Hard Light Productions Forums
Off-Topic Discussion => General Discussion => Topic started by: Bobboau on September 05, 2005, 01:51:47 am
-
what with Google's gmail, and now makeing an IM (http://www.google.com/talk/), and Goolge makeing it's own seperate internet (http://www.business2.com/b2/web/articles/print/0,17925,1093558,00.html), and mocrosoft planning to "fu (http://www.smh.com.au/news/technology/microsoft-ceo-im-going-to-fing-kill-google/2005/09/03/1125302772214.html)cking kill google (http://www.smh.com.au/news/technology/microsoft-ceo-im-going-to-fing-kill-google/2005/09/03/1125302772214.htm)"
how long is it intill Googlezon "devourer of networks" makes it's own opperateing system and totaly owns everything?
and might it be cool?
-
Believe it or not, this is a conversation my wife and I have regularly. Both of us are big fans of Google, but we come down on different sides of this question.
She's very pleased with all that Google does, and thinks that GooglOS will be the bestest thing ever.
I am vrey pleased with everything from Google (especially Google Earth, but I'd like to see Google Flickr), but I think they are really good at making individual things and extending them. I don't think I want to see it all incorporated into one big GooglOS.
-
Good to see I'm not the only one who thought of that mad idea...
Installing a bit of software is one thing. Changing over to another OS is another thing entirely. The majority of users would stick with their current OS because they don't know how to change it. The others are probably quite happy with their current OS.
So if Google were to make an OS, it'd only be used by about half a dozen people at the most.
-
I'm not so sure. Google does seem to make good products. I think the key feature that any OS they produced would have to have is the ability to not have it integrated to all things Google when it connects to the internet. That level of security to choose what applications you run is still very important indeed.
-
If Google were to develop their own OS, they most likely would use linux and use existing distros as an example. Developing a completely new OS from scratch would take immeasurable time and resources, not to mention it would require a full team of developers who know what they are doing.
If the GoogleOS would be based on linux, what really would make it different from the two dozen of other distros? I don't really see them developing an OS from scratch anyway. Them developing a new OS from BeOS is as unlikely as a completely new OS.
I believe Apple officially releasing their OSX to all x86_64 computers is far more likely to happen than GoogleOS.
-
With the existance of very good open source OS kernels out there I don't think Google really needs to do a new OS from ground up. If they want to do something of their own they can take an existent open source kernel and build on it (at the application level, ie to offer you a "eye candy" interface because it seems most people confuse the OS with this) and/or extend if kernel if needed.
-
Maybe.
-
I don't see the point in Google making an OS. Their strength is in applications and making the OS doesn't really help that. It's not as if Google needs to build special kernel features or low-level userspace apps.
If anything I think it's more likely that Google will create a web browser. If Google can control what browser technologies are used by the masses they can serve more interesting content. SVG, XHTML, CSS, Javascript all depend on browser support. Google's been big on AJAX solutions and it's far more beneficial to them if they control the client side than Microsoft.
-
Actually, what would be a logical next step for Google is to start providing free webspace to everyone, to host files or whatever. The more the better. I think that as time goes by people will naturally host more of their files on the Net for easier access, until one day hard drives will be at best a seldom used redundancy measure. Imagine getting 2 gigs webspace for free. That would just blow everyone else out of the water.
That, or they could always buy Ebay or Amazon. Or Mozilla.
-
I wouldn't use webspace for anything except redundancy. I doubt many people on this board would either. WAN speeds might be increasing but I doubt they'll ever come close to equalling the speed you can get from local storage.
-
EDIT: How the hell did I post in the wrong thread?! I apologize.
-
Wrong thread.
-
Hmm, whatever hurts Microsoft is good for 99.9% of us, regardless of whether you're on-side with the company or not. That article about Google poaching its top staff gave me a laugh. :D
-
I think Google's a great idea but it must be kept under control or "Googlezon" may just become a reality, I mean who wants another microsoft? Well now that I think of it that sounds kind of good. Maybe while Google and microsoft duke it out other smaller companies can emerge and add a little more competition to the market ie. linux.
-
GoogleOS sounds like AOL waiting to happen to be honest.
-
Originally posted by karajorma
I wouldn't use webspace for anything except redundancy. I doubt many people on this board would either. WAN speeds might be increasing but I doubt they'll ever come close to equalling the speed you can get from local storage.
