I know, I know, this is Slashdot, but you really need to read articles before commenting on them, because/. article summaries, are, as a rule, always incorrect in some important detail (yes, hopeless, I admit, and I also admit to making the same mistake in the past, but. ..).
From the fine article:
"The proponents of tallow-based fuel admit that raising livestock in order to burn their corpses for energy would be a very carbon-intensive way of making biofuel. Rearing cattle or pigs involves the emission of lots of greenhouse gases. But that's not the idea: rather, the thinking goes, people will raise livestock anyway in order to eat it. Thus it makes sense to use the waste products for energy.
Nobody is suggesting that it makes sense to raise livestock exclusively for fuel use. What they are proposing is using the millions of gallons of waste fat and oil created as a byproduct of food and leather production worldwide, every year, as a fuel source. This is, really, a very ancient idea. Almost all peoples around the world (except those few ethnic groups, such as Hindus, that may have been almost exclusively vegetarian) used the fats and oils from the animals they used for food and hide, as a source of light and heat - whale blubber, bison fat, caribou fat, etc. have been rendered into lamp oils and candles for thousands of years.
But, it's true that this is not really a 'solution', because there is not enough waste oil to provide all of the heat that is necessary for buildings in cold climates the world 'round. However, it does make sense to use those waste products to the extent that they can be, to heat some buildings.
My main problem with a lot of O/S'es and Linux distros these days as that too much functionality is 'default on'. If a user needs MySQL, or network printing, they can turn it on, but it seems to me that having the OS install with as few background services as feasible running, is a great way to get OS'es both more secure, and more scalable. In addition, a little bit of engineering might be able to go a long way - for example, I've noticed over the last few releases of Ubuntu that the Gnome environment seems to be taking up a lot more background processes and memory than it used to. Is all that stuff in the background really necessary? Ok, I realize some of it is no doubt necessary (sound daemons, etc), but couldn't a lot of that stuff be loaded 'on-demand' as it were, and unloaded after a period of inactivity? For example - if I'm not sharing a printer on the netwrk, and I'm not currently printing any documents, does CUPS or any other printing system need to be loaded in memory? Why not load it when I actually try to send a print job from an application to the printer (this does, I realize, imply that there is a different background process extremely similar in concept to inetd which is monitoring for activity and loading the appropriate process on demand - but really, for services which aren't heavily used, what is wrong with the inetd model; I do realize that under heavy usage, the inetd approach becomes inefficient due to the overhead of starting and stopping processes, but I think that on a lot of 'personal' desktop/laptop/netbook situations, the usage would only be very occasional)?
Anyhow, you might be right that no real progress will be made on this front, but I still hold out hope - even on modern systems with lots of RAM, there is a benefit to keeping the memory usage low - it leaves more memory available for the actual applications you are using, whether that is a large database, a CAD system, 3D-or-2D graphics apps (Blender, Gimp, etc), video/audio editting, games, whatever. I believe that keeping a minimum 'background' memory profile is always a good idea for O/Ses, because people don't use O/Ses - they use applications.
Man, I gotta assume that programs running as Windows Services must be excluded because, if services were included, in addition to the things the parent mentioned. ..
* CD Burning packages (Roxio, Nero, et al) often put 1 or more services in the background (like a service to let you burn files 'on-the-fly' to a CD instead of mastering the Disc; recent versions of Roxio include a network media sharing system to share music and videos with other users on your LAN, etc)
* iTunes installs 2 or 3 background services (bonjour, iPod service, mobile device service)
* Some applications like Java, Firefox, and OpenOffice have 'Quick Start' apps which run in the system tray, whose purpose is to keep the core of the app loaded into memory for faster 'startups'.
* etc, etc.
I suspect that if a 'quick start' application is NOT running as a service, but instead is running in application mode, it will count against this 3 app limitation.
Basically, though, I have to agree with other posters - I think Microsoft is royally shooting itself in the foot here. I mean, I understand Microsoft wanting to create some differentiation between different editions of Windows, so that they have the ability to create different price levels. But multi-tasking is a core feature of the O/S which has been around since Windows 1.0 in 1988 (or whatever - I don't remember the exact dates). Limiting the number of applications artificially is just plain dumb.
If Microsoft wants differentiation, they should base it on the apps which are included in the base install of the O/S. Like, maybe the Starter edition doesn't get any games (or only 1 game, like solitaire). Of course, the problem with that route is that as soon as Microsoft removes that stuff, people will just download free versions off the Internet. I was trying to think of things MS could remove from the base install, and all of them, except for, *maybe* Windows Media Player, could easily be replaced by third-party programs. The only reason WMP couldn't easily be replaced is because of WMP audio/video streams and files on websites, which other players cannot (legally) support. But removing WMP from Windows Starter is a losing proposition for Microsoft - the only reason that anyone uses WMP is because they know it is likely to be installed on user's computers. If Microsoft unbundles WMP, suddenly MP3/Mpg, RealAudio/RealVideo, iTunes AAC, etc become just as likely (or possibly more likely, in the case if iTunes) for a user to have the capability to play, so why would websites bother with WMP support any more? So, Microsoft just cannot unbundle WMP - it's bundled status is the only thing it has going for it, really.
The only other things I can imagine microsoft pulling out of the starter edition are things they already pulled out of the Home Basic version in XP, or the equivalent Vista SKU - things like file encryption, IIS, various other network services - which most home users wouldn't miss anyhow.
Actually, there is *one* thing I can think of Microsoft pulling out of a "Basic" edition which *might* drive home users to upgrade - Windows networking support - that is, the ability to access network shares, or share directories from the local computer to the network, and the ability to print to a network printer or share a network printer.
However - the problem with that is that, I suspect, third-parties will step in with their own solutions to that problem (like a network file browser based on a port of Samba, or a third-party device driver to map a network drive which looks like a local drive to Windows, or a printer driver that appears to be a local printer to Windows, but really is printing to an SMB shared printer, or even just prints using something like the old lpd protocol, or IPP.
