Maybe the full name and meaning of “BIOS” explains it, already?
Basic Input Output System
It’s there precisely for basic stuff like USB, etc. It’s the lowest level software you could theoretically write arbitrary programs for. It’s the interface between hardware and relatively high level software.
I wish everyone would use the correct acronym:
No up-to-date end-consumer product uses BIOS. Some UEFIs on the market don’t even feature legacy BIOS support, anymore.
I thought the UEFI was only the interface between the OS and the firmware (BIOS)? I’m not really shocked to hear about a BIOS update when it comes to fixing an USB problem on a motherboard, even on an UEFI system. And there are still legacy BIOS systems.
The operating system does not access the BIOS or UEFI. The BIOS was and UEFI is responsible for making the hardware “work” and booting operating systems. There is also an infamous setting in many UEFIs: “CSM”, i.e. Compatibility Support Module. It’s more or less like a compatibility mode for the UEFI to make it work like a BIOS, in some ways.
Reading my comment again you seemingly insufficiently quoted, you perhaps were referring to the following part.
The operating system is relatively high level software. But it doesn’t have to be an operating system. It can theoretically be an arbitrary program, as previously mentioned.
It just happens to be the case that it’s almost exclusively used for operating systems. It does not have to be, though.
No, there aren’t. If a shop still sells a product with pure BIOS rather than UEFI, then it’s quite old second hand hardware. There is no up-to-date hardware with BIOS. If there were, it wouldn’t be up-to-date.
Think about what computers were like before BIOS or UEFI existed.
Before the PC, large mainframe computers had ‘peripheral processors’… ie separate small CPU’s just to handle various input and output devices, like line printers, magnetic tapes and card readers. Booting was an entirly human affair consisting of toggling in a bootstrap in binary using switches on a console.
When you consider that, BIOS was a great innovation, combining all those functions one chip
And before that again, early mainframes only had one CPU, and it had to drive everything. So when you ran something like Fortran, it had to link in from its libraries all the drivers for all the I/O devices on the system. When your program ran, it controlled the hardware, directly and its speed was limited by I/O limitations.
There is such a thing as progress, even if it comes down to taking a heap of housekeeping tasks away from the CPU so it could operate more efficiently.
Had to think of these guys, when I read that sentence. This is why I despise some people feeling smart saying “we have always done it like this” or “it has always worked like this” or “why is this new thing there, which makes everything a million times more efficient and we would only need to learn it for a week to accomodate to it and yet we still complain about it as if it were a bad thing”. Admittedly, the last quote example is my interpretation of what these people mean when they say, they have always done something a certain way to have an excuse for keeping things as they are.
New is not always better! But it very often is.
I’m going to make myself even more enemies (which I love to do):
Saying “don’t fix it if it ain’t broken” is wrong, too. If there is a better, more efficient way to do something, it should be changed, even if the old system still “works”.
Back then, when people said stuff like that, I was always the smartass responding with “well, communism ‘worked’, too, you know”.
The OS relies on the BIOS, so yes, it does access it. It’s not there to make the hardware work, though it’s part of its job, it’s also used by boot loaders and the OS until its own drivers can take over.
And the UEFI is an interface to that layer, so again, it’s not incorrect to use that acronym, the semantic is the same.
It was a PS and I was targeting that part, so not insufficiently quoted or it would have been too general. Let’s not get too fussy here, shall we? And no, I was referring exactly to the part I quoted.
Yeah, I know perfectly what type of software can access the BIOS, I’ve made enough programs using it.
Yes there are, I have a few devices that are still have only a BIOS, and others who have both. You are mentioning “up-to-date” (which I didn’t) and implying I was talking about what is built today (which I didn’t) so maybe your confusion comes from that. But in the end anyway, “up-to-date” is all a matter of perspective, I have a 3-year old NUC which is still very much up-to-date and has a legacy BIOS.
I was explicitly referring to “up-to-date end-consumer product[s]”.
You reply to that part of the comment saying “there are still legacy BIOS systems.”. Since it’s a direct reply to the comment of which I quoted the part relevant to the topic you replied you, you must’ve talked about up-to-date systems, or your comment wouldn’t make sense, because you would need to shoe-horn in older systems which I wasn’t talking about to begin with.
So, I made clear, that, as I said initially, there is no up-to-date product which uses pure BIOS rather than UEFI. As a side effect, pointing out, that I was never talking about any hardware on earth, but only up-to-date hardware.
Now, you say that I am confused, because you shoe-horned in all legacy devices, when I was specifically talking about up-to-date devices, only and you specifically replied to me saying that.
So, either you must’ve talked about up-to-date devices or your comment didn’t make any sense, because I didn’t talk about legacy devices. Obviously, there will always be legacy devices left, that have BIOS running somewhere in the world. Wow, how surprising.
Who exactly is confused, again?
I was reading your comment several times and tried to make sense of it. So, I figured you did not sufficiently quote me, since your comment wouldn’t make much sense, otherwise.
