At last count (a couple 2-4 months back?) the $100 laptop was able to be made at ~$135-140. Large production, further refinement of production techniques and the improvement of technology will bring that down. It would be a miracle if they could actually make it a "$100 laptop", but there real goal is to get as close as possible to the "$0 laptop" over time.
Since the specific locations have never really been specified, AFAIK, everything is just speculation... The most logical reason for Google, a company with multiple massive datacenters that store the same information would be for them to... keep their data in sync!
It makes much more sense that Google has their own fiber to connect most/all of their datacenters together for fast syncing of newly added information (web crawl, new service rollouts, etc, etc).
Or, you know, support Multicasting so that you can have N viewers of a live webcast only only have the bandwidth of 1 user across the backbone and core Internet.
My understanding is that this (very accurate) point is academic since the vast majority of ISPs don't support multicasting.
Multicasting appears to be the ideal solution to base any significant streaming or multi-source downloading technology (with the required upper-layer error-checking to compensate for UDP).
If the TFA is believed to be true, this shouldn't be a problem. It says the Apple will add QOS to the networking stack, so that background AppleTorrenting would not hurt your current download speeds or latency.
Further, with this in somewhat widespread use, the ISP's would save on upstream costs because you wouldn't be hitting the Apple (or Akamia, in some cases) servers that are likely on the other side of a backbone pipe from you. Rather, your download would likely be MUCH more geographically close and that bandwidth is cheap. ISPs would be grateful if their upstream bandwidth got cut by a few percent in exchange for their switches and internal routers to take up the load.
While the first 3 octets would tell you the vendor, it won't be the address you want. MAC address are changed per-hop. So the MACs this NTP server would be getting are either 1. his local router or 2. other machines on the same subnet. Only the IP address stays consistent from end-to-end.
The OSL is the Open Source Lab ran out of Oregon State University. Among other things, it provides hosting to open source software. Things like Gentoo, Firefox and Linux, for example.
Re:Like paying extra to fix leaks in brand new roo
on
No Anti-Virus in Vista
·
· Score: 1
And if they had included Anti-virus software, you would be complaining about the anti-competitive actions that will push Norton and Symantec (and Grisoft, and, and, and...) out of business.
You get 1 TB (yea, Terabyre, 1000 gigs) of transfer.
On top of which, the added bonus of bookmark, address book, and other first-party application integration are either not available, or taking some light/moderate hacking to setup all the services for your own hosted server to mimic.
100 bucks seems much more reasonable, esspecially if you do use those first-party syncing services, if not, you are still getting a pretty good deal.
(and I'm sure they've got someone working on it with their 20% free time).
The 20% idea is for, like you said, free time. I think you are right, that they will work on some sort of spam control for third-party server's, but it will be an 80% project, something the company assigns, not a free time project.
Though, having a preference that says a user will only accept (or be bothered with requests from) user's already in their buddy list would seem to solve this for the most part. These settings have been around for years (on AIM, atleast, where you can have only Buddy List members IM you, as a form of spam/parental control).
Re:There's also the itsy bitsy license change...
on
Nessus 3.0 Released
·
· Score: 3, Informative
Yea, they do actually. It is a revenue source for slashdot, paid stories. No kidding.
the 500MHz AMD processor is a custom building, using something like.5 watts of power (the system is expected to use 2-3 watts total) as this system is meant to have very low battery usage.
the 1 GB is for flash memory, used in place of a harddrive for nonvolatile storage. There will be 128 MB of RAM for the actual running system. 4 USB ports, which may change. Wi-fi, and some form of cell phone wireless connection, i believe. They support some customized ad hoc wireless mesh networking system, as well.
The laptop using up to 4 C batteries (believed to be natively available in every country in the world), or using a handcrank. Which both goes back to the very low power usage. The crank is expected to get 1:10 ratio (1 minute crank, 10 minutes of battery). In eBook + black & white mode, they expect to get up to 1:45 ratio. Perfect for cuddling up with your favorite electronic (text)book for school or play.
Redhat is using a customized Redhat Linux (Fedora, presumably) made to be fully and specifically supporting the hardware capabilities of the laptop. The entire laptop is IP free, all the technological developments are going to be 'open sourced', so no Intellectual Property, Trademarks, etc. They are inviting other manufacturers to learn from the $100 laptop's advances in efficiency.
These things will be hacked on faster than the original Xbox. It will be cheap, ubiquitious, reliable and natively run Linux.
