UEFI Formed to Replace BIOS
anonymous cow-herd writes "Businesswire reports that several leading technology companies including Intel, AMD, Microsoft, IBM, Dell and HP and others have formed the Unified EFI Forum. The non-profit corporation will assume responsibility for the development and promotion of the EFI specification, a pre-boot interface originally developed by Intel that is intended to replace the aging PC BIOS."
I've said before, and I'll say it again: Why not OpenFirmware/OpenBoot?
Let's go through the list and see what EFI has compared to OpenFirmware, shall we?
1. EFI has a built-in bootloader. (Check)
2. EFI has built-in device drivers. (Check)
3. EFI has a shell environment. (Check, except that OpenFirmware isn't so laughable.)
4. EFI is cross platform. (Check)
5. EFI maintain *some* of the old PC BIOS calls. (No Support in OpenFirmware. Boo hoo.)
6. EFI adds trusted computing. (No Support in OpenFirmware. OF believes in computers being controlled by their owners.)
So why EFI and not OpenFirmware? Could it be a Not Invented Here Syndrome, or something more sinister? Is this the beginning of Trusted Computing for all? How do they expect to get customers to purchase an EFI system when a PC BIOS one is still well supported? Will they try to make an exclusive contract with Dell and invite the wrath of the justice department?
Only time will tell.
Javascript + Nintendo DSi = DSiCade
So.. Is there really any doubt whether Apple will use EFI in their machines? Seriously.. they can't use BIOS now!
- Henrik
- when the Shadows descend -
Intel,
Maker of overpriced, underperforming processors...
AMD,
Leading manufacturer of budget CPUs.....
Microsoft,
Singlehandedly proved that breaking antitrust law can be worth the hassle....
IBM,
Services provider de jour....
Dell
Master of manufacturing, jack of no other trades.
HP
Titanic 2000.
Wow, what a dream team.
... make it about as hard as possible, if not impossible, to impliment a completely free open source operating system. I reckon that is all but guaranteed.
My bet wpuld be on some weird and wonderful, not very good, patented DRM technology that will be forced on it by one of the partners and cross licensed to the others for peanuts. Of course those won't be the licensing terms given to other people
Thinking of licensing terms I have another grumble but I think I'll spare you that one for now [walks off to grumble elsewhere]...
I used to have a better sig but it broke.
there already is too much of a demand for Linux, either UEFI will accept Linux or some motherboard MFGer's will continue to produce mainboards with the old PC BIOS, i don't like the sound of UEFI and will probably go out of my way just to not purchase boards with it...
Politics is Treachery, Religion is Brainwashing
- Written in Assembly
- Not modularized
- Extremely craptistic source code
- Stuck with ancient ways of doing things
- At the mercy of the board manufacturer if you need features outside of what is provided
- etc, etc.
Believe me, I love assembly, and use it at any chance I get, but for something that is as complicated as a BIOS has become, it just isn't the right way to do it.-Jesse
Nothing says "unprofessional job" like wrinkles in your duct tape.
I'm sorry but do you people take the time to read up before you complain? This is a wonderful opportunity for the open source movement. EFI makes booting multiple operating systems like a thousand times easier. Instead of having a single boot record on the hard disk boot information is stored in a data table and given as an option to the user who selections the OS they want.
This means that Linux can be installed without breaking the existing installations or screwing with the boot loader at all. The DRM is a problem but there is not too much information about if there is going to be a lot of DRM in this new bios replacement.
Please correct me if I'm wrong but isn't the BIOS of today backwards compatible with a lot of obsolete hardware that require the BIOS to still behave in a certain way? I belive there were hardware components that for example required that BIOS waited for a certain amount of time before processing some commands due to their startup time. And as years has passed by new features have been added while the old ones are kept and at some point it's a unnecesarily messy code.
Zere vere zwei peanuts valking down der Straße, and von vas assaulted...peanut
If it ain't broke....
tear apart until it is.
Arguing with an engineer is like wrestling a pig in mud. Soon, you realize the pig is dirty, and he likes it.
What language would you put it in? The bios seems like the perfect application for assembly code. The problem is that the bios needs to be kept simple.
Forth, with the Forth virtual machine/interpreter written in assembly. This is the sort of application that Forth excels at.
Cyric Zndovzny at your service.
Seriously, why would they need to join the group promoting EFI? Apple just has to decide to use it, put it in every Mac and that's it. There's not a bunch of motherboard and chipset makers to convince.
As with everytime I post a comment about DRM, someone has to come along and say, "but see, there's a way around it!" Wrong.
DRM'd OSs will not work if the hardware they run on isn't DRM'd as well. This initiative (along with others that may flurish if this doesn't work -- i.e. Phoenix BIOS) is to make certain that the hardware is protected as well so that people won't be able to easily circumvent the restrictions.
Why would they bother to go through all of this if it didn't matter?
I'm going fully Mac when the x86 powermacs come out anyway so Windows is just going to be something I use for emulation purposes.
An obvious troll but I'll respond anyway: Windows will not run in emulation because of DRM. Sure, they might get an emulation layer up and running but it certainly won't be able to do anything that you would be able to do w/the "appropriate" hardware/software... Software will be trusted. Trusted software will not run on emulation layers.
Sorry, welcome to the future.
Anything that aims to remove the rubbish PC bios which is 20 years past it's used by date can't be anything other than a good thing.
And AMD / Intel / Dell / IBM make far too much money selling linux servers or chips that run OS OSes to try and curb that market.
fast as fast can be. you'll never catch me.
