Slashdot Mirror
Why Do Computers Take So Long to Boot Up?
An anonymous reader writes "Computers take too long to boot up, and it doesn't make sense to me. Mine takes around 30 seconds; it is double or triple that for some of my friends' computers that I have used. Why can't a computer turn on and off in an instant just like a TV? 99% of boots, my computer is doing the exact same thing. Then I get to Windows XP with maybe 50 to 75 megs of stuff in memory. My computer should be smart enough to just load that junk into memory and go with it. You could put this data right at the very start of the hard drive. Whenever you do something with the computer that actually changes what happens during boot, it could go through the real booting process and save the results. Doing this would also give you instant restarts. You just hit your restart button, the computer reloads the memory image, and you can be working again. Or am I wrong? Why haven't companies made it a priority to have 'instant on' desktops and laptops?"
hibernation?
~~~ Paf. Le chien.
It is because until now, you haven't submitted your question to Slashdot.
... so what little embedded hardware they have doesn't take long to boot up.
On the other hand, my Toshiba HD-A1 HD-DVD player, which I believe runs Linux, takes as long to boot from its flash ROM as my XP box takes to boot from a cold start. I imagine the Blu-Ray players aren't instant-on, either. This is something we will just have to get used to, I suspect.
It's been done, laptops do it all the time.
Personally I don't reboot my desktop enough for it to be an issue.
It's called hibernation. Use it. Laptops and desktops have had it for years. I never turn my Windows machine down. I always send it in hibernation. My computer has been running for a month like this. Next time I actually reboot is probably when I BSOD or install a new driver.
As for TVs, and VCR, they are wayyyyyyyy simpler. That's the reason they turn on so quickly.
I think a large portion of the delay is initializing and setting states for all the hardware. Reducing the kernel and libraries to an image might speed things up, but not by much. It'd be about as fast as starting up from hibernation mode.
If you want a quick start, just use sleep mode. Takes very little power and you're up in seconds.
And no - people don't care *enough* for manufacturers and software writers to do it. Plus even if you power-down and store memory state on the hard drive, it takes a minimum time to load it back in.
"Why can't a computer turn on and off in an instant just like a TV?"
My new HDTV takes about a minute to boot. Something about an ATI bios
Possibly now that we have the hybrid flash/magnetic hard drives this could be possible. A highly compressed start image could be put together by windows much like how you could create a Linux kernel with just the drivers for the things you have in it. This could be loaded into the flash area of the drive and pulled into ram very quickly. A flag could be set to that if the hardware had changed and the system was unbootable, the next restart would attempt to load the OS in the normal way.
I have absolutely *NO* idea what is and isn't feasible regarding boot up times with current methods, but hybrid drives should improve this by a lot. With all of the required boot-up bits written to non-volatile memory with practically zero seek-times, computers with hybrid drives will hopefully boot up almost instantly. And from what I hear, some motherboard makers will soon be including flash memory on the mobos themselves for this very purpose.
IBM gave something like this a whirl back in the day (late 90s) with their Aptiva product line (at least - possibly with others), called "Rapid Resume." If I recall correctly, it just dumped the contents of all memory to an image file on the hard drive, and preserved the swap and shut down. The theory was, on start-up, it would just load the state right back into RAM, and away you went.
The problems I recall it having for certain wer that the time to dump to disk was waaaaay too long, and I don't think it was compatible with FAT32, among other issues. Anyone else remember this thing?
Karma: Excellent, but still won't get you laid.
I triple boot NetBSD, FreeBSD and x86 Solaris on my old desktop with an Athlon XP processor, and 512 MB of RAM. I don't recall off-hand the exact processor speed.
Regardless, NetBSD is the fasted of the three. It takes a little over 6 seconds from power-on to the login screen. FreeBSD takes 11 seconds. Solaris is a bit longer, clocking in a 14 seconds. I know these times because I was curious of this question as well, and so I did the timings. All three systems are basically the default installs, plus whatever initialization file changes there have been from installing various pieces of software.
Solaris does start into X, so that may be why it takes longer. Still, adding the 2 or so seconds it starts to get X running, NetBSD and FreeBSD are still less than Solaris.
Hibernation, as almost everyone's pointed out, works fine, unless you *need* to reboot for an install, crash, etc. In taht case, geez, 30 seconds? Cry me a river, how often do you actually need to restart?
There are two reasons why your suggestion won't work.
First, let's say that you upgrade some hardware. There will be no way for the OS to know that there's new hardware unless it goes through the hardware detection and configuration stages of bootup, which is what takes most of the time. Worse, if it doesn't do this, the system will probably just crash, as the memory image loaded will have the wrong set of drivers installed and they'll be pointing at the wrong set of hardware addresses.
Second, and this is more of a recent issue, there is a lot of work that's going into randomizing memory addresses to increase security. In the event of a security hole, randomized memory addresses make it far more difficult to take control of the machine as a hacker, virus, or worm can't use a hard-coded memory address during the attack. With a pre-built boot-up image, the memory addresses will not be randomized, which defeats a lot of the gain of this security benefit.
That said, you could just use hibernation on your computer. That is essentially the same thing as what you're asking for. A desktop is just as capable of sleeping or hibernating as a laptop is. The only thing is, if you want to make any hardware changes, you must remember to turn on the machine and do a complete shutdown first.
Also, there are companies who are focusing on bootup speed. In fact, every major Linux distro has been focusing on it for the last year or two. It's unfortunately just not that easy to speed things up without sacrificing stability or functionality.
You have just described the concept behind hibernate (suspend to disk) and sleep (suspend to memory).
I have not "booted up" in over 6 months now; I updated my kernel 6 months ago.
The time from opening the lid to being able to use it is under 5 seconds.
There should be no reason why the same can't be done for desktops.
Suspend To Ram.
If you need to reboot, you're rebooting for a reason - likely because something in that "50 to 75 MB" has changed.
Of course, if your box doesn't support suspending to ram, then hibernation is an ok alternative. But sometimes hibernate can be just as slow, if not slower than rebooting.
end of line.
It's just Crap.
back in the day... they used to have this ramdisk thing...
You're whining about 30 seconds? My god, get a life.
When I get up in the morning I go into my office and power up everything then hit the bathroom, start the coffee, walk the dogs then back in to grab a cup of now brewed coffee on my way to my desk.
By then everything is not only booted up but it's warmed up and happy.
You Generation M (microwave oven) kids need to quit demanding everything on demand.
"I want it and I want it NOW!"
You could put this data right at the very start of the hard drive. Whenever you do something with the computer that actually changes what happens during boot, it could go through the real booting process and save the results. Doing this would also give you instant restarts.
Windows pretty much does that already.
it appears that some, including samsung do care and are looking to the future of speeding up boot times.
personally, i think this is a fantastic idea. i really love the fact that my powerbook can go from sleep mode to on in under a second. however this takes quite a bit of battery power to accomplish. wouldn't it be much better if they wrote out the memory to flash when the lid closes? then instead of sleeping, you'd be able to shut down and re start very quickly.
I've spent a lot of time using Windows in virtual machines. For VM platforms that provide on-demand block allocation for virtual disks you can see a typical Windows boot do wild things like write to 250MB worth of blocks that were previously unused (i.e. the virtual disk grows by 250MB). NB: I'm talking about an ordinary boot, not one following installation of anything. It gets harder to see as virtual disk occupancy increases but it's an eye opener.
Some will say hibernation gives the same facility, but (at least with Windows) a clean boot needs to be done fairly often (when using a Windows development box, I reboot it daily).
You can enable Hibernation for your Windows (under Control Panel, Power). This essentially dumps the ram to disk, and then reloads from that image on next bootup.
The added benefit is that not only are you back to Windows sooner, your Windows session is exactly where you left off.
Mock Tech Interviews & Free Resume Review
I'd want to know why Windows takes so long to shut down. It can be as long as booting up, sometimes longer. Seriously, wtf?
hybrid hard disk will put the part of the os in flash ram. But still you still need to do a full flush of ram to get rid of all of junk and leeks that get stuck in ram.
Also in pc there is a lot more hardware then a tv that has to have it's own bios load. Also with a tv the first boot is lot slower then when you trun it off and on later on.
Indeed.
In the beginning, say from Edison's development of the electric lighting system, through the invention of the fractional-horsepower motor which enabled the development of home appliances such as vacuum cleaners and washing machines, most things started up in a fraction of a second.
Then came vacuum-tube-based electronics, which took a minute or two to warm up.
Then came the "solid state" revolution, and, once again, things started up instantly. WIth the exception of television sets, which had a vacuum-tube-based "picture tubes" in them. However, manufacturers soon developed circuits that kept a small amount of current flowing to keep the filament partially warm while the set was "off," producing "instant-on" televisions.
Early hobbyist computers were instant-on, too. Before diskette drives were common, the machine had everything it needed to boot stored in ROM and was up displaying some kind of welcome prompt within a fraction of a second. Even when the serpent entered Eden in the form of "operating systems," startup was quick. When you turned on an 48K Apple ][+ with a diskette drive and spiffy Apple DOS 3.3, there was a brief "whish" as the disk spun and loaded a few K of code into the processor, and there you were.
It seems to me to be lazy design that says that booting consists of more than loading code into RAM and establishing state for the internal hardware. I have no idea why OSes must churn away for big fractions of a minute _running_ code. Why can't it just load a snapshot of the desired final state of RAM?
What really gripes me is that lately Windows and Mac OS X have taken to presenting an empty _illusion_ of a faster startup. What seems to be happening is that all the minute-long processes still churn away, but the processes that present the UI run in parallel. The result is that the visible desktop gets into a displayable and interactive state quickly. But while the UI seems to be ready, nothign else is... particularly anything to do with the local network. If you actually try to do anything on that desktop, you still encounter minute-long delays.
"How to Do Nothing," kids activities, back in print!
I think you are perhaps using the wrong OS. If you want a linux console within 3 seconds of power on, use this:
http://linuxbios.org/index.php/Main_Page
The truth is, this subject is an old one. The main answer is that it just takes time to get a DHCP lease, set up a few dozen services, and generally get all of the "junk" you need up and working without crashing the system. If your main complaint is Windows XP there are a few tuning guides that can reduce your boot time dramatically.
It all stems from the human brains inability to deal with the orders of complexity inherent in the Vonn Neumann architecture.
(I hope I have this story right...this is from memory)
The story goes that the engineer working on the boot sequence for the original Mac was working late one night when Steve Jobs wanders past and asks how long the thing takes - the engineer is pretty happy that he's gotten it down to around 30 seconds (or however long it was) and that's probably good enough. Jobs then comments that they'll probably sell at least a million of these things - and each one will probably be booted a couple of times a day - and the machines will last maybe five years - so if he can save just one second more from the bootup time - that's equivelent to 113 years from the lives of Mac owners. So if you can save just one more second - that's like saving someone's life.
Talk about pressure!
But it's a serious point. The amount of human lifetimes that are wasted waiting for PC's to reboot is pretty horrifying - and there's a lot more than a million of them. Someone should take this seriously.
www.sjbaker.org
It's marked as an Intel something or other, and it's also marked as an engineering sample. I bought it off ebay years ago. It is something like a 500MHz P3. From turning it on to getting the freebsd login prompt is 15 seconds. It doesn't mess around. Note that by "login" prompt, I don't mean boot prompt.
Why this isn't more common, I don't know. My dell and valinux boxes take a minute or so just to get to the boot prompt, very annoying.
Do you have ESP?
And about 3 seconds after I hit a key on the keyboard to wake it back up.
If I do a complete shutdown, powerup to the Mac OS X login screen is about 15-20 seconds.
Yes, this is on a new Intel iMac with the Core 2 Duo.
Computers do a lot of stuff behind the scenes which most people are unaware of. Your computer has more than one operating system on it. Your computer will first be initialized by the Basic Input/Output System (BIOS) and perform a Power On Self Test (POST) to initialize your hardware and to do some basic checks. Many manufacturers hide this step by displaying their logo instead of what's actually going on behind the scenes. The BIOS then continues its boot sequence and searches for an operating system. There is a list in the BIOS of where to search for the OS. It will go in a specific order until it finds an operating system and will proceed to boot. All of this is done before WinXP even begins to be initialized. WinXP does many checks behind the scenes such as the registry and system files. It also initializes virtual memory, etc. etc.
This is the way most computers have operated for the past 20 years. The bootstrap sequence is much faster than it was in the early days when it took a few minutes to check the ram. Now it is checked so fast that you can't even see it happen.
Hate to imagine the amount of human lifetimes lost on slashdot...
James Tiberius Kirk: "Spock, the women on your planet are logical. No other planet in the galaxy can make that claim."
As a friend of mine used to say ...
I used to boot up my computer when I had nothing to do.
Now I have nothing to do when I boot up my computer.
All I ask is a warm bed, a kind word, and UNLIMITED POWER
Around 30 seconds?
I work for a large Fortune 500 company which does IT consulting. My work-issue laptop comes with a lot of baggage, including anti-virus, anti-spyware, automatic backup & disaster recovery, a special system update program, et cetera, et cetera.
How bad is it? It's like this: I can start my computer, and within about a minute, I get a standard XP pro login screen. After entering my username and password, I immediately get up and walk away, down a flight of stairs, out the door, and about a hundred yards to our campus cafeteria, where I'll buy a coffee. By the time I get back, my coffee is cool enough to drink, and my laptop is usually in a useable state.
Recently someone at work got one of the Mac Pros. Chime to login 12 seconds. I was just plain floored.
One of the things that takes much more time on windows than anything else ive seen in the 'tray march' all the misc junk that loads after you login. That can be the really painful part, particularly if you have insufficient memory for windows XP
What gripes me more than slow startup is the idea that a computer can't be shut off quickly.
The last time we had a power failure at work, I tried to shut down my Windows machine, which was on a UPS. For some reason, the machine decided at that very exact instant... apparently _after_ I selected shutdown... that it would be a good idea to download and install a system update first! There did not appear to be any way to interrupt the process. Knowing that the batteries on the UPS weren't what they usta be, I quickly turned off the CRT to reduce the load, crossed my fingers, and hoped for the best.
It took the machine the better part of ten minutes to shut down. Fortunately the batteries held out. Heaven only knows what would have happened if power had been interrupted while it was in the middle of installing a system update.
Years ago the science writers used to tell us that we needn't be afraid of computers taking over the world because, after all, we could always shut off the power. Yeah, right.
"How to Do Nothing," kids activities, back in print!
Is a good thing.
If Windows didn't go through the complete boot process each time how would it come up with random reasons to crash?
Professional Politicians are not the solution, they ARE the problem.
Lurn2...FreeBSD! Different hardware and operating systems...work different. If boot up time was crucial, it could be hammered down. Windows is bloated with crap, so it takes a long time. And I also bet that those who use windows like it to take a bit, otherwise they don't think their computer is "doing enough work." My windows machine boots in about 30 sec, 20 sec of that is getting past my raid prompts, 5 sec is pagedefrag (unless it actually defrags somthing, then another 5), and 5 more for windows to login.
he demonstrated by A plus B minus C divided by Z that the sheep must be red, and die of the rot
30 seconds! When I was a kid computers took 5 minutes or more to boot. We'd have given our eye teeth for a 30 second boot time.
I was thinking the same thing...but then forgot about it as I was launching WoW.
...if you have extreme problems with regards to patience.I mean seriously, this guy is complaining about 30 seconds?In what world is 30 seconds (what, once a day?) so significant that it must be eradicated without prejudice?
1) IBM Mainframe's used to have an idea called 'savesys' - the OS was smart enough to bouut itself to a point it could easily restart from later, and a copy of memory was saved at that point. It was very helpful when the boot process took 10 MINUTES.
2) If you are running windows there is already a lot of pre-indexing being done to reduce start up time. That is why it only takes 30 seconds.
4) As a large number of poster have already replied to you; 30 Seconds? Get a life.
Is there anyone with a fast SCSI or RAID setup who can chime in?
Jef Raskin, creator of Macintosh and Canon Cat (the latter embodied his instant-on ideal), also complained about the time it takes a computer to start up.
Startup times have not changed in several decades. Here are some data points I collected a while ago:
you had me at #!
I don't think the issue here is for people who know computers very well, or even decently.
I think the idea here is this: Ask your mother what Hibernation is. She probably won't know. And she is also probably wondering why her DVD player turns on instantly but her computer takes a whole (yes, I know) minute or two in order to function. For me or you, we understand why. But I'm sure the vast majority of users (who ask "How do I shuffle the cards in Solitare?"), this sort of thing is unreasonable.
The last computer that could do this was the Amiga, and it was so far ahead of its time that they still haven't caught up with it...
Someone's obviously never forgotten to sleep their laptop before entering a big presentation...
Me failed English...
FreeBSD over Linux. If my comments seem odd, this may explain...
I used to have a Macintosh Performa 475 (My first Mac), one of the things I was able to do on OS 7.5 is set up a RAM disk that acted as a real disk (It was mounted on the desktop as the real thing).
One time, I copied the OS to the disk and selected it as a Boot disk (I knew the RAM should be erased when the power goes off, but I was hoping that a restart wouldn't kill it)
And I was right!!, the booting took les than 5 seconds, it was amazing, the computer was as quick as I could wish for.
This was in 1995...
http://www.winsupersite.com/reviews/winvista_05c.
-William Brendel
Mod me down, i don't care.
The person who asked this question is a moron. A computer is a piece of electroniuc equipment that can do billions of operations every second. We're down from a few minutes of boot time (remember how long it took an 8086 to boot? Counting through *all* four meg of ram!) to some seconds and he wants to know why it isn't instant.
Go read a book on how computers work. An old one such as Norton's Inside the 8086 should suffice (or whatever the name is.) The documentation is out there, if you really care, find it and read it.
But please, don't ask stupid questions. What next? Will people ask what acronyms stand for?
Have you read my journal today?
I have a Dell 2600 server that takes over 30 seconds just to complete it's BIOS POST. Then it starts to spin up the disks. It only has 1.5Gb of RAM, yet it just slobs along.
I don't know what manufacturers are doing in the BIOS, but they sure take their time.
I have some current tech 3.2GHz IBM's - they aren't particularly fast either.
How did this anonymous masterpiece make it to the Slashdot front page? Have the reader and the editor never used a windows operating system in years? Hibernate was obvious even to the first poster. Doesn't linux have suspend to ram features as well? Why did this useless near troll of a paragraph end up as an ask Slashdot?
