You see, my two year old Macbook already has an Instant-On technique. When I open the lid, I see my desktop in 2 seconds or less. That probably beats Splashtop. Also, I can then access all my applications rather than a restricted subset. What is this amazing technique, you may wonder? It is simple: ACPI! You see, when you have a frequently used computer with a stable operating system and ACPI support, you can go a long time without rebooting. Basically, the only limits are hardware failures and your need to install security updates. In this case, the boot time does not matter. Sure, it drains a little power, but if you use it more than once a week for a long amount of time, you will have to carry a power cord around anyway. Ultimately, working ACPI support would be a much more viable 'instant-on' technique than Splashtop.
What about viruses? Alright, what about them? Sure, if you use Splashtop, it might be harder for malware to maliciously modify your computer, but if a remote vulnerability or privilege escalation vulnerability is discovered in your software, that vulnerability will remain viable for a long time. There is a tradeoff.
Also, how many people actually use these media-BIOS things anyway? My mother's Dell laptop comes with 'Dell Media Direct', and it has been nothing but a nuisance to her; she sometimes accidentally boots into it and fiddles with it for a few minutes to return to Vista. If Splashtop is used in this fashion, then millions will hate Linux just as much as we do.
Ahhh, the smell of victory!
12 comments:
Simple fact people don't have a stable OS in windows.
The mac book method is suspend to ram. When suspend to hard disk it takes longer to start.
Splashtop is currently depending on funding from people wanting to bundle applications with it until its market gets large enough.
Splashtop's roadmap is 3 steps we are currently in second step. That Dell media thing was its first form. Final form thrid party applications.
Part of it they are trying to pull MS into embedded there OS. MS dumb enough to do that they are so dead. Splashtop is 1 to 2 dollars per machine. If does enter embed most they will be able to sell for is 5 dollars when splashtop changes to a full OS.
Linux is basically being used to do a number of MS. Who would not want there next OS and Office suite for 5 USD.
I have the sense that this sounds like an abysmal failure. Ok, great, I can have an environment in 2 seconds. How long until the e-mail application starts up. And then what happens when I checked my POP mail server with whatever the linux mail application dujour is, then need to get real work done so boot into windows. Ooops... those mails aren't there, unless I use thunderbird and by some trickery the linux version can be configured to use the available windows settings (which I doubt).
By the way Windows and Mac both restart for secuirty updates way more than they have to. There are ways of fixing problems in place.
Long term Linux goal is once running Zero reboots. Prototype projects have shown that is more than doable.
Its about time people start asking everyone for this too.
radiantfire you just talked about what http://pim.kde.org/akonadi/ is working on.
Yes currently splashtop does not provide a email client instead uses web mail interfaces.
So called cloud computing works in the direction of splashtop.
The game is a foot. Really hard to point at a weakness and say hey I have you done to rights here.
"Simple fact people don't have a stable OS in windows."
The same could be said about any flavour of GNU/Linux.
Not really. There are stable forms of Linux. Just they are not popular. Forms currently used in phones, routers and the like are highly stable. X11 is currently a dog that is going to change.
Windows will fail being brought back from hibernate or suspend on all hardware in time. Windows is triple killed. 1 issue is memory management others are is ACPI and closed source driver issues ie drivers not built to support being hibernated like linux.
Linux is only killed on 2 fronts and 1 of them it can avoid.
Linux has less closed source drivers. Linux can do hibernation and suspend to ram without ACPI support just cannot match 2 seconds or less in the release kernel. Hardware re int is taking to long. Fastboot branch can get a Linux back up from hibernation without ACPI support in under 7 seconds. That is restored from harddrive.
Restored from fast read media under 2 seconds is doable. Yes laptop fully powered off to restored no ACPI support.
Windows is sadly lacking in restore tech thinking everyone really knows how bad ACPI is they should have provided non ACPI was around it.
Linux having two hibernation systems is handy for those boards with defective ACPI. Windows answer is either run with defective ACPI and have trouble or don't hibernate.
There are many places were Windows could do way better than it does. Just like Linux can still do way better.