"-Apple is no longer what they were when they started out, and now their proprietary software sucks, the 3rd party Apple MAC software is great, and the Apple Macintosh software is great -Open source is good for companies that would like it, but Apple software is still better"
So not to nitpick, but MAC is either a Mandatory Access Control, or a MAC address, used for networking. A Mac, is a brand of computers (with accompanying OS, yada yada) made by Apple computers.
"their [Apple's] proprietary software sucks... and the Apple Macintosh software is great."
Seems to contradict themselves, so I am not sure what you meant.
"Oh and one more thing. What exactly did you mean that half the internet can be taken down with a backhoe?"
Many 'fat' internet connections share a single tunnel. Long haul fiber outages and what not can have a huge sweeping blow to thousands of websites if properly planned. Yes, there are redundant links, but if you cause a large enough chunk of traffic to be routed through alternate paths, you will cause those paths to get flooded and DOS not only the originally effected sites, but also the sites that WERE using the alternate paths but now have to share.
It isn't a single backhoe, though I don't doubt certain peering points could be FUBARed with a single snip, that we should be worried about. But a small coordinated attack on a couple (2, 3?) peering points, well planned, and you could take down much of the internet in a chain reaction.
they are looking into plans to sell it to the general public for around $200, but sell to governments and large education entities (in bulk) for the famous $100 price tag.
No. Zero-day doesn't depend on a patch being released. Zero-days are exploits that are unknown to the publisher, as in "Microsoft has known about exploit XYZ for zero days." This is not the case, Microsoft has known about the hole. So it is just an exploit, nothing too terribly special about it beyond being able to ram a large rod up Window's backside if you view a rogue website in IE.
1. Microsoft promising something 18 months down the road is meaningless. Hell, ANY tech company promising something 18 montsh out is meaningless.
Apple will ship Intel macs and finish a transition within 18 months. They are by all accounts on and or ahead of schedule. Timetables in tech aren't meaningless. rediculous timetables are.
Other than that, I agree. They don't have to stay with any 'open format' they use, nor do they have to let the entire world enjoy it(Europe vs. US, in this case, or other Software Patent-friendly countries). We shouldn't be swayed by this distraction of half attempt. Use a complete, fully open, fully usable standard in OpenDocument now, not the promise of a partially openned, patent protected, format 18 months from now.
At last count (a couple 2-4 months back?) the $100 laptop was able to be made at ~$135-140. Large production, further refinement of production techniques and the improvement of technology will bring that down. It would be a miracle if they could actually make it a "$100 laptop", but there real goal is to get as close as possible to the "$0 laptop" over time.
Since the specific locations have never really been specified, AFAIK, everything is just speculation... The most logical reason for Google, a company with multiple massive datacenters that store the same information would be for them to... keep their data in sync!
It makes much more sense that Google has their own fiber to connect most/all of their datacenters together for fast syncing of newly added information (web crawl, new service rollouts, etc, etc).
Or, you know, support Multicasting so that you can have N viewers of a live webcast only only have the bandwidth of 1 user across the backbone and core Internet.
My understanding is that this (very accurate) point is academic since the vast majority of ISPs don't support multicasting.
Multicasting appears to be the ideal solution to base any significant streaming or multi-source downloading technology (with the required upper-layer error-checking to compensate for UDP).
If the TFA is believed to be true, this shouldn't be a problem. It says the Apple will add QOS to the networking stack, so that background AppleTorrenting would not hurt your current download speeds or latency.
Further, with this in somewhat widespread use, the ISP's would save on upstream costs because you wouldn't be hitting the Apple (or Akamia, in some cases) servers that are likely on the other side of a backbone pipe from you. Rather, your download would likely be MUCH more geographically close and that bandwidth is cheap. ISPs would be grateful if their upstream bandwidth got cut by a few percent in exchange for their switches and internal routers to take up the load.
While the first 3 octets would tell you the vendor, it won't be the address you want. MAC address are changed per-hop. So the MACs this NTP server would be getting are either 1. his local router or 2. other machines on the same subnet. Only the IP address stays consistent from end-to-end.
The OSL is the Open Source Lab ran out of Oregon State University. Among other things, it provides hosting to open source software. Things like Gentoo, Firefox and Linux, for example.
And if they had included Anti-virus software, you would be complaining about the anti-competitive actions that will push Norton and Symantec (and Grisoft, and, and, and...) out of business.
You get 1 TB (yea, Terabyre, 1000 gigs) of transfer.
On top of which, the added bonus of bookmark, address book, and other first-party application integration are either not available, or taking some light/moderate hacking to setup all the services for your own hosted server to mimic.
100 bucks seems much more reasonable, esspecially if you do use those first-party syncing services, if not, you are still getting a pretty good deal.
