Slashdot Mirror
Try Out Chrome OS In a Virtual Machine
itwbennett writes "Some very generous Alpha OS geeks have snagged the Chrome OS source code and compiled a version to share with the rest of us, writes blogger Peter Smith. 'The build comes in the form of a virtual machine, which means you'll need VMWare or VirtualBox running, and of course the image of Chrome OS itself. The folks at gdgt are distributing the latter, and they've set up a page with all the links you'll need. You'll need to create a gdgt account if you don't have one yet. The Chrome OS image is only a bit over 300 megs, so it's a fast download. If you need a little more handholding, TechCrunch has a step-by-step guide to getting Chrome OS installed and running using VirtualBox, and a Chrome OS torrent they link to.'"
The only novelty is that the lack of a "shutdown" option seems to be intentional; the local machine is supposed to be stateless in the sense that it commits all transactions remotely before announcing their completion. Plan 9 also tried to achieve that goal, at least initially.
Kudos to the people who put these images together, though--they've saved many of us significant time.
I got really excited until I realised that this wasn't for the DEC Alpha processor. Shit.
There's a lot of handwaving about how Chrome is not Windows, how it won't let you use photoshop on the netbook, as if you would. Here's a hint: if you're trying to run Photoshop on a 10" screen, you're doing it wrong.
Look for disastrous reports from Gartner, Forrester and of course the Rob Enderle / Maureen O'Gara flackalyst duet on how Chrome is the worst thing since smallpox. These are your clues that this is the real thing. They said the same things about the When Google says they released the source, people build it and publish virtual machines the same day.
Netbooks are stepping up in performance, as this four-threaded model shows, and will soon be able to do many more things. Yes, VDI is starting to ramp. There is still a place for Chrome. It's the dead-simple desktop interface that many of the technology impaired need. It's a point on the graph twice the distance on the line from Debian to Ubuntu.
A bunch of people are going to whine it doesn't support disk. It's a next-generation operating system and solid state is the storage of the next generation. It has local storage - just not the slow kind you're used to. There is no more reason to support the legacy spinning disk on this platform than there is to support tape storage or floppy disk. Moving parts are so 2008.
Help stamp out iliturcy.
Torrents ought to be choice #1 in distributing files of any decent size. As an example, I thought it was utterly retarded that the recent Ubuntu 9.10 release didn't have the download torrents front and center. Why the hell not? Obviously they didn't have the bandwidth to handle all the direct downloads, as I started one just to see how slowly it would go. It crawled along at less than 1 KB/s for hours. They had the torrents advertised on the forums, at least, but they clearly made the launch harder on their servers than it should have been.
Torrents let you do more with less bandwidth. Take advantage of that! I understand some people may not be able to use them because of their ISP being a douchbag or whatever, and those people will need normal HTTP/FTP transactions and mirrors... but everyone else can use torrents and share your burden with you.
Do what thou wilt shall be the whole of the Law
before RMS crashes the party!
Torrent and Info: http://pastie.org/706872 http://dl.dropbox.com/u/457451/ide.vmdk.torrent Because making an account on some shady website that's exploiting the situation seems wrong.
http://gdgt.com/google/chrome-os/download/
Yes it requires an account to download.
However it is not verified. AKA login immediately after.
I pounded the keyboard a few times and dloaded no problem.
I'd tell u my user and pass if I could.
User: DotSlash
Pass: Slashdot
Dunno if it allows multiple logins....
User confusion, really. Users are fragile and easily puzzled creatures. Every link you put on your website is a link the user can frantically flail onto and accidentally click. Then you end up with people on your forum asking why the OS isn't working. After all, they burned the file right onto the CD!
In the case of a beta OS meant to be run inside a VM, yeah, user competence is probably not a huge issue. In the case of an OS which is trying to be a mass-market OS, you want it to be as easy as humanly possible, and adding a torrent link to the homepage does not make things any easier.
Breaking Into the Industry - A development log about starting a game studio.
Torrents are actually pretty inefficient(I don't mean to knock the design of the protocol; but making heavy use of last-mile upload capacity, which is some of the shittiest and most expensive pipe on the whole network, cannot be made as efficient as a conventional client/server setup). Further, they tend to promote the energy-inefficient situation of having large numbers of small servers, all bandwidth starved, spending hours nearly idling(but not able to sleep or shut down) as they wait for the bits to trickle in.
