My 2011 17" is still chugging along, and it's been through a lot, including being dropped on a corner that slightly dislodged the video cable. I put in 16GB RAM years ago (try that with the 8GB soldered-RAM models they are selling now!) and only recently did the trackpad start to get flaky. Fortunately I had just picked up a first-gen unibody 15" a month before and found out about the problems with old trackpads, so I already knew about the adjustment screw.
At least they still sell the 13" with optical drive that can presumably be upgraded to 16GB. I suspect the main reason they dropped the 17" was because a retina display that size would have been too expensive. I really don't even want a retina display anyhow.
All servers got deleted and the offsite backups too because the remote storage was mounted just before by the same script
Clearly a case of a fool thinking that a sync (copying data to another place regularly) is a backup. It's not a backup if you can easily copy corrupted data to your only copy. Or, in this case, if you can easily delete the data from your "backup" copy.
I missed the part of TFA where it said they intended for this to be built in Iceland, completely in the middle of fucking nowhere! The only good part is that Iceland probably has some decent bandwidth from transatlantic cables landing there. How you connect it (reliably!) to the middle of a big ice field is another matter. There's probably also earthquake issues to consider in Iceland, yet another reason not to make a high-rise.
Of course the place to build a high-rise is in the middle of nowhere where land is plentiful! The usual reason for a high-rise building is because land and property tax is expensive. (For an example of a high-rise on its side, check out the USAA building.) In this case, the only reason is to get the chimney effect, which others have pointed out probably wouldn't even work properly as conceived.
The task of getting sufficient bandwidth to a data center on a glacier is of no concern to such lofty exalted beings as architects. They are only there to design the buildings. It is the job of the filthy occupants to connect up their little boxes and blinky lights.
How are you going to keep the pod powered while you lower it 65 stories to the ground level? A really long extension cord? A third rail? (Mind the gap!) Also, wired Ethernet is limited to 100m, right? So you would have to use fiber, and be sure to not violate the minimum curve radius, wouldn't want to break it!
They also don't seem to understand the concept of humidity. Open air cooling by vertical convective flow may be great for humans, but is it a good idea for electronics? Even if you don't get condensation, if you are near an ocean (the architects are Italian, you work it out) you will have salty humidity in the air. ("The hot air inside the tower goes up and sucks the cold air from the outside. The outside cold air, to enter, is obliged to pass through the pods, and in this way cools the servers.") It also is useless in places where it gets hot in the summer, such as Texas.
Another problem I see is that they intend for the modules to be brought down to ground level for maintenance. This might be fine if you are Google and have thousands of basically redundant servers, but if this is a data center with lots of customers, how are you going to explain to a customer about the need for downtime because the server in the next rack of the module needs to be upgraded?
This is basically "stunt architecture" so that they can wow the judges of a competition that is if no consequence to the real world. The goal isn't to make a real building, it's to make the most awesome looking imaginary building without having to consider if it has fundamental flaws.
Ditto for landing on shore. I remember seeing a landing profile for the successful pad landing. It aimed for a point away from the landing site during the re-entry burn, then restarted the engines and zigged over when it was almost down. That ensures a failure to restart the engines results in a splash a few miles away, not a boom at your landing site.
Blue Origin has been mocking them in the other way. "Hey, look at what we just did! What took you so long?" Sure, you had a sub-orbital launch profile (almost no horizontal velocity), popping off a tin can that came straight back down. Boy Scouts recover their Estes rockets all the time. SpaceX already did the landing thing with their Grasshopper rocket (and DC-X long before either of them), and the only reason they didn't take it higher was because they didn't have clearance to go higher at McGregor.
Falcon 9 has been on an orbital launch profile every time, sometimes even GTO, which is a lot harder to come back from. Even hitting the drone ship and falling over was harder than what Blue Origin did. A side-effect of having an actual useful launch profile is engines that can't throttle down to hover (Blue Origin can), so they have to do the much harder "hoverslam" maneuver. (zero vertical velocity at the same moment as zero altitude)
I will, however, give Blue Origin a few points for doing quick turnarounds. Their short-term objective is space tourism, and they're doing exactly what they need. It's just not nearly as hard as what SpaceX is trying to do.
