Researchers at IBM are testing software that would let you tote your home or office desktop around on an iPod or similar portable device so that you could run it on any PC. The virtual computer user environment setup is called SoulPad, and consumers install it from a x86-based home or office PC. SoulPad uses a USB or FireWire connection to access the network cards for connecting to the Internet, the computer’s display, the keyboard, the main processor and the memory, but not the hard disk.
I read about this earlier, and I probably missed a few things, but are they really doing anything new? Aren’t they just usb/firewire booting onto a knoppix distribution stored on a usb/firewire harddisk/device, such as the iPod in the video on their website?
It’s an interesting idea but:
It seems to me that I already have almost everything required for this. Under OS X, all I have to do is copy my Home folder onto a mobile disk that’s big enough. Then to use it, I just need to convince my second PC to use this folder as a users Home directory and voila, I have my documents and settings wherever I go.
Admittedly, the ‘convinceing’ part of the process is fairly tricky at the moment. But I am sure, given a couple of days of a bit of poking around, I could find out which files on the host computer it is necessary to alter n order to achieve this.
install Mac OS X on an iPod, walk up to any mac, boot off your iPod by holding down the option key…I’ve even started using a Mac mini like this recently….
Off topic, but it’ll be interesting to see how Apple move their flexibility with firewire to an x86 BIOS…
but it will only boot on a Mac, and there are only a few thousand of those in existence.
most likely this is why they went with intel rather then amd. intel have a boot system standing by that is quite similar to the openfirmware that apple uses with their os x machines.
so, all they have to do is implement a similar feature in intels bios-killer and presto, there you have it.
This is all good and well if you only use ‘bodies’ (terminals) that you trust. If you ever come across a hacked terminal though it will make identity theft look like childsplay.
This isn’t as far fetched as it might sound btw: a while back there was a gang in Belgium that ‘hacked’ the payment terminals at certain gas stations. The machine would copy you card details while a camera captured your PIN.
> This isn’t as far fetched as it might sound btw: a while back there was a gang in Belgium that ‘hacked’ the payment terminals at certain gas stations. The machine would copy you card details while a camera captured your PIN.
I’ve heard about similiar gang in Poland.
But I don’t see how would it be possible to create software that is secure enough to be run on every, even hacked hardware. If the computer’s keyboard had built in keylogger, or network traffic was captured by software independent network device, there is no software “secure” enough to handle this. If hardware is “hacked” itself, if it records user’s actions, there is no good software protection and will never be.
The thing is: this is not necessarily meant for public terminals/internet cafe’s.
The best use for this technology is perhaps to carry your work with you around between your home computer, work computer and travel notebook.
As such, it’s a pretty neat idea. Storing your saved work would no longer require seperate storage mediums nor synchronising. Instead, you lug with you a portable harddisk. Another advantage is that your configuration settings won’t need to be replicated across computers, which is still a seperate issue with USB sticks. Also: more storage space than USB sticks.
You can already do that by mounting the usb stick as your home directory.
Would be nice if the same technique could be used for one Linux (package) standard.
..and it’s somewhat cheaper, because certain hardware devices won’t have to be in each computer, like storage devices.
Depening on the price of the newly required hardware of course.
I investigated this heavily about a year ago. As it turns out, it is not really a good option because USB/Firewire is a poor interface to be driving ones entire OS through. Until eSATA is common this won’t fly. Well that is if the eSATA port turns out to be small and powered. I hope they don’t f* the up.
The fact is, this is a marketing job (not that that’s necessarily a bad thing). We’ve all had the thought “Wouldn’t it be nice if I could take my desktop with me werever I go?” This is just the next logical extension.
I still don’t buy how you can make it’s secure on untrusted terminals, though. The real problem I see, is that if they market it as secure, (or, more likely, if people just don’t fully consider the security implications), my insurance agent, my HMO, or any other innocent might use this technology to access my personal information in an insecure way, and what’s worse, without me even knowing it or being able to do anything about it. I’m not worried about me violating my privacy. I’m worried about them. I’m worried about all the people I need to share my personal information with, some of which may have no real clue when it comes to identifying real exposure points to information security.
take the palm livedrive, a pda with a hardrive. maybe stuff wifi and phone electronics into it. hell, lets stuff a video output into it to (or maybe have it as a plugable addon) so that you can play presentations straight to the projector.
need more horsepower or a bigger screen? hook it into a laptop or desktop and boot the basic os from a cd or dvd. most livecd linux distros today allready allow you to store home to a external device. hell, get a 1gb usb key and you should in theory be able to put the cd on that
still, i kinda like what you can do with a mac today. boot the openfirmware into a special mode so that it pretends its a firewire storage device. hook it up to a diffrent computer and have this boot from said external device. only reason we dont see anything similar on a x86 is the bios. it basicly cant handle it. it would be fun if say amd fired up support for openfirmware
to bad that intel have a competing system to openfirmware that it wants to force down every x86 users throat.
only real diff i can see with IBMs solution is that they use something like xen or vmware so that one can store the state of the system at any time and then resume the state on a diffrent computer at a diffrent time. as in, you no longer need to log out
Now that IBM has sold the ThinkPad, we have the “SoulPad” for capturing the “soul” of a PC?
Like a PC has a soul?
Like people really are coming crying to IBM, “I need to capture the soul of my PC!”.
It is pathetic not to mention somewhat disrespectful to creatures of Nature that do have souls vs. a bunch of jacked up files.
One could argue the Mac had this feature back in 85 or when ever the SCSI HD was added, any drive could boot almost any willing Mac. Not exactly keychain though but USB drives now store many times what those 1st scsi dives did, wasn’t that long ago that 20MByte was a big thing.
>>>> One could argue the Mac had this feature back in 85 or when ever the SCSI HD was added, any drive could boot almost any willing Mac. Not exactly keychain though but USB drives now store many times what those 1st scsi dives did, wasn’t that long ago that 20MByte was a big thing. <<<<
I was just going to say that I was doing this three years ago using the removable hard drives of my Dell laptop. I got the IDE adapter for my desktop computer and Larry got one for his. At that point I could move the hard drive between and three machines and be booting my enviroment in less than twenty seconds. Add in the machines work which at first required me to move the monitor first and I was using 6-8 diffirent machine with the same boot drive.
However, your message reminded me about my own Mac servicing days. And yes I knew more than one Graphic Artist who moved an external SCSI hard drive between the machine at home and the machine at work. A few had no choice as the copy protection only allowed one installation of thier software to be made and if they wanted to use the software at home too it was either buy two copies or install the entire OS, FontSets and needed software on an external hard drive.
The posters making the suggestion that this is already possible with current hardware and software are absolutely correct, which is why we should have this already. However, the issue is that we can only make our mobile storage devices and PCs cooperate in this manner by some amount of configuration, and we can only integrate them *well* by using some special-purpose software. To make these changes, however trivial, we need demand. Right now, users are quite happy with the (to our tech-elite circle: crippling) limitations of personal computing, which is why ideas like this remain possibilities, rather than realities.
Visit IBM’s Pervasive Computing Lab if you are in the Austin, TX area and have an interest in these things. It’s open to the public, and it’s chock-full of prototypes that are more-or-less viable if there was only an appropriate level of demand. As you might guess from TFA, there’s a lot of OSS leverage going on in the lab.
For the time being, this sounds like a job for the iPod Linux Project…
Please look again at the article. The `soul’ that is being referred to is the entire _virtual_ machine state.
Booting from external media isn’t new, (hell, look at floppy disks). The thing that is new here is the concept of wrapping an auto-configuring OS around a virtual machine that has a more static config.