Part of why they are a Sun shop is that Sun systems (in their experience) have longer lifecycles, thus 6 years of use.
Servers cost money, and getting more bange for your buck is a *good* thing.
So they happen to sell adult movies on DVD. People spend big money on that.... so I hear.
That is one way to look at it.
Ok, so it doesn't ship anymore, but there USE to be an IE for UNIX, which is pretty close.
The 20% idea is for, like you said, free time. I think you are right, that they will work on some sort of spam control for third-party server's, but it will be an 80% project, something the company assigns, not a free time project.
Though, having a preference that says a user will only accept (or be bothered with requests from) user's already in their buddy list would seem to solve this for the most part. These settings have been around for years (on AIM, atleast, where you can have only Buddy List members IM you, as a form of spam/parental control).
Yea, they do actually. It is a revenue source for slashdot, paid stories. No kidding.
Needs to be clarified:
.5 watts of power (the system is expected to use 2-3 watts total) as this system is meant to have very low battery usage.
the 500MHz AMD processor is a custom building, using something like
the 1 GB is for flash memory, used in place of a harddrive for nonvolatile storage. There will be 128 MB of RAM for the actual running system. 4 USB ports, which may change. Wi-fi, and some form of cell phone wireless connection, i believe. They support some customized ad hoc wireless mesh networking system, as well.
The laptop using up to 4 C batteries (believed to be natively available in every country in the world), or using a handcrank. Which both goes back to the very low power usage. The crank is expected to get 1:10 ratio (1 minute crank, 10 minutes of battery). In eBook + black & white mode, they expect to get up to 1:45 ratio. Perfect for cuddling up with your favorite electronic (text)book for school or play.
Redhat is using a customized Redhat Linux (Fedora, presumably) made to be fully and specifically supporting the hardware capabilities of the laptop. The entire laptop is IP free, all the technological developments are going to be 'open sourced', so no Intellectual Property, Trademarks, etc. They are inviting other manufacturers to learn from the $100 laptop's advances in efficiency.
These things will be hacked on faster than the original Xbox. It will be cheap, ubiquitious, reliable and natively run Linux.
"-Apple is no longer what they were when they started out, and now their proprietary software sucks, the 3rd party Apple MAC software is great, and the Apple Macintosh software is great
... and the Apple Macintosh software is great."
-Open source is good for companies that would like it, but Apple software is still better"
So not to nitpick, but MAC is either a Mandatory Access Control, or a MAC address, used for networking. A Mac, is a brand of computers (with accompanying OS, yada yada) made by Apple computers.
"their [Apple's] proprietary software sucks
Seems to contradict themselves, so I am not sure what you meant.
You keep using that word. I do not thing it means what you think it means.
Many 'fat' internet connections share a single tunnel. Long haul fiber outages and what not can have a huge sweeping blow to thousands of websites if properly planned. Yes, there are redundant links, but if you cause a large enough chunk of traffic to be routed through alternate paths, you will cause those paths to get flooded and DOS not only the originally effected sites, but also the sites that WERE using the alternate paths but now have to share.
It isn't a single backhoe, though I don't doubt certain peering points could be FUBARed with a single snip, that we should be worried about. But a small coordinated attack on a couple (2, 3?) peering points, well planned, and you could take down much of the internet in a chain reaction.
they are looking into plans to sell it to the general public for around $200, but sell to governments and large education entities (in bulk) for the famous $100 price tag.
No. Zero-day doesn't depend on a patch being released. Zero-days are exploits that are unknown to the publisher, as in "Microsoft has known about exploit XYZ for zero days." This is not the case, Microsoft has known about the hole. So it is just an exploit, nothing too terribly special about it beyond being able to ram a large rod up Window's backside if you view a rogue website in IE.
1. Microsoft promising something 18 months down the road is meaningless. Hell, ANY tech company promising something 18 montsh out is meaningless.
Apple will ship Intel macs and finish a transition within 18 months. They are by all accounts on and or ahead of schedule. Timetables in tech aren't meaningless. rediculous timetables are.
Other than that, I agree. They don't have to stay with any 'open format' they use, nor do they have to let the entire world enjoy it(Europe vs. US, in this case, or other Software Patent-friendly countries). We shouldn't be swayed by this distraction of half attempt. Use a complete, fully open, fully usable standard in OpenDocument now, not the promise of a partially openned, patent protected, format 18 months from now.
All the more reason for informed geeks to push its adoption and not wait for the big corproations to do it.
Why not just used the bayesian spam filters and controls in email clients and apply it to newsreaders?
You forgot the most important one!