They also have a fixed hardware platform - not so with a colocation datacenter where the operator is going to need to accommodate a wide mixture of equipment from unrelated vendors that customers bring in.
One thing I found interesting that seems to be popular with new facilities like this one is omitting the clean agent fire suppression systems that used to be all the rage. Specifically it says:
4.10 Fire Alarm and Protection System
Pre-action fire sprinkler system uses nitrogen gas in lieu of compressed air to eliminate pipe corrosion.
Online nitrogen generator.
A VESDA air sampling system is provided for early detection for fire/smoke detection.
For those who don't know, a pre-action system is a water system except that the pipes are charged with compressed air (on in this case, nitrogen) with an interlock system that requires smoke detection followed by loss of pressurization before releasing water. The benefit is preventing accidental discharges; if someone breaks a head without the smoke detection, the loss of pressure will be seen as a trouble event and water will not be released into the pipes. If you get a smoke detect event and *then* a head fuses, water is released. It's all still a fancy water system though.
What I noticed is the guy's hands were covered in oil, and when he opens the lid of the server it's resting against his shirt, likely covering it in oil as well. This kind of idea seems like a unholy mess: oil everywhere. I rarely remove my servers once racked, but I do occasionally plug a USB drive or put a DVD in them.
Oil immersed servers are probably likely to end up in the trash straight away rather than end up on the used market, donated, or reused elsewhere.
Some things require this. 802.1x with PEAP and MSCHAPv2 is one that comes to mind. When forced to do this, just run the LDAP server alone on its own server/VM so that nothing can get to it.
- set up sudo... to allow everyone to do everything
I do this (on a per-user basis) because when I do a root action I want it to be intentional. I normally work as an unprivileged user and perform the few required root actions via sudo. There's no point in restricting sudo commands to anyone who already knows the root password, but I believe it is a good habit to work as a user, not as root. (For people who don't know the root password, sure, restrict away to only commands they need to do their job.)
VLANs are an ethernet (or equivalent) layer concept. A specific VLAN can have an arbitrary number (including zero) of IP4 or IP6 networks on it.
That's not really relevant; the example was just to show how one can use IPv6 addressing to create a shorter address. If you can remember your prefix, something tangentially related (like a switchport access VLAN number), you can probably come up with an IPv6 addressing scheme that is easy to remember.
You don't have to make long addresses if you don't want to. You can drop leading zeros and the:: compression replaces any range of zeros, not only one set. So a prefix you might get from your ISP becomes:
2001:DB8:A::/48
I can remember that easily and then make up a plan such as "/64 corresponds to VLAN". Say you have VLAN 5 and a statically assigned host 9 on that VLAN.
2001:DB8:A:5::9/64
Although it still has scary A-F in the number. Or you can stick with the crazy long addresses if that's easier.
If someone can't afford to buy a router that's hundreds of dollars, at least look at MikroTik (routerboard) hardware. Similar price range without the brain dead functionality of the typical D-Link or Linksys.
Thats not a normal person solution. I have boxee on a home media server, used to do mythtv but I moved out of a place where I had provided cable and without cable it wasn't worth the hassle. Its certainly possible, and its great once its set up, but it requires active maintenance.
My personal desires are for a single, low-power, easy-setup box that can: - Stream from Netflix - Stream from Hulu/Hulu Plus (to be legit it probably requires hulu plus) - Stream from Pandora
The Roku does all of those, and more (like Amazon video).
- Run local/LAN-shared video and audio with good codec support
Roku has channels that say it will do LAN streaming, but I haven't tried because I don't really care to use it for that.
- Extensible to help future proof it (i.e. easy to integrate some new streaming service)
Roku has an open SDK, assuming you or someone else writes a channel for said new streaming service.
I've had mine since the early days when it was "the Netflix box" and they've released a whole lot of firmware upgrades since then. It was most certainly worth the $99 I spent on it, especially since its functionally continued to increase with each update. The only thing it can't do that the current hardware can is 1080p (mine does 720p).
On the other hand, Stallman brings up worthwhile points. You may lose certain legal rights -- in the USA, for example, you may lose your 4th amendment rights. You do not have control over web applications -- the provider can change things, yank out features or add new features you do not want, and you have no recourse (how many times has Facebook done this?). You may even lose your access entirely.
Likewise with the Facebook example, we can see that such changes do not cause the users to seek alternatives, either out of apathy or because there is none.
Why is this amazing "that the controller works at all"? There was a time before microprocessors, you know, and they did fun things like travel in space without them.
If redirecting NXDOMAIN to partnered search results pages
VeriSign != ICANN
And why didn't ICANN start the process of "firing" VeriSign immediately after the incident?