Not very well know fiction relating to this: The Man Whose Name Wouldn't Fit: Or, The Case of Cartwright-Chickering
One-line synopsis: Arthur Duane Cartwright-Chickering, is fired from his job because the new computer that processes employee files cannot handle his long name.
I have a copy of it somewhere, under stuff. I'd read it if I knew where it was right now.
I take the time to find out where the hell I'm going before I ever leave. I get on Google Maps, find out where I'm going, then figure out the route myself. If the route is complicated enough (like 200 miles across west Texas, where the map looks like a graph from a CS textbook), I may even make a little skeleton map of various waypoints. I may also use the street view so that I know what a place looks like before I get there. I know the main routes around most of the big cities in the main DFW/Houston/San Antonio triangle of Texas because I used to do a lot of day-trip thrift store raids about ten years ago.
In San Antonio there are quite a few major streets that do this. In the northeast, just try tracing where Nacogdoches Road goes. A lot of these happen because of roads that had an intersection built where there used to be a curve, so now you have to make a turn at the intersection to stay on the same road.
In one neighborhood where I grew up, Google street maps has the house numbers increasing southward when they should be increasing northward. It's the only block of the street, the numbering starts with 1, and it's pretty long, with at least 20 half-acre lots along both sides. I'm going to guess that the other parallel streets in the neighborhood are similarly misnumbered.
In my experience, when a PBS station puts up special programming during pledge drives, it is often programming (such as rock concerts) that you never see any other time of the year. So even if you like them, your money won't go toward more of them appearing regularly. Fortunately my MythTV will happily record the unmolested regular content that shows up at 1AM.
It's a microcontroller, so it should be compared to the Arduino, not the RasPi. In that comparison it's not bad, with an ARM Cortex M0 vs an Atmel AVR. These days the only "cool" thing that AVR has going for it is the availability of DIP packages.
And this really isn't news until they actually get delivered, because we've already known for months that they were going to give them to school kids.
Example: I am still running OS X 10.6.8 on my main laptop (yeah, I know, but I only recently got rid of the last PPC stuff I was using), and the reason you can't run newer versions of Minecraft on it is because it hits OpenGL so hard that the display driver freaks out. You can still SSH in, but the display driver requires a reboot. And that's (probably) not even a DMA issue, it just lets itself run out of resources.
That was due to Rovi (formerly known as Macrovisiion, so no stranger to annoying consumers) decided to shut down the OTA version of the TVGOS service. And not just shut it down, but yank it from the world. They demanded the hardware back from stations, who might otherwise have left it running just for the clock. So while it wasn't technically bricked, the usefulness of a DVR without a working clock is very near zero.
The problem was that Sony, in their infinite arro...wisdom decided that TVGOS would be the only way to set the clock for their DVR. There was no manual clock setting option. Nobody discontinues products like Sony, and their DVR was already discontinued when this happened, so no updates were provided.
I personally had a ChannelMaster CM7000PAL DVR at this time. I can say that the TVGOS service with its 14 days of guide info was awesome, but that particular DVR also supported using the ATSC guide data, and the clock could be set manually. It was flaky (show descriptions would go to the wrong show, which also happens on my MythTV, but I insist on using the OTA guide data and someday I'll figure out the problems) and most stations only put up 12 hours of data because apparently some TV sets had crappy guide implementations that would freak out with too much data. But it still worked, unlike the Sony.
Oh, now I see that you were talking about the "get more comments" link or something like that. My comment still stands as something at the bottom of the page that should have something done about it, or at least reviewed as to whether it should be kept.
Slashdot Mirror
My 2011 17" is still chugging along, and it's been through a lot, including being dropped on a corner that slightly dislodged the video cable. I put in 16GB RAM years ago (try that with the 8GB soldered-RAM models they are selling now!) and only recently did the trackpad start to get flaky. Fortunately I had just picked up a first-gen unibody 15" a month before and found out about the problems with old trackpads, so I already knew about the adjustment screw.