There is almost nothing Microsoft can remove from Windows, that third parties won't replace either for free, or ver
If they are not licensed, then by what right do you expect them to not get interference? One unlicensed user has just as much right to the spectrum as another. If the mike's were digital, they could, I think, happily co-exist with other digital users of that white-space spectrum. Outside of ham-bands, I begin to think that analog radio devices will quickly become a think of the past - the problem with analog stuff is that basically only one user (or one group of users) can use a certain frequency at a time. With packetized digital communications (and/or spread-spectrum techniques), multiple groups of users can share bandwidth at the same time.
Shared use, seems like a much more equitable use of limited spectrum resources, than the old analog 'dedicated channel' model. Still, if you are someone with analog A/V equipment, I suppose it would kind of stink. I suppose an equitable compromise might be to start by making sales of such 'legacy' equipment illegal, but continue to allow it's use for some number of years, to allow people and companies who've invested in such equipment to 'get their monies' worth' out of it, with a plan to make the continued use of such equipment illegal after a cutoff date in like 5 or 10 years.
I've been kind of wondering the same thing. I tried configuring Firefox to report itself as IE7 on Vista, to see if Netflix would work, but then I got an error page that ActiveX support must be enabled. Haven't had time to chase that any further. It occurs to me that Wine might be able to be used as the basis for for a Firefox ActiveX plugin under Unix/Linux/*BSD, but I don't know if anyone has actually tried to create such a beast.
Really simple way to illustrate the problem to people - If you were to use the full capacity of your connection, you would only be allowed about 15 hours a month of activity. Now, to be somewhat fair to Charter, et. al., most people don't come close to tapping the actual capacity of their internet connection anyhow - in my experience, most file servers and web sites won't download files to me at close to the full connection speed. Streaming HD video only needs about 2.0Mbps, I think. VoIP/Webcams take maybe 1Mbps. Online games need low latency, but I don't think they actually use up that much bandwidth - maybe what, like 100-200 Kbps?
15Mbps service is nice, though, because it does mean I can be using Teamspeak, downloading files in the background, and watching an HD movie all at the same time.
I suppose these caps will hit people who heavily use P2P file sharing though. You know, I wonder - do these caps include same-network traffic? I think it would be tremendously intelligent (so it probably won't happen) for ISPs like Charter, Time-Warner, Comcast, etc, to try to work with P2P software providers to develop clients that preferentially 'bias' P2P traffic to stay within the same ISP network where possible. That is, if a torrent or other p2p client where trying to find a source to download the files from, it would try to pick other clients on the same network to transfer from - of course, that's only helpful when you have a lot of people seeding.
Still, the point is, I think that the ISPs could find P2P technologies to be tremendously beneficial for themselves and their customers, if they implemented things correctly. You could extend the P2P concept to cache-ing mechanisms (something like a p2p version of squid, perhaps). I really think that the benefits of p2p technologies are dramatically under-utilized, currently, because most bean-counters hear p2p and think 'copyright thieves'.
Blizzard is one company that has at least 1/2 a clue - it's my understanding that their updated for WoW uses bittorrent to distribute updates to their millions of users - what a great way to help make sure that users don't have to wait for hours for their computer to be able to download updates on patch day because of an overloaded update server (an experience I've had on several occasions with other MMO's that used a 'centralized' update server that simply couldn't handle 20000+ users all trying to connect simultaneously and download a 100MB+ file).
Can someone explain to me under what legal doctrine this does *not* violate the 4th Amendment? Does this fall under 'probable cause' since the person has been arrested? I could see taking a DNA sample upon arrest and checking it against a database of crime-related DNA samples (that is, DNA samples taken from a scene of a crime or victim [like skin under the fingernails of a person who was assaulted and scratched the assailant]), but it shouldn't be stored in a general purpose DNA database unless the accused is actually convicted (it would probably be stored, of course, in the case files for the case under investigation).
While off-the-shelf PC games might not work that great on this combo, I suspect that there could be really good, beautiful looking games created and fine-tuned just for such a platform. Someone could possibly use this as the basis for a portable entertainment system to compete with the Nintendo DS, PSP, etc. XBox Portable anyone? If you consider that Atom-based systems would typically have smaller screens, so that you might be looking at a resolution of maybe 640x480 or 800x600 (or perhaps wide-screen aspect ratios, but with similar vertical resolution), I could see this being a sufficient solution for such a portable entertainment system.
Yeah. . . first you need some clouds to seed. I'm not sure, but I believe that you will not often find the confluence of clouds suitable for seeding at the same time and location that you have a wildfire?
Students at the University of Nitwitshire make history by transmitting digital data wirelessly using a radio transmitter and receiver they built themselves.
"Considering that you can deliver a ~1kt warhead from off of a satellite platform..."
Is that an actual explosive warhead, or is that just dropping an object from several miles in the sky? Seems to me that if you have a sufficiently massive object (which doesn't necessarily have to be very large - maybe 100 or 200 kg), and it doesn't burn up during re-entry, it's gonna do some substantial damage just from kinetic energy at impact. It'll probably reach terminal velocity before it reaches the ground, so we probably cannot use E = m g h for the energy of the projectile, but we can use E = m v^2 (if we knew the terminal velocity of the projectile). I don't know the terminal velocity, but I can guess it would be moving pretty fast if dropped from a satellite, so it's gonna have a pretty substantial v^2 value. It might not have the energy to destroy a whole city, but I bet it could take out a couple city blocks. . .
Of course, all this assumes that the Iranians have the materials science and engineering capability to actually build something that won't burn up during re-entry, and can be accurately delivered to a target from the satellite.
The main problem I could see with cell jamming during a terrorist or similar criminal situation is that there is a small possibility that maybe, one of the victims could be trying to secretly call 911 (or whatever the local equivalent is) to try to give police information about the situation inside the building (or vehicle, etc).
Thank you for this reply. It seems to be somewhat rare to get such a well reasoned rebuttal, and I do appreciate it, because I get to learn new things.
