For most applications where CPU power matters, hyperthreading actually hurts performance, because the scheduler assumes that the hyper cores are real cores, when in fact they are not. A hyperthreaded core is not a full core, and will bottleneck waiting for the bits it doesn't have on the real core. I prefer not even having hyperthreaded cores unless the os and applications are aware that some cores are gimpy hyperthread cores and takes that into account; i.e does not schedule on them, or only schedules tasks that are not actually crippled by dependency on the real cores on the hyper cores.
Try running a database on hyperthreaded cores....it blows chunks. It's better if you just disable them in BIOS.
Yep. I have dental implants (titanium/ceramic composite teeth screwed into the skull) and a titanium plate with six screws in my arm. It may not be electronic, but I do have artificial replacement parts.
Whoever designed the system is terrible at computer science.
These are home users, using a government provider scanner, and id card, and a key.
Would be pretty easy to build a rootkit filter driver that steals the data off the card during legitimate transactions, along with a keylogger. At that point, you can pretty much remotely impersonate anyone whom you've rootkitted. Doesn't matter how secure the back end is because you can easily dupe the scanner side.
Terrible, terrible design by idiots....you can't trust home user systems to be clean of rootkits.
#1 You teach people that software is usually deterministic; if I follow known sequence of steps A, result B will occur. If not, something is usually wrong and its time to ask an expert for assistance.
#2 The way to learn technology is to play with it before crunch time. That means that when you get a new application in front of you, you should explore the menus, see what they can do and how the software behaves so that you can later on be creative when using it under pressure.
#3 Basic computer architecture; e.g. RAM is temporary working space, the hard disk is where your files are saved, etc. How to identify which part of the computer needs to upgrade; e.g. when to add RAM, when to replace the video card, when to get a new machine. How to navigate directories so you actually know where you saved your files (people over 40 have issues with this).
#4 Basic networking. What a NAT is, how to hook up a broadband router. How to determine how fast it really is and reason about whether you are getting what you pay for.
#5 Basics of what kind of problems are done well by computers and what kind of problems computers suck at. Basics of why some algorithms are faster than others at doing the same job. One of the biggest problems with business using computers is that people do not understand anything that does not scale linearly. Examples of applications used for various common types of tasks.
#6 How incredibly complex the stuff is under the hood....that when something is actually wrong with the software, it's extremely difficult to track it down.
Duh, the extra whitespace helps you parse each sentence seperately.
It's much more mentally taxing to read a singlespaced block of mooshed together text. Also, typography concerns are nowadays about how to wrap text around oddly shaped illustrations in magazines and advertisements without strange gaps, not making actual text readable. The typography nerds may claim that the original standard before the advent of typewriters and modern computers is "more correct", but I rather think it's not due to objectiveness in analyzing which one is actually better.
At some point, the hassle of working with old junk and making it work, putting up with how slow it is, dealing with failing electronics, and so forth isn't worth it.
I have 17 Pentium 3 class systems in my basement in a render farm. Sure, it's neat to have so many systems. But for my purpose, a single $300 quad core box literally has more compute power, more memory, more memory bandwidth, and uses way less electricity. Plus you don't have to maintain a billion systems. And it takes up less space. And there's no heat problem. I haven't replaced the pile yet, because I'm not doing that much 3D lately, but I will, and it will be awesome to be rid of so much clutter. I also have a bunch of Sun boxes. They were fun to get working, but they use too much power, and it's an absolute hassle to fiddle with them, maintain software on several platforms, and so forth. My free time is valuable; I don't want to waste it doing menial maintenence on crappy hardware.
Off brand low end consumer gear is barely designed to last 3 years, let alone past life expectancy. Most of that 802.11b gear is pretty limited in what it can do, and barely worked when it was new. It's not like you can install dd-wrt and turn them into a mesh.
Best case scenario is probably hooking up somebody who has no wireless and no resources up, like your local church or whatever. If it breaks, meh, they had low expectations to begin with. It may not even be worth doing that though, because a lot of older consumer routers break when subjected to the network behavior of newer versions of Windows because they can't handle scaling window sizes with the default settings, and it's a support chore to dink around with the settings on every machine that comes along in a non-enterprise environment.
Bottom line is that old junk starts costing you more to use than buying new stuff would.
Yeah, I went to get firmware updates for some older Sun hardware I wanted to fix up and ran in to this too. Got the same "support contract is required for firmware updates" crap.
My 11 sun machines are headed to the dumpster, as a direct result of this policy. It's incredibly stupid. It's not like Sun was winning any new customers these days anyway, and now they'll bleed out the few they had. Obviously their intention is to kill off their hardware business, because no one in their right mind would decide to implement a policy like this.
Their model is vertical market with machiavellian control over customers. That model broke in the 1990's. Ridiculous to keep hanging on to it like a moldy wet blanket.
Actually, "the computer", and by extension "the internet" would be just another device to control as far as these interfaces go. In the various cybernetic monkey experiments, the robotic parts ultimately become an extension of the body; a two way digital communication link to the outside world would probably be adapted to in the same way; eventually the user would sort of "think at" the link to be able to read and write information. Would work a lot like io completion ports and device registers, which is how current drivers communicate with current devices. Bit of a challenge to work out a protocol for meaningful communication, but this part does get bits in and out.
This class of device is quite close to the minimum requirements for a direct neural interface. (Assuming the infection bits get worked out and so forth.) Would not want to be a beta tester though.