IETF, IANA, the RIRs, etc. all seem to work well without to have some legal entity with a bunch of corporate bullshit. The suits have managed to royally fuck things up at ICANN.
I agree with the GP: money has generally corrupted the process which should be a simply technical matter of updating a simple list of TLDs as countries have come and gone according to ISO 3166-1 alpha-2. The bureaucracy has started to serve itself.
ICANN did assert that they overstepped their authority, and VeriSigned later sued ICANN.
The property manager for the office park I'm in let us put Ethernet switches on the buildings so we could offer fast symmetric speeds like our non-US friends enjoy for a fraction of the average cost for cable or DSL. It was interesting for a few reasons:
* The people that signed up were amazed that such speeds were possible; they had been stuck with DSL, cable, or mobile broadband that topped out at 1.5Mbps downstream on a good day. * They were even more surprised that it was symmetric (the most expensive cable package was only 3Mbps up) and included a subnet standard. * One of the tenants accused us of trying to scam them with a service that could not be possible for a company of our size.
So I'd say that the expectation of the average US consumer is to pay a high price for slow service from the local big cable company or big telco, then pay extra for add-ons like a static IP. Some of them won't even believe you or accuse you of lying if you go outside of that expectation.
Basically, they charge $200 because they can. It has nothing to do with the actual cost of service.
Correct; it is the price a residential customer will bear for what they feel is "advanced" pricing for a "fast" speed the customer probably doesn't fully grasp. These customers are also willing to pay for truck rolls, play "reboot the router" games with scripted tech support, and wait until the next business day for someone to fix it if it breaks. Many business class accounts (like cable) are the same quality as residential except the pricing is higher and you're allowed to pay extra for a static IP; that's how Charter does it where I am.
When you move into the land of T1, T3 and OC-x then you're dealing with a whole different class of customer. One that won't accept anything less than the technical capacity of the line, a comprehensive SLA, a first line tech that doesn't read a script, and real 24x7 service should there be trouble or an outage. Other factors are terminating your line to a much more expensive router that does lots of BGP rather than some cheap thing that only does static routes with a default. This customer is also willing to pay for it. One can argue the "bandwidth" should still be cheap, and it is, but the human aspect is certainly not.
Ah, the Reno 911 AVR. It was applied to cell phone calls specifically because there was an unmanageable amount of pocket dialing, and our dispatch center is not super huge. I called 911 once when the system was in place and either you dialed * or supposedly it would connect the call anyway if it heard "loud noises". I believe it's since been removed, or at least the last time I called in concrete chunks on the freeway from my cell I was connected to a dispatcher immediately.
Lets be honest though. Would you really want to do excel sheets on a tablet? I mean, to bring an acceptable level of excel functionality to a tablet you'd need an insane amount of gestures/menus to compensate for the desktop interface. By that point it doesn't seem worthwhile. How many unrecognized gestures would it take before you hurled the thing across the room?
Probably not beyond simple input. I can see it being great for generating quick estimates, quotes, or something along those lines using a spreadsheet created on a computer.
I have one from Sprint at the office. After arguing that I might as well cancel since it's not my problem and I don't want to pay for their coverage hole, they sent me one for free. It has its bugs, but it works more often than no signal at all.
Except to compete with google, you need not only as good or better algorythms, as good or better other services (e-mail etc). you need as good or better database too.
And you have to offer it for free: that's the expectation now.
Also, if you consciously try to keep your code as efficient as possible and run it on something that isn't abstracted 10 frameworks deep, scalability suddenly becomes much less of a problem.
Now-- Stargate Universe- I'm liking so far. There are some dumb things (like why they are not training the doctor during some of her at home time- but at least they addressed that by bringing in specialists). seems like they should be working on a forge and a machine shop too. And at some points the guns have got to start breaking and running out of ammunition. They should find/use some energy weapons.
The thing that's been bothering me about SG:U is that they are on this massive ship and don't seem to be exploring it much outside the tiny area they occupy. The ship likely has manufacturing facilities onboard and a full hospital with diagnostic equipment. It was intended to be ready for the arrival of a large crew for long term travel, yes? So it goes to reason everything they need is already there, including repairs to the ship. And how could only Rush know the ship has a bridge? Even the advanced Atlantis era ships that could be run by one person still had a bridge, and they've brought up maps on consoles before.
Slashdot Mirror
They also have a fixed hardware platform - not so with a colocation datacenter where the operator is going to need to accommodate a wide mixture of equipment from unrelated vendors that customers bring in.
One thing I found interesting that seems to be popular with new facilities like this one is omitting the clean agent fire suppression systems that used to be all the rage. Specifically it says:
4.10 Fire Alarm and Protection System
Pre-action fire sprinkler system uses nitrogen gas in lieu of compressed air to eliminate pipe corrosion.
