Unfortunately, this sort of thing - "Proles! Welcome Big Brother into your home! Allow our cameras/microphones/tracking devices into your lives! We will give you shiny things!" - is becoming ever more common.
Where I am (Central KS) the radio stations are running an ad, trying to drum up "volunteers" to for "an exciting study of a potential new way to fund the highway system". Guess what: they want people to voluntarily submit to putting a GPS logger on their car, to track where they drive, for the purpose of "usage-based taxation". The very idea of which "makes me ill and angry" to quote The Outer Limits! But the idea that people, rather than rising up with metaphorical pitchforks and torches, are VOLUNTEERING to have this... Folks, we are on the downhill slide - get used to it.
Or rather, DON'T get used to it! Fight it at every step!
OK AMD, I understand some of your reasoning, but as an embedded designer, I cannot stress this enough:
If you don't support Linux, I won't design your parts in.
From the linked slick: "Delivers a Microsoft® Windows Vista® Premium experience, handling the rich Aero user interface and extremely demanding application workloads with ease"
OK, I understand why you have to add that - and in and of itself, that's OK, as many of the target users will be running Windows.
However, what I *don't* see is anything like:
"Supports EXA, OpenGL, V4L2, and GStreamer under Linux."
Indeed, I don't find "Linux" anywhere on the slick. Sorry, but the gear I design needs a longer lifespan for the software - including the OS - than Microsoft is able to give me (yes, I know they "promise cross our hearts and hope to die pinky swear" to support their 'embedded' OS'es for a long time, but history shows otherwise).
"Senior manager for product marketing Xavier Lauwaert stated that the QA engineers did this to make the systems more profitable by creating an artificial differentiation we can use to charge more money for basically the same thing."
The S/PDIF system standard used over those connections tops out at a bitrate that's too low for the newer Dolby Digital TrueHD and DTS HD audio formats used on Blu-Ray discs.
You are both right and wrong. True, there is no way S/PDIF can handle the full rate 5.1 uncompressed format that is one of the many formats that HDMI can support.
However, my Blu-Ray player is QUITE capable of driving a DTS signal across S/PDIF to the stereo, because that is exactly how it is connected right now.
You can configure the PS3 to send the DTS audio across the HDMI connection without change - Hence my statement "pass [the] audio unmolested....": in that case all that is needed is for the TV to get out of the way and pass the audio stream it is receiving over to the optical link.
"I have a Logitech wireless mouse and keyboard set that is 6 years old. Works just fine 6 feet away from the receiver."
Emphasis mine. The older stuff was usefuf; the newer stuff is not. However, since the thrust of this story is "buying new stuff", unless somebody can find a source for 6 year old gear....
I bought a new HDTV, so let me share some of my experiences and things to look out for:
1) DLNA. This is *supposed* to be a standard built on top of uPnP, allowing a TV to access a media server on the local network. It sounds good in theory: let your MythTV box stream directly to your TV. In practice, it is not worth very much: the set of video formats that a DLNA TV supports is VERY MUCH smaller than the set of formats you see in practice on your media files. In fact, my Samsung TV cannot even play back content streamed over the network that it can play if that same file is placed on a USB flash drive and plugged directly into the TV. My advice is don't plan on using DLNA, plan on hooking up a real computer.
2) HDMI inputs. Again, my TV has 4 inputs - 3 more than I need. The TV will NOT take the digital audio from an HDMI source - for example, Blu-Ray audio from my PS3 - and pass that audio unmolested through to the optical output connecting the TV to the amplifier. As a result, all I would get from any game or from most Blu-Ray disks was the left and right channels passed on to the stereo - no sub, no surround, no center channel. And the TV does NOT have a six channel audio output - only 2. So I end up having to do all the switching at the stereo, and then pass everything on to the TV - so I really only need one HDMI input.
3) HDMI-CEC. In theory, this allows the TV's remote to control other things, like a Blu-Ray player or a stereo, by passing the command data down the HDMI cable. The PS3 does not support this, nor does my lower-end (but brand new) Sony stereo. Maybe if you buy all of your gear from the same manufacturer, and you buy higher end gear this works, but beware. Plan on either having many remotes, or buying a smart remote and training it (and because the PS3 uses Bluetooth rather than IR, plan on your universal remote not controlling the PS3).