So I'm a bit bitter about this: if we can't get enough people to talk with their wallets, we will soon truly have two internets: one for the masses, all EFI'd and bright-shiny-new, and one for the geeks who run ten year old hardware, because that's the last pieces which rolled off without EFI.
Wait a minute... Isn't it us geeks who buy the "bright-shiny-new" hardware before everyone else does? Or maybe are people being duped into buying 256mb $500 video cards to do word processing (hell from my understanding perhaps they are).
So if no one is taking the "first buy" leap then what will happen? Will someone come along and fill in the gap?
You know.... This might make the internet just like the TV was in the 90's and we'll have to come up with another BBS type of system.
"I am the king of the Romans, and am superior to rules of grammar!"
-Sigismund, Holy Roman Emperor (1368-1437)
Yeah, but that's with that particular PC and laptop model. Abilities can vary wildly from one manufacturer and model to the next. Even if some PCs share the same BIOS features, they can still differ in the implementation.
It's the 21st century, and IMHO those features need to be standard across the board-- hell, until a few years ago I couldn't even count 100% on every PC I encountered in the field being able to boot from a CD, much less do any of the other stuff I mentioned in my previous post.
And like you said, having the right boot disk matters for you. For me, I can put a hosed Mac into target mode and connect it as an external drive to any Mac with a FireWire port, to attempt to repair it and/or retrieve data. With a PC you've still got to crack it open and recable at best, or take the whole drive out and put it in a new machine at worst. I've tried out BartPE and a couple other useful boot disks, but having to chase down all the components I need is a pain, and it's hard to make one that is truly universal when it comes to NIC drivers, etc (my company supports a *lot* of different machines).
Blowing the twenty year-old cruft out would be nice, like you said, but I still say the addition of useful features as a standard is what's needed the most.
~Philly
The moral of the story is that, boot diagnostics are cool because you spend less time on the phone. I've never had such an experience with a PC, where if you're lucky you get a couple LEDs. I guess that's what you get for $30k.
Mmm. That would be nice, but you see the problem is that SGIs didn't generally have to cope with a lot of third party hardware. Everything that the firmware would ever communicate with was pretty much known before the box left the factory.
PCs have a huge amount of (often obscure) third party hardware available for them. What makes this even worse is that lots of the standards are often developed after the bios was shipped. How many motherboards do you have which were bought before SATA was widespread? Firewire? It ain't gonna be very clever when you add an SATA card.
In the past, BIOSes have coped with this by being fairly abstract to these things - as a consequence they're pretty dumb and don't have (m)any clever diagnostics.
Buy it's not so simple when you have a world where there are x hundred IDE chipsets, y hundred ethernet controllers, z thousand graphics chips and 100,000 UNKNOWN DEVICES.
Malike Bamiyi wanted my assistance.
Having your debugging be dependent on having a working display (so you can see said menu) is emphatically ~NOT an advantage. If the screen dies on your mac laptop, you can boot while holding 't' (stands for "target disk mode") to turn it into a firewire drive. Plug it into another mac, power up holding the 'option' key (stands for "option") and you can select to boot from the broken machine's drive. Copy your files over to the second machine or even just keep working like nothing happened until you can get your broken machine repaired. What will you do when the graphics break on your wintel laptop? Will retrieving your files be as easy as holding down a key and plugging it into another machine?
All five would be more than happy to have "Linux" be redefined as a cryptographically-signed binary supported by a "responsible" company such as Novell or Red Hat.
The first four, because it suits their corporate customers. Debian, Gentoo, etc. just divert efforts away from supporting the two major distributions that Really Matter.
Microsoft, of course, because they know how to "deal with" corporate entities.
From Microsoft's point of view, F/OSS really is like terrorism. Honest. Like national armies, they know how to wage war against similar entites with known addresses, but have a hard time getting traction against amorphous movements which won't stay put for the ICBM treatment.
Lacking <sarcasm> tags,
There is also another problem from Microsoft's perspective: Trusted Computing that prevents "unauthorised" software from running breaks the "mostly backwards-compatible" guarantee that has helped MS maintain a virtual stranglehold on operating systems. One of the things that keeps people on Windows instead of migrating to the Mac, Linux, etc. is the fact that they can usually re-install their existing software and data on a new machine, and begin using it straight away. When they break this with things like XP Service pack 2, the result is that business users especially stay away in droves, and business users are very important to MS.
This does not of course mean that MS won't try and foist DRM onto everyone, but like some other unpopular things that they came up with (software rental, changing bulk business licensing terms for more complex and expensive ones, Clippy, etc.), it probably won't last because it will adversely affect their bottom line in a big way once the word gets around that you have to buy new software that won't work with some existing data.
And if DRM looks grim on the desktop, it's even more so on servers. There are a surpising number of Windows-based servers that run FOSS software such as Apache, Python, Perl, PHP, FOSS DBMS, etc., and few if any of these programs will end up being authorised. This means not only replacing the entire infrastructure for what are in some cases large and complex web systems, but also rewriting the whole lot to run under ASP.NET on SQL-Server. The sheer cost and time-scales involved would mean that replacing DRM-infested Wintel servers with Linux on a nice IBM PowerPC box that will run your existing web infrastructure without change would be cheaper and quicker by several orders of magnitude (IBM would not be silly enough to DRM-encumber servers that are specifically designed to run Linux).
With the above in mind, I believe that operating systems and software that require hardware DRM authorisation will be a short-lived phenomenon if they ever appear (which is by no means certain). IMO it's more likely that DRM will be used to progressively replace today's software code-signing, e.g. for certain types of embedded web applets, signed drivers that are guaranteed virus-free, and content that already carries intrusive and annoying DRM schemes.