At least they still sell the 13" with optical drive that can presumably be upgraded to 16GB. I suspect the main reason they dropped the 17" was because a retina display that size would have been too expensive. I really don't even want a retina display anyhow.
If they weren't on alt.religion.kibology back around '92 or so, they don't know shit about trolling.
Because he is a retard.
All servers got deleted and the offsite backups too because the remote storage was mounted just before by the same script
Clearly a case of a fool thinking that a sync (copying data to another place regularly) is a backup. It's not a backup if you can easily copy corrupted data to your only copy. Or, in this case, if you can easily delete the data from your "backup" copy.
I missed the part of TFA where it said they intended for this to be built in Iceland, completely in the middle of fucking nowhere! The only good part is that Iceland probably has some decent bandwidth from transatlantic cables landing there. How you connect it (reliably!) to the middle of a big ice field is another matter. There's probably also earthquake issues to consider in Iceland, yet another reason not to make a high-rise.
Of course the place to build a high-rise is in the middle of nowhere where land is plentiful! The usual reason for a high-rise building is because land and property tax is expensive. (For an example of a high-rise on its side, check out the USAA building.) In this case, the only reason is to get the chimney effect, which others have pointed out probably wouldn't even work properly as conceived.
The task of getting sufficient bandwidth to a data center on a glacier is of no concern to such lofty exalted beings as architects. They are only there to design the buildings. It is the job of the filthy occupants to connect up their little boxes and blinky lights.
How are you going to keep the pod powered while you lower it 65 stories to the ground level? A really long extension cord? A third rail? (Mind the gap!) Also, wired Ethernet is limited to 100m, right? So you would have to use fiber, and be sure to not violate the minimum curve radius, wouldn't want to break it!
I just want to know how Simon and the PFY fit into this story. Then it would be perfect.
They also don't seem to understand the concept of humidity. Open air cooling by vertical convective flow may be great for humans, but is it a good idea for electronics? Even if you don't get condensation, if you are near an ocean (the architects are Italian, you work it out) you will have salty humidity in the air. ("The hot air inside the tower goes up and sucks the cold air from the outside. The outside cold air, to enter, is obliged to pass through the pods, and in this way cools the servers.") It also is useless in places where it gets hot in the summer, such as Texas.
Another problem I see is that they intend for the modules to be brought down to ground level for maintenance. This might be fine if you are Google and have thousands of basically redundant servers, but if this is a data center with lots of customers, how are you going to explain to a customer about the need for downtime because the server in the next rack of the module needs to be upgraded?
This is basically "stunt architecture" so that they can wow the judges of a competition that is if no consequence to the real world. The goal isn't to make a real building, it's to make the most awesome looking imaginary building without having to consider if it has fundamental flaws.
Can we all just agree that csh/tcsh is stupid?
The SSRBs were little more than empty pipe after all the fuel burned. They had a lot less parts to damage than a liquid-fuel engine.
Ditto for landing on shore. I remember seeing a landing profile for the successful pad landing. It aimed for a point away from the landing site during the re-entry burn, then restarted the engines and zigged over when it was almost down. That ensures a failure to restart the engines results in a splash a few miles away, not a boom at your landing site.
Blue Origin has been mocking them in the other way. "Hey, look at what we just did! What took you so long?" Sure, you had a sub-orbital launch profile (almost no horizontal velocity), popping off a tin can that came straight back down. Boy Scouts recover their Estes rockets all the time. SpaceX already did the landing thing with their Grasshopper rocket (and DC-X long before either of them), and the only reason they didn't take it higher was because they didn't have clearance to go higher at McGregor.
Falcon 9 has been on an orbital launch profile every time, sometimes even GTO, which is a lot harder to come back from. Even hitting the drone ship and falling over was harder than what Blue Origin did. A side-effect of having an actual useful launch profile is engines that can't throttle down to hover (Blue Origin can), so they have to do the much harder "hoverslam" maneuver. (zero vertical velocity at the same moment as zero altitude)
I will, however, give Blue Origin a few points for doing quick turnarounds. Their short-term objective is space tourism, and they're doing exactly what they need. It's just not nearly as hard as what SpaceX is trying to do.