Ever? Try to be a little more forward-looking that that. A mere decade ago, connecting to the Internet was a slow, ackward affairs where you modem would screech at you and take 5 minutes to load a single picture. I have little doubt that within twenty years at the most, probably a lot less, Net speeds will rival those of your local drive.
Besides, think of the advantages. Any game, movie or program would be instantly accessbile from anywhere, most likely free of charge. You could access your personal files from any computer, not just your own. I think it's more or less inevitable, it's just a matter of when.
-
Originally posted by Bobboau
AHHH! Googlezon is atacking!
In other words, I don't care.
-
You know, it's easy to go into the rebellious geek mode and say it'd be great to have another OS out there to rival Windows, but the reality of it is - the more OS' out there, the higher software development costs will go (esp. multimedia marketing tools) due to cross-platform development requirements.
This in turn increases the price of such projects, and thus starts to dampen the market.
-
Originally posted by Swantz
Maybe while Google and microsoft duke it out other smaller companies can emerge and add a little more competition to the market ie. linux.
Google was a small company not too long ago...
-
Originally posted by Rictor
Ever? Try to be a little more forward-looking that that. A mere decade ago, connecting to the Internet was a slow, ackward affairs where you modem would screech at you and take 5 minutes to load a single picture. I have little doubt that within twenty years at the most, probably a lot less, Net speeds will rival those of your local drive.
Really? And hard drive speeds would stay static for some magical reason? Local access will always be faster than remote access if only simply because of the fact that the server you're connecting to has to be doing it's own local access anyway.
Originally posted by Rictor
Besides, think of the advantages. Any game, movie or program would be instantly accessbile from anywhere, most likely free of charge. You could access your personal files from any computer, not just your own. I think it's more or less inevitable, it's just a matter of when.
meh. I think you're the one not thinking this through.
All my data instantly accessible. To everyone not just me. What about pivacy. Who else can access that data? The government are going to love your idea. Makes it incredibly easy to monitor everything since you've now given them one easy access point to everyones data.
What about hacking and viruses. I'm pretty safe from it cause who the hell is interested in data stored on my local drive. But one of these big google storage drives. Now that's an enormous target right there.
What happens when I've got no ISP. That's it. Computer's completely dead. All I've got is BIOS since if you're getting rid of HDs I see no reason whatsoever to keep CD and DVD drives after all.
And you're dreaming if you think it would stay free of charge forever. Maybe for a year or two to pull in the fools.
-
Correct on every point, kara.
[edit: I'm gonna go shorten my sig now...]
-
Originally posted by karajorma
Maybe for a year or two to pull in the fools.
Speaking of which, isn't that what Google Email BETA is all about?
I'd figure it was.
-
Originally posted by karajorma
Really? And hard drive speeds would stay static for some magical reason? Local access will always be faster than remote access if only simply because of the fact that the server you're connecting to has to be doing it's own local access anyway.
meh. I think you're the one not thinking this through.
All my data instantly accessible. To everyone not just me. What about pivacy. Who else can access that data? The government are going to love your idea. Makes it incredibly easy to monitor everything since you've now given them one easy access point to everyones data.
What about hacking and viruses. I'm pretty safe from it cause who the hell is interested in data stored on my local drive. But one of these big google storage drives. Now that's an enormous target right there.
What happens when I've got no ISP. That's it. Computer's completely dead. All I've got is BIOS since if you're getting rid of HDs I see no reason whatsoever to keep CD and DVD drives after all.
And you're dreaming if you think it would stay free of charge forever. Maybe for a year or two to pull in the fools.
No ISP? Do you really believe that the number of people accessing the Internet is going to decrease at any point? Almost one third of the world has a cell phone, which is a lot more than just the rich countries. Give it a few years, and Net access will be universal. I did say that pysical drives would serve as a redundancy, for those unable to access the Internet.