Hibernation is fine, but it is not what the OP is talking about. There are many times you need to reboot, unless you're lucky. E.g., in many corporate settings, everyone has to reboot every day, to get new patches, updates and whatnot; anyone with any flakey device driver will run in to problems from time to time that will need to be cured with a reboot; there are (still) software installs that require reboots, etc. IANASP (I Am Not A Systems Programmer [I'm an applications programmer]), but I've always gotten the impressions that the bulk of the wait time comes from device drivers initializing. It would seem like this would be a solvable problem, as device drivers would not necessarily need to test their hardware and initialize themselves on every boot if they could each save their states on shutdown. (So the OS doesn't hibernate, but the individual device drivers do.) Of course, you'd also need a way to tell the OS that you'd like to do a "complete" reboot w/o saved states. I remember there have been pushes for speeding up boot times in the past, and they have indeed gotten better (I remember when it was common for the BIOS to require a FULL memory check on boot). But for something like this to happen it would have to be percieved across the industry as an important thing -- or else Microsoft would have to push for it.
Hibernation with Windows seems to work okay. I actually use it most often, rather than shutting down the machine.
That being said, I totally enjoy using Centos on my old Dell C640. One thing I don't get, is either a functioning suspend, ultra-low-power modes, or hibernation. I do realize I should try using the BIOS-supported suspend to disk, but I don't want to repartition at this point. It is annoying to setup, with Dell's DOS-based utilities.
Zhrodague.net - I do projects and stuff too.
My T40 Thinkpad takes at least 3 minutes minimum from a cold start to completely finished booting. More if you include the 2-3 minutes of Windows contacting Redmond to authenticate WGA and check for updates. This last step takes 80% CPU and flat out disk while it runs. I could fix that, but WTF.
Of course, it's nothing like saving someone's life. It's like saving 1 second twice a day. Meaningless.
Because the last moving part inside our computers is the hard drive. For the last week I've been seriously looking into the I-Ram. Basically you plug a few RAM chips into a SATA converter and you max out the SATA bus. Granted you can only do 8GB in RAID 0 but the next gen is around the corner and it'll have SATA2 speeds and a 16GB limit. Not cheap but then nobody said performance was. Check out this Google Video to see what it's capable of.
What if Digg added local news and a Slashdot inspired comment karma system? ---
http://houndwire.com
I have a laptop, not a desktop, and yes I don't turn it off. Mostly this is because I have Centos installed, and I haven't figured out how to configure either a hibernation method (via Dell's BIOS), low power modes, or a functioning suspend. So, I leave it on all the time. This is very annoying. I do get much better battery life with Windows.
Zhrodague.net - I do projects and stuff too.
Am I the only one who finds it ironic that this topic should show up on the same day as a topic about Vista hibernation issues?
honestly, this is like the dumbest possible way to ask why we can't have faster boot times.
Ok, maybe not. The dumbest possible way is probably something like:
"why can't the compujigger turn on faster, like the whatchamavision?"
but still, it's pretty damn close.
there is no need to sign your posts. this isn't usenet. your username is right there above your post. stop it.
Man, that's smokin'. I think we should be asking you how you do that.
Not all computers run Windows XP, so not all computers should instantly load it into memory. It is possible, and has been achieved, to boot up an OS almost instantly using flash memory. The only real bottleneck is hd seek speed.
I hate sigs.
I found an interesting oddity in SP2. If you do a plain install of XP SP0, the boot time is impressive... but then you install SP2 over that, and the boot time increases exponentially. However, if you slipstream SP2 onto an XP installation CD (IE make an XP install CD which installs SP2 for you) you get fast boot up times again! Weird... but I am thankful for it. :) My boot up time now only takes a couple of minutes, where before it would take maybe double that time.
Why are you turning a computer on at all? Because you turned it off in the first place. *There's* your problem - PEBKAC.
The reason they don't care is because you turn your computer on at most once a day. I turn mine on once every N months. ie I don't turn it off.
Finally, I cannot believe you compared a computer to a television and said "My TV turns on quicker than this".
Oh wait.
Trolled. Well done, you got me.
What you're describing is mostly solved by hibernate; save the entire state to disk, and when you start up again, some mini bootloader restores the state. Doesn't mean others haven't tried different stuff. WinXP took some steps to make a first boot go faster. Put files accessed on boot somewhere special for faster access, etc. Cut boot down to.
But the basic problem is one of disk throughput and memory usage. There's a hell of a lot of stuff used on boot. CPU usage is secondary to pulling things off of disk. Unlike other computer systems, your desktop isn't intended to run programs directly off of ROM. It's intended to run a variety of applications, and accept a variety of underlying hardware. Since neither nor the hardware is designed to run a specific application from ROM, you can't just start with an assumed operating system or program.
Also, to bring up a nit, your TV only starts up instantly because it's halfway started most of the time. Turn it off for a long time or unplug it and you'll see it take a while to "warm up". This uses quite a bit of power. If you felt like it, you could pay extra to build a motherboard etc that supports suspend for desktops, but it takes a lot of effort to get the software right, so its primarily done for laptops. It'd be a nice comprimise between booting/hibernating and "instant on" that you want.
I Browse at +4 Flamebait
Open Source Sysadmin
When you close the lid, the computer goes to sleep, waits a couple of seconds, writes RAM out to the hard drive, then commences sleeping. When you open the lid later, the computer is instantly on. If your battery runs down before you open the lid, it will be exactly like waking from hibernation. I'm pretty sure Windows Vista has a setting like this as well, it's truly the best of both worlds.
At first it looks like you're asking a question about computers. Then we see you're really just asking a question about some specific product (MS Windows).
Eh, when I get up in the morning it takes me more than 30 seconds to boot up.
I have a computer CRT and I've plugged it in and turned it on cold and it doesn't do what you talk about. It just comes on and starts to get brighter as the phosphors warm up. I don't think that it's a function of a continual biasing, I think it's a change in the circuitry and the way it works.
Either way, it's not the same thing. All a CRT does is get signal information form it's input and move it's guns in response to that. That's rather simpler than all the stuff a computer has to deal with. For that matter the display part of my computer is instant on. It's the tower itself that takes time.
I haven't put a stopwatch to it, but my Xbox 360 boots up much faster than my Windows XP PC.
You want some cheeze to go with that whine?
Doesn't someone read these before they post them?
Must be a slow weekend...
Three quick stories:
1. The Atari ST would boot in under two seconds (it's been a while -- definitely less than five seconds, even from hard disk). This was fast enough that, after several years with Macs and PCs, I had a visceral sense of "something is wrong" when I hauled my ST out of the closet. Hit [reset] and SPROING! the GUI is there in front of you a couple of seconds later. Wow.
2. When I joined Apple in 1987, the systems software folks were getting lots of flack about how long it took for a Mac to start up. There was a sign on the top floor of Deanza 6 saying "Time to boot tripled!" (I think this was System 6.0, memory fades). Through public shaming and general introspection boot times got better, but even today they still stink.
3. Visa asked IBM for the source code to the OS of whatever big IBM iron they were using to process credit card transactions. It took many, many minutes to reboot their machines, and *any* off-line time to Visa is pretty serious (thousands of dollars a second, probably). IBM said, "Are you crazy? You'll just break it. No way do you get the sources." Visa talked to the banks, the banks Had A Talk with IBM, and IBM said, "What were we thinking, here you go, need any help?" Visa hacked on the OS and got the boot time down to under a minute.
Any sufficiently advanced technology is insufficiently documented.
No, not SDRAM, SRAM. Have a chunk of Static RAM that will survive power-downs (e.g some fast Flash RAM). Place boot-files and drivers here, and the whole machine starts in no-time.
For us with OS X and decent power management, suspend will work, even for days. So booting becomes less of an issue. My spanking new Macbook Pro starts in less than 3 seconds from S3.. But next year, Apple machines will contain Flash RAM which will make this quicker boots happen (among other things). As usual, the PCs will probably follow suit.. With suspend finally working, I dont see the point in waiting though.
Tip: The biggest bottleneck of the boot sequence is the harddrive.
Next one is crappy BIOSes. Try turning off boot-checks. The memory checks and POST are usually draining many of those seconds away. Again, Apple is leading with its EFI proprietary BIOS. Its QUICK!
Anyone whos complaining about slow XP boots have probably installed too many programs. Try deleting entries in Run in the Registry.
I hate those MRI and CAT scan computers...they've never saved any one by taking less time to boot...
Money is the root of all evil?
http://www.google.com/search?client=opera&rls=en&q =Why+Do+Computers+Take+So+Long+to+Boot+Up
Today I have an HP Jornada 820 built in 1999. It runs Windows CE, and it turns on faster than anything. You hit the on/off button and you are either on or off just like that. --Best of all, it holds open all of your documents and programs exactly as you left them. I feel confident not saving stuff because it's so rock-steady reliable. The little critter is run on Flash memory; no hard drives.
My PC. . ? Well now. . , that beast is slow. Very slow.
I thought electrons moved at the speed of light, so what's the hold up? I refuse to blame the hard drives; those things are usually faster than Flash memory. So what's up? Bloat-ware? Too much hardware to configure? Poor programming? All of the above?
I don't know, but I suspect that if engineers had their act together and were not constrained by the ridiculous way of doing things which are currently in place, we'd have much better machines available.
-FL
assuming you never actually *do* anything with your PC.
How would booting from a persistant image work with file tables, data caches, network stacks and general data and metadata that change rapidly and unpredictably?
Oops! You loaded your antivirus definitions from last month into RAM for no reason because you forgot to update your boot image.
Or perhaps you add an additional thirty seconds to the load time because your boot image is for tcp/ip settings that conflict with your LAN since you set up DHCP last week and forgot to tell Windows Persistant Image Boot Manager(tm).
Unless of course you form a new boot image every time you shut down, but that will increase your shut down time, and may very well lead to problems if you have power outages or you get impatient and hard power down your system because your OS now takes twice as long to shut down properly.
My MacBook boots in about 10 seconds, and comes out of sleep nearly instantly. This is among the lamest Ask Slashdot questions I've seen.
My (5-year-old) Mac laptop suspends to RAM whenever I close the lid. In about two seconds. When I open the lid, it is usable in 4 seconds, 3 of which is waiting for the hard drive to spin up. It will last in a suspended state for at least a week on a full battery charge.
I suppose it's mostly a benefit of Apple doing the hardware and software together, which is quite a monopolistic thing in itself.
Sorry guys, but all this hibernation talk is way off base. Using hibernation in the future isn't going to help me now, when firefox is leaking memory like crazy, programs aren't working right and I need to reboot to get things running smooth again. I think the best way to help boot times right now is by getting a 15k (or faster?) rpm hard disk...main bottleneck these days is hard disk read speed. Or am I wrong?
I mean, all those gears and counterweights can't be that slow, now can they? Wait...
But a TV is a specialized electronic; basically a computer designed for one job. And because it only needs to do one thing, it only has one task to do as a startup procedure and it can be optimized for that task.
A computer can be used to handle an infinite number of tasks. Considering that, I'd say 30 seconds is a pretty meager time to wait for a machine that will do anything you tell it to.
Given the above comments, very few on Slashdot are virtuous.
Cheers,
Dave
They that can give up essential liberty to obtain a little temporary safety deserve neither safety nor liberty.
Ben
Well, if everybody just stared at their screens and drooled while they booted, I guess you could say something was being wasted. Except for all the quality drooling time, of course.
I was just reading about fast reboots earlier today: Here
did this , so did the vic 20
its not like its impossible
its already been done
back in the day we didnt have no old school
Most computer systems companies do actually spec a "maximum boot time" as a design requirement to engineering. Usually for server grade systems it's ~45s to the windows splash screen.
What goes on during boot is quite often extensive diagnostics of components that commonly fail, or that if they are corrupt, would cause your system to potentially lose data. Hence memory testing, some quick tests of network and pci components etc. While your machine may be doing the same thing every time it boots, it may not be in the same condition every time. You may add in a new card, or your cat may have pissed on it, or a small amphibian may have hopped into the fan duct (yes, both of these happen and have resulted in warranty repairs). Sometimes the effects of these actions are not felt until it is too late, unless diagnostics are run.
Once that's done, the OS has to start figuring out which parts of itself it needs to load. That's not always fast either, because they can't guarantee a consistent hardware set to work from. If you want it to "boot" fast, hibernate, but remember thanks to windoze it's not always a bad thing to reset and blow its brains out on a daily basis with a nice reset.
Well, I've seen UIDs approaching 1,100,000 around here. Sounds like a lot of lives to me. : p
I work at a huge company. My 512MB RAM, P4 CPU notebook takes about 10 minutes to startup due to huge logon scripts shit, corporate antivirus, firewall, antispyware, and lots of additional useless services activated by logon scripts and unable to turn all this crap off. This is really annoyng, sucks and pisses my off way too much. And i do have to turn it off to avoid hardware problems. BTW, an old RS6000 server in the company took 15 minutes to bootup.
- This can't be... - Be what? Be real?
I'm surprised it hasn't been mentioned yet (or maybe it has, and the FireFox search feature has failed me), but Intel's new Santa Rosa platform, combined with Vista's ReadyBoost, should bring boot times down to ~10 seconds by middle of next year. At least for those of us who buy notebooks.
Use a startup utility to remove all the crap you have when you start up. My XP pro install boots in around 5 seconds just because I don't have anything running but a basic environment when I startup - not even the themes module. Another way to increase your boot time would be to press the delete key after restarting you computer to go into the BIOS options and enabling quick boot to disable some of the less useful stuff that hinders your PC boot time.
uhh.. When was the last time I had to boot... Hmm... My system is usually up 99% unless I leave on vacation or something.
logout is about as far as I'll go.. Oh, I'm sorry, I forgot about when MICROSOFT FORCES MY MACHINE TO REBOOT EVEN THOUGH
I TELL THE #$%!@#$@#$% UPDATE PROMPT NOT TO REBOOT!!!!!!!!!!!!!!!!!
Sorry, Forgive me... Thats just one of my Pet Peeves with my XP Box.. My linux mythtv box on the other hand never reboots.
nit-picking... I know...
--- Relax, that mass muderer is just trying to reduce our carbon footprint, one fetus at a time...
"Why Do Computers Take So Long to Boot Up?"
Yes.
This is a question based on lack of research. There are many things you can do.
1. Dump c:\windows\prefetch, and lock it where it can't read/write.
2. Use winlite to strip out crap like WMP, IE, ActiveX, and everything else you don't use, NO OS shouldn't require these things to function.
3. Use a winlogon crack. Believe it or not this makes a speed difference on bootup.
4. Use XP pro not home, so u can do some of the stuff mentioned above.
5. use toolbarcop, and remove everything in there.
6. Don't use Norton, or Mcafee. They bloat the system down like there WAS a virus on it. Norton is considered a virus by my repair shop, and is removed upon exiting the building.
7. Turn off system restore, and dump c:\windows\system32\dllcache
8. Don't use cheapo HP printers and install their garbageware.
9. The result? My build of XP. It will be on piratebay in a few days. It is 450mb. Upon install it is 900mb.
10. Avoid venders like HP, compaq, and Dell that load spyware on the machine out of the factory. Don't use factory disks. FTF (Format the Fucker) and load XP from scratch with a Winlite modd'd copy, and load the drivers. Then load your software.
11. Avoid walmart PCs. Especially the compaqs that take 20 minutes to boot up, and are substandard to a 10 year old machine. Then charge $80 an hour to work on them. The cheaper they are. The more i charge.
12 WHEN YOU ARE DONE. Ghost it to DVD or CD. Then boot from that CD/DVD, and re-ghost (I use ghost 7 corporate) the hard drive from the CD / DVD. This makes a huge difference because it aligns the data on the HD correctly. Try it, and you will see.
My largest argument is WHY shoould anyone have to do all the crap I mentioned above to make Windows run decent? My 7 year old mac takes 10 seconds to boot from a clean install of Tiger.
When government fears the people, there is liberty. When the people fear the government, there is tyranny. - Jefferson
Linux on an embedded system configured for fast booting(without plug and play peripherals etc) can boot in 2 seconds or so.
Engineering is the art of compromise.
It's true. Basically a computer that does billions of calculations per second has no excuse for taking 30 seconds - several minutes to start. Are you telling me it takes 60 billion calculations to display a start menu? I don't care for technical explainations of hardware or operating systems. Engineers who say a computer cannot start in five seconds are lazy or lying!
In this case the user says the startup is too damn slow, and it is. Five seconds should be the maximum time it takes from when the power switch is pressed to when the desktop appears.
This has always been the root of the issue for me on any laptop that doesn't work properly for suspend/resume (or any other power management issue). Severely mangled ACPI just makes things a nightmare. Acer and Toshiba are the only ones that I've had much experience with but it usually takes about two to three kernel versions before there's decent functionality for any given laptop. If you want, under linux, you can get your hands on your dsdt file (under /proc/acpi/dsdt ... it's basically the stuff built into your bios that tells the system how acpi is supposed to work) and then hop over to the intel site and grab their tools to try to debug your dsdt yourself. I did it once, and never want to do it again. It would be nice if some of these companies could just follow the standards (especially when some of them are on the comities that set them). I suspect that there's a lot more that would work properly if they did. Hey, Toshiba, if you setup your dsdt's properly so they didn't confuse the bios and the os, do you think things might boot just a little faster?
According to Anandtech, booting with the i-RAM into Windows XP takes 9.12 seconds.
I'm suprised that no one has mentioned the Linux BIOS Project yet.
From the page:
'lost'. Or masturbating. Or navel gazing. Or beating the shit out of someone. The good news is that someone occasionally gets something out of /..
Nerd rage is the funniest rage.
Don't turn the thing off.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
I have a computer that I've used daily for about the last three years, when I turn it on it is instantly (less than .4 sec) on. It has software compatible with MS Office documents, has a fast browser, a good media player and a nice selection of games. I only have to re-boot it about once a month unless I've installed new software. Oh and did I say its small enough to fit in a shirt pocket.
My point is that if Palm can do it on such limited hardware with a RISC chip emulating a 68000, why the heck can't Windows and/or Linux do the same on much more powerful machines.
Most of the comments I've read just confirm the submitters question: "Why can't a computer reboot faster?" Ummm, because that's all the faster a computer can reboot. The correct answer is: computers could be engineered to boot faster, but they would be slightly more expensive and far less flexible.