Like, I confess I didn't realize how close to epic failure the containment systems at TMI came - but, still, it's also true that the designed safety features were at least partly to credit for the containment of the radiation. (The expression "necessary but not sufficient" comes to mind).
As for the IFR, I also appreciate your critiques of that design. Honestly, I've been trying to find critical reviews of the IFR design, but it's been hard for me to find any information from groups that normally provide 'hostile analysis' of nuclear technology, wrt the IFR design. I didn't realize that the IFR only had a design lifetime of 40 years.
I think it'll be a long time before we can build a system that lasts 1000+ years, but it'd be good progress if we could be designing systems to last for 100+ years (that also helps to make them more economically feasible, as long as it doesn't cost exponentially more money to make it last longer).
I also want to point out that I didn't bring up the IFR because I'm convinced it is necessarily the solution, but really just as an example of a newer reactor design which has enough differences from old reactor designs that it needs to be criticized or praised for it's own problems or benefits, because some or all of the complaints raised by the OP did not apply to it.
I pretty much have the same feeling as you with regards to nuclear tech - continue to create new designs, build 1 or 2 prototypes as necessary to test and study the designs, but don't start building lots of new reactors until we've solved a lot more of the problems. So, from that perspective, I think it was a massive mistake for the Clinton administration to mothball the IFR project, simply because we should have finished the R&D so that we could learn as much as possible from that project. The IFR 'base design', I bet, could have been improved and refined (like, for example, replacing Sodium coolant with some other cooling system to improve the safety of the system, etc). Go through multiple iterations of refinement and improvement, and possibly down the road you come out with a design that is appropriate for large-scale, safe, economical, commercial implementation.
As for Solar and Wind, I'm all for that too - I just think that in our future energy 'mix', we'll likely need some Nuclear to provide a solid foundation of 'baseline' power, and even if we don't, I still don't think trying to contain something for 100,000 years is feasible. Mankind has no example of anything it has created that has lasted for 100k years. That doesn't necessarily mean it can't be done, but I'm quite skeptical of the feasibility of such an undertaking. I think it makes more sense to 'burn' the waste in something like an IFR, to deplete the transuranic elements (which, from all the reading I've done about this - I'm not a physicist, but I've tried, and continue to try, to educate myself on these topics - is what makes nuclear waste so bad for so long), so that we only have to contain the remaining waste in your 'granite mountain' facility for a few centuries instead of thousands of centuries. I'm fairly confident we could secure nuclear waste for 2 or 3 centuries.
Don't get me wrong, I'm all for increasing solar, wind, ocean-based, and other 'passive' power systems. But, all the people talking about wind and solar seem to always leave off some important problems - That solar panel which is producing 1kw (or whatever) at noon on a clear sunny day might only be producing 150W on a cloudy day, and nothing at night. That wind turbine which is producing 1kw on a nice windy day might be producing nothing on a stagnant day.
Now, I believe the counterargument to that is the idea that if you have enough wind farms and enough solar panels around the country and around the world, "It's always sunny/windy *somewhere*", but then you have problem of transmitting power from where you can produce it, to the place where it needs to be employed, and until we have high-temperature superconductors, that means we are suffering somewhat significant power losses during transmission. Also, you still need some way to ensure a stable baseline of power - power that you can count on producing a minimum amount, all hours of the day or night, every day of the year. Coal, oil, nuclear, and geothermal offer that - wind and solar really do not - they are spikey.
Finally, have environmentalists considered the impact of the land use necessary to produce electricity on the scale our nation needs using solar and wind? We're talking about putting arrays of turbines and solar panels (or other solar technologies - PhotoVoltaic panels aren't the only solar energy generation systems) on very large areas of land - what affects will that have on birds, insects, animals, native ecologies? How many birds will be hacked to death by wind turbines, or cooked alive by thermal solar systems? Maybe bird migrations will be confused by all the glare from PV panels? I mean, who knows what the impact will be of putting the enormous amounts of land necessary to power our nation to use as wind and solar farms?
Also, have you considered that, while the USA maybe has lots of undeveloped land in sunny deserts that are ideal for solar power, maybe other nations don't have such good conditions for solar power? Where are the UK, France, Germany, etc going to build their solar and wind farms? India? China? I suppose you can probably put lots of solar panels on roofs of buildings , so that does mean that you can use some already developed land as part of your solar farms, but I'm not sure you can get enough panels in place just doing that? Maybe, but I suspect that buildings will not be sufficient alone (I think about it this way - my understanding is that a solar panel on the roof of a typical house, commercial building, or skyscraper cannot provide enough power for that house, building or skyscraper, so it stands to reason that panels on the roof of every building cannot provide enough power for every building).
"[1st and 2nd generation Thermal] Nuclear power [are] a dead end. No new [1st or 2nd Generation Thermal] nuclear plants have been built in 25 years". . .
There, fixed it for you. Yes, old reactor designs are a dead end. They are prone to a risk of melt-down (though that risk has been, mostly, successfully managed for the past 30 years; yes, Three Mile Island was a problem, but, keep in mind that even with the TMI incident, the safety features of that reactor design prevented an escape of radiation when the melt-down did occur), they only extract a miniscule amount of the potential energy available in the fuel, and they create waste that "would have to be kept under armed guard forever".
Nuclear physicists and engineers have continued to do R&D for the past 30 years, and they are proposing *new* ideas. When new ideas are presented, you can't just assume that the same arguments that were valid criticisms of the previous designs continue to be valid for the new designs.
We have, right now, a Nuclear Waste problem, because of those previous generations of dead-end reactor designs, that must be dealt with. Putting the stuff into storage for 100000 years is not really a solution. The only real solution to the nuclear waste problem is to further process it to make the waste 'safe' and short lived.
