Fedora Looking To Offer Better Upstream Solution For Hiding/Showing GRUB Menu

[Anvil]

what i dont understand, is why not Revive LILO ? surely Redhat/IBM could do that an Get rid of Grub i know it dont support BTRFS but surely some Dev could code that extra stuff into it

[starshipeleven]

what i dont understand, is why not Revive LILO ?

What's the point of reviving a project that has none of the features they need, then paying developers to add stuff to it? They might as well add them to the same bootloader that everyone uses too.

[JustinTurdeau]

What's the point of reviving a project that has none of the features they need, then paying developers to add stuff to it? They might as well add them to the same bootloader that everyone uses too.

Because some people think at the level of "oh I heard about this other project, but have no idea about any of the important details. Why not do X, Y and Z stupid ideas guys? g-guys?".

[Anvil]

What's the point of reviving a project that has none of the features they need, then paying developers to add stuff to it? They might as well add them to the same bootloader that everyone uses too.

correct me if im wrong, but didnt Lennart revive a Dead PulseAudio Project before making it the default Sound Server?

[starshipeleven]

correct me if im wrong, but didnt Lennart revive a Dead PulseAudio Project before making it the default Sound Server?

Yeah, and did the same for other stuff that ended up in Systemd repo like udev. Because it had features they needed. Note how I said

What's the point of reviving a project that has none of the features they need.


Assuming that by "LILO" you meant ELILO (the EFI-capable LILO), it has no feature they need, and it lacks A LOT of features they are using and need.

[Anvil]

Yeah, and did the same for other stuff that ended up in Systemd repo like udev. Because it had features they needed. Note how I said

What's the point of reviving a project that has none of the features they need.


Assuming that by "LILO" you meant ELILO (the EFI-capable LILO), it has no feature they need, and it lacks A LOT of features they are using and need.

whats the point? i do believe over time one can Code features into a program, kinda like the way systemd is now, its still getting features with each release. like the kernel also. an Gnome an KDE . , an even XFCE .

[starshipeleven]

whats the point? i do believe over time one can Code features into a program,

Yes but they are not a charity, they do that only if there is no other good alternative to fit their needs.

It is easier to contribute to a project that has already done some work in the same direction than starting from scratch. Or take over a dead project.

GRUB2 supports all Linux filesystems (more or less), ZFS (with a slightly reduced feature set as it's updated slower than the ZFS development), all types of Linux RAID, LVM, encrypted partitions, on both BIOS and EFI, also network boot, it is fully scriptable, similar to u-boot...

ELILO... is a much more basic bootloader that has limitations for even basic things like RAID " It comes with basic GPT and RAID support, but has some limitations (i.e. with BTFS, GPT, RAID)."

https://www.joonet.de/lilo/

RHEL even supported/used it in the past (when Itanium was still a thing) https://web.archive.org/web/20120615...ing_ELILO.html
but eventually droppedit in favor of GRUB2


The reason why it was eventually abandoned by most/all distros is:
-those that wanted a bootloader with all possible features moved to GRUB2
-those that wanted the maximum simplicity way just dropped their kernels/initramfs in the EFI partition and used extremely basic and simple bootloaders like gummibooot (which is the now called systemd-boot)

To recap ELILO is not as small and simple as gummiboot/systemd-boot nor as big and fully-featured as GRUB2.

To drop GRUB2 and use ELILO instead you would need to develop A LOT of stuff for it.
It is much easier to just add the single new feature you want on GRUB2.

[Anvil]

Grub2 is a Mess an takes ages for any fix's let alone Features, an ya wonder why Fedora implements there own into Grub cause upstream is so slow. to go downstream. , Fedora an possibly others need something thats easier an less to maintain . no idea when Grub2 2.06 will come out, shoiuld have come out by now? still nothing , so whens a new release likely to happen?

Redhat/IBM have plenty of Developers that could do the jobor resviving Lilo , even Microsoft are now only using a UEFI based bootloader for windows there moving away from legacy support like EFI. https://lists.fedoraproject.org/arch...LIYAJE66TYTCD/

[starshipeleven]

Grub2 is a Mess an takes ages for any fix's let alone Features, an ya wonder why Fedora implements there own into Grub cause upstream is so slow.

keeping a fork of GRUB with some patches is still orders of magnitude cheaper and faster than adding all features GRUB2 supports to LILO, or just using systemd-boot.
Which is what everyone is doing btw.

no idea when Grub2 2.06 will come out, shoiuld have come out by now? still nothing , so whens a new release likely to happen?

Who cares, the code is in a repo and anyone interested can compile it. Opensource means you are not locked to upstream's whims.
Distros compile from source anyway, and they commonly compile whatever GRUB2 code is latest (regardless of release tags), with their patches added on top.

Redhat/IBM have plenty of Developers that could do the jobor resviving Lilo ,

Quit posting this bullshit, even in the proposal you linked none cares about LILO. The proposal even talks about possibly dropping GRUB alltogether and moving to systemd-boot

This post is just to gather feed back why Fedora should still continue to support legacy BIOS boot as opposed to stop supporting it and potentially drop grub2 and use sd-boot instead.

even Microsoft are now only using a UEFI based bootloader for windows

No they are not. I commonly reinstall or service BIOS machines or VMs with Windows 10, I would notice if they would NOT BOOT, you know.

Microsoft is starting to pull the plug on 32bit installs https://www.slashgear.com/windows-10...port-17620909/ which is stuff that is nearly 2 decades old (or crap hardware with Atoms) at this point.