Video and image transfer requires a lot of bandwidth to transmit, but audio streams are nowhere near as intense. Won't take too many generations of the technology to get bandwith rates that are high enough to sustain a data stream the size of a voice data stream; the brain will pretty much automagically learn to interpret the data.
It is a long way from something like the Matrix, with total override of perception, though.
Sorry, not quite... sort of on the 120hz, no on the "free" stereo 3d from the console. Most "120hz" TV's actually just interpolate the intermediate frames in their processing chip. They can't actually display real incoming data at 120hz, they just fake it.
"All it requires" for free stereo 3d as described above is about 2x the work. Game console hardware already has to drop resolution to 720p to maintain 30fps for the usual level of detail in games these days. Make the graphics subsystem render 2x frames in the same time, and at best it's going to manage 15fps. Sure, you get to re-use some of the setup work (mostly moving data into the graphics subsystem) from the first angle, but the graphics subsystem still has to do the matrix transforms to render the scene, as well as run the shaders, which are likely not designed to look good from multiple angles, and may involve destructive calculations.
Usually in games there are distinct tasks that are parallel in nature, but within a given task there is an inherent degree of serialization. Most games have a simulation thread to update the game world state, and a seperate render thread that periodically draws a frame of the current state at say 30fps or so, depending on system load. But the render thread doesn't actually do the drawing; instead it gathers up the required 3d models, textures, scene data and whatever else, and feeds them to the GPU to draw. Rendering is embarrassingly parallel. Unfortunately, snapshotting the game world state to determine what to draw in a given frame is inherently serial. Recently, shaders have added the capability to do additional manipulation of the data on the GPU itself, which is faster because of the i/o bandwidth bottleneck. (talking to the slower main system bus). Anyway, a free core on the system does precisely nothing for you unless the game architecture is designed to parallelize the render thread to some degree, which most engines do not do. Instead, that core should be used for better AI or whatever. While game consoles are designed to have more bandwidth between the graphics subsystem, main memory, and the cpu, it's still the bottleneck, just not as bad as on PC hardware.
There are other engine differences you would have to build to do this as well, so it's hardly free, and that's not even thinking about memory constraints in the graphics subsystem, and whatever requirements there are for a 3d signal to the 3d TVs (presumably some sort of alternating frame format at a higher framerate; I haven't really looked in to it lately).
Gonna be a couple generations of the 3d display technology before the kinks are worked out, too. Probably the next generation of consoles will source it reasonably well. That's still 5-7+ years out though. Which is about the same timeframe that the no-glasses 3d TV tech is anticipated to appear. That means several years of crappy rapidly changing tech are in store.
Existing deployment tools from Microsoft already do this. You need the WAIK, which is a free download from Microsoft.
You need to create a generalized image. If you get all the required drivers for all your hardware into the driver store, the drivers will be found during install. You can also deploy from PXE boot using WDS with a generalized image...
There are a few caveats around a few drivers that aren't designed properly for Sysprep, and applications that aren't designed with sysprep in mind, but otherwise it's quite slick. You can script the installation of these exceptions to occur later on during deployment using unattend.xml and RunSynchronous commands though. You can also supply your licence key in the unattend.xml file.
About 90% of all Windows deployments are sysprepped by OEMs or by corporate IT folks....
Please read the documentation, the tools are quite flexible.
A decent NAT router will do all the internet control stuff you are looking for, with an interface that isn't too terrible to grok. The rest of what you're asking for starts to crank up the cost and complexity extremely fast, especially for a single parent with three teenage kids...
Really, you don't want to try to set up draconian enterprisy stuff, it's not designed for consumers, will take time to administrate, and will break on you anyway. Network control via NAT router should be sufficient. If the parent wants to be able to physically control computer use too, then set up the clients up with the hard disk in a removable bay carrier. Yank it when you want to deny access to the machine totally. Much cheaper than setting up a domain and controlling access with accounts, and more reliable.
Seriously though, today's popular computing tasks pretty much require network. The NAT router gets you enough control. All other security measures are pointless because the kids have physical access to the machines; it doesn't take much to get Ubuntu running from a USB key with a spoofed MAC address...which negates almost anything you might set up on the clients anyway (and can bypass some NAT restriction configurations anyway).
A decent NAT router will have web access logs, so if the parent is paranoid, they can check up on what websites the clients have been going to, and also block specific sites. If necessary, the NAT router can block communications by port, too, to deny specific applications from working on the network, such as msn messenger, XBOX, bittorrent, or specific game protocols. In practice though, it's a pain to change that stuff all the time.
Technology isn't going to solve the parenting problem of the parent teaching the kids what is and is not appropriate. That requires the parent doing *parenting*. You've already failed if you have to resort to logging, blocking, and physical denial to reinforce consequences for going outside what is acceptable, more than once or twice.
Running a wireless router is like hosting a kegger at your house with music blaring to the whole block and a big sign saying "FREE BEER!", standing at your front door, and handing a cup of beer to anyone who shows up.
Your SSID broadcasts the presence of your network. (the loud music) It's unencrypted, and is handing out DHCP leases; (your agent handing out the beer at the front door to anyone)
The key is that when connecting to a wireless router, first you negotiate communications over the wireless. The devices negotiate the communication, and both devices agree to communicate using certain parameters; therefore the device has authorized you to communicate with it by negotiating the parameters. Then you request a DHCP lease, and by granting you one, the router has granted you permission to use the network. The device has negotiated communication on your behalf (you configured it to do so) and granted explicit permission to communicate over the network by doing so.
Sequence is: 1. Hey, can I talk to you wireless device? [device answers, yes, let's agree on a speed to talk at] 2. Hey wireless device, can I have an IP to talk on your network? [device answers yes, here is an ip so that you may communicate with any device on the network].
Looks an awful lot like explicit authorization to use the network to me..... The wireless device is your agent, if you don't want to explicitly allow all connections from random strangers then configure the device not to allow them on your behalf. If the device is configured to allow access like this then any reasonable person must conclude that the owner intends to share the device and the resources it is connected to with anyone who wants to use it.
Of course, the story changes if encryption is turned on, or if the device is configured to block certain mac addresses. Such a configuration clearly indicates that the intent of the owner is not to share.
Most CRM systems (big databases) *NO NOT* allow free query of the production database by the users because random, unoptimized queries will grind the database server to a halt. In fact, even on production servers, usually the template queries are so badly written that the thing runs terribly even without randomization from free queries.
Read only access to the database to someone writing poorly optimised queries is extremely dangerous. As an example, I've optimized poorly written queries before which previously ran in 45 minutes down to running in 45 seconds, on the same system. If someone runs a stupid query against the live database it can kill the performance for hours, and also may cause denial of service, timing bugs, and race conditions to be exposed in the applications dependent on the database due to the poor performance.
What most folks do in this situation is to allow free query of a second copy of the database which is not updated in realtime. This allows random queries for research to run against something that is not in production and will not have side effects for the other live database clients. This should be quite acceptable for random one-off research requests.
If Comcast does this, it won't piss off too many people *today* because 250GB/month is somewhat reasonable, *today*.
However, in the big picture, bandwidth usage rates are expected to continue to rise at the same rates as they have historically, which means that in not too long 250GB/month will be unreasonable. I doubt that Comcast has it in their plan to raise the caps to something more reasonable as typical usage starts to exceed any limits set now, say in three years.
Also, what is the UPLOAD cap? Typical game servers (with 30-60 users) such as NWN upload about 300gb/month. And don't forget about services such as MMOs..and VOIP....
They are a telecom; their whole business revolves around charging premium rates for decades old service technologies.
What the world really needs for the last mile problem is a cheap scaleable wireless router mesh project which would establish an Internet completely outside the control of anyone. You want access, get a router and join the community..... (would work very well in high population density areas, assuming the radio engineering side of it can be worked out and that you can get a decent range on the things, like 1-5 miles) Oh, and yes, no service fees because the hardware ownership is distributed....
Depends on the genre, or course, but CRPG genre games have a tendency to have ridiculously predictable plots stretched out over about 40 hours of play. They have no real surprises, and theatrically weak characters.
Your garden variety film is NOT really about *what* happens and the universe, it's about the persona of the characters and how seeing them move in a 2 hour scenario takes you on a trip. For a lot of folks, as John suggests, yeah, that's about the sex appeal and escapism, and simply "I wish I was that guy/gal". Your typical actor/director/producer isn't going to grok the feel that *you* got from a game character because *they don't play games*. Besides, the joy of getting your ass kicked into the dirt in a video game isn't terribly compatible with the "hero always wins" formula for a garden variety film.
And also, there is a need to "explain" the game world to "those who don't get it". That's a bunch of crap. Plop right in. The gamers will already know the backstory and don't want to have their hand held, and the rest of the folks aren't interested in the back story other than the feel for it that they pick up as the film develops, so they won't notice if you don't hold their hand.
The main thing that keeps *you* interested in the weak story of these games is that you are distracted by controlling the game and advancing your avatar. Making this worse, the majority of games these days have gotten away from focusing on character, story, and feel and instead pour most of the energy into the engine, specifically around cranking up the visual effects.
Your typical "game movie" is more about milking franchise icons than trying to write a good, satisfying story in the game universe. The way to go with gaming films is with small vigniettes in the universes. Tell a story in the universe, maybe about the "main characters", maybe not, but stay away from depicting the same exact events as implemented in the games.
Personally, I think the studios ought to look in to pen and paper role playing materials. Take a typical Shadowrun game (a cyberpunk pen and paper rpg). The game basically forces you to play interesting, shady characters with dirty backgrounds. The typical gameplay goes something like this: Meet up with the shady Johnson and take a run (a crimnal job for money); the players then spend an hour planning how to pull off the job; Then the action starts. Of course things never go according to plan, and it always goes south. That's where the fun is. Of course, the actual action is always short, intense, and would make for incredible cinematics. I've *never* played a shadowrun game session where I didn't leave thinking that it would have been awesome on screen. Then again, maybe my Shadowrun GM rocks =).
While the realtime gameplay for this usually takes about 6 hours, it typically simulates about ten minutes of "game time". Some decent writing to introduce and develop the characters, and a string of related runs would make a rather compelling film or series. And it can be about the characters and the choices they make rather than trying to explain the whole of some fantasy world in 2 hours while the generic "hot hero" type bumbles through.
Seriously, why ever close anything? Until your OS can't handle the number of open windows, anyway..... (or you launch a video game or 3d modeller that needs the resources)
I have an SGI Octane with dual MIPS R12k 300 mhz and it outperforms a P4 2ghz on floating point ops all the time. Integer performance isn't as good, but that's the way the things are built. Some of the reason it's faster at floating point than a much newer P4 is because each chip has 2MB cache, and MIPS chips have way more registers than intel+friends, but still...it's a ten year old machine.
Even if the MIPS implementation these guys are using is dated and has a teensy cache, 400-600 mhz MIPS would be roughly in the ballpark of a P3... and 64-bit to boot. And have a lot more registers, which makes it easier to write fast code because you dont have to swap things out of your primary (what, four? =P) registers to do anything, like on Intel + friends.
Certain people would like you to believe that because your video file is "digital" that you are able to make a perfect copy of it forever. This is simply not true most of the time.
Digital data does not degrade the same way that for example, generational VCR-dubbing degrades a signal (four generations is pretty crappy on VCRs), but there are similar gremlins which make it much less bulletproof than popular belief holds.
Take your average 1 gigabyte video file from the net.
Once you convert the thing into a compressed movie file (that's two generations of loss, first the a/d conversion then the first compression process.. three if you are cracking a typical DVD since you typically recompress afterwards to get smaller) the file is ready to go. Now, here's where people want you to think the losses stop. And they are right, under ideal conditions, namely you copying the file to another spot on your own hard drive. The probability of incorrectly copying without loss is pretty low in that case, unless you have a broken hard drive, which unfortunately, Joe Public who bought his PC at a department store, usually has because the manufacturers of those things use low bid hard drives.
Anyway, once you start transmitting that gigantic file over the internet, you introduce transmission errors, especially when the transfer is interrupted and restarted several times. Not all file transfer software *correctly* resumes files. Most assumes that the last byte recieved was in fact whole, when in fact it could have been less than all 8 bits....you have to use rollback to be *sure*.
Optical media such as CDs and DVDs are in fact an analog medium, in that the data is stored as a sequence of dots burned on the surface of the disk. If you *lose* some of that surface, you lose the file. (yes, you can pull most of it off with the right tools, but most people don't know Norton utilities from their ass)
I don't know about you, but my ten year old CD-Rs that I burned in 1995 are getting CD rot pretty bad. Some of them have the foil flaking off. I lost my original OEM Windows 98 CD to CD rot as well. There is a pinhead sized hole on the surface of the cd where the data layer flaked off. Naturally, that hole is right in the middle of a cab file, and well, you get the picture.
My twenty year old video tapes may be a little fuzzier than they were when they were new, but they are still watchable.
Also, if you burn at higher speeds, and most people do, the error rate for the data written to the cd is much higher. That means that as the CD ages and God knows what the color change chemicals on the data layer do over time, the error bits on the CD will protect your data much less of the time.
Formats such as video cd and DVD are more error tolerant because if a chunk of the data is missing or unreadable the playback device can happily corrupt the display until it hits the next keyframe in the file it is playing....becuase the encoding method is relatively simple.
More complicated compression formats are smaller, but not so error tolerant, hence you get things like those divx green-screen ghost-trip video files that are found on the p2p networks. Those can be repaired sometimes, but since the file format is corrupt most player software poops out when regular people try to watch.
IIRC, the video-cd and video-DVD formats actually use the error correction bits on the disc to store more movie data, because it's considered to be no big deal if the picture is fubared for a second or two. (that's what causes those rainbow squares and pauses on a mildly scratched DVD)
And I don't even want to go into all the things people do to video files to make them fit on media that is too small, like truncation and recompression.
The bottom line is that while yes, under the right conditions you can make perfect copies of video data after the initial two generations of loss, the media on which you usually store them is *less reliable* than video cassette....
Re:U U D D L R Start Select
on
Got Game
·
· Score: 2, Informative
U U D D L R SELECT START you insensitive clod!
You push select before start in order to get two player mode.....otherwise there's no reason to push select....
Well i dunno..... have you ever had to write a 4 page SQL query pulling data out of a bajillion tables? Had to rewrite it the hard way (using prepositional calculus) because your database didn't support the kind of joins that would have made the query easy? (in a database with properly normalised schema instead of one big table with all the data) It CAN take days to verify that the logic is correct and that the record set you are getting back is in fact what you think it is....
We as TV watchers want shows. We hate ads, but are willing to put up with them in order to get the "free gimmick", the show.
The shows are created merely as a side effect of advertising; the goal of the television industry is *NOT* to create quality programming for you. The goal is to sell advertising to corporations. What you as a TV watcher want is *completely* irrelevant.
At least in the US, without advertising, there would be no tv shows. Period. The show is just there to entice you into watching the advertising.
There is no real *money* to be made from the shows themselves directly; they are merely a means to an end for generating vastly more significant ad dollars.
As such, you can't tell *great* stories on TV like you can in books because you are beholden to the customer (corporate advertisers) to produce bland inoffensive content that leaves the consumer with a vaguely positive feeling after watching it.
If you controlled your copy of the show, you would not be generating new ad dollars every time you watched it, which is the business model these things operate on. You would also be less likely to watch it (and the ads) "just because it's on", further detracting from the ad revenue. DVD's of shows are a new concept, and they only come out after the up front ad dollars have been sucked out of the system.
While there may be a quite a few people who would like to own a copy of a tv show for their collection, there is no way the sales for copies can compare to the royalties and advertising revenue from broadcasters.
Even if the producers of shows *could* cut out the middlemen and offer their content as a free download with inserted advertising, they would be foolish to move to that model because they would only get paid for the show *once* (from the ads). With broadcasters, they generally get paid royalties each time the episode airs, and it is up to the broadcaster to make a profit on the ads.
One wonders why people are willing to both pay for cable/satellite TV service *and* deal with advertising. The advertising revenue should more than pay for the service; paying service fees on top of the inconvenience of be advertised to is outrageous.
Slashdot Mirror
Layer your firewalls like the design of a medieval keep. Exterior curtain wall, plus defensible keep.
You don't know whether threats come from inside or outside; therefore when in doubt firewall everywhere.
For most applications where CPU power matters, hyperthreading actually hurts performance, because the scheduler assumes that the hyper cores are real cores, when in fact they are not. A hyperthreaded core is not a full core, and will bottleneck waiting for the bits it doesn't have on the real core. I prefer not even having hyperthreaded cores unless the os and applications are aware that some cores are gimpy hyperthread cores and takes that into account; i.e does not schedule on them, or only schedules tasks that are not actually crippled by dependency on the real cores on the hyper cores.
Try running a database on hyperthreaded cores....it blows chunks. It's better if you just disable them in BIOS.
Yep. I have dental implants (titanium/ceramic composite teeth screwed into the skull) and a titanium plate with six screws in my arm. It may not be electronic, but I do have artificial replacement parts.
So there. Cyborg.
Whoever designed the system is terrible at computer science.
These are home users, using a government provider scanner, and id card, and a key.
Would be pretty easy to build a rootkit filter driver that steals the data off the card during legitimate transactions, along with a keylogger. At that point, you can pretty much remotely impersonate anyone whom you've rootkitted. Doesn't matter how secure the back end is because you can easily dupe the scanner side.
Terrible, terrible design by idiots....you can't trust home user systems to be clean of rootkits.
#1 You teach people that software is usually deterministic; if I follow known sequence of steps A, result B will occur. If not, something is usually wrong and its time to ask an expert for assistance.
#2 The way to learn technology is to play with it before crunch time. That means that when you get a new application in front of you, you should explore the menus, see what they can do and how the software behaves so that you can later on be creative when using it under pressure.
#3 Basic computer architecture; e.g. RAM is temporary working space, the hard disk is where your files are saved, etc. How to identify which part of the computer needs to upgrade; e.g. when to add RAM, when to replace the video card, when to get a new machine. How to navigate directories so you actually know where you saved your files (people over 40 have issues with this).
#4 Basic networking. What a NAT is, how to hook up a broadband router. How to determine how fast it really is and reason about whether you are getting what you pay for.
#5 Basics of what kind of problems are done well by computers and what kind of problems computers suck at. Basics of why some algorithms are faster than others at doing the same job. One of the biggest problems with business using computers is that people do not understand anything that does not scale linearly. Examples of applications used for various common types of tasks.
#6 How incredibly complex the stuff is under the hood....that when something is actually wrong with the software, it's extremely difficult to track it down.
#7 Algebra. No, really.
Duh, the extra whitespace helps you parse each sentence seperately.
It's much more mentally taxing to read a singlespaced block of mooshed together text. Also, typography concerns are nowadays about how to wrap text around oddly shaped illustrations in magazines and advertisements without strange gaps, not making actual text readable. The typography nerds may claim that the original standard before the advent of typewriters and modern computers is "more correct", but I rather think it's not due to objectiveness in analyzing which one is actually better.
At some point, the hassle of working with old junk and making it work, putting up with how slow it is, dealing with failing electronics, and so forth isn't worth it.
I have 17 Pentium 3 class systems in my basement in a render farm. Sure, it's neat to have so many systems. But for my purpose, a single $300 quad core box literally has more compute power, more memory, more memory bandwidth, and uses way less electricity. Plus you don't have to maintain a billion systems. And it takes up less space. And there's no heat problem. I haven't replaced the pile yet, because I'm not doing that much 3D lately, but I will, and it will be awesome to be rid of so much clutter. I also have a bunch of Sun boxes. They were fun to get working, but they use too much power, and it's an absolute hassle to fiddle with them, maintain software on several platforms, and so forth. My free time is valuable; I don't want to waste it doing menial maintenence on crappy hardware.
Off brand low end consumer gear is barely designed to last 3 years, let alone past life expectancy. Most of that 802.11b gear is pretty limited in what it can do, and barely worked when it was new. It's not like you can install dd-wrt and turn them into a mesh.
Best case scenario is probably hooking up somebody who has no wireless and no resources up, like your local church or whatever. If it breaks, meh, they had low expectations to begin with. It may not even be worth doing that though, because a lot of older consumer routers break when subjected to the network behavior of newer versions of Windows because they can't handle scaling window sizes with the default settings, and it's a support chore to dink around with the settings on every machine that comes along in a non-enterprise environment.
Bottom line is that old junk starts costing you more to use than buying new stuff would.
Yeah, I went to get firmware updates for some older Sun hardware I wanted to fix up and ran in to this too. Got the same "support contract is required for firmware updates" crap.
My 11 sun machines are headed to the dumpster, as a direct result of this policy. It's incredibly stupid. It's not like Sun was winning any new customers these days anyway, and now they'll bleed out the few they had. Obviously their intention is to kill off their hardware business, because no one in their right mind would decide to implement a policy like this.
Their model is vertical market with machiavellian control over customers. That model broke in the 1990's. Ridiculous to keep hanging on to it like a moldy wet blanket.
Actually, "the computer", and by extension "the internet" would be just another device to control as far as these interfaces go. In the various cybernetic monkey experiments, the robotic parts ultimately become an extension of the body; a two way digital communication link to the outside world would probably be adapted to in the same way; eventually the user would sort of "think at" the link to be able to read and write information. Would work a lot like io completion ports and device registers, which is how current drivers communicate with current devices. Bit of a challenge to work out a protocol for meaningful communication, but this part does get bits in and out.
This class of device is quite close to the minimum requirements for a direct neural interface. (Assuming the infection bits get worked out and so forth.) Would not want to be a beta tester though.
Video and image transfer requires a lot of bandwidth to transmit, but audio streams are nowhere near as intense. Won't take too many generations of the technology to get bandwith rates that are high enough to sustain a data stream the size of a voice data stream; the brain will pretty much automagically learn to interpret the data.
It is a long way from something like the Matrix, with total override of perception, though.
Sorry, not quite... sort of on the 120hz, no on the "free" stereo 3d from the console. Most "120hz" TV's actually just interpolate the intermediate frames in their processing chip. They can't actually display real incoming data at 120hz, they just fake it.
"All it requires" for free stereo 3d as described above is about 2x the work. Game console hardware already has to drop resolution to 720p to maintain 30fps for the usual level of detail in games these days. Make the graphics subsystem render 2x frames in the same time, and at best it's going to manage 15fps. Sure, you get to re-use some of the setup work (mostly moving data into the graphics subsystem) from the first angle, but the graphics subsystem still has to do the matrix transforms to render the scene, as well as run the shaders, which are likely not designed to look good from multiple angles, and may involve destructive calculations.
Usually in games there are distinct tasks that are parallel in nature, but within a given task there is an inherent degree of serialization. Most games have a simulation thread to update the game world state, and a seperate render thread that periodically draws a frame of the current state at say 30fps or so, depending on system load. But the render thread doesn't actually do the drawing; instead it gathers up the required 3d models, textures, scene data and whatever else, and feeds them to the GPU to draw. Rendering is embarrassingly parallel. Unfortunately, snapshotting the game world state to determine what to draw in a given frame is inherently serial. Recently, shaders have added the capability to do additional manipulation of the data on the GPU itself, which is faster because of the i/o bandwidth bottleneck. (talking to the slower main system bus). Anyway, a free core on the system does precisely nothing for you unless the game architecture is designed to parallelize the render thread to some degree, which most engines do not do. Instead, that core should be used for better AI or whatever. While game consoles are designed to have more bandwidth between the graphics subsystem, main memory, and the cpu, it's still the bottleneck, just not as bad as on PC hardware.
There are other engine differences you would have to build to do this as well, so it's hardly free, and that's not even thinking about memory constraints in the graphics subsystem, and whatever requirements there are for a 3d signal to the 3d TVs (presumably some sort of alternating frame format at a higher framerate; I haven't really looked in to it lately).
Gonna be a couple generations of the 3d display technology before the kinks are worked out, too. Probably the next generation of consoles will source it reasonably well. That's still 5-7+ years out though. Which is about the same timeframe that the no-glasses 3d TV tech is anticipated to appear. That means several years of crappy rapidly changing tech are in store.
Existing deployment tools from Microsoft already do this. You need the WAIK, which is a free download from Microsoft.
You need to create a generalized image. If you get all the required drivers for all your hardware into the driver store, the drivers will be found during install. You can also deploy from PXE boot using WDS with a generalized image...
There are a few caveats around a few drivers that aren't designed properly for Sysprep, and applications that aren't designed with sysprep in mind, but otherwise it's quite slick. You can script the installation of these exceptions to occur later on during deployment using unattend.xml and RunSynchronous commands though. You can also supply your licence key in the unattend.xml file.
About 90% of all Windows deployments are sysprepped by OEMs or by corporate IT folks....
Please read the documentation, the tools are quite flexible.
A decent NAT router will do all the internet control stuff you are looking for, with an interface that isn't too terrible to grok. The rest of what you're asking for starts to crank up the cost and complexity extremely fast, especially for a single parent with three teenage kids...
Really, you don't want to try to set up draconian enterprisy stuff, it's not designed for consumers, will take time to administrate, and will break on you anyway. Network control via NAT router should be sufficient. If the parent wants to be able to physically control computer use too, then set up the clients up with the hard disk in a removable bay carrier. Yank it when you want to deny access to the machine totally. Much cheaper than setting up a domain and controlling access with accounts, and more reliable.
Seriously though, today's popular computing tasks pretty much require network. The NAT router gets you enough control. All other security measures are pointless because the kids have physical access to the machines; it doesn't take much to get Ubuntu running from a USB key with a spoofed MAC address...which negates almost anything you might set up on the clients anyway (and can bypass some NAT restriction configurations anyway).
A decent NAT router will have web access logs, so if the parent is paranoid, they can check up on what websites the clients have been going to, and also block specific sites. If necessary, the NAT router can block communications by port, too, to deny specific applications from working on the network, such as msn messenger, XBOX, bittorrent, or specific game protocols. In practice though, it's a pain to change that stuff all the time.
Technology isn't going to solve the parenting problem of the parent teaching the kids what is and is not appropriate. That requires the parent doing *parenting*. You've already failed if you have to resort to logging, blocking, and physical denial to reinforce consequences for going outside what is acceptable, more than once or twice.
Running a wireless router is like hosting a kegger at your house with music blaring to the whole block and a big sign saying "FREE BEER!", standing at your front door, and handing a cup of beer to anyone who shows up.
Your SSID broadcasts the presence of your network. (the loud music)
It's unencrypted, and is handing out DHCP leases; (your agent handing out the beer at the front door to anyone)
The key is that when connecting to a wireless router, first you negotiate communications over the wireless. The devices negotiate the communication, and both devices agree to communicate using certain parameters; therefore the device has authorized you to communicate with it by negotiating the parameters. Then you request a DHCP lease, and by granting you one, the router has granted you permission to use the network. The device has negotiated communication on your behalf (you configured it to do so) and granted explicit permission to communicate over the network by doing so.
Sequence is: 1. Hey, can I talk to you wireless device? [device answers, yes, let's agree on a speed to talk at] 2. Hey wireless device, can I have an IP to talk on your network? [device answers yes, here is an ip so that you may communicate with any device on the network].
Looks an awful lot like explicit authorization to use the network to me..... The wireless device is your agent, if you don't want to explicitly allow all connections from random strangers then configure the device not to allow them on your behalf. If the device is configured to allow access like this then any reasonable person must conclude that the owner intends to share the device and the resources it is connected to with anyone who wants to use it.
Of course, the story changes if encryption is turned on, or if the device is configured to block certain mac addresses. Such a configuration clearly indicates that the intent of the owner is not to share.
Done.
Most CRM systems (big databases) *NO NOT* allow free query of the production database by the users because random, unoptimized queries will grind the database server to a halt. In fact, even on production servers, usually the template queries are so badly written that the thing runs terribly even without randomization from free queries.
Read only access to the database to someone writing poorly optimised queries is extremely dangerous. As an example, I've optimized poorly written queries before which previously ran in 45 minutes down to running in 45 seconds, on the same system. If someone runs a stupid query against the live database it can kill the performance for hours, and also may cause denial of service, timing bugs, and race conditions to be exposed in the applications dependent on the database due to the poor performance.
What most folks do in this situation is to allow free query of a second copy of the database which is not updated in realtime. This allows random queries for research to run against something that is not in production and will not have side effects for the other live database clients. This should be quite acceptable for random one-off research requests.
Cheers!
It's a long term strategy.
If Comcast does this, it won't piss off too many people *today* because 250GB/month is somewhat reasonable, *today*.
However, in the big picture, bandwidth usage rates are expected to continue to rise at the same rates as they have historically, which means that in not too long 250GB/month will be unreasonable. I doubt that Comcast has it in their plan to raise the caps to something more reasonable as typical usage starts to exceed any limits set now, say in three years.
Also, what is the UPLOAD cap? Typical game servers (with 30-60 users) such as NWN upload about 300gb/month. And don't forget about services such as MMOs..and VOIP....
They are a telecom; their whole business revolves around charging premium rates for decades old service technologies.
What the world really needs for the last mile problem is a cheap scaleable wireless router mesh project which would establish an Internet completely outside the control of anyone. You want access, get a router and join the community..... (would work very well in high population density areas, assuming the radio engineering side of it can be worked out and that you can get a decent range on the things, like 1-5 miles) Oh, and yes, no service fees because the hardware ownership is distributed....
As if people up to this sort of nefarious thing would carry bluetooth devices, or use non-disposable cell phones.....
Obviously prisons are a poor solution to the "undesirable waste persons" created by our economic system.
Perhaps more resources should go into fixing the socio-economic situation that drives the behavior that gets people into the prisons....
Depends on the genre, or course, but CRPG genre games have a tendency to have ridiculously predictable plots stretched out over about 40 hours of play. They have no real surprises, and theatrically weak characters.
Your garden variety film is NOT really about *what* happens and the universe, it's about the persona of the characters and how seeing them move in a 2 hour scenario takes you on a trip. For a lot of folks, as John suggests, yeah, that's about the sex appeal and escapism, and simply "I wish I was that guy/gal". Your typical actor/director/producer isn't going to grok the feel that *you* got from a game character because *they don't play games*. Besides, the joy of getting your ass kicked into the dirt in a video game isn't terribly compatible with the "hero always wins" formula for a garden variety film.
And also, there is a need to "explain" the game world to "those who don't get it". That's a bunch of crap. Plop right in. The gamers will already know the backstory and don't want to have their hand held, and the rest of the folks aren't interested in the back story other than the feel for it that they pick up as the film develops, so they won't notice if you don't hold their hand.
The main thing that keeps *you* interested in the weak story of these games is that you are distracted by controlling the game and advancing your avatar. Making this worse, the majority of games these days have gotten away from focusing on character, story, and feel and instead pour most of the energy into the engine, specifically around cranking up the visual effects.
Your typical "game movie" is more about milking franchise icons than trying to write a good, satisfying story in the game universe. The way to go with gaming films is with small vigniettes in the universes. Tell a story in the universe, maybe about the "main characters", maybe not, but stay away from depicting the same exact events as implemented in the games.
Personally, I think the studios ought to look in to pen and paper role playing materials. Take a typical Shadowrun game (a cyberpunk pen and paper rpg). The game basically forces you to play interesting, shady characters with dirty backgrounds. The typical gameplay goes something like this: Meet up with the shady Johnson and take a run (a crimnal job for money); the players then spend an hour planning how to pull off the job; Then the action starts. Of course things never go according to plan, and it always goes south. That's where the fun is. Of course, the actual action is always short, intense, and would make for incredible cinematics. I've *never* played a shadowrun game session where I didn't leave thinking that it would have been awesome on screen. Then again, maybe my Shadowrun GM rocks =).
While the realtime gameplay for this usually takes about 6 hours, it typically simulates about ten minutes of "game time". Some decent writing to introduce and develop the characters, and a string of related runs would make a rather compelling film or series. And it can be about the characters and the choices they make rather than trying to explain the whole of some fantasy world in 2 hours while the generic "hot hero" type bumbles through.
Seriously, why ever close anything? Until your OS can't handle the number of open windows, anyway..... (or you launch a video game or 3d modeller that needs the resources)
I have an SGI Octane with dual MIPS R12k 300 mhz and it outperforms a P4 2ghz on floating point ops all the time. Integer performance isn't as good, but that's the way the things are built. Some of the reason it's faster at floating point than a much newer P4 is because each chip has 2MB cache, and MIPS chips have way more registers than intel+friends, but still...it's a ten year old machine.
Even if the MIPS implementation these guys are using is dated and has a teensy cache, 400-600 mhz MIPS would be roughly in the ballpark of a P3... and 64-bit to boot. And have a lot more registers, which makes it easier to write fast code because you dont have to swap things out of your primary (what, four? =P) registers to do anything, like on Intel + friends.
Certain people would like you to believe that because your video file is "digital" that you are able to make a perfect copy of it forever. This is simply not true most of the time.
Digital data does not degrade the same way that for example, generational VCR-dubbing degrades a signal (four generations is pretty crappy on VCRs), but there are similar gremlins which make it much less bulletproof than popular belief holds.
Take your average 1 gigabyte video file from the net.
Once you convert the thing into a compressed movie file (that's two generations of loss, first the a/d conversion then the first compression process.. three if you are cracking a typical DVD since you typically recompress afterwards to get smaller) the file is ready to go. Now, here's where people want you to think the losses stop. And they are right, under ideal conditions, namely you copying the file to another spot on your own hard drive. The probability of incorrectly copying without loss is pretty low in that case, unless you have a broken hard drive, which unfortunately, Joe Public who bought his PC at a department store, usually has because the manufacturers of those things use low bid hard drives.
Anyway, once you start transmitting that gigantic file over the internet, you introduce transmission errors, especially when the transfer is interrupted and restarted several times. Not all file transfer software *correctly* resumes files. Most assumes that the last byte recieved was in fact whole, when in fact it could have been less than all 8 bits....you have to use rollback to be *sure*.
Optical media such as CDs and DVDs are in fact an analog medium, in that the data is stored as a sequence of dots burned on the surface of the disk. If you *lose* some of that surface, you lose the file. (yes, you can pull most of it off with the right tools, but most people don't know Norton utilities from their ass)
I don't know about you, but my ten year old CD-Rs that I burned in 1995 are getting CD rot pretty bad. Some of them have the foil flaking off. I lost my original OEM Windows 98 CD to CD rot as well. There is a pinhead sized hole on the surface of the cd where the data layer flaked off. Naturally, that hole is right in the middle of a cab file, and well, you get the picture.
My twenty year old video tapes may be a little fuzzier than they were when they were new, but they are still watchable.
Also, if you burn at higher speeds, and most people do, the error rate for the data written to the cd is much higher. That means that as the CD ages and God knows what the color change chemicals on the data layer do over time, the error bits on the CD will protect your data much less of the time.
Formats such as video cd and DVD are more error tolerant because if a chunk of the data is missing or unreadable the playback device can happily corrupt the display until it hits the next keyframe in the file it is playing....becuase the encoding method is relatively simple.
More complicated compression formats are smaller, but not so error tolerant, hence you get things like those divx green-screen ghost-trip video files that are found on the p2p networks. Those can be repaired sometimes, but since the file format is corrupt most player software poops out when regular people try to watch.
IIRC, the video-cd and video-DVD formats actually use the error correction bits on the disc to store more movie data, because it's considered to be no big deal if the picture is fubared for a second or two. (that's what causes those rainbow squares and pauses on a mildly scratched DVD)
And I don't even want to go into all the things people do to video files to make them fit on media that is too small, like truncation and recompression.
The bottom line is that while yes, under the right conditions you can make perfect copies of video data after the initial two generations of loss, the media on which you usually store them is *less reliable* than video cassette....
U U D D L R SELECT START you insensitive clod!
You push select before start in order to get two player mode.....otherwise there's no reason to push select....
Well i dunno..... have you ever had to write a 4 page SQL query pulling data out of a bajillion tables? Had to rewrite it the hard way (using prepositional calculus) because your database didn't support the kind of joins that would have made the query easy? (in a database with properly normalised schema instead of one big table with all the data) It CAN take days to verify that the logic is correct and that the record set you are getting back is in fact what you think it is....
We as TV watchers want shows. We hate ads, but are willing to put up with them in order to get the "free gimmick", the show.
The shows are created merely as a side effect of advertising; the goal of the television industry is *NOT* to create quality programming for you. The goal is to sell advertising to corporations. What you as a TV watcher want is *completely* irrelevant.
At least in the US, without advertising, there would be no tv shows. Period. The show is just there to entice you into watching the advertising.
There is no real *money* to be made from the shows themselves directly; they are merely a means to an end for generating vastly more significant ad dollars.
As such, you can't tell *great* stories on TV like you can in books because you are beholden to the customer (corporate advertisers) to produce bland inoffensive content that leaves the consumer with a vaguely positive feeling after watching it.
If you controlled your copy of the show, you would not be generating new ad dollars every time you watched it, which is the business model these things operate on. You would also be less likely to watch it (and the ads) "just because it's on", further detracting from the ad revenue. DVD's of shows are a new concept, and they only come out after the up front ad dollars have been sucked out of the system.
While there may be a quite a few people who would like to own a copy of a tv show for their collection, there is no way the sales for copies can compare to the royalties and advertising revenue from broadcasters.
Even if the producers of shows *could* cut out the middlemen and offer their content as a free download with inserted advertising, they would be foolish to move to that model because they would only get paid for the show *once* (from the ads). With broadcasters, they generally get paid royalties each time the episode airs, and it is up to the broadcaster to make a profit on the ads.
One wonders why people are willing to both pay for cable/satellite TV service *and* deal with advertising. The advertising revenue should more than pay for the service; paying service fees on top of the inconvenience of be advertised to is outrageous.
Food for thought...