4) Internet through your TV. Two words: Flash Javascript. Your TV will likely not support EITHER of them very well. Again, plan on an external computer. And DON'T plan on using a wireless mouse or keyboard - those things are so range-crippled now that unless you are within a couple of feet of the receiver (and I mean that literally: less than 4 feet!) they won't work (and that's not some no-name keyboard: that's a Logitech).
5) Linux. Yes, my Samsung runs Linux. It is cool to see the GPL in the manual. Other than that - it really matters not at all. The TV application itself is NOT FLOSS, nor is the Linux any kind of a standard distribution: You aren't going to do a "apt-get install $FOO" here. The system doesn't implement any of the "standard" things you might want, like SSH or X. It doesn't even support any file system on external devices other than VFAT (so no larger-than-4G files using EXT2).
All in all, my advice is: treat the TV as a monitor - it's job it to show pixels. Don't expect the TV to make sound - you'll have an external amp or receiver for that. Don't expect the TV to surf the web - you'll want a real computer with a real operating environment for that. Don't expect the TV to play media files - again, you'll want a real computer for that.
Cut those features from the TV, and spend the money you save putting together a media PC as a companion for the TV.
"We know relatively well how much mass supergiants loose[sic], and how it ends up in the interstellar medium as planetary nebulae," Pierre Kervella of the Paris Observatory tells Astronomy Now.
Emphasis mine.
IT IS "LOSE", NOT "LOOSE"!!!!!!
This "infection" of the world with slashdot-itis MUST BE STOPPED!
Drives the government further into debt, at a time when the "value" of a government bond is approaching junk bond status.
Causes people to go further into debt, at a time when consumer debt is at an insanely high level, and there is still the possibility of more people losing their jobs.
Gives (borrowed) money to the car manufacturers, many of whom are NOT US entities (follow the money).
Removes money from programs (like renewable energy) that WILL create wealth in the US.
Yes, this sounds like a brilliant idea to me.
And on the subject of "improving efficiency of the fleet" - look at the relatively low mileage targets the program has: they consider 26MPG highway to be an improvement? If they REALLY wanted to improve the fleet mileage, they would have insisted upon any car being purchase having at least 40MPG highway.
Sorry, this is just the "bread" part (with the ongoing MJ crap being the "circuses" part).
To have a such dumb programmer that expects that memory allocations can be entirely managed by other layers ?
That's funny because any good C++ programmer will take advantage of auto_ptrs which effectively do manage memory for you automatically.
However, the idea is that an auto_ptr "knows" much more about what the programmer is doing than many of the "garbage collection" systems that try to infer it by reading tea leaves^W^Wthe stack frames.
So you are both right: garbage collection that doesn't get enough data from the programmer sucks, and programmers that rely upon it suck - but auto_ptr and Boost's smart_ptr's don't belong to that category.
"Do you really want to hold back all the other very important features like lambda, rvalue references, variadic templates, type deduction etc. just because of concepts?"
Unfortunately, without concepts, many of the templates that would make features like those REALLY powerful aren't implementable due to silly things like the compiler insisting upon being able to instantiate member functions that don't make sense for a class and won't be used, just because there isn't a means to tell the compiler "and if this member doesn't make sense, just don't instantiate it, and throw an error IF AND ONLY IF somebody tries to use it." (and yes, I know about SFINAE, but that gets REALLY UGLY to do).
I've been bit by this, where I ended up having to create two completely separate template classes, one for objects that don't have sub-members and one for objects that do, just because I couldn't tell the compiler "Look, if operator . doesn't exist for this method, then don't worry about it!"
That said: I will say, I felt that some of the implementation details of concepts felt "forced" to me, in the same way that the streams library feels "forced": they "hacked" (in the bad way) the library in using language semantics that weren't a good fit.
<sigh/> - I hope the extra time will allow them to put a bit more polish on how you actually express things, and make it feel less "forced"....