Linux is moving on to support hibernation support even when hardware has changed. Remove a ram stick then try to restore windows or mac from harddrive hibernation. Linux will be able to cope perfectly with that change.
Same with a lot of other device changes. Hot pluging video cards is also on the roadmap for the end of 2009. So yes in time it will come as long as the machine has a close enough CPU on a board that Linux will run on and the same filesystem restore from hibernation will be able to happen.
Lock to motherboard Windows does is going to become a limitation.
Its a simple fact without the tech in your OS you don't stand a chance in particlar markets.
Ie X11 not having DRI2 and KMS kinda kicks Linux hard.
Windows not having non hardware requiring restores kicks them hard on incompatible hardware.
Mac is lucky they set there hardware so they can do far more stuff simply.
Who would not want there next OS and Office suite for 5 USD.
Uhhhh . . . . . everyone who is not already using Linux? Remember, Linux has been 'free' (i.e. zero price) for its entire lifetime, and it STILL has less than 1% of the market. Apparently, your $5 operating system is still too expensive for most people.
Yeah they are stable forms of Linux, In tightly controlled hardware. Just like a mac. Where GNU/Linux has to compete with Microsoft in the same market aka mass-market pcs it fails in the most visible ways; 1) wireless encryption support, 2)graphics card support. (And yes I know M$ has cozy relationships with hardware venders to allow better drivers. GNU\Linux devs and promoters should take note of that and stop making false representations about the abilities of GNU\Linux. What you can do with published specs pales when you have a non-disclosure-agreement to build a driver from).
ATI and Intel pay developers to build open-source drivers and the results suck. Only the closed source ATI driver ever gave me good results.
Why does the open source OSS driver not work after suspend to ram? And yes ALSA sucks compared to OSS.
kerensky, I have never seen a linux system that was as tightly integrated as the average Macintosh. Sure, there are some distros like Ubuntu that make an attempt to integrate various parts into a cohesive whole suitable for an end-user. There are also companies that sell pre-installed Linux systems. What I have not seen is any company that combines these two impulses. If a company offered both a distribution and the hardware to run it, Linux would have a seriously viable system. I do not think any of the companies listed above are capable of creating their own distros, so maybe the major distros should create their own hardware. If Canonical wants to become profitable, it should sell hardware alongside software. I think I will elaborate on this in a future post.
I was quite surprized to know that there was no proper way of connecting to WPA wifi networks on Kubuntu 8.04 (without going through a whole lot of muck). I upgraded to Intrepid Ibex & not only did the wireless break, even the LAN doesn't work now!. I don't know what the fuck is wrong with this shit. Man Linux sucks.
the question begs: who buys a fancy new computer to use a hokey embedded OS?
Any new computer from the store comes with a full OS.
And no geek who is building his own from new parts is going to bother with it.
It's a gimmick.
And cloud computing, in one form or another, has been heralded as the Next Greatest Thing many times.
Splashtop and cloud computing? Well shit, son, that sounds like the same old Thin Client song we've been hearing about for years.
There may be a bit more OS on the client but it's the same damn thing.
And who is going to want their applications on the other end of a wire?
After their network connection dies a few times: no one :P
Hard drive makers could pre-load Linux on their products. Doesn't mean people will use it. The only use I could see for an embedded OS would be to troubleshoot the hardware or main OS.
Just another tool, in other words.
And boot time arguments are just plain stupid. If 1 minute is too long to wait for a system to cold boot, man, you've got issues.
And, for the record, I can't remember the last time I shut down my old T41 XP system. Which means that I can't remember the last time it failed to wake up from it's twice daily suspend(to RAM)/resume.
Guess it works-for-me(tm)
When all of these wonderful new abilities Linux will have finally condense from the vapourware state, and actually prove to work, we'll all be there to take notice.
Until then it's all dreams and promises. And a fellow can't run a computer on those.
@ linux haters redux: I was thinking of the router with linux on them and the various cell phones w/ linux. But that's not the same league as a mac.
Post a Comment