Horrifying, eh? By that logic, we can save bucket loads of life by taking down this website.
Fuck systemd. Fuck Redhat. Fuck Soylent, too. Wait, scratch the last one.
Sure, it's kind of trite, and it's no substitute for having a better-designed OS with better-designed apps and hardware, but it really does help.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
113 years? That's nothing. If those fuckers at the grocery store would actually STAFF all those empty registers we could save millions of years. Seriously, why the fuck do you have 20 registers if you're only going to staff 4 of them?
But the plug and play stuff was all detected and configured previously. Why go through the whole detection nonsense again?
When our name is on the back of your car, we're behind you all the way!
And in the GNU/Linux world we have http://upstart.ubuntu.com/ and http://www.initng.org/
However it's not so very interesting on this side of the fence because we don't BSOD or reboot to finnish an install...
If everyone did a few situps, pushups, jumping jacks, whatever while waiting for a shutdown + restart to happen, I wonder what the overall health impact would be? 1 second could turn into 1 jumping jack, which would be 113 * 365 (days in a year) * 24 (hours in a day) * 60 (minutes in an hour) * 60 (seconds in a minute) jumping jacks. 31536000 jumping jacks. How many lives would be so much more prolonged by that amount of jumping jacks? What impact could that make on the high obesity rates in America (guilty as charged...)?
Perhaps those lifetimes aren't wasted by necessity but by negligence, laziness, and choice.
Comment removed based on user account deletion
but i am pretty habitual about shutting off my windows tower when i go to bed. I have often been told that not shutting down your computer can slow things down, so it is rare when i leave it on for a day or two. Reading this conversation, i thought to myself that maybe i shouldnt shut it down every night, to save boot time. is it ever bad that a computer is shut down too often?
Back in '99-00, I was contracting out as a Sun admin for a global corporation, which had set their maintenance window for Sunday mornings from 7 till noon.
My startup times were pretty quick, but I felt for the AIX guys. They were working on an SP cluster, they would have to be done with their work and ready to reboot by 11 am to ensure that the cluster would be up by the end of the window. It generally took around 45-50 minutes.
I laughed, until my next gig which used IBM RS/6000's. 15 minute boot time, each and every time. And THAT was streamlined down from 20 minutes.
I just spent 30 seconds reading your post.
YOU BASTARD!
I can tell you weren't born in the early days of TV when it took several minutes for the tubes to warm up enough to show a picture on the screen. Patience sonny; instant-on computers are probably only a couple decades away!
The bottom line is that the fraction of wasted time stays the same no matter how many people you consider. I hate it when people try to prove a point not worth proving by considering a large population. (That's not what you were doing, it's what Steve was doing. I'm criticizing his and other's logic, not your post.) It's usually an argument used by alarmists to try to get something shutdown (ironically enough). For example, did you know that every year HUNDREDS of lives would be saved if we outlawed backyard pools? How can you let hundreds of people die?!? Well, if a hundred people die in America from something, I consider that an incredibly safe activity. Anyway, to bring it back to topic, I rather enjoy the down time I spend waiting for my PC to boot. It's like a free few minutes where nobody expects anything from you.
On a good day I can boot the Encore 32/67 machines at work in under a minute, but at least with a Windows PC I don't have to punch raw machine code into the front panel to clear memory and run the IPL.
Of course, it's nothing like saving someone's life. It's like saving 1 second twice a day. Meaningless.
But... but... it's a Mac thing! It's a Steve Freakin' Jobs thing! It has to have iMeaning.
Actually, it reminds me of the old saw about this sort of thing: "We need a baby in a hurry. We'll get nine women working on it, so let's call it one month."
Heh. That one never gets old.
Don't disappoint your bird dog. Go to the range.
The big delay in returning from either of these modes is waiting for wireless 802.11g internet connection to wake up and set all its parameters correctly, get an IP address from the WAP, etc. If I know I'm going into my office, I can avoid that by turning off the wireless before closing the lid, and Ethernet is finally smart enough to just ask for DHCP every time it gets connected or wakes up.
The other issue I have is that I normally use a VPN to connect to work, and the VPN tunnel doesn't like getting shut down and restarted, especially with a different IP address, so I still have to re-authenticate by typing in my security token code to the VPN client.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
You gotta see my system. Boot it's nothing comparing to what I have to go through every day.
After my box is done with booting, it starts compiling. Bye the time is done with compiling I have to turn it off before going to sleep.
...and the same story every day.
"Why Do Computers Take So Long to Boot Up?"
Man, you ever watch all that stuff it's doing when you turn it on? That's some serious COMPUTIN' goin' on dere!! It's checkin' the hard drive, the soft drive, the jiggly drive, your momma's board, all that stuff, even activatin' the RAM tough stuff that makes it super powerful and all, and flicking them switches on as fast as it can! You try to figger all that out and then type all that stuff up as fast as your computer does it! Bet it's a lot longer than 30 seconds!
Back in 2000, Intel made an effort to investigate the boot cycle for a PC contrasted with how people actually used their computers. By eliminating things like floppy seeks and redundancies. By streamlining the BIOS, they enabled a POST time of 25 seconds, down from their 60 second benchmark.g /it08001.pdf
1 041_3-5897993.html
http://www.intel.com/technology/magazine/computin
Although I can't find any articles right now, I do remember the term "Zero Boot Time PC" from an article, or maybe it was a Comdex buzzword. Bottom line is that PCs still take time to get to the OS, but it's a hell of a lot faster than it used to be. Hibernation doesn't get me up and running any faster, since I'm running 2gb of ram and XP seems to take forever to save/restore this data. The other downside of XP's hibernation is that when you boot the PC, you have to use it within a certain amount of time, or it shuts off again.
Intel's got a Flash RAM solution in the pipes, it seems: http://news.com.com/Intel+cuts+PC+boot+time/2100-
As to the reason that the article's poster might have such a long bootup time, I'd chalk it up to SATA, RAM counting/testing and the BIOS in general. After googling "zero boot time" I didn't get any hits for AMD, or off brand chipset makers. That leads me to believe that Intel is probably leading the pack, in terms of making PCs POST/Boot faster.
Often in Error, Never in Doubt.
You have to be talking most of the problems in computing of course are windows related. Most OS's out there are faster at booting, some taking less than 10 seconds on ***OLD*** hardware. Depends a lot on the storage, if you use flash, or you use Winchester (HD) tech, or you do a 'RAM drive' I remember the old old days of my Apple II/e/+ systems that had RAM hard drives with battery backup, and they took about 3-4 seconds to boot. A lot of issues are at hand, hardware handshaking, syncing and bootstrapping (thus the term boot'in), loading your daemons, com-stacks, user applications, most of which is bloatware. A nicely designed FORTH system will come up in seconds, most Mac users know by seeing the OpenFORTH of which is the bootstrapping software monitor in all macs since 1994-era. If one uses cars, most of them are quickboot systems, turn the key and the system is almost 'instantly' available. The reason, as I believe I can safely say, your system is slow in booting is due to a very very old concept of a generalized computer designed to do specific tasks. Instead, you should use a generalized computer that is set up in a very specialized way.
When I was in Germany in the late 80's we went to Koln for an Amiga convention. Me and another guy bought a multi-eprom board, some eproms, Amiga OS V2.0.4, a multi-rom board and burned the Amiga OS onto Eprom (heck it is only 4 880K floppy disks). and then put the multi-boot rom board in an Amiga 2000 and when the Amiga was powered on the OS was up instantly! No waiting for the hard drive to boot up. It rocked! Windows could NEVER do that! I would LOVE to see a Linux distro that could do it, I'm sure it could since there are floppy based distros like Pocket Linux that come in around 1.4 meg. Of course the Amiga OS included a GUI....
The Truth is a Virus!!!
I've seen a few Dell's get out of BIOS extremely quickly and into Windows boot loader part. My custom built PC with an Abit motherboard takes forever to get out of BIOS. It's a blank screen, it's a 'press F1 to run setup', it's a hardware IRQ screen and THEN shows the windows boot loader. Geez. How can I make it go as fast as the Dell PCs. What do they do differently?
If it's digital tuning, it may be waiting to have enough of both video and audio data to synchronize. Digital signals as broadcast are (say) MPEG-2 data streams, with video and audio muxed together. However, the audio may predate the video by 1-5 seconds (or vice versa), and for certain tuning methods that pretty much means you have to wait until the earliest packets have their matching data from the other stream.
It gets worse with the HD signals, especially moving to the newer encodings (MPEG-4), as audio and video may be separated by up to 5-10 seconds in typical broadcast streams.
I suppose it's solvable if you have spare tuners doing pre-tuning, but that starts bumping the manufacturing cost and chip or board design costs up. There's not a huge incentive to throw the additional hardware (and resulting complexity) at it. And that would only fix tuning up/down, not entering a channel number. And if you flip too fast, you'd be right back waiting on the sync issues.
The reason it can whip through channels when scanning is because it doesn't need any audio/video sync, it just needs a valid signal.
One of the joys of going digital...
An increasing amount of hardware is using firmware. To save cents, many of these devices are being sold without flash to store this firmware, and are relying on the OS to load the firware into the device on every boot.
This means that the OS must upload the firmware on a restart, or full hibernation. While it is conceivable that a system could be implemented to do this, and leave the device in a conistant state, it sounds like a tedius, error prone setup, that is likely to cause no end of problems.
Of course, you could do away with the problem by making us all pay an extra quater-cent for a few k of flash, like a sensible hardware vendor.....
Prediction for end of Universe #42: Fencepost error in Quantum_bogosort.cpp
My Commodore 64 boots instantly.....
What I don't understand is the last few versions of Windows I keep hearing "my computer turns on in x seconds with OS-1, why does OS-2 take y seconds?" My computer takes 3 1/2 minutes to get through the BIOS to even hit the hard drive. The OS could boot in a millisecond from when the MBR gets read, and it's still long enough that I hit the power button and walk away. That may be part of why I never intentionally turn it off (sometimes the baby, power company, or drunk neighbor has other ideas).
I agree that rebooting your PC is slow, but to be honest I only restart my PC once every couple of weeks, and normally because I have installed something that requires it. I only ever find it really frustrating when I am doing something that requires multiple reboots....but it would be great if it were faster
Hard work is just an accumulation of the easy things that you didn't do when you should have.
Yeah, and if broadband would be at 100Mbps rather than 1.5Mbps or slower, we would save Gigawatts of electricity/energy don't have to import oil/gas to burn, and cut down on polution.
I use a mac powerbook G4 laptop. After a quick scan of the wtmp.x files, my average time between reboots is about 7 or 8 days. Let me translate: I reboot my laptop once a week. Outside of reboots, it goes to sleep, and wakes up in 1-2 seconds. I almost never wait for my computer any more (since I got my new 2g -o- ram).
I think the real question here is not "why do reboots take so long?", but why do you need to reboot so often. The people who design your OS are working to minimize reboot time, but at some point you will have to do a fresh cold boot to set the system up from scratch.
The tools to save that state are not good on windows (see title).
Why does so much of normal proceedure in Microsoft require a reeboot? (see title).
Why are windows OS's so unstable? The answer to this is clear - see title above.
Windows Vista, despite fluff about the power options, does exactly this. By default the Off button is a hybrid suspend/hibernate. Then, it is like a TV with instant on if power doesn't go out. If power does go out, then it is as if the computer hibernated. You get the advantages of instant-on suspend without the lack of reliability in a power outage.
Brant Gurganus http://gurganus.name/brant
The simplest answer to speed up windows is put your virutal memory on a different drive than the system directory is on. The problem with windows is it has to load up the data and then the very next thing it does write most of it back to the drive in the virtual memory. If you add a drive it will not only reduce the startup time but it will reduce the i/o usage of that peticualr drive, increasing system preformance. I've done this on older systems and it can increase the performance by 100% or more. For instance an old dell system would hardly run until I did this, then it worked like a charm. Right now I have two drives on my system - both 250 gigs and I have my c:\windows on one and my caches on the other drive. It's decreased my boot time from 20 to 10 seconds, not to mention that there are no 'hickups' all windows load smoothly on my 3800 amd 64.
- A 'normal' cold boot of my PC takes about a minute
- 'Hibernate' takes about 20 seconds
- S3 Standby takes about 6 seconds.
One catch is that by default most systems use 'S1' mode for standby, which keeps the machine semi-alive including the CPU fan, power supply fan, etc. You can often go into the BIOS, change the default standy mode to 'S3' -- this will shut down the entire machine (including fans, etc.) but keep proviging a minimal power charge to the RAM in your machine so it won't lose its contents.
Since all the content remains in RAM that way, your machine will behave the same as if you did a hibernate, except it doesn't have to spend the additional ~25 seconds writing everything to disk first when you shut down, and also doesn't have to spend that time to read it back into RAM on bootup... Resulting in the ~6 second bootup time.
(While it takes some power for the RAM to keep its information, it is negligible compared to a complete shut down, since any modern PC still provides some power to the motherboard after it is 'powered off'. Case in point: See the LED on the main board indicating the power status on a machine that's supposedly turned off)
It's been a long time since I truly shut down my PC.
Note: the one catch is that if you do lose power to your machine while it is in standby mode, any contents that were in memory at the time will be forgotten again, and it will do a 'full' bootup next time you start. Hibernate doesn't have that problem, but takes significantly longer to shut down and boot up.
Maybe you should GOOGLE the subject before bothering to submit this a Hot Slashdot News of The Day (TM). Both Microsoft and Apple have made great strides towards reducing the boot-up times...
I don't know, but I suspect that if engineers had their act together and were not constrained by the ridiculous way of doing things which are currently in place, we'd have much better machines available.
What makes you think we're constrained? Believe me, if I could figure out some new way to build a better computer, I'd do it. And I'd be a billionaire, too.
You think there is some cabal of engineers or business owners keeping that from happening? Nope. Nobody has figured out a better way yet. What we've currently got is the best we can do.
Soon as someone figures out a better way, you will see it.
Weaselmancer
rediculous.
"They are called Macs."
You Mac guys and your superiority complexes.
My bottom-of-the-line Toshiba A65 laptop can do the same thing, in the same amount of time, using Windows XP in its standard configuration.
The time you spent typing your post would have been better spent trying to understand why Steve Jobs is right.
> you shut the lid on your iBook, and five seconds later, it's in zzz mode (with a battery life of about two weeks - I tested that once)
Welcome to standby, where the contents of your RAM are kept in RAM and the devices are still initialised.
Hibernate copies the RAM to hard disk along with the device state information. This is a lot slower, but lasts 'forever'.
--I thought I was wrong once, but I was mistaken.
Linux, Microsoft, MacOSX would have done it by now. Otherwise, not big enough feature to hit their radar. Besides, faster disk (or disk like) access, and memory will get so fast the it may not be an issue one day.
Well, first off, the comparison between a TV and a Computer is misleading. TVs for the most part, remain nothing more that big Audio Video amplifiers. If I could post a block diagram, you'd have the receiving section (UHF/VHF etc), the audio and video amplifiers with a little bit of tuning capabilities etc, and the presentation (the screen, audio output etc.) There's not much going on in terms of what the device needs to know to be able to boot.
Fast forward to the newer TVs with a lot of digital "intelligent" boxes in them and you can already start to see bootstrapping time.
Computers (circa80s and so on) have almost always required a lot of time to discover their environment, whether it be the associated hardware to discovering the network they're on.
Nonetheless, the question is a good one. Why not? Part of the reason is that in making devices modular, one incurs a certain need to exchange data to make the device work. The interfaces (e.g., CPU to Video card or CPU to hard disk) continue to remain slow... so at boot up time, there is considerable time taken to repeat these very same actions each time. The second reason has to deal with the operating systems we got out there - Why must they control every aspect of the hardware beneath them? Why couldn't it just be a set of modules where they can send a unified data stream and have the device deal with it. This rant ranges from the IO buffering required for some devices to the management of actual devices for consuming data by the OS. I'm appalled everytime I see how many queues get involved in just sending data in and out of a modern OS.I'll readily grant that this is just an off the cuff reply - many here have given equally good reasons and the topic deserves much more careful study. Just my humble 2 cents.
Cheers!
Netbios shared printers and (especially) folders are frequently the cause of long boot-up times for Windows PCs; there's lots o money to be made 'magically' decreasing boot-time delays when this is the cause.
Boottime is dependent on your hardware and which programs start at bootup. Disable any program that is set to start at bootup that you don't use. Look at the benchmarks of your hard drives. Keep in mind that having a slow hard drive and a fast hard drive on the same ribbon(which is often the case when you have more than one hard drive) it can slow it down. RAM speed and the amount of RAM. If your severily low on RAM often the system will have to do a lot of swapping out of memory to the hard disk. Processor speed determines how fast OS, drivers and software loads. Hibernation is kind of like a dumb bootup. Take everything from RAM and put it on the hard disk. Shut off the computer. Hard drive speed is the key factor for this. When you turn it back on take everything from the harddisk and put it into RAM. Device manufacturers must write their drivers to be able to handle what I like to call the "dumb" boot. A lot of devices can prevent the system from going into hibernation. The trick is to figure out where your bottle neck is. The easiest bottleneck to fix is put more RAM in your computer because many computers ship from the manufacturer without enough RAM in my opinion.
The computer 'turns on' just as fast as any electronic appliance, - instantly. Bootup is a process whereby the appliance or PC is placed in a state where one is 'using' it is in some state that provides an environment for interaction. Even though the TV is on and a picture is on screen along with sound, one must go through several processes; to wit, look at the on screen menu, select a channel to watch or 'surf' and finally select a channel. Feed the cat, adjust sound level, etc. etc. When you are finally watching some show, the TV 'boot' process is complete. Likewise the PC must go through similar, albeit electronic processes to reach the environment for interaction. Load the OS, cleanup any outstanding transactions, connect to the internet, restore the last environment, etc. etc. After these processes ar complete and we are in the 'environment for interaction'. This is what takes time. I believe what is wanted here is the ability to present the 'last using' environment first, then continue with boot process. In other words, start the last application ASAP, say the word processor, then continue the boot process, ie connect to web, restore environments, etc. etc. This will require wome work and thought but is doable. An intelligent boot like this should be on the order of starting the word processor plus os load, a reasonable time. These two operations could also be trimmed by similar 'intelligent' loading. Good Luck
A man spends the first half of his life accumulating stuff, the second trying to get rid of it all.