That is a good philosophical question. First, I'd say that in raising that question, you in many ways validate my statement, because in many ways a file system IS a database - so the "put everything in a database" types are simply using the wrong sort of a data base.
Using a relational database to store a "path to blob" mapping vs. using a file system is a question of the overhead of the actual mapping: locating a file in a directory structure vs. doing the SQL pattern matching. Due to the issues of concurrency and locking, the SQL check is frequently more expensive, especially when you have to do an out-of-process operation (passing the message from the web server over to the database program) vs. an in-process file system lookup. Add to that the cost of moving a large amount of data (the blob itself) from the database process to the web server process to the kernel TCP stack vs. being able to do a simple send_file() operation on the inode from the web server process, and the conventional database vs. a file system becomes the difference between being Slashdotted vs. surviving.
And THAT is the real problem: rather than using a very cheap sort of database (a file system) they use an expensive data base, for no real advantage: they aren't saying "give me all items that match this pattern" but rather "give me THE SINGLE ITEM that matches this pattern, where it is guaranteed that there will be zero or one match, not a multiplicity of matches", they aren't saying "give me all items that match THIS pattern AND THAT pattern", etc.
Now, if we were to see the adoption of means to use relational database operations to find inodes (as an alternative to a simple tree structure like directories), where the web server could do a data base lookup without any context switches, get an inode, then use normal operations (like send_file() ) on that inode, then that would make a great deal of sense.
You're right, but it's not the AJAX that is the cause. It's the incompetent "web developers" who think AJAX is going to solve all of their problems....
Almost, but I would also lay the blame on the "when all you have is a hammer, all the world's a nail" developers who insist upon putting EVERYTHING into a database, because "that's how you make websites DUH LOL KTHXBYE". So every image, every bit of text, every CSS file, all require accessing a database, rather than a simple pull of a file.
Sure, you can have a database table that maps images to URLs, so that the code that IS accessing the database to build the page can find what URL to embed, and you can change which image to use by altering the database table, but the ACTUAL IMAGE FETCH should be nothing more than pulling a file from the web server's file system. Nice, simple, and it allows the web server to correctly handle IF-MODIFIED-SINCE and Expires.
And I have to wonder if having the database execute a command on modification, to build simple static HTML which can then be served out ad infinitum to the world until the next time the database changes wouldn't also help speed things up.
After all, things like that work for Slashdot - why NOT look into them for other sites and use them where appropriate?
I'm rather disappointed that The Kansas Cosmosphere and Space Center isn't listed: there is NO other museum where you can see an authentic V1 and V2 under the same roof, let alone an SR-71, Apollo 13, plus more Russian gear than anyplace outside of Moscow.
"If I had an access point that could reliably do virtual SSIDs... I would likely set up a separate open network that used a transparent proxy to do logging so that anyone could use it."
That is an INCREDIBLY BAD IDEA. Do NOT log anything - for 2 reasons:
1) There is the possibility of logging information that could be considered "private" - in the (admittedly unlikely) event that somebody caught wind of it, you might find yourself on the receiving end of a civil suit and/or a criminal charge. The fact such a charge would likely be found unwarranted in no way mitigates the expense and hassle of dealing with it.
2) Should the police trace something to your connection, they WILL subpoena those logs. If and when they don't find what they are looking for, they will assert you have what they want and didn't surrender it, and they will tear your place apart looking for it. If you keep the logs for $TIMEPERIOD, and the event happened at $TIMEPERIOD+$POSITIVEDELTA ago, they will STILL want the logs - and the statement "I deleted them" will be portrayed at "COLLABORATION with TERRORISTS and PAEDOPHILES!" (again, it doesn't matter if they ultimately succeed in that, the cost of defending against it will be staggering).
Ultimately, the rule to follow is "Record NOTHING you don't want to see on the 6 o'clock news, or have used against you in a court of law."
'finite but unbounded" - in terms of the Turing machine this is a meaningless statement. Yes, you can say the surface of an orange is finite but unbounded, but for storage, you cannot just keep reusing what you have, like retreading the surface of an orange.
Thus, you have to have storage that is NOT finite - storage that cannot run out. In other words, infinite.