As for free of charge, look at the various filesharing program available today. I certainly didn't mean free of charge in a legitimate, authorized way, but you can't stop filesharing and widespread online storage would make it even easier. Hell, just look at how much BitTorrent has advanced filesharing, within a period of maybe 2-3 years. For example, I woke up this morning with a hankerin' for some TV shows. About two hours from now, I'll be watching both Real Time with Bill Maher and the new Rome. ;) ;)
Now the privacy issue I do agree with to some extent. But it stands to reason as people start moving more of their files online, which I think will be the case regardless of anyone's opinion on the matter, security would likely improve. A dedicated hacker could access almost any PC right now, but they don't. Because browsing through someone's pictures of their dog, or listening in on their MSN conversation with their aunt in Ohio just isn't worth it. What would change in an online storage system? The vast majority of data would still not be worth accessing, and supposedly those who need to protect themselves from intusion would find ways to do it, as they do now. Anything worth getting, like games or movies that are stored on a person e-drive would either be avaialble elsewhere, or would be freely shared by the person themself.
-
I just don't see the point to an online drive, for everything anyway. haveing the ability to easily share files and such, yes, would be hugely useful, but asside from the novelty I don't see why anyone would want to put ALL there stuff on a remote drive, do you realy think guys are going to put there porn colection up in an exposed place were anyone could concevably find it?
-
I can't see the utility of this. If I need to access my data, I can just use my computer directly. There would be no point in having the files any more accessible than that. The handful of things that I might need to use on another machine (my laptop, basically) fit easily on a USB flash drive.
-
Originally posted by Rictor
No ISP? Do you really believe that the number of people accessing the Internet is going to decrease at any point? Almost one third of the world has a cell phone, which is a lot more than just the rich countries. Give it a few years, and Net access will be universal. I did say that pysical drives would serve as a redundancy, for those unable to access the Internet.
Who said anything about net usage being reduced. I mean if you actually have no ISP because you move house, change ISPs etc. Now you don't just use net access you lose EVERYTHING on the internet share.
Originally posted by Rictor
As for free of charge, look at the various filesharing program available today. I certainly didn't mean free of charge in a legitimate, authorized way, but you can't stop filesharing and widespread online storage would make it even easier.
What the hell are you on about? P2P works cause it uses the hard drives of people online. You're saying that those HDs would be obsolete so how the hell are you going to store anything on them? You're making absolutely no sense whatsoever.
For your idea to work you'd need absolutely massive servers with gigantic amounts of storage. Who in the hell is going to pay for those?
Hell, just look at how much BitTorrent has advanced filesharing, within a period of maybe 2-3 years. For example, I woke up this morning with a hankerin' for some TV shows. About two hours from now, I'll be watching both Real Time with Bill Maher and the new Rome. ;) ;)
Originally posted by Rictor
Now the privacy issue I do agree with to some extent. But it stands to reason as people start moving more of their files online, which I think will be the case regardless of anyone's opinion on the matter, security would likely improve. A dedicated hacker could access almost any PC right now, but they don't. Because browsing through someone's pictures of their dog, or listening in on their MSN conversation with their aunt in Ohio just isn't worth it. What would change in an online storage system? The vast majority of data would still not be worth accessing, and supposedly those who need to protect themselves from intusion would find ways to do it, as they do now. Anything worth getting, like games or movies that are stored on a person e-drive would either be avaialble elsewhere, or would be freely shared by the person themself.
Accessing pictures of someone's dog isn't worth it to a hacker. Taking down a server and denying 30,000 people access to pictures of their dogs however is exactly the sort of thing that would appeal to people who like to write viruses. Although who needs to write a virus any more. No need to spread the information from one computer to another. If I hack the Google London server I can similtaniously infect the PCs of every single person connected to that. I can keystoke their bank records details when they enter them, I can overwrite their CMOS when they attempt to boot up, I can simply wipe the server and everybody loses everything.
And your having a laugh if you think that security would ever improve to the point where one person couldn't do that remotely. And even if he could you've not included the possibility of someone in the company f**king up or deliberately screwing everyone over.
And you still haven't dealt with the fact that it's slower than HD access, especially once you start including the need for encryption.
-
You know, I already have webspace that I can use as a local filesystem (through WebDAV). It's a pain to set up so I've just been using ssh access. With ssh there's already quite a bit of lag even though they're hosted in a blazing fast colocation center. So I don't see ubiquitous online storage coming any time soon.
BTW: Textdrive is a great webhost
-
I keep hearing the rumors and thoughts about a GoogleOS...it could happen.