I've been using BeOS lately because I'm working on Haiku and it boots from POST into the full BeOS GUI, ready to use, in about 9 seconds. It shutdowns in about 4 seconds. This is on an Athlon 1.33 GHz with 512 MB RAM, a system that Windows XP takes a minute or two to boot up and shutdown on.
As others have mentioned this isn't really an issue for most other operating systems.
I really have to wonder what the hell Windows is doing at boot-up to be so freaking slow. Keep in mind that BeOS (unlike Windows) can quickly adapt to hardware changes at boot-up, so it isn't like some crazy hardware driver caching is happening or anything.
I used to always joke that the Windows boot-up included a couple million no-ops to be so damn slow. Anyhow this is one reason of many I'm helping with Haiku.
I also find it fascinating that the new "performance features" touted in Vista are really all just kludges to work around Windows' inefficiencies. The fact that other systems (like, ahem, BeOS and Haiku) don't need such kludges shows this is a Microsoft Windows problem.
This is about selling Microsoft's Vista. I don't like these sorts of disguised posts. No, there's no compelling reason to buy Vista no matter how well hibernate works.
You can lead a man with reason but you can't make him think.
Hibernation insuffient resources exist to complete the API see MSKB 909095.
Download and install to fix (you don't even have to ring PSS any more!) IMHO, this is an update they really should push out using Automatic Updates. (They probably will when the next security bug is found in the kernel.)
My cell phone takes over a minute to turn on. I never turn it off now, ever. If I am lying somewhere dying and my cell phone was turned off I would be dead before it would have started up and I would be able to call. I understand why a PC takes a while to boot up. Loading kernel, drivers, etc. And this is not just limited to Windows. Linux, OS X, Solaris, they all have their start up times. But why the heck should portable electronics, like a cell phone, take over a minute to "boot up"?
Because the memory image that you speak of is dependent upon every single bit of hardware you have, every single driver you have, every single system file, your BIOS version, and every single CMOS setting.
So, each time you get a Windows Update hotfix, you'd have to re-build the image. Every time you installed something that loaded at startup, you'd have to rebuild the image. Every time you changed a CMOS setting, you'd have to rebuild the image - and so on, and so forth.
While that's entirely doable for some folks, it is either impractical or beyond the reach of over 99% of the computer users in the world, and Microsoft (and the hardware manufacturers, BIOS writers, etc.) simply aren't going to expend the resources necessary - it would be a money-losing endeavor.
In the case of a television or other settop box, it's more practical as the various settings and hardware aren't going to change, you can build an image once, slap it on half of a million units, and ten years later, having those units run from that image will still be just fine.
Many devices with embedded Linux still don't do that, they use the regular boot process, but have the kernel and OS *HIGHLY* stripped down, so that it happens very, very quickly.
Oh, you're not stuck, you're just unable to let go of the onion rings.
Tube filaments were designed for a warm up time of 11 seconds. Since the resistance of the filaments varied as they heated, it was important, in a series string where the low voltage filaments operated from the 120 line, to keep the filament heating uniform so the voltage dropped across each tube stayed relatively constant as all the tubes got up to operating temperature.
A QD_005
I agree with the parent that there were sets where the filaments stayed on all the time for an "Instant-On" effect. Actually it was an always-on situation, but the B+ and high voltage wasn't applied until the set was "turned on". See http://www.repairfaq.org/REPAIR/F_tvfaqd.html#TVF
"Do the Right Thing. It will gratify some people and astound the rest." - Mark Twain
Randomising the memory addresses seems like a really dangerous way to secure a system, it wouldn't take long to scan though memory performing an operation to break an operating system.
Instead memory protection exists where only the operating system can access such memory locations, gates and other techniques are used to give access to services and programs.
I suspect the major hicks to having instant-on is device initalisation and simply reading the system files from the harddisk.
What's the point of shutting down a computer (except for hardware upgrades)? My desktop stays on for months at a time, so it's always ready to use. Furthermore, I can return to the 30 or so open windows (on several virtual desktops) and be exactly where I left. Also, if I'm away from my desk, I know I can ssh in. And it can run all the cron jobs at night, rather when it is restarted. [Of course, I do save power by getting the screensaver to DPMS off the monitor after 20 minutes]
So, if I only boot once every 100 days, what does an extra 30 seconds mean? Admittedly, on laptops you have more of a point, but you can leave it running for 2 hours on battery, or suspend to ram/disk.
Windows, being the user-friendly-excuse-me-while-i-go-vomit operating system cunningly hides all the background activities behind the windows flag logo. If you want to get a real sense of what goes on, watch the linux kernel spew pages of messages (and then go back and skim through them) while it boots. I'm not an expert on OS (operating system) kernels but here's a quick sketch of it...
But the digest version is: basically an operating system (be it unix or windows or whatever) loads all the drivers for your hardware and gets a sense of the rig's configuration. And in case something has changed while the computer was off, like a new video card, or RAM upgrade, or something went toast because of a power surge; the configuration cannot just be stored a blindly read again.
Aside from that some of these operations take time because of hardware limitations (harddrive for example). Network initialization, especially if there are issues on the other end like a dead router/modem, can take some serious time. Many modern OSes is a plug-and-play system, which means it's supposed to recognize hardware on demand which means it has to do a bunch of tests on the system's ports (USB, COM, PCI, etc) and if it finds a device like a joystick or a mouse it has to fetch the drivers for it. Many of these events also have to happen is sequence which means waiting.
Also some programs are run while you are still seeing the windows logo (mainly server applications) so it of course adds to the boot time.
If booting is getting under your skin, why bother turning your PC off in the first place?
Oh of course... you are running windows, crahses etc, now I understand. Sorry, Linux user here, rarely need to reboot or boot, so I leave it running all the time, not like it uses megawatts of power.
Move along, nothing to see here.
RSC
But it's not just digital TVs that do it. I have about a 15 year old TV. I also have a TV tuner for the computer. Plugged into the same analog signal, the tuner takes a second and a half or two to change channels. Many (analog) TVs I've seen are about the same. But that over-a-decade old TV... changes channels almost instantly. Maybe a quarter second.
I always get very frustrated when on a TV that takes a while to change channels.
And from what you're saying it can take up to 10 seconds on digital? Holy crap... how can you channel surf with that? Maybe I'll just not go digital. Sometimes old tech is better.
What makes you think we're constrained? Believe me, if I could figure out some new way to build a better computer, I'd do it. And I'd be a billionaire, too.
Oh, I dunno, just the 50 years of computing showing that backwards compatibility is the main tool to keep your customers... Witness the bending over backwards everyone does....
The biggest problem of booting up like this is that the contents of memory and cpu registers isn't enough. The hardware has to be properly initialized as well. Since the internal state of the drivers indicates that has already been done, a consistent mechanism to force re-initialization of all hardware has to be in place after the system reloads the image. That might take as long as a normal boot does.
[100% ISO 646 Compliant]
SVM, ERGO MONSTRO.
Hans Moleman: You took four minutes of my life, and I want them baaaaccckk!! . . . Oh well, I'd only waste them anyway . .
Depends whether saving peoples' lives is of value or not. My take is that saving two seconds a day for a million is not meaningless.
Essentially what's needed is for the computer to boot normally and get to where the user can start interacting. At this point save the ram image like hibernation.
Now, if I shut down/restart the computer, the software should be smart enough to know if the boot process has been altered and the hibernation image is valid, or if it should be truly rebooted and a new ram image dumped.
Back when I used Windows I used to reboot occasionally, and the reason I rebooted was that XP starts crap out after running for too long. Not as bad as 9x, but it still did. Or an application or video driver would lock up and the system would go down. In these cases I had no time to hibernate, or a hibernation would just save a bad memory state. If a previous hibernation image was already held, I would get fast boot AND a cleanly running system. Kind of a hybrid system, but one that I agree with the OP, has much potential.
On a K6 II 350, BeOS would boot in less time than it took to POST. Why?
You think all those fancy widgets, transparent windows, fading/sliding/animated in some assinine way menus only require a couple lines of code to work? A lot of the computer code a system loads any more is stuff that makes it more "useable" for everyday folk. You kids and your fancy UI's!
No sig for you!!
The Palm is still powered up and retaining its memory state when you turn it "off", and turning it on only restores the power to the display. When you do a reset, that actually reboots the Palm OS, and it takes MUCH longer than one second.
Mainframe/UNIX Bit Twiddler and long time Windows/Linux Hobbyist.
The Theorem Theorem: If If, Then Then.
Television sets don't have hardware that requires POST. Computers have to make sure the hard drive is there, that there is an operating system, and every other input/output device that is connected. Feel free to disable as much as you'd like in BIOS, but your computer is not a television, as much as it may act like it these days.
I use win98(dos console without explorer) and BIOS set to cache to ram and several checks disabled.My database/PIM program on other hand takes 40~55 seconds to load,slowest thing on my drive.
If you are so concern about boot time, why shutdown?
Laptop users either use sleep, which only powers the RAM, or uses hibernate, which writes the memory in blocks into the hard drive, and reads it back on boot. Win2k supported this feature a long time ago.
Putting your desktop into sleep, if properly supported, should have the feel of instant on. At work, there is this Dell that auto-sleep, and wakes up in a couple of seconds, and you can probably change the behaviour so that either it wakes on LAN on not, so if you are paranoid about malware, you can make sure that it doesn't wake on LAN.
And no, booting up computers aren't just about reading a bunch of stuff in memory (this ain't the days of DOS, and even DOS needed drivers on boot), so your reading large chunks into memory is basically waking from Hibernation.
Back when I was using an Amiga 1200 as my primary computer, I timed it as taking around seven seconds from flicking the power switch to being 'ready to go'. I, too, wonder why it takes my HP laptop more than 30 seconds to be in a usable state from hitting the power button. Sure, it probably has to do more on bootup than the old Amiga did, but the Amiga ran a 14Mhz 68020 and the HP runs a 1.8GHz Core Duo...you'd think that would make some sort of difference to things.
Even coming up from a standby or hibernate, the drivers have to re initialize (depending on the OS) to some extent.
Another thing that creates long boot times, especially in the windows world is crap. Low level anti-virus crap, and things that software should not be doing. This is one area Vista helps.
Also remember we were paging out systems with 32/64/128mb of RAM in the past, and now we are dumping Gigs of data. Vista and Xp have optimizations that compress and reference pagefile info, but this is still a lot of RAM even with the speed of the Hard Drives today.
Another area that kills performance is extra low level driver and hardware, especially in relation to BIOS. For example I have a legacy free Toshiba from 2002 that can resume from hibernate in less than 4 secs, including bios time. The extra legacy hardware not being in the systems BIOS makes a big difference on the pre-OS boot.
Another area that doesn't' get enough attention is device pairing. When I was active in OEM building, we would not only test components, but performance/driver related issues to the device in relation to each other. A system with super specifications, and one bad pair of hardware/drivers can retard boot performance tremendously, it also will slow over all system responsiveness.
The device pairing is something that doesn't get a lot of focus, but is extremely important, and it is sad when I see high end Alienware or even Dell systems that have poor pairs (even integrated chipsets can be poor pairs).
As an example, my Theatre computer is an AMD +2800 with a normal 7200RPM drive and dual pair 400/800mhz RAM. Its specs are not impressive, but because I paired every piece of the system, it can boot(not resume from hibernate) to a Vista desktop in under 15secs, and an XP desktop in under 10seconds, including BIOS time. This is in pairing low level components, as it has a ton of USB devices and TV Capture, etc.
So for pairing, for whatever reasons, the main board and chipset are very important, as well as the HD controller, Video card, and even the brand/model of HD itself.
As for pairing, it helps if you have access to several brands of components when building your own system. Although there are exceptions in the past, ASUS boards usually have the best stability and work the best in pairing with other components. However, your results will vary.
We could get into paging and hibernation and how ACPI works and advantages to legacy free and EFI, but if you can pair well and not have bunch of software in the lower levels that just shouldn't be there.
A final note, there are a lot of myths on boot times, especially in the Windows world. One I hear all the time is having too many fonts will reduce boot time. This is not true in Win2k/WinXP or Vista. The sample AMD +2800 theatre system I referenced above, has over 5000 fonts installed, and still boots just as fast as having only the base fonts installed.
I used a free utility that would create a partition in RAM to be read as a drive. I wrote a little .bat file to be called from the autoexec.bat on bootup which would 1) create the ram drive, 2) copy over the larger and most frequently used executables and system files on the drive to the ram drive, then SET the location of the programs to the drive letter of the ram drive. The idea was there are a certain number of programs that are loaded everytime the computer turns on, and repeatedly while it stays on. Immediately getting these files off the drive and into memory and resetting their location for the operating system's reference to the ram drive enabled the grunt work to get done by RAM, which was fast and silent, rather than the hard-drive (which in those days was super-slow and noisy). The result was the computer booted much quicker and would operate a heck of a lot faster. What we need is the equivalent of a gigantic RAM drive which is immediately loaded with all the most frequently accessed executables each time the system turns on. This should be the first operation of bootup after the system checks are complete. After the RAM drive is pre-populated with the major OS executables and startup programs' executables, the operating system should then run itself off its executables located on the RAM drive and look to the ram drive, 1st, as the location of all startup programs' executables when attempting to load them at bootup. Is this already being done, if not, why?
-- I'd give my right arm to be ambidextrous
My issue isn't so much the boot time (which is pretty quick if you don't count the stop for a second for the bootloader) its the time after I type my password until my desktop is available. I think it mostly has to do with the network drive I have mapped (which I use frequently and won't unmap), but I also noticed once or twice after fresh wipes once I updated Windows, the login time goes from almost instant to about 10-15 seconds or more. I.e. the little login music is playing while your desktop is appearing, then you update (installing no third party apps) and suddenly the music has played, and you're still sitting there twiddling you're thumbs waiting for your desktop. It is microsoft's doing. I have no idea which particular update it is. Next time I wipe, maybe I'll install them one at a time until I find out which one it is.
Remember those numbers vary from install to install. They were likely running a fairly new install but not completely so.
Probably 80% of the boot time is crappy drivers and helper apps that seem to accumulate over time.
I put my OS on a Raptor and a clean install boots in roughly 6 seconds. A few months later, it's up around 20 seconds. Give it a year and I have no doubt I'll be sucking up near 60 second boot times as the assorted cruft Windows picks up tries to initialize itself and happily conflicts with everything else.
Much as I love knocking Microsoft as much as the next guy, their pure OS can boot relatively quickly. Adobe's acrobat reader, VersionCue and color management, Microsoft's own office and search helpers, various IM clients, Apple's iPod agent, ATi's Catalyst Control Center, the ever useful ABit uGuru, the Windows Media connect I have for my 360, the UPS monitor, Spyware blocker, Java install and the Windows Update I chose to set to auto-download-but-not-install culmulatively kill me. The sad truth is, I know a clean OS install takes me maybe a quarter of what I currently endure due to my love of additional features.
The interesting question will be whether Microsoft has forced anything useful in to their new found love of signed drivers that actually aids this. A logical system would involve drivers that had to register how critical they really were, along with criteria to change that criticality. Arguably, VersionCue could sit there saying, "OK, load your critical stuff and get to me when you can unless an Adobe program tries to start." The same goes for the iPod agent - until the USB bus anounces the presence of a connected iPod, it can probably let the user run the rest of the boot, open iTunes, check mail, etc. My UPS monitor could function almost as well if it didn't start until a minute or two after boot - so long as I had the option to bump its priority up for fault finding. Same goes for uGuru. ATi's catalyst control center could likely load a few critical features for the desktop and worry about the 3D stuff only when an app required it or I tried opening the full control center. Most of the boot could be far more sensibly prioritized but, sadly, Microsoft likely only has signed drivers so they can promote crap like Zune better whilst hindering competition and, even if they did have it, every hardware manufacturer would likely ignore it in the name of never having users curse their software for taking an extra two seconds on those rare occasions. Oh well. A dream at least.
Now, with magnetic ram just around the corner, which keeps its charge powered or not, don't we have the answer right around the corner? And wouldn't asynchronous computing render this problem mostly solved?
Microsoft has done us a real disservice by implementing a broken ACPI, but the best response is to push forward the things that make that moot. For now, Ubuntu seems to get this right on my hardware out of the box, and my MacBook Pro handles it with ease. This can happen yet.
Ah. Obviously this is some new meaning of the words 'plain old text' that I wasn't previously aware of.
When our name is on the back of your car, we're behind you all the way!
I support a large number of HP DL380s across a large GAN; they can take a few minutes to boot and for all services to start up. Many of them are frequently shut down due to extended power outages caused by tropical electrical-storms in the area. Other sites turn off their generator every night, so the server has to be shutdown.
:)
I don't much care about the downtime, but clients are faced with an outage of several minutes every time the server so much as reboots. I'm sure everyone here knows how people get when there's a 5 minute delay before they can start repeatedly (and optimistically) clicking the 'Check Mail' button. I'm sure they'd love the faster boot times
I don't relish the idea of populating a 16GB hibernation file - even on arrays of SAS drives. It might be easy to trivialise this question in a workstation context, but it has its relevance.
Address layout randomization tackles the problem of a process that already has privileges getting some foreign code stuck into it via a stack overflow or similar exploit. Normal memory protection doesn't help, because the process already has access to memory. But the exploit code has never been linked with the rest of the system, so it either has to know exactly where everything is, or know almost exactly and use a NOP slide, or else guess and usually crash as a result.
An attacker could submit thousands of shellcode attempts with different assumptions about memory layout until one finally worked, but (1) many people will notice when a daemon restarts several thousand times, and (2) there may be a limit on how often the OS will restart a crashed daemon (Microsoft is about to add such a limit, for example).
It's just a little "groggy"... Ever wake yourself up with pasty eyes or morning breath and wonder how to get out of bed without disturbing the friends? (the beelions of mites we all take to bed with us....)
Give windoze time stop dreaming of electric sheep and throwing invisible chairs in the screensaver.
Speaking of... I wouldn't be surprised if some jokester ms engineers embedded an invisible ballmer throwing invisible chairs in an invisible screensaver...
Previously: "Linux... Toward the Sunrise..." Now: "Linux... Toward the-- No, now, part of Every Sunrise"
I found that on Windows boxes, BootVis (a small freeware) reduces average boot times. Half, pre-install, is normal.