Now, I do not really know if the design proposed in this article is "the solution" or not. Maybe it is. There was also a solution proposed in the 1990s, called the Integral Fast Reactor, which was essentially melt-down proof - not because of systems put in place to prevent a possible melt-down, but because it used a different Nuclear Reaction called a Fast Nuclear reaction, instead of the older Thermal Nuclear reaction, and was such that if the reactor increased in temperature beyond the normal operating temperature, the reaction actually choked itself, sort of like a candle sealed in a glass container. They even successfully tested the design, by purposefully cutting off the cooling to a prototype reactor that the DoE built out in the desert somewhere, and it did, in fact, shut itself down as it is designed to do.
The IFR design was also based around the concept of using our existing waste stockpiles as *fuel* for the reactor, producing hundreds of times more energy from that fuel, than older 'conventional' reactors do, which should have made it much more economically feasible.
The reason I mention the Integral Fast Reactor, is that is an example of a new design which I've studied more about than this new fission-fusion hybrid in the article, which demonstrates that the old arguments don't *necessarily* apply to new designs. Every proposal must be studied and evaluated on it's own merits - you can't just make a sweeping statement that Nuclear power is a dead end.
Unfortunately, the IFR project (which was being conducted by the Department of Energy) was canceled by the Clinton administration because of the same knee-jerk reaction to all Nuclear technology, exhibited by the parent, instead of really considering the IFR design on it's own merits or problems.
Also, in regards to this new technology, it sounds like they are not necessarily proposing to build new plants, but to 'upgrade' existing plants. If we can upgrade the already built plants in such a way as to reduce our existing waste stockpiles, where is the downside? True, this new design, as with any new design, needs to be thoroughly evaluated and proven, and also compared to other proposals (for example, we should consider if this proposed design is actually superior to the IFR design - if not, we should be restarting the IFR project instead, perhaps) before we role it out to any large scale.
*Maybe* we should have never gotten into the business of Nuclear Fission, but the fact remains that we have all this waste that we need to do something with. Why not 'burn' it in a new reactor type in such a way that we produce significantly less toxic, shorter lived waste? Environmentalists should be proponents of finding ways to deal with our nuclear waste problem, not object to every single proposal with a blanket statement that nuclear power is a dead end and re-hashing the same old tired arguments regardless of whether or not they apply to the new proposals.
You're right, but it's a shame - spectrum doesn't generally go to those who can/will use it best in the US.
It's like the spectrum auctions the FCC had a couple years ago - When Verizon, or T-Mobile pay $4 Billion (Or 6 Billion or whatever) for licenses to a chunk of spectrum, all I can see as a result of that is that those operators have to squeeze an additional $4 Billion of revenue out of their customers, which means everyone pays higher prices for mobile services. What I see is the government colluding with mobile operators to guarantee that service is expensive, and to prevent small companies from starting up competing mobile services (coming up with Billions for a spectrum license is not something any small company could ever do).
I wish more power to you in your company's pursuit to make a solid business as a Wireless ISP, and I do wish that companies that get the best results at the lowest cost to consumers would get more spectrum, but I don't see that becoming reality any time soon.
I think the Physics department at my university has done a good job of applying this principle. They used an online homework platform for most of the assignments for the General Physics classes I had to take as part of my engineering program. Now the thing about homework is, you don't really learn much from problems which you solved wrong, and then only learned you solved them wrong when they are marked incorrect by your teacher (which is the 'traditional' method). Usually, in such a traditional method, the teacher will hopefully spend some time in class going over problems students got wrong, so they could at least see the correct way to solve the problem, but it's still not as good a learning experience as when you figure out how to do it yourself.
With this online homework platform, students got up to 5 chances to get the solution correct. That is, if you solved the problem, inputted your answer in the website, and your answer was not correct, it would notify you the answer was not correct (but would not tell you what the right answer is), and give you a chance to try again. I personally found this to be an excellent teaching tool - simply knowing that my answer was not right gave me the chance to go back and look at the problem, and try to figure out where my mistake was. Almost all of the time, I could figure out my mistake, and correct it within 1 or 2 tries - something I never had the opportunity to do in classes I took earlier in my academic career. I truly believe that having those additional chances to correct my own mistakes helped me to learn the material better.
This is one example of something that is not easy to do without computers, but quite easy to do with computers.
"It comes from the (frankly quite bizarre) idea that it is somehow unethical to randomise educational methods."
But, isn't that exactly what is happening here? We have this IT director from a UK school, asking/. about ideas for how to setup computer systems to somehow improve the educational experience of their students. It sounds like the educational methods are being randomized (to the extent that they aren't necessarily in uniformity to the methods in place at other UK schools) in this situation.
I don't think this is, already, that much different, in practice, from what I've described, except to the extent that I'm suggesting that instead of setting things up in a particular way just because someone thinks it's a good idea, that it be done as part of a controlled experiment designed to validate or invalidate a hypothesis regarding educational outcomes. That it be done in such a way that some meaningful metrics be obtained regarding the effectiveness (or lack thereof), for the benefit of other schools in the future.
You know, this is one of those areas where I wish educational systems would take a more scientific approach to these types of problems. It's pretty much the same way here in the U.S.
There seems to be little 'method' to the ways we try to figure out the 'best' ways to integrate IT into education.
It seems to me that in situations like this, schools could benefit from systematically applying the scientific method - Observations, Hypothesis, Prediction, Experiment, Analysis. (Repeat as necessary.)
Start building a *theory* of education and IT, and then make your school IT decisions and budgets based upon the body of theory thus developed.
So, this means that you gather lots of ideas from all sorts of people (everyone from Education Ph.Ds, down to teachers and IT staff at the schools, even to interested members of the public who have ideas) about how IT could be better implemented in education, and start using a small number of schools as experimental test beds (and other schools as 'control data' for the experiments).
These experiments should be, first, submitted to and approved by some national 'school board', or at least something like a group of professors at a University education department, who are tasked with tracking and eventually reporting on the results. From the results, this 'school board', or university task force, or whoever is responsible, can start creating recommendations and best practices.