Just because an ideal Turing machine has access to "infinite" storage does not mean that all algorithms will use it. I could design a Turing machine that could play Tic-Tac-Toe in a finite and well defined amount of storage, for example.
Consider the following example: In theory, a Sinclair ZX-81 could do computation my 4G RAM 340G disk Core 2 Duo laptop could do (albeit no where near as quickly), given only some form of storage interfaced to it to allow it to access enough information. Now, that information is going to be larger than 340G+4G (by how much it would be hard to say), but given something like, oh, say an infinitely long tape and reader (where did that idea come from?) it could do it. Likewise, the emulated machine could run Firefox, which could run Javascript, which, again given a simulated infinite storage, could emulate the ZX-81 emulating the Core 2 Duo running Firefox running Javascript emulating....
Normally I don't see or respond to ACs, but I will second this: the OMAP3 and the Beagleboard are very interesting, but trying to get the video acceleration and the DSP going is WAY too hard, due to the closed source parts. Imagitec, if you are reading this: You make money selling the core IP to the companies, not by selling the drivers - MAKE THE DAMN DRIVERS OPEN AND GET THEM INTO XORG!
And while I am ranting, a bit pointed towards all the folks working on the Free software items, like kernel drivers: PUSH IT UPSTREAM! Get the kernel drivers into the mainline kernel source, rather than forcing me to play patch-monkey from half-a-dozen branches to get power management, 2D video drivers, DSP Bridge, and all the other features working in ONE kernel at the same time.
Slashdot Mirror
Unfortunately, this sort of thing - "Proles! Welcome Big Brother into your home! Allow our cameras/microphones/tracking devices into your lives! We will give you shiny things!" - is becoming ever more common.
Where I am (Central KS) the radio stations are running an ad, trying to drum up "volunteers" to for "an exciting study of a potential new way to fund the highway system". Guess what: they want people to voluntarily submit to putting a GPS logger on their car, to track where they drive, for the purpose of "usage-based taxation". The very idea of which "makes me ill and angry" to quote The Outer Limits! But the idea that people, rather than rising up with metaphorical pitchforks and torches, are VOLUNTEERING to have this ... Folks, we are on the downhill slide - get used to it.
Or rather, DON'T get used to it! Fight it at every step!
OK AMD, I understand some of your reasoning, but as an embedded designer, I cannot stress this enough:
If you don't support Linux, I won't design your parts in.
From the linked slick:
"Delivers a Microsoft® Windows Vista® Premium experience,
handling the rich Aero user interface and extremely
demanding application workloads with ease"
OK, I understand why you have to add that - and in and of itself, that's OK, as many of the target users will be running Windows.
However, what I *don't* see is anything like:
"Supports EXA, OpenGL, V4L2, and GStreamer under Linux."
Indeed, I don't find "Linux" anywhere on the slick. Sorry, but the gear I design needs a longer lifespan for the software - including the OS - than Microsoft is able to give me (yes, I know they "promise cross our hearts and hope to die pinky swear" to support their 'embedded' OS'es for a long time, but history shows otherwise).
So, to summarize:
BAD AMD! WRONG OS! NO BISCUIT!
"Senior manager for product marketing Xavier Lauwaert stated that the QA engineers did this to make the systems more profitable by creating an artificial differentiation we can use to charge more money for basically the same thing."
The 1990's called, and they want their business models back (if they'll give me my 401K back it's a deal!)
So, we can either:
or
and have something we can use free of charge.
Hmmmm. I wonder which one will win out.
Does anybody else see the family resemblance to the CueCat business model here?
The S/PDIF system standard used over those connections tops out at a bitrate that's too low for the newer Dolby Digital TrueHD and DTS HD audio formats used on Blu-Ray discs.
You are both right and wrong. True, there is no way S/PDIF can handle the full rate 5.1 uncompressed format that is one of the many formats that HDMI can support.
However, my Blu-Ray player is QUITE capable of driving a DTS signal across S/PDIF to the stereo, because that is exactly how it is connected right now.
You can configure the PS3 to send the DTS audio across the HDMI connection without change - Hence my statement "pass [the] audio unmolested....": in that case all that is needed is for the TV to get out of the way and pass the audio stream it is receiving over to the optical link.
"I have a Logitech wireless mouse and keyboard set that is 6 years old. Works just fine 6 feet away from the receiver."
Emphasis mine. The older stuff was usefuf; the newer stuff is not. However, since the thrust of this story is "buying new stuff", unless somebody can find a source for 6 year old gear....
I bought a new HDTV, so let me share some of my experiences and things to look out for:
1) DLNA. This is *supposed* to be a standard built on top of uPnP, allowing a TV to access a media server on the local network. It sounds good in theory: let your MythTV box stream directly to your TV. In practice, it is not worth very much: the set of video formats that a DLNA TV supports is VERY MUCH smaller than the set of formats you see in practice on your media files. In fact, my Samsung TV cannot even play back content streamed over the network that it can play if that same file is placed on a USB flash drive and plugged directly into the TV. My advice is don't plan on using DLNA, plan on hooking up a real computer.
2) HDMI inputs. Again, my TV has 4 inputs - 3 more than I need. The TV will NOT take the digital audio from an HDMI source - for example, Blu-Ray audio from my PS3 - and pass that audio unmolested through to the optical output connecting the TV to the amplifier. As a result, all I would get from any game or from most Blu-Ray disks was the left and right channels passed on to the stereo - no sub, no surround, no center channel. And the TV does NOT have a six channel audio output - only 2. So I end up having to do all the switching at the stereo, and then pass everything on to the TV - so I really only need one HDMI input.
3) HDMI-CEC. In theory, this allows the TV's remote to control other things, like a Blu-Ray player or a stereo, by passing the command data down the HDMI cable. The PS3 does not support this, nor does my lower-end (but brand new) Sony stereo. Maybe if you buy all of your gear from the same manufacturer, and you buy higher end gear this works, but beware. Plan on either having many remotes, or buying a smart remote and training it (and because the PS3 uses Bluetooth rather than IR, plan on your universal remote not controlling the PS3).
4) Internet through your TV. Two words:
Flash
Javascript.
Your TV will likely not support EITHER of them very well. Again, plan on an external computer. And DON'T plan on using a wireless mouse or keyboard - those things are so range-crippled now that unless you are within a couple of feet of the receiver (and I mean that literally: less than 4 feet!) they won't work (and that's not some no-name keyboard: that's a Logitech).
5) Linux. Yes, my Samsung runs Linux. It is cool to see the GPL in the manual. Other than that - it really matters not at all. The TV application itself is NOT FLOSS, nor is the Linux any kind of a standard distribution: You aren't going to do a "apt-get install $FOO" here. The system doesn't implement any of the "standard" things you might want, like SSH or X. It doesn't even support any file system on external devices other than VFAT (so no larger-than-4G files using EXT2).
All in all, my advice is: treat the TV as a monitor - it's job it to show pixels. Don't expect the TV to make sound - you'll have an external amp or receiver for that. Don't expect the TV to surf the web - you'll want a real computer with a real operating environment for that. Don't expect the TV to play media files - again, you'll want a real computer for that.
Cut those features from the TV, and spend the money you save putting together a media PC as a companion for the TV.
I just want to know how this bug got out of testing. You'd think "causes AIDS" would be a showstopper
Thus sayeth the Lord:
"It is no bug, yea verily, it is a feature."
Unfortunately, that is exactly what some organized religions would assert....
Which libraries? libxml2, expat, or some other library?
The last I'd checked, Python could use several XML libraries, and Sun distributed several libraries.
It would be nice if TFA had told us which libraries, or had a link to the actual report listing them.
"We know relatively well how much mass supergiants loose [sic], and how it ends up in the interstellar medium as planetary nebulae," Pierre Kervella of the Paris Observatory tells Astronomy Now.
Emphasis mine.
IT IS "LOSE", NOT "LOOSE"!!!!!!
This "infection" of the world with slashdot-itis MUST BE STOPPED!
And in East Elbonia, they are planning on dominating the world's energy market in 20 years....
So this "stimulus" money:
Yes, this sounds like a brilliant idea to me.
And on the subject of "improving efficiency of the fleet" - look at the relatively low mileage targets the program has: they consider 26MPG highway to be an improvement? If they REALLY wanted to improve the fleet mileage, they would have insisted upon any car being purchase having at least 40MPG highway.
Sorry, this is just the "bread" part (with the ongoing MJ crap being the "circuses" part).
Your rong, you looser! An apo'strophe mean's "LOOK OUT! HERE COME'S AN "'S"!!!"
'So it 'should have been:
Nobody e'scape's the apo'strophe Nazi's!
To have a such dumb programmer that expects that memory allocations can be entirely managed by other layers ?
That's funny because any good C++ programmer will take advantage of auto_ptrs which effectively do manage memory for you automatically.
However, the idea is that an auto_ptr "knows" much more about what the programmer is doing than many of the "garbage collection" systems that try to infer it by reading tea leaves^W^Wthe stack frames.
So you are both right: garbage collection that doesn't get enough data from the programmer sucks, and programmers that rely upon it suck - but auto_ptr and Boost's smart_ptr's don't belong to that category.
"Do you really want to hold back all the other very important features like lambda, rvalue references, variadic templates, type deduction etc. just because of concepts?"
Unfortunately, without concepts, many of the templates that would make features like those REALLY powerful aren't implementable due to silly things like the compiler insisting upon being able to instantiate member functions that don't make sense for a class and won't be used, just because there isn't a means to tell the compiler "and if this member doesn't make sense, just don't instantiate it, and throw an error IF AND ONLY IF somebody tries to use it." (and yes, I know about SFINAE, but that gets REALLY UGLY to do).
I've been bit by this, where I ended up having to create two completely separate template classes, one for objects that don't have sub-members and one for objects that do, just because I couldn't tell the compiler "Look, if operator . doesn't exist for this method, then don't worry about it!"
That said: I will say, I felt that some of the implementation details of concepts felt "forced" to me, in the same way that the streams library feels "forced": they "hacked" (in the bad way) the library in using language semantics that weren't a good fit.
<sigh/> - I hope the extra time will allow them to put a bit more polish on how you actually express things, and make it feel less "forced"....
That is a good philosophical question. First, I'd say that in raising that question, you in many ways validate my statement, because in many ways a file system IS a database - so the "put everything in a database" types are simply using the wrong sort of a data base.
Using a relational database to store a "path to blob" mapping vs. using a file system is a question of the overhead of the actual mapping: locating a file in a directory structure vs. doing the SQL pattern matching. Due to the issues of concurrency and locking, the SQL check is frequently more expensive, especially when you have to do an out-of-process operation (passing the message from the web server over to the database program) vs. an in-process file system lookup. Add to that the cost of moving a large amount of data (the blob itself) from the database process to the web server process to the kernel TCP stack vs. being able to do a simple send_file() operation on the inode from the web server process, and the conventional database vs. a file system becomes the difference between being Slashdotted vs. surviving.
And THAT is the real problem: rather than using a very cheap sort of database (a file system) they use an expensive data base, for no real advantage: they aren't saying "give me all items that match this pattern" but rather "give me THE SINGLE ITEM that matches this pattern, where it is guaranteed that there will be zero or one match, not a multiplicity of matches", they aren't saying "give me all items that match THIS pattern AND THAT pattern", etc.
Now, if we were to see the adoption of means to use relational database operations to find inodes (as an alternative to a simple tree structure like directories), where the web server could do a data base lookup without any context switches, get an inode, then use normal operations (like send_file() ) on that inode, then that would make a great deal of sense.
You're right, but it's not the AJAX that is the cause. It's the incompetent "web developers" who think AJAX is going to solve all of their problems....
Almost, but I would also lay the blame on the "when all you have is a hammer, all the world's a nail" developers who insist upon putting EVERYTHING into a database, because "that's how you make websites DUH LOL KTHXBYE". So every image, every bit of text, every CSS file, all require accessing a database, rather than a simple pull of a file.
Sure, you can have a database table that maps images to URLs, so that the code that IS accessing the database to build the page can find what URL to embed, and you can change which image to use by altering the database table, but the ACTUAL IMAGE FETCH should be nothing more than pulling a file from the web server's file system. Nice, simple, and it allows the web server to correctly handle IF-MODIFIED-SINCE and Expires.
And I have to wonder if having the database execute a command on modification, to build simple static HTML which can then be served out ad infinitum to the world until the next time the database changes wouldn't also help speed things up.
After all, things like that work for Slashdot - why NOT look into them for other sites and use them where appropriate?
The did well to list The Cosmosphere, but I'm surprised they didn't mention The Salt Mine Museum just down the road, or Big Brutus over in the eastern part of the state.
I'm rather disappointed that The Kansas Cosmosphere and Space Center isn't listed: there is NO other museum where you can see an authentic V1 and V2 under the same roof, let alone an SR-71, Apollo 13, plus more Russian gear than anyplace outside of Moscow.
"If I had an access point that could reliably do virtual SSIDs ... I would likely set up a separate open network that used a transparent proxy to do logging so that anyone could use it."
That is an INCREDIBLY BAD IDEA. Do NOT log anything - for 2 reasons:
1) There is the possibility of logging information that could be considered "private" - in the (admittedly unlikely) event that somebody caught wind of it, you might find yourself on the receiving end of a civil suit and/or a criminal charge. The fact such a charge would likely be found unwarranted in no way mitigates the expense and hassle of dealing with it.
2) Should the police trace something to your connection, they WILL subpoena those logs. If and when they don't find what they are looking for, they will assert you have what they want and didn't surrender it, and they will tear your place apart looking for it. If you keep the logs for $TIMEPERIOD, and the event happened at $TIMEPERIOD+$POSITIVEDELTA ago, they will STILL want the logs - and the statement "I deleted them" will be portrayed at "COLLABORATION with TERRORISTS and PAEDOPHILES!" (again, it doesn't matter if they ultimately succeed in that, the cost of defending against it will be staggering).
Ultimately, the rule to follow is "Record NOTHING you don't want to see on the 6 o'clock news, or have used against you in a court of law."
Sad, but true.
Better still, set your SSID to "AreYouUpForAnalSex" - imagine the cop going door-to-door on THAT one.
I suspect they'd just let that one slide....
'finite but unbounded" - in terms of the Turing machine this is a meaningless statement. Yes, you can say the surface of an orange is finite but unbounded, but for storage, you cannot just keep reusing what you have, like retreading the surface of an orange.
Thus, you have to have storage that is NOT finite - storage that cannot run out. In other words, infinite.
Just because an ideal Turing machine has access to "infinite" storage does not mean that all algorithms will use it. I could design a Turing machine that could play Tic-Tac-Toe in a finite and well defined amount of storage, for example.
Consider the following example: In theory, a Sinclair ZX-81 could do computation my 4G RAM 340G disk Core 2 Duo laptop could do (albeit no where near as quickly), given only some form of storage interfaced to it to allow it to access enough information. Now, that information is going to be larger than 340G+4G (by how much it would be hard to say), but given something like, oh, say an infinitely long tape and reader (where did that idea come from?) it could do it. Likewise, the emulated machine could run Firefox, which could run Javascript, which, again given a simulated infinite storage, could emulate the ZX-81 emulating the Core 2 Duo running Firefox running Javascript emulating....
Normally I don't see or respond to ACs, but I will second this: the OMAP3 and the Beagleboard are very interesting, but trying to get the video acceleration and the DSP going is WAY too hard, due to the closed source parts. Imagitec, if you are reading this: You make money selling the core IP to the companies, not by selling the drivers - MAKE THE DAMN DRIVERS OPEN AND GET THEM INTO XORG!
And while I am ranting, a bit pointed towards all the folks working on the Free software items, like kernel drivers: PUSH IT UPSTREAM! Get the kernel drivers into the mainline kernel source, rather than forcing me to play patch-monkey from half-a-dozen branches to get power management, 2D video drivers, DSP Bridge, and all the other features working in ONE kernel at the same time.
I found a paypal link off the main site page:
http://wildfiregames.com/0ad/page.php?p=6438
True, but I was trying to simplify my post a bit - any strategy game, RTS or TBS, are like crack for me. There aren't many of either for Linux....