They do, though, have a great advantage, which is why we bother: In absence of a functional micropayment system, bittorrent is pretty much the best way of allowing the bandwidth costs of distributing something to be spread across all parties who are interested in receiving it. If it were possible to transact in 1cent increments, everyone would almost certainly be better off if the distributor just dumped it on Amazon EC2 or some other big hosting service and let interested parties pay the per-megabyte download costs directly(saving themselves the upload bandwidth and power costs). Since that isn't really viable(particularly, though not exclusively, if what is being distributed isn't wholly legal), bittorrent's easy sharing of hosting duties among downloaders is the next best thing.
Am I the only one with a CPU lacking virtualization? :(
Emacs: for people who just never know when to
ChromeOS is a very good move for everyone involved. Remember, this OS and the devices it will run on are not targeting average slashdotters. I can personally vouch that I come across daily contact with people, business people not just teenagers, who don't use anything other than their browser. The worst aspect of a computer for them, is upgrading, updating all applications, viruses, malware, and general maintenance of the system. They nearly all fail in these, and after a year, they think their laptop is not usable anymore and go and buy a new one. They would LOVE this OS, and are they primary targets of it. Also, synchronisation between multiple computers is a bitch, that even they most fail at. And they hate leaving their documents here and there. Files and directories don't work for them, it's a broken metaphor for most people, and as much as love to organise my files in hierarchical directories, they simply don't care. They just want access to their information, when they need, as conveniently as possible.
I hate Web apps as much as the next guy on this forum, and even use my trusty IMAP client for fetching my emails from Gmail. But I can't deny that web apps are the future, specially when HTML 5 comes off age and becomes widespread. If you look back at what the Web looked like 5 years ago and compare it to now, you'll see that it will be irresistible in 5 years time. Have a look at http://www.chromeexperiments.com/ to get a taste of what we are looking at.
On a more general note, anyone who is comparing this to old failed projects based on thin clients, X terminals or net pcs, is missing the point. Yes, the technology behind this might be similar to those, but times are changing. On the one hand, people are getting used to ever-present always-available services. On the other hand, 3G is now widespread, affordable, and provides great utility for many. Laptops and phones are converging. 2007 was the year of netbooks, 2010 might be the year of smartbooks (running ARM processors). Smartphones are morphing into Internet tablets (e.g,, N900). These are very different, and interesting times.
Yes, this is cloud computing, and yes, it raises huge privacy issues. It is up to us the tech savvy crown to raise these issues and address them.
Slashdotters can always run their trusty Debian or Fedora or FreeBSD or on their computer. And they remain great choices. But Google is pushing applications to go online and cross browser. They are pushing for open source drivers. They are pushing for open standards and cooperation with upstream and downstream projects. This is a Good Thing (TM) for all of us, even if we are not the target consumers of this OS.
--
Depends on who you ask. If you ask me, torrents not only clogs my connection regardless of upload/download speed so no one can browse the net, but are actually slower than most http/ftp downloads that support a few simultaneous connections.
Any reasonable client lets you control how much bandwidth it uses... it's up to you to know how to configure your client so it allows you to do other things while you torrent.
That's why rapidshare et all are so popular.
Most people can't afford as much upload capacity as RS has... that isn't cheap, you know. My argument is that torrents are more useful for the uploader of the files, not necessarily the downloader.
Besides, most people don't know what torrents are. You can't have them front and center to general users.
Why is that? Because of the negative rap that torrents get. Which is why the bullshit in the original AC's post is a troll. I'm refuting this point of view.
Do what thou wilt shall be the whole of the Law
The Chrome OS image is only a bit over 300 megs, so it's a fast download.
I'm on dial-up, you insensitive clod!
R.Mo
A very good point. I thought nothing of the impact the number of links on a page could have on navigation until I watched my girlfriend or my brother try to figure out which link was the one they needed when downloading software.
Do what thou wilt shall be the whole of the Law
Call me paranoiac! Call me antique! Tell me whatever you want, but THEY MUST BE OUT OF THEIR MINDS if they think I would leave ALL my stuff on THEIR SERVERS.
It might be faster than blinking, but I simply DO-NOT-LIKE the paradigm they're trying to spread.
It reminds me the "old" ATM machines, when a mainframe did all the processing. I guess I don't have to recall it was a bank who owned the mainframe and that you must pay them periodically.