"save a cow, eat a vegetarian"
Also think about power consumption... Many devices (such as hot water systems) are actually more power efficient to leave running due to the extra power required to start up (i.e. heating 60L of water 30 each day when you turn on a hot-water booster as opposed to keeping it warm all day). Obviously PCs are a little different and consume a fair bit of power during even idle usage, so it's still wasteful to leave them running unused for days at a time but switching them off over a lunch break is probably overkill.
... and then there were none
http://linuxbios.org/
There've been discussions here and here.
The idea is to load services in parallel, and replace the ancient Init system. Windows NT supposed does this; still, it's no match for a hand-tuned minimal init system, such as this or this. I tried this back in 2003, and had my desktop open in a maximum of 5 seconds. Bios ~1 sec, Kernel load time ~ 1 sec, Services+GUI ~3 secs. This was on a 1Ghz Celeron, with 128MB RAM. It also helped that I was running Sawfish, with no desktop widgets(no menus, toolbars, anything). Just a bunch of keyboard shortcuts to launch the apps I used. I stopped installing any new distros after that - none of them could offer a better user experience than what I had. It also helps in gaining a better understanding of your system - what are each services used for, how to do file system recovery, and esoteric details.
Well that's a simple answer. It takes somewhere around 20 seconds to spin your harddrive up to speed. After that the read/writes are pretty fast. So if it takes 20 seconds to get the harddrive going, plus say 10 to get windows into memory, you get a 30 second boot. Usually there's a ram check and a few other things (such as dual booting which will usually wait 10-20 seconds before picking a default) which tack on another few seconds here and there, to give you something between 30 seconds and a minute.
As for instant boot computers, they're beginning to make them. Hybrid drives with 16 Gigabytes of flash and 200 of harddrive space (i don't have exact numbers there) are already available for some laptops, and will come into more common use soon. These hold the boot information in the flash memory which is instant access, providing for an optimized boot time of under 10 seconds, maybe even less. Hope this helps.
TVs boot fast? HAH! My old CRT takes about 30 seconds to show anything, and my new HDTV takes about a minute to START fading in. It's pathetic, and it's so frustrating that often times i dont even feel like watching TV because it takes so long to turn on..that and there isn't much good on, anyway.
If history repeats itself, why can't we study the future?
I see people posting saying that hardware detection and initialisation is bound to slow things down, and that is true, but its not the whole story.
Whenever you boot a computer, as opposed to a TV set, there are an awful lot of processes going on. Services start up, various configurations and libraries are loaded up. Lots happens and this lots happening contends for one another for the limited resource of I/O, memory and CPU.
Antivirus scans start happening; if you have AV software which scans DLLs or executables on load this will increase the resource contention significantly.
And at every boot things may be slightly different. For one thing, between last reboot and this a virus could have found its way into the system.
Computers are not exactly finite state machines. Every boot will inevitably differ from the last for oh so many reasons.
Its not like a TV where it only has so many states that it could be in at any time and where things don't change between startups.
If you want instant boots, I suggest sticking with a console and playing games and for math use a calculator.
In the free world the media isn't government run; the government is media run.
I have no idea how it would be harder to read sequentially than any other way.
Flash memory is just plain slow, period. Sequential or random access doesn't make any difference. I think what the other guy was trying to say was that the Flash APPEARS to be disproportionately slow in sequential reads as opposed to random access, because DISKS are so very slow at random access.
The main idea behind ReadyBoost is that the Flash cache can be read from while the disk is spinning up, which reduces the total resume from hibernate time. I can't imagine that that's going to make enough difference to be worthwhile.
It's not that it's not a priority... it's more that it's much more difficult than you're making it sound. Windows (Or any other modern OS, in standard configuration, for that matter) doesn't just boot up the base system and leave you. It boots up several subsystems, and checks for things like "Is the Ethernet connector connected? If so, search for DHCP. What stuff do I have in my USB ports? Do I have drivers for it? For that matter, are my PCI slots filled with the same crap as last time I booted?" The raw fact is that, especially with USB and/or Bluetooth, there are a lot of things that can change from boot to boot. With a TV... there just aren't. A TV doesn't need to worry about having new stuff plugged in, or installed, or whatever. They don't even need to worry about networking. They just need to warm up the screen, than connect it to whichever input it's set on.
Our frustration with the very concept of boot time will continue so long as the user is viewed as a peripheral device.
Ever wonder why your mouse freezes, jerks, or disappears/reappears on Windows and not Mac OS X? Look at the difference in interrupt priorities and figure out where you, the poor user, ranks.
Ever wonder why the three finger salute -- Control - Alt - Delete -- doesn't always result in the CPU acknowledging your presense in its universe?
It's because certain architects, programmers, and testers still truly don't give an ant's fart about humans.
So just get used to it or use your wallet to vote for something completely different.
+++
My Commodore 64 starts instantly! Ahh, the benefits of ROM...
I'm busy unzipping my pants.
I spend most of my time in bed, darling.
..because that's how long it takes to send your personal information and recent surfing history to the NSA.
Only if everyone in the world sits around and waits for it to happen every single time, and does absolutely nothing else with that down time. It doesn't count if you spend that time even THINKING about another issue/problem. You have to sit there motionless, stare at the screen, and do absolutely nothing but age.
Personally, I can find plenty of things to do with my time when I know I can walk away.
The more significant issue, IMHO, is the responsiveness of programs. Forget boot-up times, when you don't even have to be there. How about the delay between clicking the Firefox icon, and waiting for it to start-up so you can do useful work? How about the delay between clicking on a link, and having that link load and render? How about the ammount of time the system is unresponsive as it does something (like render a webpage) in the background?
That, IMHO, is many times more important, and something I certainly have to deal with far more often than reboots. Personally, I have a 2GHz system, with 1GB of RAM, and I still strictly stick with GTK-1 programs, because it's so much faster and more responsive than GTK-2 (or QT) equivalents (as well as not uselessly wasting screen realestate). Ever program I use has a fully functional GTK-1 equivalent, so I'm not missing out on anything by sticking with it, it's just an occasional hassle to change the default configure option, or using a different program because the new version of whatever dropped GTK-1 support (like switching from GAIM to Ayttm). It's a rare issue, and well worth the improved performance anyhow.
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
I think it would be a fairly easy hardware-software job to make non-volitile ram. I used to buy NV-ram chips with built in batteries, would allow instant computer boot ups.
Almost nobody cares for the boot time, IMHO:
andreas@andi-lap:/mnt/disc60> uptime
08:09:00 up 9 days, 22:29, 4 users, load average: 4.77, 4.01, 2.54
And that's from my personal laptop.
Now some people (with say huge server farms) do care,
but they usually do their own customizing down to the level of the BIOS.
yacc
Jeebus! We had instant on computers 20 plus years ago!
****COMMODORE 64 BASIC V2****
64K RAM SYSTEM 38911 BASIC BYTES FREE
READY.
Frob the power switch on the side, bring up the ML monitor in FastLoad or Warp Speed or SuperSnapShot and get to coding.
Guaranteed! This comment 100% Anthrax free!
I am surprised nobody mentioned Robson Technology, something proposed by Intel to speed up things including boot times... The technology will be launched sometime in 2007... i suspect middle of 2007... here's the link to some details http://www.extremetech.com/article2/0,1697,1936630 ,00.asp ... you can also google for it..
There's always Linux BIOS. It doesn't go through all the long processes the 2 commercial BIOSes do, so it boots faster. Unfortunately, it's only available for a limited number of mobos, but the list is growing quickly.
look! it's a bird, it's a plane, it's....a girl? yes, a girl browsing Slashdot on Linux
There are no 2.5 inch 10K rpm hard disk drives. You must be confused..
Computer Science Research
Censorship is obscene. Patriotism is bigotry. Faith is a vice. Slashdot 2.0 sucks.
...it depends on how you have built up your embedded system - if it uses slowly accessible flash (chances are high - because cost is low) 2 seconds is pretty much unreachable. The best you can get is probably 10-15 seconds, but without special (hardware/software) changes your kernel will definitely not boot in 2 seconds.
Slashdot: stuff for news, nerds that matter, matter for news, stuff that nerd
And yet you give no mention of the five years of life lost per person when using a Mac?
Ta da! Read it. "Plug and Play" now if it didn't re-examine the machine it couldn't be "plug and play" now could it? If this is what you want, disable "plug and play".
Well, let's see. It could quickly check that the old PnP stuff is still there, then scan for new stuff in the background. For hotpluggable devices, you wouldn't even need to change any code to allow this.
I hereby place the above post in the public domain.
Some engineers are constrained by limitations inflicted upon them. Others simply cannot see over the edges of the ever-lovin' proverbial box of small-thinking.
For starters. . .
1. Stop building computers around crappy O.S.'s (An outside constraint if there ever was one!)
2. Put the O.S. on a nice fast EPROM. All hardware drivers can similarly be put on secondary Flash-style memory chips and they would only change when the hardware is updated. Whatever needs to be loaded into faster memory can do so after start-up.
3. Don't power down by default! Put a big Lith-ion battery inside every box and keep the memory on-line, or partially on-line. When was the last time anybody powered off for more than a month? If you need to re-boot from scratch, make it a secondary option.
There. That's three ideas which would provide a lot of elbow room to solve problems, and I'm not even an engineer. Heck, give me another ten minutes and I'll think up three more useful ideas. So where's my billion dollars?
But seriously, I'm assuming that I'm not the only one who can think of useful ideas which could make life easier. No. There are lots of smart people out there. So instead I'm assuming that there is something preventing smart people from implementing useful changes; I'm going to blame the system, which I find far more satisfying than thinking that the world is populated by morons.
-FL
What we've currently got is the best we can do.
That's got to be one of the scariest comments I've seen on Slashdot... if it were true. Fortunately there are plenty of things we can do to reduce boot times. We can eliminate the RAM self test. We can put the OS on a cartridge instead of on a hard drive. We can get rid of the spinning platters and go with completely solid state permanent storage. We can re-arrange things (Why have the hard drives spin up first, then have the user log in. That makes the user wait for the machine. Have the user log in, and while doing that you spin up the hard drives. Chances are high that the drives will be ready before the user finishes the login.).
Of course, to do most of these things, you would have to abandon the current PC platform. For example, it takes 20 seconds from poweron to playing on my room-mates XBox, whereas my PC takes two minutes from poweron to Ubuntu login screen.
When our name is on the back of your car, we're behind you all the way!
I bought a 10k rpm 36gb Western Digital Raptor drive for £80ish (very steep, if you ask me), but when i installed a new windows partition, the bootup speed was incredible. About 20-25 seconds from the moment i hit the power button(provided i spend less than a second deciding which OS to use in GRUB). it's gone up a bit since then, what with all the applications and shit, so i might have to start again soon. It's just a hassle to deal with windows erasing the MBR.
I have Ubuntu dapper on the same disk and it takes about 3x as long, even though it's at the start of the disk!
sudo killall humans
Vista does this in its default sleep mode. Sleeps to RAM and to the hibernate file, so it's on fast if it hasn't lost power, and on somewhat slower if it has.
Gentoo Linux will compile to suit your hardware, but it doesn't do it automatically. What I'd love to see is an O/S that, on install, detects your bios, motherboard and peripherals. It decides the level of permanence of same and compiles twice. Once, in an 'optimal' version and the second time as a generic. Boot-up would default to the optimal one. Hickups would toggle the generic.
John Q. Public doesn't exactly go swapping processors on their Compaq Presarios every week. If you DO decide to 'upgrade' and swap out a processor or two, then the BIOS should detect it and the default kernel would boot and get you started towards compiling a new kernel.
Well, that's my opinion, and I'm sticking to it.
NOTE: "All brontosauruses are thin at one end, much thicker in the middle and then thin again at the far end." - A. Elk
*** Don't be dull.***
I've often wondered the same thing.
Most of the time used for booting up is repetitive. Scanning for changes, loading the appropriate drivers and rebuilding the runtime information needed to make windows load.
If this were all compiled once and written as an image, then windows could boot within seconds of turning on the computer.
It would require a little more interaction--to make bootup as fast as possible you would need to tell it to scan for hardware changes. Also, the install of a new "Startup" program like winzip, Quicktime, the IM clients, all the crap in your lower-right corner, along with every driver change would require a significant "Rebuild".
Every change would require a rebuild of some sort--you'd have to reboot, re-create a new image, store it off to disk in a contiguous block and then probably boot off it to test the new configuration. No more on-the-fly changes.
It would also mean a little more work for manufactures of programs that want to startup with windows because the image will be loaded then control passed. It's not overly difficult, but it is different than just running an "Exe", and every different path is another entry point for defects.
It'd totally be worth it though.
Love the Mac fanboyism. The story is actually about Bill Gates and his point of view on OS bootup speed. That happened during the development of XP while they were working on the part of the OS that dynamically wrote system files on the outer parts of the HDD so it could spin/load faster into memory during boot.
Use VMWare 'resume' for a quick boot, work only inside the VM, and never turn off your computer..
The problem is old-school linear thinking we've inherited.
There is no technical reason that a computer could not wake up, verify the keyboard, memory, hd, mouse and display are the same (in a few microseconds, probably) and be up and responding very well to the user, while (new concept, brace yourselves) the computer carefully brings up other hardware subsystems and makes them available as they become functional. You could be in a word processor, graphics editor, all manner of things that don't require more hardware until you do something like print or attempt to access the network; if those subsystems are not ready when you try to use them, the design would allow for [establishing hardware, wait or cancel] and there you have it.
There is no problem whatsoever with plug and play concepts coexisting with fast usability other than current design shortcomings end users have been forced to live with. The computer is running as soon as the HD is spinning, memory sized, and the video card is on and the KB and mouse work. Just because current operating systems don't let you begin working at that time isn't a reflection on plug and play as a concept, it's a reflection of linear thinking that descends from old single tasking systems like early DOS.
The idea that a 2...3 GHz 32 or 64 bit CPU cannot bring itself to decent usability in under a second is one that is silly right on the face of it except in that common systems are using old school thinking and layering more and more crap on top of that thinking. There is not a thing in the world that says drivers can't be loaded on demand, or after usability from boot, or separately. Nothing.
I've fallen off your lawn, and I can't get up.
The long boot time is caused by poor OS design: when I used BeOS, the computer took 14s (duration from boot loader prompt to GUI) to boot to a functional/reactive desktop (no cheat like Microsoft do in WindowsXP).
...
And this was on a Celeron333 with 128MB of RAM!
So Windows and Linux (which is even worse in this respect) long boot times are caused by poor software design.
What's amusing is that when I point this to Linux's user, they are in denial:
-"boot time doesn't matter, just leave your computer on", sorry but I like to sleep at night and I have a small flat, my computer is quiet but it still makes too much noise..
-"BeOS is dead", that's mostly true, but the point remain that it is possible to boot fast as it have shown.
-"BeOS was a single user OS, that's why it could boot fast", if Linux's multi-user management has such huge impact, then it is broken..
-"BeOS worked on a single configuration, that's why it could boot fast", not true, sure the supported hardware list wasn't very big, but this was more a development resources problem than anything else.
I think the problem boils down to making the computer ready to perform (at least in theory) a trillion different tasks, flawlessly (again, in theory), and maybe even concurrently (theory once again)......and we still expect it all to be done within a few seconds of startup time. Yes, your TV comes on immediately, but compare the work done by your TV and your PC.
Why not give some mind-blowing tradeoffs...NO NEED to have a fullblown PC when I only want to do a, b, and c. My PC should emulate my pocket calculator in terms of features, and I should expect just those features and no more from it, but at such speeds too.
Then again, ramping up your hardware should solve all these problems....(in theory).
I see plenty people point at Linux BIOS. Although it helps to have a BIOS that boots so fast it had to be delayed so the hard disk got a chance to spin up, this is NOT where the main delays lie.
From loooooong ago I remember a friend of mine speeding up the DOS that came with Apple II floppy drives and call it TurboDOS - the main improvement was to remove near insane wait states to something that was practical and reflected the improved FDD quality.
This could offer a clue here too. Does a service really need to check world+dog when it boots up on the same hardware? If the hardware on my laptop would change overnight I'd consider that a neat trick..
Not sure I'd go all the way to 'recompiling' a startup, but yes, some checksumming could be sensible.
Oh, and strip all the bootup crap you don't need. In Windows, you'll end up with a iTunes watch thing, Acrobat Reader sticks 'update watchers' in etc etc - and they're the ones that ask. Ubuntu, ditto - there's plenty you can switch off.
Insert
My personal use box, will take almost 4 minutes from the time the screen switches to graphics mode before it will give me the login prompt (XP) .. I have no idea wtf is causing it, and doing a restore to whatever was before it, didn't help one bit. :(
"Champagne for my real friends - and real pain for my sham friends!" http://ericblade.postalboard.com/
As a couple of posters have mentioned, setting up the hardware is a big part of what slows systems down during boot.
One way to increase speed this up is to perform a lot of tasks in parallel. For instance, DHCP tends to take a lot of time on some systems, particularly if the network is lousy. Most linux distros I've seen tend to block on that operation, waiting for it to complete before starting the next step of the boot script. Load time could be improved in that and other cases if this operation were performed asynchronously.
The main problem with this and most other techniques to speed up boot times, like loading a saved boot image directly into ram, is that this increases the complexity of the boot process, which increases the likelihood of failure. For instance, if services/deamons are starting in parallel, then services that depend on one another must be set up to block until the services that they depend on start up. This becomes a problem, because 99% of the time a service which depends on another could just start later naturally and testers might not even realize that this service depends on the earlier one, but on a multi proc machine, or just on a bad day, the service that depends on the earlier service might boot later.
Basically, doing things in parallel wrecks the determinism that computers normally exhibit, which as any computer scientist, philosopher, or physicist will tell you, is something that injects a lot of uncertainty into our reasoning.
At the very least if boot times are sped up with techniques like this, then computers need a "slow and safe" fall back boot script.
Now, you should go to and read carefully.
At this point you may want to step out of your computer a bit, and think about what you have just said.
Then, I strongly advise you that, as you are obviously wrong, you should start to understand that not everything you think to know is true, and that a lot of what you think seems to be base on truthiness instead of facts.
Grow up. You are using cognitive dissonance to justifiate the fact that you are running a Windows OS.