Online nitrogen generator.
A VESDA air sampling system is provided for early detection for fire/smoke detection.
For those who don't know, a pre-action system is a water system except that the pipes are charged with compressed air (on in this case, nitrogen) with an interlock system that requires smoke detection followed by loss of pressurization before releasing water. The benefit is preventing accidental discharges; if someone breaks a head without the smoke detection, the loss of pressure will be seen as a trouble event and water will not be released into the pipes. If you get a smoke detect event and *then* a head fuses, water is released. It's all still a fancy water system though.
What I noticed is the guy's hands were covered in oil, and when he opens the lid of the server it's resting against his shirt, likely covering it in oil as well. This kind of idea seems like a unholy mess: oil everywhere. I rarely remove my servers once racked, but I do occasionally plug a USB drive or put a DVD in them.
Oil immersed servers are probably likely to end up in the trash straight away rather than end up on the used market, donated, or reused elsewhere.
Hard drives are not air tight, they have breather holes.
It does. The SLA gives paying users up to 15 days of free service.
http://www.google.com/apps/intl/en/terms/sla.html
- passwords stored in the clear in ldap (WTF??)
Some things require this. 802.1x with PEAP and MSCHAPv2 is one that comes to mind. When forced to do this, just run the LDAP server alone on its own server/VM so that nothing can get to it.
- set up sudo ... to allow everyone to do everything
I do this (on a per-user basis) because when I do a root action I want it to be intentional. I normally work as an unprivileged user and perform the few required root actions via sudo. There's no point in restricting sudo commands to anyone who already knows the root password, but I believe it is a good habit to work as a user, not as root. (For people who don't know the root password, sure, restrict away to only commands they need to do their job.)
They must have forgotten that a real Mac is a general purpose computer and not a walled garden like the iThings are.
Reliable bandwidth on the scale Wikipedia must need is not.
Personally, I wouldn't mind if Wikipedia cut back on their bandwidth to reduce costs (and therefore slowed page loads) since I'm getting it for free.
Short Circuit (1986)
VLANs are an ethernet (or equivalent) layer concept. A specific VLAN can have an arbitrary number (including zero) of IP4 or IP6 networks on it.
That's not really relevant; the example was just to show how one can use IPv6 addressing to create a shorter address. If you can remember your prefix, something tangentially related (like a switchport access VLAN number), you can probably come up with an IPv6 addressing scheme that is easy to remember.
You don't have to make long addresses if you don't want to. You can drop leading zeros and the :: compression replaces any range of zeros, not only one set. So a prefix you might get from your ISP becomes:
2001:DB8:A::/48
I can remember that easily and then make up a plan such as "/64 corresponds to VLAN". Say you have VLAN 5 and a statically assigned host 9 on that VLAN.
2001:DB8:A:5::9/64
Although it still has scary A-F in the number. Or you can stick with the crazy long addresses if that's easier.
If someone can't afford to buy a router that's hundreds of dollars, at least look at MikroTik (routerboard) hardware. Similar price range without the brain dead functionality of the typical D-Link or Linksys.
Thats not a normal person solution. I have boxee on a home media server, used to do mythtv but I moved out of a place where I had provided cable and without cable it wasn't worth the hassle. Its certainly possible, and its great once its set up, but it requires active maintenance.
My personal desires are for a single, low-power, easy-setup box that can:
- Stream from Netflix
- Stream from Hulu/Hulu Plus (to be legit it probably requires hulu plus)
- Stream from Pandora
The Roku does all of those, and more (like Amazon video).
- Run local/LAN-shared video and audio with good codec support
Roku has channels that say it will do LAN streaming, but I haven't tried because I don't really care to use it for that.
- Extensible to help future proof it (i.e. easy to integrate some new streaming service)
Roku has an open SDK, assuming you or someone else writes a channel for said new streaming service.
I've had mine since the early days when it was "the Netflix box" and they've released a whole lot of firmware upgrades since then. It was most certainly worth the $99 I spent on it, especially since its functionally continued to increase with each update. The only thing it can't do that the current hardware can is 1080p (mine does 720p).
On the other hand, Stallman brings up worthwhile points. You may lose certain legal rights -- in the USA, for example, you may lose your 4th amendment rights. You do not have control over web applications -- the provider can change things, yank out features or add new features you do not want, and you have no recourse (how many times has Facebook done this?). You may even lose your access entirely.
Likewise with the Facebook example, we can see that such changes do not cause the users to seek alternatives, either out of apathy or because there is none.
Why is this amazing "that the controller works at all"? There was a time before microprocessors, you know, and they did fun things like travel in space without them.