I think the idea of avoiding the startup delay is really cool, but has a SMALL detail.. data is stored on GOOGLE servers, which means if Google powers down their servers you cannot access your data.
Tomorrow Google could say, "ok, since now you must pay to use our services.." And that's when you regret your decisions. I haven't mentioned the fact they can do whatever they want with the data in their servers (yeah.. yeah.. the data confidentiality agreement - i don't think so).
Nevertheless, I think it might be suitable for some people in some cases. Computers would require less hardware, which is a pro.
In summary, I like the idea of speed up the OS, but I think some stuff is private property and must remain as such (at least for my stuff).
try BugMeNot next time. (just looked on bugmenot for gdgt) Crap, gdgt is blocked from bugmenot. well, you can use this for other websites that you need to log into on a one-time basis
I don't know about you, but long battery life to me would be a killer app. I think that the standard six hours or less shows a peculiar lack of any progress. Sure, I can go to a coffee shop with my laptop. But I can't relax at a coffee shop with my laptop. How long will a smart-phone CPU with a notebook-sized battery last, I wonder?
I also consider a boot time of less than 10 seconds a killer app. The standard 45 seconds or more that even Windows XP (old) on my Core 2 Duo (new) gives me is baffling after 25 years of the PC. (Really, its more like two minutes before it is really ready to give me attention.) If my computer shuts down in two seconds and boots in three, l wouldn't plan my morning around it: "Time to make coffee --- no, wait, start the computer before you make coffee, then it will be ready at the same time."
Security is also a killer app. Encrypted home directory + read-only root + twin root partitions + a lot of other things = a lot more peace of mind. What if my laptop is stolen? Well, at least they're not going to find anything on it. My house guest is asking me if he can borrow my laptop. If it's a Windows laptop, I (but admittedly not the average user) will do a quick mental check --- do I have anything private on it that he might see? Is he going to accidentally download a virus on it? Etc. Sure, I can do things so that it will be less of a problem, but it's a lot easier if the computer already is set up as much as Chrome OS is for sharing.
Now that I look at them, what do these things all have in common? A less-stressed user experience. I don't have to think as much as I used to about taking care of my computer. Sure, it won't run Final Cut Pro. But I say, you should have made these the priorities --- at least with some --- any of your models. Get battery life, boot speed, and security to where you would have expected to be in the 21st century. Then branch out to fancy applications. Which is exactly what will probably happen. Browsers are only getting abler.
..which is some of the shittiest and most expensive pipe on the whole network, cannot be made as efficient as a conventional client/server setup).
Efficient in what sense? Certainly not time in release cases. If host A has a 100MB file and a 10MB/s connection, the most that host can upload that file at is 6/min. At 10,000 users this takes about 28 hours.
If the first torrent uploads to 2 people, and those two people to two more...etc it will quickly (4 hours) beat the single host case even if every pipe after the first is 1/100th the speed.
I've been waiting for the next platform shift. It's been moving towards cloud computing for about a decade, now, but aside from killing the client-server application, the Internet really hasn't caused any major change in platform.
We all still boot an O/S and run applications on the O/S, some of which are Internet-access applications. But it's struck me for some time that the browser really *should* be the next generation O/S. With plugins and all, Firefox is showing lots of signs, but it's just not stepping up to the plate - I guess the vision isn't quite there - the guys at Firefox still see the browser as a browser.
A decade ago, the idea of moving any kind of application "into the cloud" was a laughable concept that most people wouldn't dare touch. Nowadays, it's so common that perhaps 50% of all software development is now oriented around "cloud computing". I wouldn't be surprised if the number was even higher.
So Google's taking this trend to its logical conclusion: why bother with "local" at all?
It's an interesting take, and one that's sure to really upset the Winopoly if it's got any success at all. The flaws of the Winopoly are obvious and horrible - security woes too many to number, spam spewing from the many leaks, disks that crash, and an Operating System so big, complex, and cumbersome to work on that not even one of the wealthiest companies in the world can do much about it.
After investing untold billions into the Windows codebase, the result was Windows Vista/Windows 7, which is a bit prettier but certainly won't be introducing meaningful change. It might even be more secure, as much as something larger and more complicated is ever more secure than simpler, ancestral systems.
But Chromium takes us a whole new direction. My guess is that it *belongs* in a VM/application style software stack, where you can either run it alone on a netbook or something, or run it as a Win/Lin/OSX application. VMWare makes this a reality, even if it's never set up as an "application".