Seriously, not being dependent on foreign companies for critical national technological infrastructure is in the strategic national interests of every nation on earth. If you are a foreign nation, how do you know that the OS you are getting from $OS_Vendor doesn't have 'wiretaps', back-doors, remote kill switches, or other secrets in the software which $OS_Vendor, or the nation to which $OS_Vendor is based out of, can use to cripple you? Another problem is, that $OS_Vendor could simply stop providing you with necessary patches to update known problems and vulnerabilities in the OS.
One possible solution would be, if you are using a closed-source vendor, to require that vendor to provide the government with buildable source code, which could be reviewed by your own Computer Scientists, then built by your government, and distributed throughout the nation. This also allows your developers to provide your nation with patches and support if you are cut off from support from $OS_Vendor. That is not true Open-Source, but that is still, effectively, a "National Operating System". Open Source is one step better though, because you have, potentially, a lot larger base of people that are reviewing the code. That whole Eric Raymond thing to the effect that with sufficiently many eyes, all bugs are shallow.
Just saying that some foreign leader that is not well liked has something in common with another leader is sort of mis-leading, because there will often be many things in common between good leaders and bad leaders - what's important often isn't the similarities, but the differences.
I think I might be about ready to give up on PC gaming completely, lol. When I bought the computer, 2 years ago, 1 Gig of ram was quite a lot (though there were some systems selling with 2G, that was about the top).
PC Gaming is the only consumer market that thinks it's reasonable that something you buy today, will be almost completely unable to use content produced only 1 year after you buy the device, without constant upgrading. That'd be like an iPhone that couldn't play some of the songs on iTMS after a year, or a TV, which couldn't display some TV shows after 1 year.
At least with consoles, you get about 3 years before a system gets replaced with a 'new generation', and when it does, consoles are (comparatively) cheap.
That said, I'll probably at least spend the money to upgrade to 2 or 4 Gigs of RAM, as that is pretty cheap, but when I reach the day that this computer can no longer play most of the new games coming out, I think I'll just go the console route, like most other gamers.
Yes, it was about 2 years ago, I guess. I have, actually, been thinking about upgrading the RAM. The thing is, most games play OK for me, even on that system. Like I said, I know it's not top of the line, but games need to run on systems that are not top of the line, too. Historically, Valve games have run pretty well on reasonably older hardware. Half-Life 2, the Episodes, and Portal seemed to all do OK on my laptop (after I turned down the graphics some), just not TF2.
As for the nVidia, the GPU claims to have 256MB of it's own RAM. It should not be shared memory, so I should really have 1 Gig (of course, Windows XP and background services seem to use a good 400M+ when I'm running nothing else, so that is cutting me down to about 500M of useable RAM for games).
Slashdot Mirror
I know, I know, this is Slashdot, but you really need to read articles before commenting on them, because /. article summaries, are, as a rule, always incorrect in some important detail (yes, hopeless, I admit, and I also admit to making the same mistake in the past, but. . .).
From the fine article:
Nobody is suggesting that it makes sense to raise livestock exclusively for fuel use. What they are proposing is using the millions of gallons of waste fat and oil created as a byproduct of food and leather production worldwide, every year, as a fuel source. This is, really, a very ancient idea. Almost all peoples around the world (except those few ethnic groups, such as Hindus, that may have been almost exclusively vegetarian) used the fats and oils from the animals they used for food and hide, as a source of light and heat - whale blubber, bison fat, caribou fat, etc. have been rendered into lamp oils and candles for thousands of years.
But, it's true that this is not really a 'solution', because there is not enough waste oil to provide all of the heat that is necessary for buildings in cold climates the world 'round. However, it does make sense to use those waste products to the extent that they can be, to heat some buildings.
My main problem with a lot of O/S'es and Linux distros these days as that too much functionality is 'default on'. If a user needs MySQL, or network printing, they can turn it on, but it seems to me that having the OS install with as few background services as feasible running, is a great way to get OS'es both more secure, and more scalable. In addition, a little bit of engineering might be able to go a long way - for example, I've noticed over the last few releases of Ubuntu that the Gnome environment seems to be taking up a lot more background processes and memory than it used to. Is all that stuff in the background really necessary? Ok, I realize some of it is no doubt necessary (sound daemons, etc), but couldn't a lot of that stuff be loaded 'on-demand' as it were, and unloaded after a period of inactivity? For example - if I'm not sharing a printer on the netwrk, and I'm not currently printing any documents, does CUPS or any other printing system need to be loaded in memory? Why not load it when I actually try to send a print job from an application to the printer (this does, I realize, imply that there is a different background process extremely similar in concept to inetd which is monitoring for activity and loading the appropriate process on demand - but really, for services which aren't heavily used, what is wrong with the inetd model; I do realize that under heavy usage, the inetd approach becomes inefficient due to the overhead of starting and stopping processes, but I think that on a lot of 'personal' desktop/laptop/netbook situations, the usage would only be very occasional)?
Anyhow, you might be right that no real progress will be made on this front, but I still hold out hope - even on modern systems with lots of RAM, there is a benefit to keeping the memory usage low - it leaves more memory available for the actual applications you are using, whether that is a large database, a CAD system, 3D-or-2D graphics apps (Blender, Gimp, etc), video/audio editting, games, whatever. I believe that keeping a minimum 'background' memory profile is always a good idea for O/Ses, because people don't use O/Ses - they use applications.
Man, I gotta assume that programs running as Windows Services must be excluded because, if services were included, in addition to the things the parent mentioned. . .
* CD Burning packages (Roxio, Nero, et al) often put 1 or more services in the background (like a service to let you burn files 'on-the-fly' to a CD instead of mastering the Disc; recent versions of Roxio include a network media sharing system to share music and videos with other users on your LAN, etc)
* iTunes installs 2 or 3 background services (bonjour, iPod service, mobile device service)
* Some applications like Java, Firefox, and OpenOffice have 'Quick Start' apps which run in the system tray, whose purpose is to keep the core of the app loaded into memory for faster 'startups'.
* etc, etc.
I suspect that if a 'quick start' application is NOT running as a service, but instead is running in application mode, it will count against this 3 app limitation.
Basically, though, I have to agree with other posters - I think Microsoft is royally shooting itself in the foot here. I mean, I understand Microsoft wanting to create some differentiation between different editions of Windows, so that they have the ability to create different price levels. But multi-tasking is a core feature of the O/S which has been around since Windows 1.0 in 1988 (or whatever - I don't remember the exact dates). Limiting the number of applications artificially is just plain dumb.
If Microsoft wants differentiation, they should base it on the apps which are included in the base install of the O/S. Like, maybe the Starter edition doesn't get any games (or only 1 game, like solitaire). Of course, the problem with that route is that as soon as Microsoft removes that stuff, people will just download free versions off the Internet. I was trying to think of things MS could remove from the base install, and all of them, except for, *maybe* Windows Media Player, could easily be replaced by third-party programs. The only reason WMP couldn't easily be replaced is because of WMP audio/video streams and files on websites, which other players cannot (legally) support. But removing WMP from Windows Starter is a losing proposition for Microsoft - the only reason that anyone uses WMP is because they know it is likely to be installed on user's computers. If Microsoft unbundles WMP, suddenly MP3/Mpg, RealAudio/RealVideo, iTunes AAC, etc become just as likely (or possibly more likely, in the case if iTunes) for a user to have the capability to play, so why would websites bother with WMP support any more? So, Microsoft just cannot unbundle WMP - it's bundled status is the only thing it has going for it, really.
The only other things I can imagine microsoft pulling out of the starter edition are things they already pulled out of the Home Basic version in XP, or the equivalent Vista SKU - things like file encryption, IIS, various other network services - which most home users wouldn't miss anyhow.
Actually, there is *one* thing I can think of Microsoft pulling out of a "Basic" edition which *might* drive home users to upgrade - Windows networking support - that is, the ability to access network shares, or share directories from the local computer to the network, and the ability to print to a network printer or share a network printer.
However - the problem with that is that, I suspect, third-parties will step in with their own solutions to that problem (like a network file browser based on a port of Samba, or a third-party device driver to map a network drive which looks like a local drive to Windows, or a printer driver that appears to be a local printer to Windows, but really is printing to an SMB shared printer, or even just prints using something like the old lpd protocol, or IPP.
There is almost nothing Microsoft can remove from Windows, that third parties won't replace either for free, or ver
If they are not licensed, then by what right do you expect them to not get interference? One unlicensed user has just as much right to the spectrum as another. If the mike's were digital, they could, I think, happily co-exist with other digital users of that white-space spectrum. Outside of ham-bands, I begin to think that analog radio devices will quickly become a think of the past - the problem with analog stuff is that basically only one user (or one group of users) can use a certain frequency at a time. With packetized digital communications (and/or spread-spectrum techniques), multiple groups of users can share bandwidth at the same time.
Shared use, seems like a much more equitable use of limited spectrum resources, than the old analog 'dedicated channel' model. Still, if you are someone with analog A/V equipment, I suppose it would kind of stink. I suppose an equitable compromise might be to start by making sales of such 'legacy' equipment illegal, but continue to allow it's use for some number of years, to allow people and companies who've invested in such equipment to 'get their monies' worth' out of it, with a plan to make the continued use of such equipment illegal after a cutoff date in like 5 or 10 years.
I've been kind of wondering the same thing. I tried configuring Firefox to report itself as IE7 on Vista, to see if Netflix would work, but then I got an error page that ActiveX support must be enabled. Haven't had time to chase that any further. It occurs to me that Wine might be able to be used as the basis for for a Firefox ActiveX plugin under Unix/Linux/*BSD, but I don't know if anyone has actually tried to create such a beast.
Really simple way to illustrate the problem to people - If you were to use the full capacity of your connection, you would only be allowed about 15 hours a month of activity. Now, to be somewhat fair to Charter, et. al., most people don't come close to tapping the actual capacity of their internet connection anyhow - in my experience, most file servers and web sites won't download files to me at close to the full connection speed. Streaming HD video only needs about 2.0Mbps, I think. VoIP/Webcams take maybe 1Mbps. Online games need low latency, but I don't think they actually use up that much bandwidth - maybe what, like 100-200 Kbps?
15Mbps service is nice, though, because it does mean I can be using Teamspeak, downloading files in the background, and watching an HD movie all at the same time.
I suppose these caps will hit people who heavily use P2P file sharing though. You know, I wonder - do these caps include same-network traffic? I think it would be tremendously intelligent (so it probably won't happen) for ISPs like Charter, Time-Warner, Comcast, etc, to try to work with P2P software providers to develop clients that preferentially 'bias' P2P traffic to stay within the same ISP network where possible. That is, if a torrent or other p2p client where trying to find a source to download the files from, it would try to pick other clients on the same network to transfer from - of course, that's only helpful when you have a lot of people seeding.
Still, the point is, I think that the ISPs could find P2P technologies to be tremendously beneficial for themselves and their customers, if they implemented things correctly. You could extend the P2P concept to cache-ing mechanisms (something like a p2p version of squid, perhaps). I really think that the benefits of p2p technologies are dramatically under-utilized, currently, because most bean-counters hear p2p and think 'copyright thieves'.
Blizzard is one company that has at least 1/2 a clue - it's my understanding that their updated for WoW uses bittorrent to distribute updates to their millions of users - what a great way to help make sure that users don't have to wait for hours for their computer to be able to download updates on patch day because of an overloaded update server (an experience I've had on several occasions with other MMO's that used a 'centralized' update server that simply couldn't handle 20000+ users all trying to connect simultaneously and download a 100MB+ file).
Can someone explain to me under what legal doctrine this does *not* violate the 4th Amendment? Does this fall under 'probable cause' since the person has been arrested? I could see taking a DNA sample upon arrest and checking it against a database of crime-related DNA samples (that is, DNA samples taken from a scene of a crime or victim [like skin under the fingernails of a person who was assaulted and scratched the assailant]), but it shouldn't be stored in a general purpose DNA database unless the accused is actually convicted (it would probably be stored, of course, in the case files for the case under investigation).