Another problem with reboot/wakeup is that many services have to establish communication with the outside world. XP has what I consider to be a bug, every new internet link requires a new session of svchost.exe, the bug is that this totally consumes the CPU, the computer is nearly frozen for 10-30 seconds. This is part of the login process even though the login is actually finished and everything is running.
Dvorak wrote a few months ago about the problems MS has writing multitasking code. svchost.exe is yet another example.
The only OSes that are booting fast I know of are
Atari ST (almost instant, because OS was in ROM)
Amiga (same)
BeOS (Despite OS was on disk, it only needed 4 seconds on my PC to boot up!!)
Any BeOS guy out there to tell us how they were doing that?
>The other issue I have is that I normally use a VPN to connect to work, and the VPN tunnel doesn't like getting shut down and restarted, especially with a different IP address, so I still have to re-authenticate by typing in my security token code to the VPN client.
Isn't that what's supposed to happen? You've left your computer for a while, especially a portable one, it better disconnect any secure resources it has. It's comfort over security as usual, but I think this is by design.
You can try one, or more of the flavors of TinyXP. My machine boots up much faster than it would with the version of windows that came on my machine.
rhY
I hold very few opinions. I hold information based on observation and fact. If you wish to disagree, please use facts.
Yes, early computers could switch on and off quickly. Handheld computers are pretty fast. The serious computers that we use today take time though, and the reason is simple:
Objects at rest tend to stay at rest.
When you have 3 gigglebytes of RAM, 300 gigglebikes of storage, and a computer capable of nearly 3 gigglehurts; well you can't expect everything to happen right away.
Even the bullet in a gun doesn't instantly reach full speed, and when you are talking about the speed of light (approximate speed of electrons in your computer circuits) you have to make allowances.
Now if it were only one electron that you were accellerating through your computer you could reasonably expect a quick response. But when you add the mass of billions of electrons, patience is helpful. Remember that those billions are not all going in the same direction--some are going to the display, some to the ALU, some to the bit bucket. Yes, it takes time.
Next, suppose that the I/O holding platform requires 200 million bits to begin its work. You dump those bits there as quickly as possible, of course. But wait! Those bits are raining down at the speed of light; do you think they can just slam in there and stop? They need to settle in and get comfortable. Another delay.
As computers get faster and faster, the CPU chips have to be reprogrammed to allow for the mass of data and increased speeds. In assembly language there is a command to RSSC (Reduce Speed, Step Carefully) that sometimes helps with the above I/O holding platform problem. Five years ago there would have been no need for such a command.
So, without getting too technical, there are a number of reasons that big, fast computers that we live with today cannot start, or even stop as fast as we would like.
...omphaloskepsis often...
So you have to wait 2-3 minutes, is this a bad thing? Use the time to relax with a cup of green tea. Are we so soon to forget the old days, waiting 20+ minutes for our ZX Spectrum to load hunter killer/R-Type/Chuckey egg, just enjoy the free time.
In the not too distant future, next Sunday A.D.
i usually turn off the front mounted (due to modding) mains power switch with my toe. Of course it's win 98SE (also patched and modded) and scandisk is disabled via tweakUI, but it never complains when i reboot. and still boots up plenty fast*
*though some might say that sticking a stripped down install of Win98 on a 2.8GHz P4 is a waste of clock cycles, considering it runs ok on something one tenth the speed
If you don't risk failure you don't risk success.
This is just what I thought it would turn into... a degenerated discussion about macs hibernating vs. windows reboot. For those of you with slow booting systems, it's because your computer is so overloaded with crap that you put on it that it can't function easily. How can you tell? Go out and buy a new system and see how fast it boots. 20 seconds for XP systems... a Windows 2003 server may take a bit longer - but that's a server. Same goes for Linux systems - a default installation of RedHat or Fedora has so many checks and balances. In my previous life, I worked on embedded linux systems and a cold boot from a flash drive into X took less than 5 seconds. Unfortunately, starting an embedded Mozilla after that, also took another 5-10 seconds.
It's do-able. But most companies don't chose to do it. Try restarting Tivo from cold. Or PS3. it takes a while.
swapping out the hard disk. Wowzah!
"The dew has clearly fallen with a particularly sickening thud this morning"
The Plan 9 livecd boots in ~5 seconds in VMware...
I mod down anyone who says "I will be modded down for this", regardless of the rest of their comment
In 1995, I modified BSD/OS (BSDI 3.0) to boot without the network. When it was able to detect the root name servers with a =ping, it would start up the networking code (not a solution i would recommend today - horribly hard on RNS's.) People at BSDI were amazed that this could be done, and wanted to learn more. If i remember correctly, without all that networking crap in there to put interlocks on the boot process, boot-up was something quick, like 10 secs or so ...
It should not take that long for your desktop to work. Download the Startup Control Panel applet and disable everything that's attempting to boot. This tool is really nice as it has a tab for every way for a program to autostart itself.
I use then when writing auto-install scripts. For each app that tries to autostart (which is absolutley unnacceptable for any application to do) I find out how that particular one does it and disable it after the install/upgrade.
"It ain't a war against drugs.it's a war against personal freedom" --Bill Hicks
given that the nature of the boot sequence hasn't changed in the last 20 years of PC usage ?
also, I distinctly remember this being discussed in 1998 so its a dupe
So they can actually test the software on the hardware it is being used on. Frankly, it would be a disgrace if stuff like this didn't just work on an apple.
In the PC world the software gets to run on all sorts of exciting combinations of hardware, a very different world.
First of all the disk startup sector playlist OS X uses (creates a list of all the sectors you tend to use in booting and just reads them from disk at startup in the most efficient order) ends up speeding up boot up pretty impressively.
But now suppose you want to do something more than this and really just read things in from a memory image. At some time I wondered why they didn't do this and then it became clear that device initialization and dependency tracking make this impossible. The problem is how the OS boots is dependent on the results of querying devices and it must set these devices up as well. For instance you want your network to be configured as the result of values from the DHCP server not just whatever values you had last time. Also you don't want to use the image of processes that think they are still talking to some no longer existent name server or other expired data.
This having been said I think there is a lot of improvement that can be had. OS X starts up pretty quick but it could be faster and the programs still can take awhile to start. Explicit dependency graphs combined with the priority that some startup process should execute with could make things even better by allowing you to get started right away while utility processes are still being starting and executing in the background. User programs that are set to being at startup should be able to be loaded and launched with some equivalent of nice even if they are to execute at some other priority once working.
If one wants to get really fancy you could probably work out a system that accelerates launching applications in general. It should be possible to use code/data dependency analysis to create modified executables whose (detectably) non-input/state dependent initialization work is already finished. Even more ambitiously one could try to create a 'playlist' of system calls that the application makes when initializing and try to optimize the execution of these calls. To get really crazy one might try to combine this with some type of profiling and data/code dependency analysis to check if some later commonly used image can be used (e.g., cache the results of telling the application that yes the resolution is greater than 320x200).
I think the prospect of an automatic version of this is to go to an all JITed world where the OS is integrated with the compiler. Using partial analysis and clever caching techniques one could probably crazily speed up the startup times of most programs. Besides it would allow architecture independence.
If you liked this thought maybe you would find my blog nice too:
Why haven't companies made it a priority to have 'instant on' desktops and laptops?
They have. It's called the Mac.
Seriously, I had heard it before, but it was still astonishing to watch when I got my MacBook Pro this year. I was used to long boot times in both windos and Linux. I was used to long times going from hibernate and even sleep to activity.
OSX boots in less time than XP on the same hardware takes to awaken from hibernate. When it's sent to sleep, OSX is back before I've opened the lid completely. It's not quite instantaneous, but it's as close as I need.
OSX still needs to do a bunch of things at boot, and after login, and there is certainly the possibility for caching (as you suggested) or other speedups. I'm sure they will happen, because as I see it, Apple is currently the only company that actually cares about this stuff.
Assorted stuff I do sometimes: Lemuria.org
We laughed at you PC guys cause you had to "boot" your machine prior to doing anything useful with it. (Later we bought Amigas or Atari STs and were in the same camp.)
Maybe it's time to go back. Most people need a browser (file, graphics, web), an e-mail client, a DVD/CD/MP3 player. So just slam that -- together with some stripped down multitasking OS (heck, you don't even need multiuser on the average Joe's machine, they would run anything as root or in administrator mode, just because...) -- into a nice FlashROM, add a harddrive and a DVD player and that's about it.
Wait a minute.
Those things were around about 15 years ago. But neither the CDTV nor the CD32 were a success. So much for the price of being ahead of time.
"Yessir! But, as an engineer, it is difficult for me to translate the non-effect of your psychological leadership hyperbole intending to boost morale. So, let me put it to you in droidspeak. In order to monetize your guilt trip, we need a little seed capital to grease the wheels.
I think incentives should be based on real world figures. Now, as a propeller-head, I could spit out some geekery about how my currently salary of $40K per annum minus two weeks of unpaid vacation divided into the 6 day work weeks of 10 hours per day, like you encourage all us fun-gineers to 'take advantage of' in this 'exciting tech environment' works out to meaning one second is worth roughly 22 cents.
So, to calculate the bonus you need to offer me in order for me to continue ignoring everything for the sake of squeezing another drop from a stone, how many seconds did you want me to save customers? 113 years? Lessee, now... that's what? 3,563,568,000 times $0.22... hey, can you hand me that calculator over there?"
As most slashdotters know, a very large amount of boot time is the result of your systems hard disk speed. Replacing the hard disk that the OS boots from with something faster speeds up boot time dramatically. There are some products out there already that emulate ATA and SATA drives, but using DRAM (with battery) or Flash to store the actual data instead. PQI have some flash based ones here: http://www.pqimemory.com/products-storage.asp [pqimemory.com] Gigabyte have a product called i-ram which is the DRAM based solution here: http://www.anandtech.com/storage/showdoc.aspx?i=24 80 [anandtech.com]
The down side to the RAM solution is it obviously needs a backup battery installed on the card in the case your system is unplugged from the wall socket. If OSs were written to be "RAM-boot-device-aware" this problem could be reduced by storing copies of the necessary boot files and restoring them from the actual hard disk should they go missing.
As for why computers are still slow to boot today, my best guess is costs. Right now, adding one of these devices to your pc costs more than some people are willing to pay just for a faster boot up time. As memory continues to get cheaper id be willing to bet we will gradually see more pcs with some form of dedicated boot drive. (This is assuming the hard disk isn't replaced by the magical holographic storage we keep hearing about any time soon). The second half of the battle for faster boot speeds is the software support to negate the down sides. RAM based boot drives would need to be backed up to hard disks and flash based drives may need the OS to be aware of the limitation on writes.
EDIT: Woops, forgot to select the right formatting
4 80
As most slashdotters know, a very large amount of boot time is the result of your systems hard disk speed. Replacing the hard disk that the OS boots from with something faster speeds up boot time dramatically.
There are some products out there already that emulate ATA and SATA drives, but using DRAM (with battery) or Flash to store the actual data instead.
PQI have some flash based ones here: http://www.pqimemory.com/products-storage.asp
Gigabyte have a product called i-ram which is the DRAM based solution here: http://www.anandtech.com/storage/showdoc.aspx?i=2
The down side to the RAM solution is it obviously needs a backup battery installed on the card in the case your system is unplugged from the wall socket. If OSs were written to be "RAM-boot-device-aware" this problem could be reduced by storing copies of the nessecary boot files and restoring them from the actual harddisk should they go missing.
As for why computers are still slow to boot today, my best guess is costs. Right now, adding one of these devices to your pc costs more than some people are willing to pay just for a faster bootup time. As memory continues to get cheaper id be willing to bet we will gradually see more pcs with some form of dedicated boot drive. (This is assuming the harddisk isn't replaced by the magical holographic storage we keep hearing about any time soon). The second half of the battle for faster boot speeds is the software support to negate the down sides. RAM based boot drives would need to be backed up to harddisks and flash based drives may need the OS to be aware of the limitation on writes.
No, I think it needs to be amended to: "the software designers are arrogant and haven't bothered communicating with their public".
Anybody who makes global statements like that, well you got to check out where they are coming from. On slashdot this line regularly appears, "I am an expert with N number of years experience, (X) is obvious, (X) is easy, anybody who doesn't think so is stupid, come back when you have 10, 20 years experience before daring to complain".
I've also had problems with Firefox blocking timed shutdowns on my Windows box, whether initiated through Nero or the console shutdown command. Although I don't remember the exact text of the error message, it's apparently a widely known issue.
To be fair to Mozilla, I believe the problem may be caused by extensions to Firefox.
F_T
Or you could just install the free StartUp Manager from Gaffers: http://www.soft-ware.net/system/steuerung/systemst art/p00388.asp or from other places easily found by searching. Works fine with XP and other Win versions. Even shows you what programs have been added since last time you ran it. Hasn't been updated in years, but no reason to....
Rather than making excuses about why faster booting is not possible, you might want to think about ways of making it possible. Of course, in the rare cases when there has been a hardware change, it will take longer, but then with plug-and-play it is possible to work that out in the middle of normal operation. So why not boot on the assumption that everything is the same but run a few quick checks such as that the amount of RAM is the same, and then do the plug-and-play thing or in extreme cases do a full reboot. But it should be possible to write the operating system so that in normal cases it takes only a few seconds to reach full operation (plus the time to enter userid/password).
Regards, Martin IT: http://methodsupport.com Personal: http://thereisnoend.org
Ok, here's maybe a naive question. Why can't the entire system state be saved to a large enough flash memory(they're getting larger and larger these days), and the system would read it's info from there. As far as new hardware, why is everyone saying that "a system needs to reboot to recognize/install new hardware?" What about USB/other hotplug technology?
-T
a commodore 64 =) instant on!
Your computer should boot in less than 10 seconds with DOS.
So say we all
Love the closing sentence...
"How to Do Nothing," kids activities, back in print!
In 1984, my C64 didn't need to boot. You could switch it on, and instantly the operating
system and an interactive programming language interpreter (CBM BASIC) kicked in.
Since then, things have been degrading, and technically, there is absolutely no need to
have "booting" machines. While there are various forms of hibernation (suspend to disk,
suspend to RAM etc.), hibernation is not designed to be the normal mode of operation,
and eventually, today's personal computers will need a reboot to start afresh.
Users have been indoctrinated to accept the boot, and Microsoft users even its nasty
cousin, the re-boot, over the decades, but it is indeed time to get rid of it.
Another question that factors in is whether machines were designed to always run
(like mainframes or UNIX workstation) or used as consumer devices; in my view,
today we should adopt an environmentalist attitude and design machines that require
minimal energy - whether they be on or off.
here is is "why are you having to reboot your computer all the time?" I regularly go weeks without a restart on my main machine, and the workhorse and server computers don't see a reboot more than every few months. It's almost always a security update that forces me to restart. It takes maybe 2 minutes to restart to the login window and another minute so to finish its background loads after login, but even if it took 20 minutes to boot it's not like it would affect me that much. If boot time is bothering you then you are either very impatient or you have a stability problem?
I work for the Department of Redundancy Department.
I remember a moron I once lived with telling me my Opteron added $50 to the electric bill. (This had nothing to do with the two massive, ancient air conditioners she installed right after my moving in, of course.)
Computers, even modern whiz-bang look I can play Oblivion at decent resolutions, everything maxxed and without any frame loss computers, take up very little power. More, of course, if you're an idiot and think a bunch of flying windows is better than turning your monitor off at night. (My UPS can keep my box powered for about twenty minutes. That drops to about five with the monitor powered by it as well, heh.)
But let's throw logic to the wind, and say $50 a month, even. That's $600 a year. That's going to pay for a new computer? Maybe if you're buying bargain basement PCs from Dell. When people replace hardware, they buy new, and they buy it for a reason. Why replace an old, say, processor, with another old processor, when for a bit more, you can once more be 'current' and extend the life of your system even longer?
At any rate, the only catastrophic (IE, unrecoverable without expensive services that usually aren't worth it, and certainly cost more than the few cents a day it costs to keep a computer running 24/7) hard drive failures I've seen have been on cold boot. When a system is running and you start getting read/write errors, you know it's time to migrate the data. When the system comes up from cold boot and the drive is no longer there, you're fucked. And quite frankly, hard drives are the only thing I've ever had fail on boot. I can't remember the last time I had a non-peripheral (I've gone through plenty of $10 el-cheapo mice, of course) piece of hardware other than a hard drive fail inexplicably and suddenly. I've had video cards fry due to fan failure. That's it.
Now, I suppose turning off your computer might be worth it to you.
Can you put a price on memories (digital pictures, saved e-mail, et cetera)? As more and more of our lives are stored on our computers, the cost associated with hardware failure increases. For some of us, the paltry amount of money it takes to keep a box powered 24/7 is worth the security of not being totally boned when a drive dies.
Get this: My Tektronix TDS2014 oscilloscope (that's right, an oscilloscope) takes FORTY-FIVE seconds to "boot"! A stupid scope, taking so long to get it's act together. What kind of crap is that? How is it that anyone at TEK thought that 45 seconds was not going to piss me off every time I turn the thing on? Next time I'm in the market, startup time will be one thing I will definately look at. When I got this one it never occured to me to even ask. Not to mention my Honda's navigation system.... first I have to acknowledge their disclaimer screen after several seconds of watching their logo (EVERY f---ing time!), then wait & wait until it's ready for me to do anything. Guess I've been using my Mac's sleep feature way too much and forgot what the rest of the world has to go through.
If I didn't have absolutely NOTHING to do, I wouldn't be here.
Give windoze time stop dreaming of electric sheep and throwing invisible chairs in the screensaver.
Speaking of... I wouldn't be surprised if some jokester ms engineers embedded an invisible ballmer throwing invisible chairs in an invisible screensaver...
Philip K Dick managed to write works of complete genius whilst taking loads of very strong drugs, it's true. But you, my friend, are no Philip K Dick.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
use the nifty little tool provided with XP, msconfig...
try Start, then Execute, then type "msconfig" and enter.
Then you remove all the services you don't need at startup.
My parents newly delivered, Dell installed XP went from 90s boot time to a bit less than 50s, login included.
And I couldn't believe how much crap Dell installs on a virgin system.
Best of luck
D.
It takes 40+ muscles to frown, but only four to extend your arm and bitchslap the motherfucker
Then you can install Linux or Windows without trashing the other OS - there are a whole pile of HOWTOs available on the web to help you do this.
we see things not as as they are, but as we are.