My guess? It's going to succeed, but in about 5 years' time. Google really needs to unify Chromium and Android. They should be virtually identical platforms. Microsoft is going the other way with IE - trying to pound the web, kicking and screaming, back into Windows proprietary extensions.
They *still* haven't figured it out...
I have no problem with your religion until you decide it's reason to deprive others of the truth.
First of all, it's not an OS, so please don't call it an OS. That term has an actual prior meaning that should not be hijacked in an attempt to sound geeky-cool.
Err.. you sure? If it isn't an OS, why do I need a virtual machine to run it? And I'm pretty sure the specs I've seen say it's based on a modified Linux kernel, which suggests to me that it's an OS.
My babysitter is a nursing student, [...] She doesn't need anything but a basic word processor...
I suspect you are mistaken. Either that or you have found the only student on the entire planet that has neither a digital camera, nor an ipod. Both of which need a real computer, unless iTunes has a ChromeOS version or she likes uploading 4GB of photos into the cloud at consumer DSL speeds.
She is a self-described techno-idiot, and loves the idea of a cheap computer with limited moving parts.
I do too. They are called netbooks, and they've been around for a few years already.
It will replace terminals (either traditional dumb terminals or Citrix) in call centers, at least at first. These things literally run one and only one application all day long.
The call centers that have reached this point are ALREADY booting off a network image of a locked down OS running that one app in kiosk mode. ChromeOS offers nothing new or compelling here, the PCs they are already using are cheap as dirt and don't have any disk drives.
Plus even call centers run 17" and 19" screens. The staff are staring at that screen all day, they aren't going to use $100 netbooks.
But I do think that the availability of the Google platform and ChromeOS may push applications that have in the past been PC based onto the cloud.
Sure that's generically true. But LOTS of applications can't move to 'the cloud'. That lathe calculation utility I mentioned will NEVER be on the 'cloud'. That VB6 front end to the enterprise SQL will never be on the cloud, etc, etc.
The real story here, though, is that whether Chrome OS wins or loses, the web has reached the point that Bill Gates feared ten years ago: it is now "the platform" for many apps. [...] It will replace desktop PC's...
No. It won't. Desktop PCs will evolve. Some of the 'stuff' that is a "Windows application" today, yes, will become a web app and run in a suitable browser. But some won't. Some can't.
What happens to AppleTV and iTunes store sales when you just stream your movies and music off Amazon when you want to watch them? This technology is already here.
Will amazon let my kids stream Mary Poppin's 40 times for a one time cost? Will they let them do it on a plane? A boat? A minivan? Will it work on a trip to Mexico? And what happens if Amazon goes under?
Seriously, streaming is set to kill the rental video. I agree with that. But replace buying movies? I don't see it. Today we can't stream 1080p, a bluray disc is better than anything you can get streamed. Sure 20 years we'll have more bandwidth... we'll also have ultra-hd. And I bet in 20 years you still won't be able stream 1080p to your laptop in a van in the middle of the rocky mountains.
Pirated? How does one pirate open source software?
Your math is flawed.
In the prefect case where everyone has 1MB/s upload it would in theory take 100 seconds for any number of people. This is because any one peer starts sharing what it's downloaded as soon as it has one single block of the file.
Of course blocks are not infinity small, people don't have 1MByte upload speeds and it's not a 'big start'.
The best models depend on the state of the swarm. During the initial seed then there is one slow seeder first order approximation is that everybody in the swarm is at the same percentage level. Transfer rate is limited by the upstream of that first seeder.
If the many of peers disappear once they have the file then the swarm is in a seeder starved state. The download rate of any one peer will be about the same as it's upload rate because of the 'tit for tat' like sharing rules in most clients when they aren't seeding.
If the swarm has lots of seeders then in addition to the 'tit for tat' rate a peer will get a 'fair share' of the total upload bandwidth of all the seeders. This is what can fill your downstream rate.
The vast majority of swarms are in the seeder starved state but at the moment the ChromeOS VMs are seeder rich; go for it.
Talk to the busybox developers. They have plenty of experience of being the victims of piracy. You pirate it by distributing it in a way that doesn't conform to the licence agreement you have for it. Generally by not releasing the source code and not letting people know they can distribute it themselves under the terms of the licence in question.
I got really excited until I realised that this wasn't for the DEC Alpha processor.
Good news everyone! If you have a CPU, you can run debian on it.
From the list of supported architectures: