Being blunt has gotten me a lot further than being polite ever did.
There was this one customer/client/coworker (yeah, fucked-up business relationship), who just did not get what was going on. I tried being polite. I tried using all his lingo, "actioning" this and whatnot. I tried. It got me nowhere.
One phone call where I straight-up said "that problem was *your* fuckup, and I am tired of cleaning up your mess then getting blamed by you for it because you weren't even aware of the problem until I took care of it", and that got me further than months of politeness.
Hell, we still seem to get along. I think we've been communicating even better now that I've stopped "artificially limiting" my communications. I actually just made a note to myself to yell at him to check his code before he checks it in - there was a SQL file with an *obvious* syntax error, one that our standard IDE (which he uses) highlights...
The last Atom/Android phone I read about had an ARM emulator for running NDK apps. Performance suffers, but it's better than not running at all (particularly since Atom is more powerful than the common Cortex-A9 cores, so it has some performance to burn).
That's how it works now, yes. But if Dropbox is becoming a "hard drive replacement", then yes, it *is* going to require that network access. Or, possibly, a local cache of *some* files, but the rest will need that network access.
The high-end i7s are now quad-channel, not triple-channel.
There's basically two types of i7s - i5s with hyperthreading enabled (880, 2600K, 3770K, 4770K), and Xeons with fewer cores, higher clocks and no multi-socket (960, 990X, 3930K). They're completely different - the former have a dual-channel memory controller, fewer PCIe lanes, often have integrated graphics left in, and max out at quad-core. The latter use a different socket (sometimes Xeon-compatible), have triple- or quad-channel memory (depending on generation), and max out at six cores. They're even on different release schedules - while Haswell "low-end" i7s are out, the next "high-end" i7s are the Ivy Bridge-E series.
I actually think they should change the naming - call the high-end ones "i9" or something. It's odd that they have i5 and (low) i7s, which differ only slightly, as separate brands, but combine the two types of i7s into a single category despite being so different.
Hard drives are currently the greatest bottleneck in 95% of systems. Why do you think "get an SSD" is the new "add more RAM"?
A good hard drive will have average latency around the 5ms range, and throughput around 200MiB/s (in actual usage, not benchmarks). Cheaper ones will be closer to 10-15ms latency and 100MiB/s throughput.
I just tried pinging dropbox.com - 98ms latency, round-trip. And my bandwidth peaks around 400KiB/s, orders of magnitude below even a slow hard drive. And that's for download! Upload, you're looking at maybe 100KiB/s. I've gotten faster transfers over USB (and not that fancy new USB 3.0).
You may be saying that "users don't need that much speed for most stuff - give them an SSD for OS+Apps, and everything else goes in THE CLOUD".
Perhaps you're right. Perhaps many users could be satisfied with such an arrangement. But until Flash is nearly as cheap per gigabyte as spinning rust, there will remain plenty of tasks that need more capacity than a (reasonably-priced) SSD can provide, but more speed than a cloud solution can physically provide.
The latency is the biggest killer. For sequential access, a high-end hard drive can keep up with common SSDs - from the slowest HDD to the fastest SSD is perhaps an order of magnitude, probably less. But the latency is the killer - it's easily two orders of magnitude between discs and flash, and even more on the high end. You can easily feel that - I stuffed an SSD into a half-decade-old workstation, and it went from sluggish and unresponsive to smooth and lightning-fast (and that with a slow SSD and 3gbps SATA). My laptop boots in seconds, and is the snappiest computer I've ever used.
Cloud storage, just by physics, are another order of magnitude below local hard drives, just because of speed-of-light. As I mentioned, I get 100ms ping times to dropbox. And that's just for pings - if they actually have to pull my data up, you're adding the same latency as disk (because seriously, are they going to use Flash?). I don't even want to think about how slow that's going to feel.
A blog I once read provided a useful metaphor. Imagine a read from RAM takes one day (this was high-latency/high-bandwidth GDDR5; DDR3 latencies would be around 3 hours or so). Depending on your processor, you'd be executing instructions in the scale of minutes. Accessing a hard drive takes around fifty years. Reading from the cloud would take nearly six centuries.
*That* is how slow the cloud is. And that's why I use it, at most, for backups, or for running cloud servers - NOT as a replacement for local storage.
We're not doing agile or scrum or anything - honestly, we have a development process that's like waterfall but with no falling. Nobody's charted out which parts are dependent on any other part.
But even with the complete lack of project management, scope creep is still a bigger problem.
The initial specs for the project ("R2" of a project we did earlier) started with about half a dozen items on the scope list. When the contract was signed (our company is technically contract working for another company), it had expanded to about ten.
Now it's around thirty, maybe forty depending on how you count things. We're about a month from the due date (we started in March), and we're horribly behind. Some of the things still don't even have specs. I'm trying to get them to trim scope - we've cut half a dozen things just this week, after I blew up on a phone call with the person in charge of managing the project (a combined VP/marketing lead/programmer with commit access, the most dangerous combination I've ever seen as he will sell a customer something, code it in a sleepless weekend, then expect us to help him when it breaks or needs more functionality).
Actually, thanks to the magic of Hollywood accounting, I would bet that Card won't see a single dollar of ticket sales. He may not even have gotten a writer's fee, but in any case, he's probably going to be effectively paid a flat rate.
Funny how one evil organization is able to cancel out the evil of a kinda-evil individual.
Physicists get to perform thought experiments in a frictionless vacuum with spherical objects of uniform density. Why can't sociologists work in a similarly ideal environment?
Perhaps. But I also think a lot of the "training" in low-skill jobs is redundant.
It probably took about an hour for me to learn how to do the important parts of my job, back when I was a high-school kid working at Dairy Queen. One hour to learn the register, how to make a Blizzard, and where the mops and brooms were kept. It took a lot longer to learn all the meaningless bullshit - the proper greeting to say every time someone walked in the store, which items we were supposed to push on people, how to make an ice cream cone solid enough to do that flip thing with.
People are pretty flexible. And remember, those lower-end jobs are also getting eaten by robots. Temps like this could fill the jobs that need human flexibility, but not huge amounts of human training.
Not to mention, the temp agency could handle some of that. Have a pool of people trained in certain skills, so the individual jobs need to only go over what is specific to *their* job (in my DQ analogy, just how to make the ice cream). It then becomes even more likely that one worker may keep cycling back into certain jobs. Many things are seasonal - toy stores have massive booms during the winter, while many theme parks only operate in the summer. Even out the demand by filtering through a temp agency, and you can add some stability to the whole economy.
Yes, and if the temp agency is working "properly" (ie. the way I think the world should), you would immediately go to some other job.
The "job" is the virtual instance, the "employee" is the hardware that runs it. You think Amazon takes a server out back and shoots it every time you kill an AWS instance?
Maybe this is a good thing. Or at least, could be a good thing.
Imagine, for a metaphor, that workers are computer servers. This would be like virtualization - since the amount of work needed is often variable, being able to quickly "provision" workers could be a benefit. And having an agency that employs these people could provide more stability for the workers, in the way that Amazon and other cloud providers get more heavily-utilized servers. And, as with the computer cloud vs. dedicated server debate, employees they *need* to have, or who provide some function that interchangeable employees cannot, can be hired full-time as they currently are.
In an ideal world, these workers would get all the benefits of permanent employment (medical coverage, unemployment benefits, even regular promotions and wage increases) via the temp agency. However, in the "anything that reduces corporate profits by one iota is COMMUNISM" economy we have, something tells me this isn't the case.
You know how the Korean War, although ostensibly a war between North and South Korea, was basically a war between the US and China? Yeah. The American Revolution was that, with Britain and France. Of course, our "AMERICA! FUCK YEAH!" school system and remnants of Manifest Destiny keep most people from thinking of it in those terms, but yeah, that's how it was. A small American rebellion persisted long enough to sap the British strength until some heavy aid from France was enough to shove them out of a war they no longer really cared for.
Or you can have a BIOS that addresses the decades of accumulated legacy bodging that is the PC, without UEFI. Just put a BIOS that removes all the old cruft of the old BIOS, adds some new features, but is totally minimalistic.
That's what UEFI is - it drops old cruft (mainly ISA, AGP and such, IIRC), ups the minimum requirements (UEFI can assume some level of graphics support, so no more MDA text mode; likewise, it no longer runs in 16-bit mode), and extends functionality (booting off 2TB+ drives). They broke compatibility in a few places, but they did so, in part, to speed up boot times by moving functionality from the BIOS/UEFI to the OS.
UEFI, itself, is a big step forward. The only problem is the "Secure Boot", and honestly, the problem is currently theoretical (at least on x86 - ARM is a different story). Secure Boot itself is fine - as long as the user is allowed to add and remove keys, and can enable/disable it, it's at worst unneeded functionality.
A MUCH smaller subset actually wanted the old start menu back. I know I don't. There are elements of the old start menu that I liked, but most of it was a bad idea. Start -> All Programs was a complete disaster -- lets put a hierarchy of everything installed on your computer in a small non-resizable popup menu. Sorry that was just awful. For anything you need the start MENU for, the start screen is a LOT better.
Pinned aps on the start menu? Use a toolbar if you want a popup menu for those on the taskbar.
The only real loss is the search box that many power users use as a quick launcher - the start screen works for this, and is better if you are actually doing any sort of real search. But a desktop widget would be more appropriate for the "quick launch task of things we already know about."
But this is a power user function / feature not something "most users" do. Personally I'm looking for good 3rd party options, that just address this small shortcoming, rather than try to recreate the disaster that the old start menu was.
I have Windows 8 on my desktop. I tried using the Start Screen. I gave it a chance. I hated it. So I installed Classic Start Menu, and basically configured it exactly like I configured Windows 7, but with a few small tweaks.
I never use the "All Programs" menu. That *is* clunky. But I frequently use the Search box, which I have configured to search only programs and control menu items, not documents. It's basically a Run dialog that uses program names instead of executable filenames.
I also pin programs to the start menu, *distinctly* from pinning to the taskbar. I have numerous programs pinned to the start menu which I do not need particularly often, but I want to take advantage of the "frequently opened items" feature (so I can immediately open things with the program - a lifesaver with PuTTY, where I have a dozen connections I commonly use). Stuff on the taskbar is stuff that I have open 99% of the time anyways, and stuff that I can treat as "there" even if it's not running (like WMP).
The reason I don't like the start screen is that it's inefficient. It takes up the whole screen, but provides relatively little functionality over a start menu that takes up 5% of the screen. The icons are massive - my 1440p screen still holds only 72 icons, which makes me suspect they do no scaling between screen sizes. Since there is not a *single* Metro app I use, every "pane" is just a grey square holding an icon that is far too small for it, making it ugly. And it doesn't seem to play well with multiple monitors.
The start screen is a specific instance of a general problem with Windows 8 - every "Metro" program has a "classic" program thatworks better. The start menu works better than the start screen. Alt+Tab/Win+Tab/the taskbar works better than whatever they call that left-side "swipe" menu. Windows Media Player works better than the Music or Video apps. Preview works better than the Photos app. Internet fucking Explorer works better than most of the Apps that just access specific web data. I think the only thing that actually was an improvement is the Reader app, and that's because they didn't have *any* PDF viewer before.
You don't really know what a development kit is, do you?
A devkit is not an SDK. It's the same hardware and software as the retail product, but with additions/modifications that enable debugging (adding debugging ports, using libraries with debug symbols, etc). They also get the ability to run "unlicensed" software, since you can't go to Sony/Microsoft/Nintendo every time you compile in order to have it certified. And, finally, early devkits may not have the final case/board, since launch titles need to start development well before the case or even motherboard are finished (famously, the early Xbox 360 devkits used Power Mac G5 cases and motherboards).
So if the devkit is running a FreeBSD kernel, the final product will be running a slightly different version of the same kernel.
I don't think they're shills. Fanboys, perhaps, but not shills. Honestly, the nuclear industry just doesn't seem big enough to warrant forum shills. Talking heads or TV experts, yeah, possible shills, but not Slashdotters. We're not that important.
I, for one, think nuclear is something we need to be using more, but I'm advocating a nuclear+hydro+geothermal+solar+wind+tidal as a replacement for coal+gas+oil, not as a pure nuclear solution (at least, until we get fusion working - if fusion delivers on its promises, I would have zero issue with a pure-fusion power grid). But if you want to advocate a pure-renewable system, I wouldn't downmod you (I've actually got mod points right now).
Just a suggestion, though? Saying "we need more studies" or "what's the *real* cost?" tends to come across more like FUD than actual debate, particularly when you're coming from a position that is just as questionable in those areas. Maybe they're thinking *you're* the shill?
A somewhat-common solution to the inflexible nature of nuclear power is to pair it with hydroelectric power in artificial lakes. During the night, or other low-demand periods, the excess electricity can be used to pump water upstream, filling the reservoir. When peak times hit, that water can be let back down to generate additional power.
This has certain additional advantages as well. Nuclear plants need cooling water, so building them next to a lake is already fairly common. And as yet another added bonus, the dam provides some level of safety in event of a leak - it can be closed off completely, for a time, and it provides a chokepoint where filters could be installed in a leak scenario.
Many have already pointed out that making an unreadable font would really only protect against physical letters (as in, mail, not email) being read, or perhaps text being distributed through raster images. After all, 0x446561746820746F20416D6572696361 means the same thing, whether it's displayed in Helvetica, Times Roman or this new font.
We have measures that are better against machine interception (such as encryption), but those still have one flaw - they're obviously hiding something, and apparently "having something to hide" is now a crime in and of itself. There are steganographic techniques to hide one message inside another, but as soon as they become commonplace, they too will be scanned for.
What we need is something machines cannot adapt to. We need language. Come up with a system of code phrases that can easily be confused for inane, "safe" chatter. Either they don't scan for it, or grabbing it gets too much, and any actual messages get hidden within the noise. Make it so that only a human could reliably determine whether it's an actual "terrorist" message, or if it's a regular email.
For extra protection, base it off a somewhat-obscure set of jargon, so that even the average person wouldn't find anything suspicious about it.
Slashdot Mirror
That's just what they want you to think!
Being blunt has gotten me a lot further than being polite ever did.
There was this one customer/client/coworker (yeah, fucked-up business relationship), who just did not get what was going on. I tried being polite. I tried using all his lingo, "actioning" this and whatnot. I tried. It got me nowhere.
One phone call where I straight-up said "that problem was *your* fuckup, and I am tired of cleaning up your mess then getting blamed by you for it because you weren't even aware of the problem until I took care of it", and that got me further than months of politeness.
Hell, we still seem to get along. I think we've been communicating even better now that I've stopped "artificially limiting" my communications. I actually just made a note to myself to yell at him to check his code before he checks it in - there was a SQL file with an *obvious* syntax error, one that our standard IDE (which he uses) highlights...
The last Atom/Android phone I read about had an ARM emulator for running NDK apps. Performance suffers, but it's better than not running at all (particularly since Atom is more powerful than the common Cortex-A9 cores, so it has some performance to burn).
That's how it works now, yes. But if Dropbox is becoming a "hard drive replacement", then yes, it *is* going to require that network access. Or, possibly, a local cache of *some* files, but the rest will need that network access.
The high-end i7s are now quad-channel, not triple-channel.
There's basically two types of i7s - i5s with hyperthreading enabled (880, 2600K, 3770K, 4770K), and Xeons with fewer cores, higher clocks and no multi-socket (960, 990X, 3930K). They're completely different - the former have a dual-channel memory controller, fewer PCIe lanes, often have integrated graphics left in, and max out at quad-core. The latter use a different socket (sometimes Xeon-compatible), have triple- or quad-channel memory (depending on generation), and max out at six cores. They're even on different release schedules - while Haswell "low-end" i7s are out, the next "high-end" i7s are the Ivy Bridge-E series.
I actually think they should change the naming - call the high-end ones "i9" or something. It's odd that they have i5 and (low) i7s, which differ only slightly, as separate brands, but combine the two types of i7s into a single category despite being so different.
Hard drives are currently the greatest bottleneck in 95% of systems. Why do you think "get an SSD" is the new "add more RAM"?
A good hard drive will have average latency around the 5ms range, and throughput around 200MiB/s (in actual usage, not benchmarks). Cheaper ones will be closer to 10-15ms latency and 100MiB/s throughput.
I just tried pinging dropbox.com - 98ms latency, round-trip. And my bandwidth peaks around 400KiB/s, orders of magnitude below even a slow hard drive. And that's for download! Upload, you're looking at maybe 100KiB/s. I've gotten faster transfers over USB (and not that fancy new USB 3.0).
You may be saying that "users don't need that much speed for most stuff - give them an SSD for OS+Apps, and everything else goes in THE CLOUD".
Perhaps you're right. Perhaps many users could be satisfied with such an arrangement. But until Flash is nearly as cheap per gigabyte as spinning rust, there will remain plenty of tasks that need more capacity than a (reasonably-priced) SSD can provide, but more speed than a cloud solution can physically provide.
The latency is the biggest killer. For sequential access, a high-end hard drive can keep up with common SSDs - from the slowest HDD to the fastest SSD is perhaps an order of magnitude, probably less. But the latency is the killer - it's easily two orders of magnitude between discs and flash, and even more on the high end. You can easily feel that - I stuffed an SSD into a half-decade-old workstation, and it went from sluggish and unresponsive to smooth and lightning-fast (and that with a slow SSD and 3gbps SATA). My laptop boots in seconds, and is the snappiest computer I've ever used.
Cloud storage, just by physics, are another order of magnitude below local hard drives, just because of speed-of-light. As I mentioned, I get 100ms ping times to dropbox. And that's just for pings - if they actually have to pull my data up, you're adding the same latency as disk (because seriously, are they going to use Flash?). I don't even want to think about how slow that's going to feel.
A blog I once read provided a useful metaphor. Imagine a read from RAM takes one day (this was high-latency/high-bandwidth GDDR5; DDR3 latencies would be around 3 hours or so). Depending on your processor, you'd be executing instructions in the scale of minutes. Accessing a hard drive takes around fifty years. Reading from the cloud would take nearly six centuries.
*That* is how slow the cloud is. And that's why I use it, at most, for backups, or for running cloud servers - NOT as a replacement for local storage.
We're not doing agile or scrum or anything - honestly, we have a development process that's like waterfall but with no falling. Nobody's charted out which parts are dependent on any other part.
But even with the complete lack of project management, scope creep is still a bigger problem.
The initial specs for the project ("R2" of a project we did earlier) started with about half a dozen items on the scope list. When the contract was signed (our company is technically contract working for another company), it had expanded to about ten.
Now it's around thirty, maybe forty depending on how you count things. We're about a month from the due date (we started in March), and we're horribly behind. Some of the things still don't even have specs. I'm trying to get them to trim scope - we've cut half a dozen things just this week, after I blew up on a phone call with the person in charge of managing the project (a combined VP/marketing lead/programmer with commit access, the most dangerous combination I've ever seen as he will sell a customer something, code it in a sleepless weekend, then expect us to help him when it breaks or needs more functionality).
Actually, thanks to the magic of Hollywood accounting, I would bet that Card won't see a single dollar of ticket sales. He may not even have gotten a writer's fee, but in any case, he's probably going to be effectively paid a flat rate.
Funny how one evil organization is able to cancel out the evil of a kinda-evil individual.
Physicists get to perform thought experiments in a frictionless vacuum with spherical objects of uniform density. Why can't sociologists work in a similarly ideal environment?
Perhaps. But I also think a lot of the "training" in low-skill jobs is redundant.
It probably took about an hour for me to learn how to do the important parts of my job, back when I was a high-school kid working at Dairy Queen. One hour to learn the register, how to make a Blizzard, and where the mops and brooms were kept. It took a lot longer to learn all the meaningless bullshit - the proper greeting to say every time someone walked in the store, which items we were supposed to push on people, how to make an ice cream cone solid enough to do that flip thing with.
People are pretty flexible. And remember, those lower-end jobs are also getting eaten by robots. Temps like this could fill the jobs that need human flexibility, but not huge amounts of human training.
Not to mention, the temp agency could handle some of that. Have a pool of people trained in certain skills, so the individual jobs need to only go over what is specific to *their* job (in my DQ analogy, just how to make the ice cream). It then becomes even more likely that one worker may keep cycling back into certain jobs. Many things are seasonal - toy stores have massive booms during the winter, while many theme parks only operate in the summer. Even out the demand by filtering through a temp agency, and you can add some stability to the whole economy.
Yes, and if the temp agency is working "properly" (ie. the way I think the world should), you would immediately go to some other job.
The "job" is the virtual instance, the "employee" is the hardware that runs it. You think Amazon takes a server out back and shoots it every time you kill an AWS instance?
Maybe this is a good thing. Or at least, could be a good thing.
Imagine, for a metaphor, that workers are computer servers. This would be like virtualization - since the amount of work needed is often variable, being able to quickly "provision" workers could be a benefit. And having an agency that employs these people could provide more stability for the workers, in the way that Amazon and other cloud providers get more heavily-utilized servers. And, as with the computer cloud vs. dedicated server debate, employees they *need* to have, or who provide some function that interchangeable employees cannot, can be hired full-time as they currently are.
In an ideal world, these workers would get all the benefits of permanent employment (medical coverage, unemployment benefits, even regular promotions and wage increases) via the temp agency. However, in the "anything that reduces corporate profits by one iota is COMMUNISM" economy we have, something tells me this isn't the case.
Sarcasm punctuation. Like that's never been done before~
Honestly, France beat the British. Not America.
You know how the Korean War, although ostensibly a war between North and South Korea, was basically a war between the US and China? Yeah. The American Revolution was that, with Britain and France. Of course, our "AMERICA! FUCK YEAH!" school system and remnants of Manifest Destiny keep most people from thinking of it in those terms, but yeah, that's how it was. A small American rebellion persisted long enough to sap the British strength until some heavy aid from France was enough to shove them out of a war they no longer really cared for.
This is why, on any production box, I tend to write updates as a SELECT first, only rewriting it into an UPDATE if the selected row count looks right.
Transactions would probably be a good idea, though.
Or you can have a BIOS that addresses the decades of accumulated legacy bodging that is the PC, without UEFI.
Just put a BIOS that removes all the old cruft of the old BIOS, adds some new features, but is totally minimalistic.
That's what UEFI is - it drops old cruft (mainly ISA, AGP and such, IIRC), ups the minimum requirements (UEFI can assume some level of graphics support, so no more MDA text mode; likewise, it no longer runs in 16-bit mode), and extends functionality (booting off 2TB+ drives). They broke compatibility in a few places, but they did so, in part, to speed up boot times by moving functionality from the BIOS/UEFI to the OS.
UEFI, itself, is a big step forward. The only problem is the "Secure Boot", and honestly, the problem is currently theoretical (at least on x86 - ARM is a different story). Secure Boot itself is fine - as long as the user is allowed to add and remove keys, and can enable/disable it, it's at worst unneeded functionality.
Shouldn't that be a screen window?
A MUCH smaller subset actually wanted the old start menu back. I know I don't. There are elements of the old start menu that I liked, but most of it was a bad idea. Start -> All Programs was a complete disaster -- lets put a hierarchy of everything installed on your computer in a small non-resizable popup menu. Sorry that was just awful. For anything you need the start MENU for, the start screen is a LOT better.
Pinned aps on the start menu? Use a toolbar if you want a popup menu for those on the taskbar.
The only real loss is the search box that many power users use as a quick launcher - the start screen works for this, and is better if you are actually doing any sort of real search. But a desktop widget would be more appropriate for the "quick launch task of things we already know about."
But this is a power user function / feature not something "most users" do. Personally I'm looking for good 3rd party options, that just address this small shortcoming, rather than try to recreate the disaster that the old start menu was.
I have Windows 8 on my desktop. I tried using the Start Screen. I gave it a chance. I hated it. So I installed Classic Start Menu, and basically configured it exactly like I configured Windows 7, but with a few small tweaks.
I never use the "All Programs" menu. That *is* clunky. But I frequently use the Search box, which I have configured to search only programs and control menu items, not documents. It's basically a Run dialog that uses program names instead of executable filenames.
I also pin programs to the start menu, *distinctly* from pinning to the taskbar. I have numerous programs pinned to the start menu which I do not need particularly often, but I want to take advantage of the "frequently opened items" feature (so I can immediately open things with the program - a lifesaver with PuTTY, where I have a dozen connections I commonly use). Stuff on the taskbar is stuff that I have open 99% of the time anyways, and stuff that I can treat as "there" even if it's not running (like WMP).
The reason I don't like the start screen is that it's inefficient. It takes up the whole screen, but provides relatively little functionality over a start menu that takes up 5% of the screen. The icons are massive - my 1440p screen still holds only 72 icons, which makes me suspect they do no scaling between screen sizes. Since there is not a *single* Metro app I use, every "pane" is just a grey square holding an icon that is far too small for it, making it ugly. And it doesn't seem to play well with multiple monitors.
The start screen is a specific instance of a general problem with Windows 8 - every "Metro" program has a "classic" program that works better . The start menu works better than the start screen. Alt+Tab/Win+Tab/the taskbar works better than whatever they call that left-side "swipe" menu. Windows Media Player works better than the Music or Video apps. Preview works better than the Photos app. Internet fucking Explorer works better than most of the Apps that just access specific web data. I think the only thing that actually was an improvement is the Reader app, and that's because they didn't have *any* PDF viewer before.
That's not how car analogies work, and you know it.
You don't really know what a development kit is, do you?
A devkit is not an SDK. It's the same hardware and software as the retail product, but with additions/modifications that enable debugging (adding debugging ports, using libraries with debug symbols, etc). They also get the ability to run "unlicensed" software, since you can't go to Sony/Microsoft/Nintendo every time you compile in order to have it certified. And, finally, early devkits may not have the final case/board, since launch titles need to start development well before the case or even motherboard are finished (famously, the early Xbox 360 devkits used Power Mac G5 cases and motherboards).
So if the devkit is running a FreeBSD kernel, the final product will be running a slightly different version of the same kernel.
At this point, I think the NSA watch lists are a superset of "list of all humans currently alive". I think I'm on there are least twice.
I don't think they're shills. Fanboys, perhaps, but not shills. Honestly, the nuclear industry just doesn't seem big enough to warrant forum shills. Talking heads or TV experts, yeah, possible shills, but not Slashdotters. We're not that important.
I, for one, think nuclear is something we need to be using more, but I'm advocating a nuclear+hydro+geothermal+solar+wind+tidal as a replacement for coal+gas+oil, not as a pure nuclear solution (at least, until we get fusion working - if fusion delivers on its promises, I would have zero issue with a pure-fusion power grid). But if you want to advocate a pure-renewable system, I wouldn't downmod you (I've actually got mod points right now).
Just a suggestion, though? Saying "we need more studies" or "what's the *real* cost?" tends to come across more like FUD than actual debate, particularly when you're coming from a position that is just as questionable in those areas. Maybe they're thinking *you're* the shill?
A somewhat-common solution to the inflexible nature of nuclear power is to pair it with hydroelectric power in artificial lakes. During the night, or other low-demand periods, the excess electricity can be used to pump water upstream, filling the reservoir. When peak times hit, that water can be let back down to generate additional power.
This has certain additional advantages as well. Nuclear plants need cooling water, so building them next to a lake is already fairly common. And as yet another added bonus, the dam provides some level of safety in event of a leak - it can be closed off completely, for a time, and it provides a chokepoint where filters could be installed in a leak scenario.
Many have already pointed out that making an unreadable font would really only protect against physical letters (as in, mail, not email) being read, or perhaps text being distributed through raster images. After all, 0x446561746820746F20416D6572696361 means the same thing, whether it's displayed in Helvetica, Times Roman or this new font.
We have measures that are better against machine interception (such as encryption), but those still have one flaw - they're obviously hiding something, and apparently "having something to hide" is now a crime in and of itself. There are steganographic techniques to hide one message inside another, but as soon as they become commonplace, they too will be scanned for.
What we need is something machines cannot adapt to. We need language. Come up with a system of code phrases that can easily be confused for inane, "safe" chatter. Either they don't scan for it, or grabbing it gets too much, and any actual messages get hidden within the noise. Make it so that only a human could reliably determine whether it's an actual "terrorist" message, or if it's a regular email.
For extra protection, base it off a somewhat-obscure set of jargon, so that even the average person wouldn't find anything suspicious about it.
Dammit, I just figured out how to re-synergize our core strengths by thinking outside the box, and they go and change all the buzzwords on me.