Comcast can just give me a free month of Playboy channel. . .
While off-the-shelf PC games might not work that great on this combo, I suspect that there could be really good, beautiful looking games created and fine-tuned just for such a platform. Someone could possibly use this as the basis for a portable entertainment system to compete with the Nintendo DS, PSP, etc. XBox Portable anyone? If you consider that Atom-based systems would typically have smaller screens, so that you might be looking at a resolution of maybe 640x480 or 800x600 (or perhaps wide-screen aspect ratios, but with similar vertical resolution), I could see this being a sufficient solution for such a portable entertainment system.
Yeah. . . first you need some clouds to seed. I'm not sure, but I believe that you will not often find the confluence of clouds suitable for seeding at the same time and location that you have a wildfire?
Students at the University of Nitwitshire make history by transmitting digital data wirelessly using a radio transmitter and receiver they built themselves.
They are eating chicken wings, ribs, and 8-Meat pizzas.
"Considering that you can deliver a ~1kt warhead from off of a satellite platform..."
Is that an actual explosive warhead, or is that just dropping an object from several miles in the sky? Seems to me that if you have a sufficiently massive object (which doesn't necessarily have to be very large - maybe 100 or 200 kg), and it doesn't burn up during re-entry, it's gonna do some substantial damage just from kinetic energy at impact. It'll probably reach terminal velocity before it reaches the ground, so we probably cannot use E = m g h for the energy of the projectile, but we can use E = m v^2 (if we knew the terminal velocity of the projectile). I don't know the terminal velocity, but I can guess it would be moving pretty fast if dropped from a satellite, so it's gonna have a pretty substantial v^2 value. It might not have the energy to destroy a whole city, but I bet it could take out a couple city blocks. . .
Of course, all this assumes that the Iranians have the materials science and engineering capability to actually build something that won't burn up during re-entry, and can be accurately delivered to a target from the satellite.
The main problem I could see with cell jamming during a terrorist or similar criminal situation is that there is a small possibility that maybe, one of the victims could be trying to secretly call 911 (or whatever the local equivalent is) to try to give police information about the situation inside the building (or vehicle, etc).
Hi,
Thank you for this reply. It seems to be somewhat rare to get such a well reasoned rebuttal, and I do appreciate it, because I get to learn new things.
Like, I confess I didn't realize how close to epic failure the containment systems at TMI came - but, still, it's also true that the designed safety features were at least partly to credit for the containment of the radiation. (The expression "necessary but not sufficient" comes to mind).
As for the IFR, I also appreciate your critiques of that design. Honestly, I've been trying to find critical reviews of the IFR design, but it's been hard for me to find any information from groups that normally provide 'hostile analysis' of nuclear technology, wrt the IFR design. I didn't realize that the IFR only had a design lifetime of 40 years.
I think it'll be a long time before we can build a system that lasts 1000+ years, but it'd be good progress if we could be designing systems to last for 100+ years (that also helps to make them more economically feasible, as long as it doesn't cost exponentially more money to make it last longer).
I also want to point out that I didn't bring up the IFR because I'm convinced it is necessarily the solution, but really just as an example of a newer reactor design which has enough differences from old reactor designs that it needs to be criticized or praised for it's own problems or benefits, because some or all of the complaints raised by the OP did not apply to it.
I pretty much have the same feeling as you with regards to nuclear tech - continue to create new designs, build 1 or 2 prototypes as necessary to test and study the designs, but don't start building lots of new reactors until we've solved a lot more of the problems. So, from that perspective, I think it was a massive mistake for the Clinton administration to mothball the IFR project, simply because we should have finished the R&D so that we could learn as much as possible from that project. The IFR 'base design', I bet, could have been improved and refined (like, for example, replacing Sodium coolant with some other cooling system to improve the safety of the system, etc). Go through multiple iterations of refinement and improvement, and possibly down the road you come out with a design that is appropriate for large-scale, safe, economical, commercial implementation.
As for Solar and Wind, I'm all for that too - I just think that in our future energy 'mix', we'll likely need some Nuclear to provide a solid foundation of 'baseline' power, and even if we don't, I still don't think trying to contain something for 100,000 years is feasible. Mankind has no example of anything it has created that has lasted for 100k years. That doesn't necessarily mean it can't be done, but I'm quite skeptical of the feasibility of such an undertaking. I think it makes more sense to 'burn' the waste in something like an IFR, to deplete the transuranic elements (which, from all the reading I've done about this - I'm not a physicist, but I've tried, and continue to try, to educate myself on these topics - is what makes nuclear waste so bad for so long), so that we only have to contain the remaining waste in your 'granite mountain' facility for a few centuries instead of thousands of centuries. I'm fairly confident we could secure nuclear waste for 2 or 3 centuries.
Don't get me wrong, I'm all for increasing solar, wind, ocean-based, and other 'passive' power systems. But, all the people talking about wind and solar seem to always leave off some important problems - That solar panel which is producing 1kw (or whatever) at noon on a clear sunny day might only be producing 150W on a cloudy day, and nothing at night. That wind turbine which is producing 1kw on a nice windy day might be producing nothing on a stagnant day.
Now, I believe the counterargument to that is the idea that if you have enough wind farms and enough solar panels around the country and around the world, "It's always sunny/windy *somewhere*", but then you have problem of transmitting power from where you can produce it, to the place where it needs to be employed, and until we have high-temperature superconductors, that means we are suffering somewhat significant power losses during transmission. Also, you still need some way to ensure a stable baseline of power - power that you can count on producing a minimum amount, all hours of the day or night, every day of the year. Coal, oil, nuclear, and geothermal offer that - wind and solar really do not - they are spikey.
Finally, have environmentalists considered the impact of the land use necessary to produce electricity on the scale our nation needs using solar and wind? We're talking about putting arrays of turbines and solar panels (or other solar technologies - PhotoVoltaic panels aren't the only solar energy generation systems) on very large areas of land - what affects will that have on birds, insects, animals, native ecologies? How many birds will be hacked to death by wind turbines, or cooked alive by thermal solar systems? Maybe bird migrations will be confused by all the glare from PV panels? I mean, who knows what the impact will be of putting the enormous amounts of land necessary to power our nation to use as wind and solar farms?