So what we just need to do is put more into infrastructure, remove all copper and replace with fiber optic..Voila no more copper thefts!
Unfortunately they'll still cut it first to look inside, they just won't bother to pull it out.
If redirecting NXDOMAIN to partnered search results pages
VeriSign != ICANN
And why didn't ICANN start the process of "firing" VeriSign immediately after the incident?
IETF, IANA, the RIRs, etc. all seem to work well without to have some legal entity with a bunch of corporate bullshit. The suits have managed to royally fuck things up at ICANN.
I agree with the GP: money has generally corrupted the process which should be a simply technical matter of updating a simple list of TLDs as countries have come and gone according to ISO 3166-1 alpha-2. The bureaucracy has started to serve itself.
ICANN did assert that they overstepped their authority, and VeriSigned later sued ICANN.
If redirecting NXDOMAIN to partnered search results pages
VeriSign != ICANN
The property manager for the office park I'm in let us put Ethernet switches on the buildings so we could offer fast symmetric speeds like our non-US friends enjoy for a fraction of the average cost for cable or DSL. It was interesting for a few reasons:
* The people that signed up were amazed that such speeds were possible; they had been stuck with DSL, cable, or mobile broadband that topped out at 1.5Mbps downstream on a good day.
* They were even more surprised that it was symmetric (the most expensive cable package was only 3Mbps up) and included a subnet standard.
* One of the tenants accused us of trying to scam them with a service that could not be possible for a company of our size.
So I'd say that the expectation of the average US consumer is to pay a high price for slow service from the local big cable company or big telco, then pay extra for add-ons like a static IP. Some of them won't even believe you or accuse you of lying if you go outside of that expectation.
Basically, they charge $200 because they can. It has nothing to do with the actual cost of service.
Correct; it is the price a residential customer will bear for what they feel is "advanced" pricing for a "fast" speed the customer probably doesn't fully grasp. These customers are also willing to pay for truck rolls, play "reboot the router" games with scripted tech support, and wait until the next business day for someone to fix it if it breaks. Many business class accounts (like cable) are the same quality as residential except the pricing is higher and you're allowed to pay extra for a static IP; that's how Charter does it where I am.
When you move into the land of T1, T3 and OC-x then you're dealing with a whole different class of customer. One that won't accept anything less than the technical capacity of the line, a comprehensive SLA, a first line tech that doesn't read a script, and real 24x7 service should there be trouble or an outage. Other factors are terminating your line to a much more expensive router that does lots of BGP rather than some cheap thing that only does static routes with a default. This customer is also willing to pay for it. One can argue the "bandwidth" should still be cheap, and it is, but the human aspect is certainly not.
Ah, the Reno 911 AVR. It was applied to cell phone calls specifically because there was an unmanageable amount of pocket dialing, and our dispatch center is not super huge. I called 911 once when the system was in place and either you dialed * or supposedly it would connect the call anyway if it heard "loud noises". I believe it's since been removed, or at least the last time I called in concrete chunks on the freeway from my cell I was connected to a dispatcher immediately.
Lets be honest though. Would you really want to do excel sheets on a tablet? I mean, to bring an acceptable level of excel functionality to a tablet you'd need an insane amount of gestures/menus to compensate for the desktop interface. By that point it doesn't seem worthwhile. How many unrecognized gestures would it take before you hurled the thing across the room?
Probably not beyond simple input. I can see it being great for generating quick estimates, quotes, or something along those lines using a spreadsheet created on a computer.
I have one from Sprint at the office. After arguing that I might as well cancel since it's not my problem and I don't want to pay for their coverage hole, they sent me one for free. It has its bugs, but it works more often than no signal at all.
Except to compete with google, you need not only as good or better algorythms, as good or better other services (e-mail etc). you need as good or better database too.
And you have to offer it for free: that's the expectation now.
Also, if you consciously try to keep your code as efficient as possible and run it on something that isn't abstracted 10 frameworks deep, scalability suddenly becomes much less of a problem.
Now-- Stargate Universe- I'm liking so far. There are some dumb things (like why they are not training the doctor during some of her at home time- but at least they addressed that by bringing in specialists). seems like they should be working on a forge and a machine shop too. And at some points the guns have got to start breaking and running out of ammunition. They should find/use some energy weapons.
The thing that's been bothering me about SG:U is that they are on this massive ship and don't seem to be exploring it much outside the tiny area they occupy. The ship likely has manufacturing facilities onboard and a full hospital with diagnostic equipment. It was intended to be ready for the arrival of a large crew for long term travel, yes? So it goes to reason everything they need is already there, including repairs to the ship. And how could only Rush know the ship has a bridge? Even the advanced Atlantis era ships that could be run by one person still had a bridge, and they've brought up maps on consoles before.