What would be different about a Linux distro from Google would be a few things. Firstly, the interface would probably be very simple and the name Google has made for itself might allow people to switch.
Google would have to make Linux easy to switch in as many ways as possible. Easy to download or order the CD, easy to install, and easy to get rolling. Possibly toss in some search engine stuff and tada...
You could hit MS hard with something like that.
-
Originally posted by karajorma
Who said anything about net usage being reduced. I mean if you actually have no ISP because you move house, change ISPs etc. Now you don't just use net access you lose EVERYTHING on the internet share.
What the hell are you on about? P2P works cause it uses the hard drives of people online. You're saying that those HDs would be obsolete so how the hell are you going to store anything on them? You're making absolutely no sense whatsoever.
P2P works by making remote data accessible. Whether your data is stored locally or remotely is irrelevant to the person trying to access it, except that having it stored remotely (from your point of view) would probably indirectly serve to make it easier to get to (from their point of view) because of the necessary advancement of remote-acess technologies.
As for ISPs, the future is WiFi. I'de be surprised if within maybe 15 years or so you could find a single point on the planet not recieving a signal.
Originally posted by karajorma
For your idea to work you'd need absolutely massive servers with gigantic amounts of storage. Who in the hell is going to pay for those?
Not necessarily. Who's to say that this theoretical e-drive would be free? You pay for you physical drive, let's say about $100 for 3 years of service. With a $10 monthly service charge, which doesn't seem to high to most people, the storage and access could easily be handled. If Google can offer 2 gigs of email storage space to anyone who wants it (though most people will never fill it up) then why can't they offer several gigs of general purpose storage space a few years down the road?
Originally posted by karajorma
Accessing pictures of someone's dog isn't worth it to a hacker. Taking down a server and denying 30,000 people access to pictures of their dogs however is exactly the sort of thing that would appeal to people who like to write viruses. Although who needs to write a virus any more. No need to spread the information from one computer to another. If I hack the Google London server I can similtaniously infect the PCs of every single person connected to that. I can keystoke their bank records details when they enter them, I can overwrite their CMOS when they attempt to boot up, I can simply wipe the server and everybody loses everything.
And your having a laugh if you think that security would ever improve to the point where one person couldn't do that remotely. And even if he could you've not included the possibility of someone in the company f**king up or deliberately screwing everyone over.
And you still haven't dealt with the fact that it's slower than HD access, especially once you start including the need for encryption.
Alright, I admit that remote access will always, as far as I can see, be slower than local access. But that's not to say that remote access will not be fast enough to accomodate the need, as fast as today's local drives.
About security, the don't accept the arguement that something is not worth doing because the solution to make it function flawlessly is not available yet. I don't see why a remote drive system would be any more vunerable than local drives are at present.
The main point is that, from what I see anyway, mobility is the big trend. That's just the way things are going. Laptops, cellphones, and newer things like WiFi, Internet and TV access over cellphone bla bla; that's all indicative of the desire to have all the functionality of your living room in your pocket. To me, even laptops are much too unwieldy to be truly mobile, to say nothing of desktops. One solution would be to increase HD storage space while making them smaller, which will certainly happen, don't get me wrong. I'm looking forward to having the functionality of an iPod in a cellphone within a few years. But at some point, people will figure that there is simply too much data out there, good stuff that you want to get your hands on at any point, that storing everything locally is redundant and inefficient.
I could be wrong, but I think that it fits quite nicely with the overall trends which have been present since home computer were first inveneted, and communication technologies more generally.
Quite coincidentally, I found this ("http://www.newscientist.com/article.ns?id=dn7943") little article today.
-
But the thing is, most of that stuff isn't really mobile. My desktop isn't meant to be used out-of-doors, nor is its entire premise designed for anything remotely like that. Not to mention, not having a hard drive would muck up your system even if we were able to hypothetically netboot everything. Memory management, and more specifically virtual memory, MANDATES that you have as high-speed local (non-volitile) storage as possible. It's one thing to have an online repository for the junk one wants to store over time, but it's quite another to talk about running from there exclusively. Also worth noting, I do have experience with netbooting, and even over a very fast (gigabit) LAN it's still absolutely crap when compared to loading off of local storage. Running apps remotely - especially when that app then accesses data on whatever drive its stored on - is even worse. Nonvolatile memory solutions will change in the long run, yes. Otherwise we'd still be using tape decks. But it's not going anywhere.