However, you now persist on quoting me sufficiently, so perhaps you are as confused as it seemed in the first place.
When I’m inside an aeroplane and I rely on the tower (i.e. Air Traffic Control Tower a.k.a. ATCT), does it mean I can access the tower? Can I go inside the tower, call some airplane pilots and greet them?
The sentence you said is a non-sequitur.
It’s not “used” by anything. The UEFI is using those, as in it being the agent.
I have the feeling you intentionally mix up terminology here, to artificially invent reasonable doubt and then in the end talk about this like “oh, it was just a misunderstanding, we both meant the same thing” or something like that.
No, it’s not. BIOS does not equal USB or SATA port. Again, what you are saying is not making much sense.
Please, show me software, which accesses UEFI functionality from within an operating system.
If you need to know more about the topic of BIOS vs. UEFI, you can look it up online.
One example bringing things to the point is the following post.
@nevj “booting” the mainframe was pretty easy 50 years ago. The first mainframe I was familiar with was the IBM 360/30. Instead of booting we call it IPL the computer. Yes, it was manual. A person press the IPL button, much like you would press the power on button on your PC. The IBM computer had 32K of “core” memory.
You can read more about IPL here;
I worked on a 360, but I never booted one.
On a 1620 you loaded the OS ( called a Monitor program) from punched cards, but you first had to tobggle in a primitive bootloader. I think all it did was set the memory address to which the OS was loaded.
Old computers were fascinating
I used a valve computer called UTECOM. It had a room full of these things, I think they were mercury vapour tubes, and they had a mean time between failure of 10 minutes. The main task of the operator was to walk around with a box of replacement tubes. If you program could not complete its calculation in less than 10 minutes, it had zero chance of success.
I also used IBM hollertih card machines. There was a tabulator which you could program by putting wires into a plugboard, and there was a reproducer and a collator, also with plugboards. We were able to do quite sophisticated calculations with this equipment, for example you could make the tabulator do sums of squares by setting up multiplication as a loop of successive additions.
Old times were fun
You go back farther then me with computers. I started in 1969. Punched Cards! I remember the machines with plug boards, but was never taught how to use the plug boards, but used the card sorter a few times. Back then, punched cards was the way programs were loaded to test with and the way data was first enter into the computer.
Besides the card reader / card punch machine and the printer, we had 3 IBM 2311 Disk drives attached to the mainframe. They disks packs had large platter disks. One disk had the DOS operating system on it and the other 2 disk drives were used for data. The disks could be dismounted and other disks mounted. I believe each disk pack could hold about 7.2 MB.
I started using computers in mid 1960,s. In Ausrralia there were still some surviving first generation computers about then, and I was lucky enough to see them.
I can remember a room with a whole wall lined with punch card cabinets. Sheep breeding data. Over a million punched cards. Today it fits on a DVD or a USB drive… I actually still have a copy of those old data cards.
Yes I programmed in Fortran using punched cards. It was good training. Made you careful.
I remember booking the computer for a whole night, to run a long calculation. When I came in in the morning the operator came out with a long face and said
" sorry the line printer jammed and it printed your whole output on one line… we will have to try again tomorrow"
Disks , when they came in, would have saved me from that
@Akito It looks like you’re here for the pleasure of contradicting people, to the point of contradicting yourself. I have more interesting things to do so I won’t bother to continue a pointless discussion and I’ll just ignore you. I’d just advise you to read a little on the subject.
My first full time IT job was a mainframe operator - 1992/93…
There wasn’t only IPL, there was IML… In fact there was a big red button, fist sized, on the wall, in the operators console area (part of the mainframe room, but not subject to freon flooding) labelled “IML”…
I wasn’t very good at it - never properly learned JCL or JES2 (or even if there were the same thing or not) - I was way more interested in UNIX and Novell Netware, and PC stuf (MS-DOS and gaming etc).
Still it was interesting stuff, IBM 4381, had two processor cabinets and big man height blue cabinets for the DASDs (direct access storage device). Only did it once, it was a scheduled IML - you couldn’t stay in that room when the DASDs were spinning up - it was like standing next to a 747 on the runway… That was before even IBM switched to CPU architecture, the processors weren’t discrete, and they could be updated from microcode, which either came via dialup to IBM, or on 8" floppy disks…
I still have two IBM 4381 “#1” badges - which IBM removed from the processor cabinets when it got upgraded to IBM 4381 revision “#2”…
Love the lively semantics debate UEFI vs BIOS - I concede @Akito’s point…
What would you call the old “open boot” thingie that most UNIX boxes and Apple “old world” systems employed? I don’t think of it as a BIOS… I guess it was more like EISA? i.e. in the case of OpenBoot you had software (e.g. Solaris) that could write stuff to EEPROM (and also some low level EEPROM writing in the very limited CLI tools in OpenBoot) - with EISA, you had to boot MS-DOS (e.g. from a floppy) and run the ECU tools for EISA - MY GOD I feel a headache coming on just thinking about EISA - what a TRAVESTY it was… THANK GOD FOR PCI!