-- anais nin
I wondered that too - I saw an article on it on IBM and CNN years ago... and it never got posted on /. or did it? :)
A TV only show a static imagen, then replace this image with the next, etc. Is a simple task, easy.
A computer load zillions of "images" you never see in-memory. Then execute most of these images, in a way that mean the computer need to understand that images (=executing his code). And once everything has been loaded, you have something much more flexible than a TV. As you see, a computer do more stuff while booting, than a TV.
You can speedup a computer boot? maybe, but that may mean loading less "images" or only one "image". Often on a computer a "memory image" can be 512 MB, while a TV "image" is only 2MB. And the PC hardware is not optimized to "load up 512MB memory images" while the TV is optimized to hardcore for years and years to show that "2 MB" image.
Apples and oranges.
-Woof woof woof!
Er, looks like JeffK has improved his English skills, but why is he posting on Slashdot?
Go take a look at http://en.wikipedia.org/wiki/LinuxBIOS. It seems that a huge amount of boot time is wasted in hardware scanning and detection, most of it with proprietary drivers, and much of it designed to work around hardware flaws and API violations which hardware vendors refuse to discuss or publish fixes for: they only provide their own custom "drivers" that the OS is forced to load from customized boot media. The result is a massive layer of often inconsistent and incompatible driver loading and scanning, and a huge waste of time and BIOS resources to find and load these expensive and unreliable workarounds.
Computers taking multiple seconds to boot up is like how a car needs a couple seconds to warm up when you start it up. All those moving bits and jims and jams. Have you ever used a PDA? those things boot and reboot in about half a second. That's cause there ain't none moving parts.
Back in my day we'd get a caining if we asked how computers worked. Nowadays every whippersnapper wants to know how the whole bloody world works. Just saying 'it's magic' ain't good enough any more.
I don't own a snook, and if I did I wouldn't leave it cocked.
Obviously you are new here...
I just tested it, and my G4 Powerbook takes less than 2 seconds to wake from hibernation, another split second to enter my password, and my apps are instantly accessible. I suppose if a user wants an "instant on" hibernation feature, he can buy a Mac.
Why power your machine off when you are done with it then? That would save you life-lovers 60 seconds a day instead of two. Oh the humanity!
It would be a better thing to push someone into finding out more about their operating system and learning about all the useless services that come installed by default on windows that take up a lot of memory and use up a lot of time & resources, instead of pushing them unto a different OS altogether that might take them twice as long to understand and also might not do what they need it to do! A better thing is always to push to learn the stuff we don't know instead of pushing what we think would be easier as we see it. Linux/Unix is not easier, actually it is a lot harder, even though more gratifying as you end up learning so much more about OS's then you do on windows. However, the key factor in this scenario is why would it take so long to boot, when I have a windows machine that boots within 15 seconds and gets out of hibernation in like 3??? Maybe one word can describe the problem.... PICNIC Problem In Chair Not In Computer!!!
Microsoft has a free utility called bootvis that visually shows you were your computer bootup is spending it's time. You can download this utility at: http://www.softpedia.com/get/Tweak/System-Tweak/Bo otVis.shtml
This utility also has some whitepapers with advice on what you can do to speed up your boot times.
I consider quick less than a minuit. 60 seconds. I am still primarily using an old computer, a dual p-pro in fact, not quite upgraded to the max. not quite worth upgrading to the max either. For a 9 year old machine it is great! Boot time into 2K is close to 30 min. Start it up and get a drink, talk on phone, tie some flies...
Boot time into Linux is not any better. The plus side of linux is uptime typically exceeds 30 days. No boot up time then.
My up-time is limited by the flakey power in my neighborhood. I loose power on average once a month, sometimes more often.
My home PC is used for surfing, some graphics, playing games, word processing. I don't need power, I need reliability. I fear the long tail of product life, and that is all.
I will not upgrade this machine on credit, it is not necessary for work. I will use it till it dies or till I can pay cash for an upgrade.
Phil
Laugh, it's good for you!
TO have the computer running when it is doing nothing is a monumental wastage. Your bonus is going down the drain there, think about it.
With hibernation available it is nonsensical do keep a machine running all the time.
IANAL but write like a drunk one.
In the year 3000, everything will be instant, but the DMV (and computer boot times) will still take like 9 fucking seconds.
OS X + Powerbook. Good bye
The crew of the Nostromo was put in hibernate while the computer was being booted on. A couple of million light years later, the computer still needed something similar to Wake-on-LAN before it started to do SOMETHING.
WARNING: Smartphones have side effects--most of them undocumented.
Yes, but your Toshiba doesn't boot up in 14 seconds. My MacBook does. What's all the fuss about wake-up/sleep mode times, when it takes a MacBook a mere 14 seconds to boot from cold?
Coincedentally, my MacBook boots up in to Win XP Professional in about 50 seconds. Maybe it is the EFI or the Bootcamp work around, but I think it is most likely XP. I haven't tried the sleep mode in XP, but I'm just going to guess it isn't stellar.
These boot times are measured by when the desktop appears. With OS X, the programs are usable in about 5 seconds, but I have to wait an additional 30 seconds or so before all the background items load into XP (with NO third party apps, just XP...it's all I have had time to install thus far, so I'm assuming it only gets slower from here as I add programs.)
They very likely have physical components that they must test to ensure that they are operating properly. Computers have no large physical devices to calibrate/test.
My computer takes longer to dehybernate than to boot.
Copyright infringement is "piracy" in the same way DRM is "consumer rape"
Computers take so long to boot because they are, well, booting.
The term 'boot' comes from 'bootstrap' as in 'to pull yourself up by your own bootstraps'.
The idea is that this is exactly what the computer is doing. By design, the computer maintains no state between one boot and the next beyond internal configuration. At the beginning of the boot process, most of the system's resources are unavailable to the system. These have to be probed and initialized. All hardware needs to be reset, ram has to be counted, etc. ad nauseum. This could probably be faster, but since most of this resided in upgradeable firmware code (e.g. BIOS, OFI, OBP), this has to take place in a programmatic fashion - it can't take place in mechanical hardware like your TV, because (unlike your TV and other 'instant-on' hardware) the hardware and software environment cannot be assumed to be identical between power-off and power-on. All of this probing has to take place in order to detect any changes and ensure that the system is in a consistent state before continuing. Depending on hardware (such as SCSI cards, RAID controllers) they system also has to load additional BIOS code before boot can continue, which has to complete its own initialization and sanity checking process.
If you had a computer made entirely of dedicated components that was impossible to upgrade, with the initial booted OS image in on-board ROM, you probably could get a computer that would boot that fast. Is this a trade-off most people would be willing to make? I doubt it. If you did not start from scratch, or guarantee an unchangeable hardware and system software environment, then you could get your hardware into an unrecoverable state across boots.
Could boot be made faster without sacrificing ALL of the flexibility of a general purpose computer? Probably, but it could not acquire the "instant-on" TV-like status you seem to desire.
OTOH - my TV takes at least 15 second to cold-start.
Personally, I think the boot time is a good trade-off for an easily upgradeable, general purpose, customizable system.
On my desktop, it's ready to rock (take my input and do what I tell it to) in 1 minute 58 seconds. On my laptop, it's 44 seconds. I hear the hard drive spin up at 10 seconds for the desktop and 3 seconds for the laptop. On both computers, I log in with a "real" username and enter a "real" password, and I made no special attempt to do that any quicker or slower than I normally would, and that is included in the startup times cited.
On the desktop, I have the following run on startup, usually via scripts that either I or an application developer created that run on startup.
A Daemon that watches for and traps any application that tries to phone home.
A font manager (I have more than 2,000 fonts on the system).
The NDIS.
A user notification daemon for the NDIS.
That damn iTunes helper.
The AntiVirus daemon.
SMART monitor.
A configurable driver for non supported USB devices (joysticks, gamepads, etc).
Some other stuff that I want to run every time (special sound card driver, etc)
A script that starts my mail application, fetches new mail, loads the browser, and loads Google News on a new browser window, and places that window on top of all other windows and makes it the active application, waiting for me to do something with the keyboard or mouse.
That takes just a hair under 2 minutes on a computer that rolled off the assembly line in the summer of 2001. I use it for "everything"; multitrack recording, some video work, web developing, graphics work, playing CDs, mail, surfing the net, and everything else.
The laptop is a little less encumbered; I do no graphics on it, for example, so it has just the regular font load, nor do I do any audio or video work on it. It does mail, the 'net, lots of text-based stuff, and whatever projects I might be working on at the time when I'm out and about. Both run OS X v10.4.8.
On the laptop, startup from a fresh install of Windows XP (it's been updated to Microsoft's current recommended state, the usual precautionary software is installed, like AV, Microsoft Defender, etc but no actual Windows user apps are on it yet) takes 1 minute 2 seconds, and that's with automatic login (which I will be changing, but apparently that must be the default on installation).
So, I guess I would have to say that OSX does a much better job managing the startup. Especially since, once the XP desktop is up and I can get the cursor and keyboard to do what I say, it nags me 3 or 4 more times, making me click on stuff to go away. In fact, the 1 minute 2 seconds is a restart from a warm state; the first time I tried to time it, AVG began downloading an update, so the timing test had to be aborted (I let it update, then restarted with the stopwatch at the ready).
I would have booted Linux on the desktop and checked that, but it's in the middle of some reworking right now and wouldn't be truly reflective of what it's capable of doing (normally, I would have a working YellowDog Linux there). I did boot Knoppix on the laptop, but it took a while because XP refused to let any other OS have at it on the first 3 tries with the normal procedure of using the "C" key (two restarts and one cold start).
So, I was forced to Option-Boot and select the Knoppix CD manually. From boot to full KDE desktop complete with Konqueror loaded and the Knoppix site up, was 2 minutes 40 seconds. That includes the time for the MacBook to poll the busses and search for bootable Operating Systems (1 OSX, 1 XP, and Knoppix 5-someting LiveCD) and for me to select one. It probably would be just under 2 minutes without that overhead (ie if I had preselected the CD as the boot device in the Startup Disk control panel).
Just some real-world not-so-scientific tests to put some perspective into the discussion. It does seem that XP has some problems with boot times, though, especially since you can hear that the HD has spun up in, like, 3 seconds from hitting the power key. There's a 7200 rpm Seagate drive in the notebook as well, so you have to think that HD issues aren't the only thing going on with that.
How I long for the days when I had a "BBC-C" micro-computer with 32K of RAM that was available the moment I turned it on ... ... NOT!!!!
I'll put up with longer boot times as long as the computer becomes more useful in relation to the boot time ....
I have an idea until solid state HD's become standard. We'll put the OS on a ROM chip. Oh, wait, my Tandy did that years ago...
If you're complaining about it taking 30 seconds for your rig to boot up your friggin' impatient! Mine takes ~30 seconds from the time I hit the switch to the time Windows have everything loaded. 30 seconds is definitely fast enough, a lot faster than my laptop, and way faster than my old K6-2 with a bare bones Win98 install so my parents could surf the net.
MRAM http://en.wikipedia.org/wiki/MRAM
011100110110100101100111
It's called a Commodore 64. ;-)
Turn it on and it is ready to go instantly. heh
"Never give up, for that is just the time and place when the tide will change." -Harriet Beecher Stowe ^_^
My first PC came with 10k of ROM and 32k(woohoo!) of RAM. All you need for faster boot times is to burn Windows onto a 1G ROM chip. Problem solved. Of course, for patches and service packs you'd need to go and buy a new chip to install in your computer, so there may be a few machines out there that miss some critical updates...
I use an old Amstrad NC100 "notebook computer" from 1992 to write on sometimes. If all you want is to produce text while sitting on the subway, it rules. It starts instantly, you never need to "save" anything (the working memory is the permanent storage memory), and it starts up exactly where you shut it off. You got to love solid state solutions. It stores the text on a battery-powered 1MB PCMCIA. Serial modem is used to transfer files to the PC. Best of all, it runs on 4 AA batteries for 40 hours. :-)
P.S.
Why are there tags like "idiot" and "lame"?? Must Slashdot serve to the flamer community?
As for Adobe, they just don't like you.
"Her idea of wit is nothing more than an incisive observation humorously phrased and delivered with impeccable timing."
I would suggest setting the computer to go into suspend mode. Most (all) modern television sets never really turn off. The CRT is kept charged so that it looks like it is coming on instantly. I know you are going to say the an LCD comes on instantly but I'm just going to ignore that for now. Once hybrid drives (HDs with flash memory caches) become common you very well could see near instant start ups as the boot information can then be stored in flash memory.
Ah, the yonder days of yore, when you could boot up a TI-99/4A or a Commodore 64 in a mere 2-3 seconds.
Sure, you only had 64 K of total RAM (16 K with the TI; unless you had a PE).
Sure, the Commodore 1541 floppy drive was only slightly faster than handwriting machine code and typing it into the machine.
Sure, the OS wasn't upgradable (by any software means).
Sure, my cell phone's processor could run circles around their CPUs.
But you gotta admit, they booted FAST!
Someone at Microsoft is getting a commission on the reboots, and someone else is getting a cut for the time spent rebooting.... it's all a game to see how much we'll put up with before revolting!
I hate call waitin`~+~~~
NO CARRIER
First, a computer will probably never be "instant on" because there's a lot of stuff that needs to happen to get it going. Especially concerning pluggable hardware- USB disks, cameras, scanners, etc.- where the computer has to basically reset it, wait for it to answer back, load drivers, or what have you. Also, consider networking- it's notoriously slow and inherently unreliable. Even if you know that your DSL router is up every day, the computer has to assume that it may not be there, so it has to allow for certain timeouts. This sort of timeout logic takes place more than you think.
That said, a lot of the time is because the manufacturers have cobbled together stuff over time, and it's really poorly done. That, to me, is the bottom line. Look at Linux and UNIX- its startup routines basically were designed 20-30 years ago. The programmers took the easy way out, "Let's see... first I gotta do this, then I gotta do that... oh, and this depends on that so I'll start that first and make sure it's up and running before I move on..."
But in the latest version of Solaris (a version of UNIX), for example, they realized that a lot of the startup IS slow and inefficient, and they redesigned it. In this way, the computer can be up and you can be logging in while services are still starting.
To my mind, the only thing that needs to be up on the computer is the hard drive and monitor. And those don't change too often. For everything else, the machine should quietly start them, while still allowing you to login (of course, that would be difficult to do without a keyboard :-) ). I think startup on most machines could be reduced to 10 seconds if it was properly designed.
Someone correct me if I'm wrong, but wasn't the BeOS really fast on startup?
-Mike Schwager
Pretty much all PC computers use the BIOS, the BIOS is 20-30 years old. At startup the BIOS performs the POST (Power-On Self-Test) stage. I heard with LinuxBIOS the record is about 3 seconds.
Most modern operating systems such as Linux and Windows have Plug-n-play, but before operating systems didn't have that, and the BIOS probed all devices and such at start-up.
Microsoft have with every release of their operating system promised faster start-up.
When booting the operating system, there is lots of data being read from the hard disk and hard disks are very slow. Flash disks are slowly coming, and in a few months we will see so called "hybrid-disks", they are hard disks with like 1 gb solid-state memory for store of important stuff such as boot things.
With fast solid-state disks, booting will be faster.
This topic has come up before, back in the day before eternal september finally came to slashdot... and was answered much more intelligently at that point in time (of course)
... be careful, some things *do* have to run before others!
I can't provide any answers for windows, but the problem on linux was:
1. init starts up all tasks in sequence
-- So start up tasks in parralel by using ye old '&' in your boot scripts
2. the init script is typically written in bash, a notoriously slow interpreted language
-- If you are macho, rewrite your init script in C
Of course, many linux users just leave their computers on all day and night, so it's less of a problem then, but still!
It's not a hardware issue as such. The BIOS zips past as you boot your PC.
:-)
I remember BeOS taking about 15-20 seconds to fully usable on an old Pentium 800 MHz, whereas Win2000 would take 3 mins.
From my experience, WinXP dramatically reduced the time to boot versus Win2000. From minutes to a sometimes less that a minute to see the logon screen. Logging in is another matter entirely and is slow on Windows due to startup objects, etc.
A custom Linux is extremely fast in showing the log prompt. But Linux distros seem to take a long time to boot.
Standby/Hibernate is not a option for most desktop PCs. People forget that we are trying to REDUCE energy use, not increase it. Leaving PCs on 24/7 is not acceptable to most people at home. They will turn off the PC at night, and then experience boot up time the next time they use it.
The fault lies with the OS people, not hardware people. They do not tune for Boot up speed. We get faster PCs but boot time "seems" long. Although I bet it's shorter than in the 386/486 days
There are two areas you need to look at on this issue, and to not look at each, you miss possible solutions.
The first area is the BIOS boot time, which has nothing to do with how long it takes Windows to load. On many systems, it takes a while for the BIOS to check for hardware, hard drives, and even devices that may or may not be there before booting. Now, this is an area that many can improve, but in some cases, the BIOS itself will never allow you to tweak things properly.
In the "old days", you had the option to set the drive parameters yourself in the BIOS, and to avoid having the BIOS scan for devices. Or you could do a one-time scan, and it would put in the parameters for you so the system doesn't need to check them every time you boot. I honestly miss that since on most systems today, you need to wait for the system to check what type of hard drives and CD drives you have every time you boot the machine.
Next, you have support for all sorts of devices that many people never use. If you disable the integrated firewire controller(IEEE 1394) for example, there should improve performance. You should be able to disable the integrated audio if you have a sound card as well. Or the floppy controller. There is a LOT of this junk that gets checked every time you boot the system. The sad thing is that it seems the BIOS STILL checks all this unused stuff at boot time. It may be a little faster, but the BIOS still knows the features are there, and it bloats the process. More features are generally not better when it comes to the time it takes the BIOS to finish checking.
So, the BIOS finally lets you boot.
And now, Windows starts to check every last feature in your system so it can set it as a device for Plug and Play. Even if you disable the feature in the BIOS, Windows will find it, and try to get a driver working for it. You may be able to avoid this by using jumpers on the motherboard to disable features rather than the BIOS/CMOS setup, but not always.
Microsoft ships Windows XP and Vista with all sorts of features enabled by default, which increases the load time. You have to love how Microsoft auto-enabled Wireless Zero configuration for every copy of Windows XP, even if you are looking at a desktop system without a wireless card in it. Windows Time, and many other services get turned on by default, and that is where most of the problem comes from as well. If you have Norton Anti-Virus, that will add to it, and Internet Security is worse still.