EDIT: Actually, if we're supposed to be extrapolating current trends...
Originally posted by karajorma
Hell, just look at how much BitTorrent has advanced filesharing, within a period of maybe 2-3 years. For example, I woke up this morning with a hankerin' for some TV shows. About two hours from now, I'll be watching both Real Time with Bill Maher and the new Rome. ;) ;)
Distributed networks FTW. Rather than your hard-drive moving online, I'd actually see parts of the web infiltrating our PCs.
EDIT2: actually, no matter how you do it, netbooting and remote file systems always have the same bottleneck: the server. It's actually not the speed of the connection that's most important, but rather the amount of time each individual connection gets allotted in a given server's client queue and that's just never going to get fast enough to be useful on a large scale.
-
Originally posted by StratComm
Distributed networks FTW. Rather than your hard-drive moving online, I'd actually see parts of the web infiltrating our PCs.
Right. Centralised data storage is not where technology is going. There used to be things called NCs (actually, they're still used) which had no local storage and were basically dumb terminals.
Why do most businesses use PCs? Because if the network goes down, all the NCs go down too and no one gets any work done. No so with PCs.
Centralised storage has one major disadvantage: if it goes down, all the data stored on it becomes inaccessible.
Redundant data storage is inefficient, is it? What's a RAID array for, then? What's the point of backups?
The way technology is going is this:
1) Distributed redundant storage, using a system such as BitTorrent to retrieve a particular piece of data.
2) Servers acting as switchboxes, temporary storage and indexes rather than actual data storage.
3) Increasing the rate at which data can be accessed remotely, without relying on bandwidth, using methods such as caching, multiply-redundant storage, etc.
Finally, local access will always be many times faster than remote access. Most hard drives can transfer data at about 90MB/sec. SATA can deliver up to 300MB/sec (theoretical), which is already faster than a 3-disk RAID-0 array.
Contrast that with Gigabit ethernet. How often does that come even near theoretical speeds? Cat6 cable may be capable of those speeds when it's still in the carton, but once it's been bent around corners, run alongside power cables and cabletied to ducting its signal-carrying abilities have decreased. Furthermore, even 1Gb/sec is only 125MB/sec, half the speed of SATA. And SATA only provides for one node: the PC it's in. Gigabit backbones have to cope with traffic for many users, decreasing speed even more. Assuming a 2Gb/sec connection to a village, that's probably 600 people making use of a single connection. Including network overhead you're looking at about 1Mb/sec each, or 125KB/sec. Pathetic.
Face it: remote access is not going to be faster than local access, period. The logistics alone make it impossible. Technology can only go so far towards mitigating that.
-
Originally posted by Rictor
P2P works by making remote data accessible. Whether your data is stored locally or remotely is irrelevant to the person trying to access it, except that having it stored remotely (from your point of view) would probably indirectly serve to make it easier to get to (from their point of view) because of the necessary advancement of remote-acess technologies.
I know how P2P works but you can't distribute your boot files unless you tie everyone down to using a very limited number of OS configurations and I don't see that happening any time in the near future.
Originally posted by Rictor
Not necessarily. Who's to say that this theoretical e-drive would be free?
You did.
Besides, think of the advantages. Any game, movie or program would be instantly accessbile from anywhere, most likely free of charge.
That was one of the two things I had the biggest problem with in your original comment. First that it would replace HDs as a form of primary storage and second that it would be free. At least you've finally conceeded the point that there aren't people out there willing to give up the stupidly huge amounts of bandwidth that this plan would require.
Originally posted by Rictor
You pay for you physical drive, let's say about $100 for 3 years of service. With a $10 monthly service charge, which doesn't seem to high to most people, the storage and access could easily be handled. If Google can offer 2 gigs of email storage space to anyone who wants it (though most people will never fill it up) then why can't they offer several gigs of general purpose storage space a few years down the road?
You've answered your own question. Most people don't use all 2gigs. Google gave out a stupendous amount of storage space knowing that they couldn't afford it if everyone used it all. What they knew of course was that very few people would ever use up that much space rapidly. By the time any average person even comes close HDs will have increased in size to be able to handle that
Your plan is several orders of magnitude more ambitious though. 2GB of storage! If you're using Windows you couldn't even run the OS and a couple of applications in that! You're going to need an enormous amount more than Google ever gave out (and remember. Google knew people wouldn't use all that space and took that into account. For this plan they will have to take that into account).