You can fix that issue very simply by only buying name brand memory.
Or maybe not so simply.
So when does it get Nasalus Rift support? It's just not a proper paint watching without the volatiles in the air.
It was decades ago, so I expected things like that were fixed.
They're probably double now. Most people stopped using fixed a long long time ago.
Not very well know fiction relating to this: The Man Whose Name Wouldn't Fit: Or, The Case of Cartwright-Chickering
One-line synopsis: Arthur Duane Cartwright-Chickering, is fired from his job because the new computer that processes employee files cannot handle his long name.
I have a copy of it somewhere, under stuff. I'd read it if I knew where it was right now.
I take the time to find out where the hell I'm going before I ever leave. I get on Google Maps, find out where I'm going, then figure out the route myself. If the route is complicated enough (like 200 miles across west Texas, where the map looks like a graph from a CS textbook), I may even make a little skeleton map of various waypoints. I may also use the street view so that I know what a place looks like before I get there. I know the main routes around most of the big cities in the main DFW/Houston/San Antonio triangle of Texas because I used to do a lot of day-trip thrift store raids about ten years ago.
In San Antonio there are quite a few major streets that do this. In the northeast, just try tracing where Nacogdoches Road goes. A lot of these happen because of roads that had an intersection built where there used to be a curve, so now you have to make a turn at the intersection to stay on the same road.
In one neighborhood where I grew up, Google street maps has the house numbers increasing southward when they should be increasing northward. It's the only block of the street, the numbering starts with 1, and it's pretty long, with at least 20 half-acre lots along both sides. I'm going to guess that the other parallel streets in the neighborhood are similarly misnumbered.
In my experience, when a PBS station puts up special programming during pledge drives, it is often programming (such as rock concerts) that you never see any other time of the year. So even if you like them, your money won't go toward more of them appearing regularly. Fortunately my MythTV will happily record the unmolested regular content that shows up at 1AM.
It's a microcontroller, so it should be compared to the Arduino, not the RasPi. In that comparison it's not bad, with an ARM Cortex M0 vs an Atmel AVR. These days the only "cool" thing that AVR has going for it is the availability of DIP packages.
And this really isn't news until they actually get delivered, because we've already known for months that they were going to give them to school kids.
Example: I am still running OS X 10.6.8 on my main laptop (yeah, I know, but I only recently got rid of the last PPC stuff I was using), and the reason you can't run newer versions of Minecraft on it is because it hits OpenGL so hard that the display driver freaks out. You can still SSH in, but the display driver requires a reboot. And that's (probably) not even a DMA issue, it just lets itself run out of resources.
Also, the lobbyists can subvert almost anyone. If not through wining and dining, then by exploiting ignorance.
That was due to Rovi (formerly known as Macrovisiion, so no stranger to annoying consumers) decided to shut down the OTA version of the TVGOS service. And not just shut it down, but yank it from the world. They demanded the hardware back from stations, who might otherwise have left it running just for the clock. So while it wasn't technically bricked, the usefulness of a DVR without a working clock is very near zero.
The problem was that Sony, in their infinite arro...wisdom decided that TVGOS would be the only way to set the clock for their DVR. There was no manual clock setting option. Nobody discontinues products like Sony, and their DVR was already discontinued when this happened, so no updates were provided.
I personally had a ChannelMaster CM7000PAL DVR at this time. I can say that the TVGOS service with its 14 days of guide info was awesome, but that particular DVR also supported using the ATSC guide data, and the clock could be set manually. It was flaky (show descriptions would go to the wrong show, which also happens on my MythTV, but I insist on using the OTA guide data and someday I'll figure out the problems) and most stations only put up 12 hours of data because apparently some TV sets had crappy guide implementations that would freak out with too much data. But it still worked, unlike the Sony.
Oh, now I see that you were talking about the "get more comments" link or something like that. My comment still stands as something at the bottom of the page that should have something done about it, or at least reviewed as to whether it should be kept.