Now, it's not always the fault of Microsoft, Dell is one of the worst offenders when it comes to "adding features". When you get a Dell, unless you specifically ask them not to pre-load all their garbage, you get Dell support agents, and software to "make it easier for them to help you". What it really does is make them money because they get to sell more hardware when angry customers want to throw their monitor across the room because of stupid garbage like this.
The speed of the hard drive is another thing to look at when it comes to how long it takes to boot. If you have a 4200RPM hard drive, that's slow by the standards of today, but most systems come with either a 4200RPM or 5400RPM hard drive in them, rather than the faster 7200RPM or 10,000RPM drives. You can have two identical machines except for the hard drive, and you will find that a faster hard drive can trim 25 percent off the time it takes to boot.
System memory is the final area that can slow down your boot speed. If you have under 512 megs of system memory, that's too little for Windows XP, and generally you want 1 gig or more of system memory to make Windows XP load and run well. With Vista, this jumps to 1 gig really being required to load/run decently, with 2 gigs for what you really want.
If you have multiple physical hard drives, because of all the above issues, it really does pay to consolidate down to a single drive. I have recently done such a consolidation, going from 4 physical hard drives down to a single drive, and it has done wonders.
There's a tradeoff with saving electricity and money. Both which can be measured in terms of "life". I don't see the problem.
Yeah right. Theres no such thing as a 10k rpm laptop hard drive. That would be such a waste of the battery no one would do that. Even a "gaming" laptop I would have trouble seeing anything over 7200..
I'll just use my special getting high powers one more time...
http://en.wikipedia.org/wiki/Hybrid_drive
This technology was recently promised as a way to help eliminate long boot-times for operating systems, and I think it's coming really soon, as a "stop gap" between traditional hard drives and entirely flash memory based drives.
LINUX BOOT PROCESS
/etc/modprobe.conf like fedora/gentoo/etc do.
/tmp and /var/tmp as a RAM filesystem so apps can write small temporary files somewhere. Turn off logging too; I find it's a waste of space except on my servers.
/etc/modprobe.conf to do pre- and post-install stuff like mixer levels and lose a minimal amount of time (1-2 secs).
/etc/hosts can resolve network stuff for apps that want it. With any luck, you'll have an IP by the time you open your browser.
I use Slackware, which is a great example of a Linux boot system which can be minimized. Everything is in one directory (/etc/rc.d) in a dozen shell scripts. If you read it all, you can put the parts you use into one big shell script and it will save you 80-90% of your boot time.
* Step 0 - Booting
Bootloader comes up, wastes some seconds and loads the kernel.
The kernel loads up and does its thing, this takes some seconds too. For a generic machine this is good.
* STEP 1 - System Initialization
( Note: Depending on the software on your system more steps may come before or after this, but this is the stuff that needs to be on all systems... )
First, the OS must initialize itself. It connects all the disk partitions it wants and starts checking filesystems; you HAVE to check your filesystem or you'll severely fuck your data, and you know that backup you've been meaning to run for 2 years now is going to become painfully obvious once a few clusters go missing. After disk-checks, make sure it's mounted as root, continue.
* STEP 2 - Hardware check
OK; now we need to find our hardware. Before we can do things like mount network filesystems, we need to find all our hardware (mostly network cards) and then run the networking sys-init stuff. This usually involves running hotplug or kudzu and windows probably has a kudzu-like tool. Here's where we can save some time and kudzu does a fair job of it - save the hardware specs in a file and if it's checksum changes, scan for new hardware again. Once new hardware has been found it should be fairly simple to save the module names for that hardware and we can keep the hardware init stuff in
Now that hardware is found, initialize the network (grab DHCP, set IP addresses, start network/nfs/samba/etc daemons). DHCP and some initial network connections will waste some time and you can't do anything about it. Once we're online, mount network filesystems.
* STEP 4 - User Software
Next run system logging, clean up tmp, re-link libraries and perform misc actions. At this point we can finish running any crap the user wants at start-up and run our graphical interface.
SPEEDING UP THE PROCESS
By taking a look at your currently running system you can really minimize the time it will take to boot up next time. First of all, most applications won't like this, but if you boot up in read-only mode and don't initialize the network you can shave off 50% of the boot time. Keep a
Next you can take your currently-running modules (/proc/modules), reverse the order and save them to a file, and have a small shell script shoot them into the system with insmod at boot time. However can use modprobe and
Finally you should start your network stuff right after the graphics, and just keep your hostname 'localhost' so
SUMMARY
Why do OS's boot so slow? Nobody really cares enough to speed it up. Overall it's better to keep your computer running slow, because slow means stable and stable means no bug reports to the OS maintainers. It's like an economy car: sure you can squeeze more performance out of it, but at the risk of losing safety and control.
My wife just got an "updated" Cell Phone - tons of nice features in a smaller package. But when you turn it on, it takes about 10 seconds to get going versus about 2 for the previous one. Just a lotta stuff getting loaded in ... and the amount exceeds the increased speed of the processors/RAM/etc.
Hulk SMASH Celiac Disease
It's not rocket science. Computers don't take long to boot. Windows takes long to boot. Look at the hard drive activity when booting - the OS and startup programs are loaded from a large number of (typically small) files, and the sluggishness is caused primarily by the seek time of the hard drive - the time it takes for the read head to move to the location of the next piece of data to be read. Transfer rate also plays a part here. When restoring from hibernation, a larger amount of data is loaded from the hard drive, but it is in one contiguous block - the contents of memory. Thus, the seek time is all but eliminated, and boot is accomplished in a matter of seconds.
My OS/2 system running on an Athlon X2 4200+ takes a whopping 4 minutes to boot up. OS/2 is probably the most leisurely booting OS ever created. OS/2 runs through a long list of drivers to load, beginning with low level drivers, then the installable file system software of choice, then 'normal' high level device drivers. Each device driver is loaded and initialized with its corresponding hardware during the boot process, unlike Windows which only does that when the driver is 'installed'. Then the workplace shell desktop is processed and loaded. Then autostart apps and network. Finally the desktop appears. There's time to make coffee, add creamer, chat with your cubicle neighbor, and go get a donut before the computer is ready for action. Fortunately, OS/2 doesn't need to boot a lot. What do you get with all of the booting action? Well, you know that if the booting went okay, then all of your hardware is working properly and ready for use which is nice if you are using the system for important data.
AmigaOS 3.1 boots in 5 seconds. Just another reason why Workbench 3.1 is better than Windows XP and Linux
My homepage: www.erkan.se
A lot of people set up applications (IM, screen eye-candy, etc.) to start during boot-up and that can lead to excessive delays in starting the computer. Also, if you know what you're doing you can fine-tune the services that automatically start when Windows boots up. Be warned that if you *don't* know what you're doing that you can brick your system pretty quickly. Other ways to speed up your boot times (especially if you've been running the same install for a longish period of time) is to reformat the hard drive and reinstall Windows. You'll be horrified at the amount of clutter that accumulates over time to slow things down (caveat: back up all data and have all of your application installation software handy as well as a day of movies/etc. while you are reinstalling).
My assumption is that you want to stay with Windows and not explore other operating systems (as has been suggested by other respondents), so I won't go down the path of suggesting the same.
Of course, your mileage may vary....
My computer boots to Windows XP's CTRL-ALT-DEL screen, from a powered-down state, in 8 seconds.
;)
Maybe upgrading from a PII is the answer to your problems?
Well, at least my post said SOMETHING. Could you inform me of exactly why saving a second of my life once every few days is worth jack?
Well then what about the Plus/4 with it's built in word processor, spreadsheet, database and graph generator? (albiet crappy ones... :p )
"There are people who do not love their fellow human being, and I _hate_ people like that!" - Tom Lehrer
...just to hit the power button... or hold the power button down to force power off... or unplug the cord... ...which is what many of these replies are saying... ...then why isn't the OS programmed to just shut off the damned power right away when I select "shut down" from the menu?
"How to Do Nothing," kids activities, back in print!
haven't seen this mentioned.
will be here shortly, and should replace all current *ram/flash. uses magnetic storage elements for data, therefore it's non-volitile, meaning it will save its state when powered-off and that same state will remain until powered-on. if the mram is large enough, the OS will not have to load anything into memory, there maybe some lag for devices to re-establish themselves, but you will have "instant" boot.
I reboot if necessary but otherwise I don't shut my machine down. ever. why?
The Autodetect shouldn't take more than a second to determine that no new devices have been added. Something lame is going on in the background of windows, it's loading a bunch of config files, and parsing them to figure out what to load up next. so a whole fun cascade occurs during each boot where all of these loads are trying to occur with some semblance of organization. But really it's a traffic jam filled with kludges to make sure that some peices can break the gridlock.
It is straight up absurd.
Reasonably with a decent set of hardware, a windows xp box should spend about 5 seconds (from the post completion) getting to the login screen. .25 seconds to rule out new hardware, and then 2.5 - 4.0 seconds loading up 256Meg of system image, and another .75 seconds initilizing the IO systems. The user authentication should come up in a flash, as it would be part of the boot image.
Unfortunatly to get it right they would need to have a major shift in the way they do what they do, and they have too much inertia to really change direction now.
Storm
My TRS-80 Model III was instant-on. But then it took 2 minutes to load Pyramid from cassette. So Windows XP is still faster overall.
What I do is put my system on standBy all the time. I run WinXP at home and it takes about 10 seconds to start up again. Works like a charm.. :)
If u would like to completly save power.. just use hibernate mode, its a bit slower than standBy, but it does the job.
When I went to the launch of Windows 95 in London - glitzy presentation etc, I was promised no more blue screens and NO BOOT UP DELAY. Instant on and instant of were the ways of the (then) future. Of course, even during the presentation this turned out not to be true! The rest, a they say, is history.
There are hundreds of comments already, and nobody has said "I use Linux, I don't have to reboot!" followed by a an uptime number. :)
(I'm at 44 days, stupid upgrade in Nvidia driver)
My beliefs do not require that you agree with them.
'' use of external command and external function modules, more commonly known as XCMDs and XFCNs. These were code libraries packaged in a resource fork that integrated into the system and/or the HyperTalk language -- an early example of the plugin concept. Unlike conventional plugins, these did not require separate installation before they were available for use ''
wikipedia Hypercard link
That's because USB is designed to be hot-pluggable. The drivers assume that devices will be (dis)connected at arbitrary times, and deal with the fact. That isn't the case for a lot of other devices. The drivers would have to include a call to re-initialize the hardware coming out of hibernation. I don't know if that has been universally done yet or not.
[100% ISO 646 Compliant]
SVM, ERGO MONSTRO.
It's not just a few cents of flash. Firmware, like any other sort of software, invariably needs to be updated. (If it didn't, it would run straight out ROM, which is cheaper than RAM or flash)
So the total really comes to.
1) A few cents of flash
2) Hardware support for programming flash.
3) Driver support for programming flash
4) Careful documentation and usability testing so that users don't botch the process, call support and complain that their hardware is dead.
I'd stick with the RAM based solution. Simpler, cheaper, and more reliable.
This thread would not be complete without at least one fond recollection of BeOS, and how it booted in just a few seconds, start to finish, even on old crappy machines.
"verucasalt"... I WANT IT NOW!!!! NOW, NOW, NOW, NOW!!!!!
Or, perhaps...
"This baby can flash-fry a buffalo in 8 seconds."
"Oh, but I want it now..."
Embedded systems (that use true embedded operating systems) do this all the time. The first boot takes longer, but subsequent boots are pre-configured and boot much faster. Heck, many of them are pre-configured in manufacturing, so other than some necessary sanity checks, are as close to "instant on" as you can get.
PCs are different, because they do necessary hardware checks. Even if you compile the drivers into the kernel, they still need to query the system to know what's there. Then there are various daemons or services to start.
Why does a PC need to check? Well for starters, what if you install new hardware? You at least have to check for that. And what if the printer attached to the USB bus were powered on while booting (or powered off it it were normally powered on)? What if the mouse was in the second USB slot instead of the first? What if your ethernet cable were unplugged? What if your wifi router were powered off? Etc, etc.
In short, a PC doesn't know what state its hardware/connections are going to be in, so it needs to check. This is also why PCs are much more flexible pieces of equipment than single-purpose embedded systems.
Don't blame me, I didn't vote for either of them!
Whenever someone says "why can't computers be more like $APPLIANCE", the correct answer is almost always "because computers are much more complicated than $APPLIANCE".
Forget how complicated your computer's hardware is (and I have no doubt that you've read the blurbs about how the latest Intel chip has 80 zillion transisters) and just think about the software. If you go through your computer and you find all the files whose names end with DLL, EXE, DRV or SYS, total up their size and divide that number by 8, that is (very, very roughly) the number of individual parts that makes up the workings of your computer's software layer.
All other human endeavor pales in comparison to that. You shouldn't be complaining that your computer is buggy or slow or bloated or that it takes too long to boot. You should be amazed that it works at all.
(And yes, people are working on making computers start up faster. Someday, someone will do it and it will be another huge, complicated effort that everyone will just sort of take for granted, just like they have every other technological miracle.)
Doesn't something seem wrong when you come up with one person doing 31 _million_ jumping jacks a year while windows is booting?
If I tried to do 113 jumping jacks a second for an entire year, I think waiting for windows to boot would be the lessor of two health crises.
For better math, how about 113 (jumping jacks/reboot) * 365 (days/year)* 2 (reboots/day)?
Are there any other "oldsters" out there who remember when TV's weren't "instant on"? When you had to wait for them to warm up? For you young whippersnappers, our televisions used to have things called "tubes" in them (or "valves" if you were in the UK). And we were thankful for them!
Frankly, instant (back) onis easy - I've designed a large number devices that do this.
It's just that Windoze was never designed to do this - it's not an important feature.
-- Butlerian Jihad NOW!
And from what you're saying it can take up to 10 seconds on digital? Holy crap... how can you channel surf with that? Maybe I'll just not go digital.
10 seconds is the worst case, but 2 seconds is not uncommon. However, if you're using a digital tuner, you quickly learn to just use the digital channel guide as well - just press the guide button and half your screen is devoted to a TV programming grid that you can quickly surf through - you're just doing your surfing based on reading instead of looking at the screen. Bad if you're illiterate, but much faster than surfing channels the old way, especially when you have lots of channels to surf. Plus, you can see what's coming up. And far better than those old "TV Guide Channel" listings that forced you to stare at the slowly scrolling list for 2 minutes before the channel you wanted finally rolled around to the display.
-BbT
Love the Mac fanboyism. The story is actually about Bill Gates and his point of view on OS bootup speed.
...plus I've never owned a Mac and I havn't used one since 1985 - so I don't think you'd class me as a Mac fanboy. I run Linux on PC hardware.
Hmmm....
Well, the story is told in the movie "Triumph of the Nerds". That came out in June 1996 - the events it's talking about were much, much older than that. But XP wasn't released in October 2001.
There is absolutely no way this could be a ripoff of a Gates-ism from the XP development era.
www.sjbaker.org
Turn off file sharing and make sure that your TCP settings do not do NetBIOS over IP. When an XP machine is booting with file sharing is on it sits there waiting for domain controllers or workgroups (master browser searching) to join. Capture the boot up process with a sniffer for some interesting insights into what the machine is attempting to do. AV software is also a huge cycle killer. Most is set to scan all of the OS files as they load ... with this is pretty much a must have but you would be surprised how much faster it loads with no AV scanning taking place during the startup process.
"There's a tradeoff with saving electricity and money."
Actually, it's between electricity and boot time. Of course, computers have low power states that they can enter into automatically (which mac people claim are superior on macs of course) and the resultant power usage increase would amount to pennies a month. Isn't that a small price to pay for the savings of so many human lives?
"Both which can be measured in terms of "life"."
Actually, neither can. No reduction in boot time nor any reduction in standby power consumption will ever amount to the production or savings of human lives. If you want to save yourself some time each day, stop picking your nose.
"I don't see the problem."
I'm sure you don't.
My provider provides a TV Guide channel for the non-digital subscribers, and the firmware in their digital boxes (at least the DVRs anyway) have a unique "feature" that makes the guide pop-up in place of the one on the TV Guide channel when you tune in to it. It's not all that noticeable if you use your box in the cable company's preset color scheme, but if you change it to one of the others it is immediately noticeable.
I, personally, don't like that feature, but I know some people who do because it auto scrolls like the TV Guide channel grid, but instead of just showing the basic cable lineup and select channels from the digital packages, it shows all the channels.
So in the theology of The Invisible Pink Unicorn, the equivalent of Satan is an invisible Ballmer?
Sounds about right.
I see even classic Slashdot is now pretty much unusable on dial up anymore.
Have a little patience! Many mainframes take several minutes to boot up. We had this one AIX machine at work that took 20 minutes to boot!
The bits on the bus go on and off... on and off... on and off...
Because your so gay
I have a 1st gen MacBook Pro (Core Duo, 2 Gb RAM, etc.).
It takes between 5 and 10 minutes to boot, from power up to my every default application started (all *5* of them: Mail, Firefox, Adium, iTerm and QuickSilver) and usable (no spinning wheel of death, etc.).
Can you believe it?
--
Stefane Fermigier, CEO, Nuxeo SAS
Open Source Enterprise Content Management (ECM)
Nuxeo 5 EP is out! - Now Java EE based, standards compliant
Web: http://www.nuxeo.org/
You can always try multitasking instead of staring at the screen waiting for it to boot. Grab some coffee, adjust your chair, read your snailmail, chat with your cubicle neighbors, whatever. I learned to not wait for the computer back when I would query the tape drive on my C-64.
sleep function, been there on all macs for ages (probably on all dozers for ages if you know where to look)
"I hope you like Guinness, Sir. I find it a refreshing substitute for, er... food." Col. Jack O'Neil, SG-1
Actually, neither can. No reduction in boot time nor any reduction in standby power consumption will ever amount to the production or savings of human lives.
Perhaps you should state why you have this opinion. I still don't see your side.So that's what PRAM is for!
Seriously. Home Theatre MX-500. Put batteries in, and a "little man" runs across the bottom as it goes through a POST.
My Win2K box takes at least 10 minutes to reboot.
-Clio
Karma: Bad (mostly from not giving a fuck)
Blog: http://clintjcl.wordpress.com