every time I boot because Fedora didn't install correctly and can't manage to fix things. Maybe make sure the new thing works before you rip out the already working older thing for once?

[cardpuncher]

Ok... I'll bite. Why is that?

Because this won't attract new users to Fedora, nobody will switch to it for no BIOS support whereas some people might leave as someone on Fedora mailing-list said. And this could well apply to me: I recently installed Fedora on an old HP laptop and it uses legacy BIOS (just fine). I was thinking about switching my main computer too from Ubuntu because I don't like being forced to use snap packages. For now, it's going to be wait and see I guess.

[SteamPunker]

All they have to do is adopt the Clover Bootloader for x86_64 platforms. It has the ability to emulate a UEFI so Fedora can shift to an entirely EFI based setup while still supporting legacy BIOS systems. They can even use the same disk partition setups outside of the tables (MBR and GPT); especially so if they go with a method using volumes like BTRFS, ZFS, Reiser, or even LVM+(Ext4,XFS).

I was thinking the exact same thing. Using Clover to add UEFI compatibility to legacy BIOS systems would be the ideal solution for this. It would make so much sense for Fedora to cooperate with the Clover developers to have the Fedora installation media offer the users to automatically install Clover on legacy BIOS systems, and then remove all legacy BIOS stuff from Fedora itself. As a nice bonus, Clover would likely see additional contributions from the Fedora developers. Also, installing Clover on such older systems would allow interesting additional features such as being able to boot Fedora from an NVMe drive (with the appropriate PCIe adapter), even with the Legacy BIOS on those systems lacking support for that. (It would still require Clover itself to be installed on a small SATA drive or a permanently connected USB stick, but that's no big deal.)

[yump]

If you think BIOS-only machines, or machines where UEFI doesn't work right, are too old and slow to be useful in the present day, consider the following:

The current "midrange" Qualcomm system-on-chip, which is used in a present generation Google product, and many other smartphones besides, is the Snapdragon 765G. These devices provide a high-quality user experience on modern websites, and low-latency interactive UI with all the slick animations. In Geekbench 5, the single thread score for this chip ranges from 570-593, while the multi-thread score ranges from 1511-1816. My own device, which provides a meh-quality user experience on modern websites and a moderate latency interactive UI with the animations turned off, scores 167-179 and 928-1025 respectively, because it's a PoS with no out-of-order CPU cores.

If we go over to the x86 processor benchmarks, we find that the 765G trades blows with high-clocked Nehalem desktop (2008) and Sandy Bridge laptop (2011) CPUs. I am also given to understand that for several years, surplus server CPUs on LGA1366 were very popular in poorer regions of the world.

People who buy laptops at retail and live in places where you can get a new high-performance UEFI-capable computer for the price of a meal in a sit-down restaurant, should keep that in mind.