But it gets worse. That's simply storage we're talking about. What about access? People with google accounts may store stuff on there but they generally download it to their PC. Most interaction with the server is just someone looking at a HTML e-mail with a few pictures in it. That's a few hundred kilobytes. And most people only check their e-mail once or twice a day.
To netboot we're talking a staggeringly larger amount. Windows on average takes up a few hundred megs even if you don't do anything with it. All this data has to be cut up into packets, transmitted to the client PC, verified, sent copies of packets that didn't reach the client and finally used. That's an absolutely enormous amount of data and all you've done is switch on and boot up. If you want to play Doom you're going to have to send over several hundred meg more. Google aren't set up for even 1/1000th of that.
On top of that you're going to have ludicrously big spikes to deal with for instance at around 9am as everyone switches their PCs on. Google may have a 9am spike but I doubt it's anything like the magnitude of the bandwidth spike you'd get in the morning from everyone switrching on.
Originally posted by Rictor
Alright, I admit that remote access will always, as far as I can see, be slower than local access. But that's not to say that remote access will not be fast enough to accomodate the need, as fast as today's local drives.
So? Who cares about the speed of todays drives? If we're talking about 10-15 years in the future you have to look at how fast HDs will be then. You're assuming that the size of games, applications and media will remain the same and I see absolutely no reason to believe that.
What's the point in having a networked PC that can do 150MB/sec transfers if games 15 years in the future are Terabytes in size?
Originally posted by Rictor
About security, the don't accept the arguement that something is not worth doing because the solution to make it function flawlessly is not available yet. I don't see why a remote drive system would be any more vunerable than local drives are at present.
Then you're deluding yourself. I've already explained why a networked drive is a more tempting target than local storage several times and you keep missing the point. Why bother with phishing scams or viruses when you can simply hack a single computer and access the data for several thousand people at once. It's an enormous target and the first people to make one would have to make it unhackable cause if they didn't who would be willing to use on after they saw the disaster it caused for the first people to get hacked?
Originally posted by Rictor
The main point is that, from what I see anyway, mobility is the big trend. That's just the way things are going. Laptops, cellphones, and newer things like WiFi, Internet and TV access over cellphone bla bla; that's all indicative of the desire to have all the functionality of your living room in your pocket. To me, even laptops are much too unwieldy to be truly mobile, to say nothing of desktops. One solution would be to increase HD storage space while making them smaller, which will certainly happen, don't get me wrong. I'm looking forward to having the functionality of an iPod in a cellphone within a few years. But at some point, people will figure that there is simply too much data out there, good stuff that you want to get your hands on at any point, that storing everything locally is redundant and inefficient.
Just because people want something doesn't mean it happens. The idea is impractical. It simply could not work and no amount of people wanting it is going to change that.
Originally posted by Rictor
I could be wrong, but I think that it fits quite nicely with the overall trends which have been present since home computer were first inveneted, and communication technologies more generally.
You're wrong. The entire concept of the home computer is a move away from the dumb terminal which is what your idea is moving us back towards.
-
Windows is extremely bloated. But the solution of keeping space down probably won't make you any happier - have the 'client' PC send the server its hardware, and the server PC sends only the drivers the client PC needs, along with the core operating system.
That would cut down a hell of a lot on the space the OS takes up, and make it not a real hassle if you wanted to add hardware.
-
What about this?
http://digi.it.sohu.com/20050904/n240337911.shtml
http://www.theinquirer.net/?article=25941
The pictures look nice, could be a lie of course but it's funny to find out that my suposition (to be based on an existent open source kernel) is confirmed in a way :)
-
Google OS would work wonders if it was compatable with all win/mac/anything else software.
-
Looks like someone's skinned XP or something similar.
-
That's what I was thinking. It doesn't look anything special at all... infact the overall design is terrible. I don't think that's from Google.
-
One of the math professors at my university is the father of one of the Google founders. Maybe I can ask him what the point of this thing is. :D
-
Er, the pics on that site make it look to fall more in the "pr0n" than "Reliable news source" category.