Also, have you considered that, while the USA maybe has lots of undeveloped land in sunny deserts that are ideal for solar power, maybe other nations don't have such good conditions for solar power? Where are the UK, France, Germany, etc going to build their solar and wind farms? India? China? I suppose you can probably put lots of solar panels on roofs of buildings , so that does mean that you can use some already developed land as part of your solar farms, but I'm not sure you can get enough panels in place just doing that? Maybe, but I suspect that buildings will not be sufficient alone (I think about it this way - my understanding is that a solar panel on the roof of a typical house, commercial building, or skyscraper cannot provide enough power for that house, building or skyscraper, so it stands to reason that panels on the roof of every building cannot provide enough power for every building).
"[1st and 2nd generation Thermal] Nuclear power [are] a dead end. No new [1st or 2nd Generation Thermal] nuclear plants have been built in 25 years". . .
There, fixed it for you. Yes, old reactor designs are a dead end. They are prone to a risk of melt-down (though that risk has been, mostly, successfully managed for the past 30 years; yes, Three Mile Island was a problem, but, keep in mind that even with the TMI incident, the safety features of that reactor design prevented an escape of radiation when the melt-down did occur), they only extract a miniscule amount of the potential energy available in the fuel, and they create waste that "would have to be kept under armed guard forever".
Nuclear physicists and engineers have continued to do R&D for the past 30 years, and they are proposing *new* ideas. When new ideas are presented, you can't just assume that the same arguments that were valid criticisms of the previous designs continue to be valid for the new designs.
We have, right now, a Nuclear Waste problem, because of those previous generations of dead-end reactor designs, that must be dealt with. Putting the stuff into storage for 100000 years is not really a solution. The only real solution to the nuclear waste problem is to further process it to make the waste 'safe' and short lived.
Now, I do not really know if the design proposed in this article is "the solution" or not. Maybe it is. There was also a solution proposed in the 1990s, called the Integral Fast Reactor, which was essentially melt-down proof - not because of systems put in place to prevent a possible melt-down, but because it used a different Nuclear Reaction called a Fast Nuclear reaction, instead of the older Thermal Nuclear reaction, and was such that if the reactor increased in temperature beyond the normal operating temperature, the reaction actually choked itself, sort of like a candle sealed in a glass container. They even successfully tested the design, by purposefully cutting off the cooling to a prototype reactor that the DoE built out in the desert somewhere, and it did, in fact, shut itself down as it is designed to do.
The IFR design was also based around the concept of using our existing waste stockpiles as *fuel* for the reactor, producing hundreds of times more energy from that fuel, than older 'conventional' reactors do, which should have made it much more economically feasible.
The reason I mention the Integral Fast Reactor, is that is an example of a new design which I've studied more about than this new fission-fusion hybrid in the article, which demonstrates that the old arguments don't *necessarily* apply to new designs. Every proposal must be studied and evaluated on it's own merits - you can't just make a sweeping statement that Nuclear power is a dead end.
Unfortunately, the IFR project (which was being conducted by the Department of Energy) was canceled by the Clinton administration because of the same knee-jerk reaction to all Nuclear technology, exhibited by the parent, instead of really considering the IFR design on it's own merits or problems.
Also, in regards to this new technology, it sounds like they are not necessarily proposing to build new plants, but to 'upgrade' existing plants. If we can upgrade the already built plants in such a way as to reduce our existing waste stockpiles, where is the downside? True, this new design, as with any new design, needs to be thoroughly evaluated and proven, and also compared to other proposals (for example, we should consider if this proposed design is actually superior to the IFR design - if not, we should be restarting the IFR project instead, perhaps) before we role it out to any large scale.
*Maybe* we should have never gotten into the business of Nuclear Fission, but the fact remains that we have all this waste that we need to do something with. Why not 'burn' it in a new reactor type in such a way that we produce significantly less toxic, shorter lived waste? Environmentalists should be proponents of finding ways to deal with our nuclear waste problem, not object to every single proposal with a blanket statement that nuclear power is a dead end and re-hashing the same old tired arguments regardless of whether or not they apply to the new proposals.
You're right, but it's a shame - spectrum doesn't generally go to those who can/will use it best in the US.
It's like the spectrum auctions the FCC had a couple years ago - When Verizon, or T-Mobile pay $4 Billion (Or 6 Billion or whatever) for licenses to a chunk of spectrum, all I can see as a result of that is that those operators have to squeeze an additional $4 Billion of revenue out of their customers, which means everyone pays higher prices for mobile services. What I see is the government colluding with mobile operators to guarantee that service is expensive, and to prevent small companies from starting up competing mobile services (coming up with Billions for a spectrum license is not something any small company could ever do).
I wish more power to you in your company's pursuit to make a solid business as a Wireless ISP, and I do wish that companies that get the best results at the lowest cost to consumers would get more spectrum, but I don't see that becoming reality any time soon.
I think the Physics department at my university has done a good job of applying this principle. They used an online homework platform for most of the assignments for the General Physics classes I had to take as part of my engineering program. Now the thing about homework is, you don't really learn much from problems which you solved wrong, and then only learned you solved them wrong when they are marked incorrect by your teacher (which is the 'traditional' method). Usually, in such a traditional method, the teacher will hopefully spend some time in class going over problems students got wrong, so they could at least see the correct way to solve the problem, but it's still not as good a learning experience as when you figure out how to do it yourself.
With this online homework platform, students got up to 5 chances to get the solution correct. That is, if you solved the problem, inputted your answer in the website, and your answer was not correct, it would notify you the answer was not correct (but would not tell you what the right answer is), and give you a chance to try again. I personally found this to be an excellent teaching tool - simply knowing that my answer was not right gave me the chance to go back and look at the problem, and try to figure out where my mistake was. Almost all of the time, I could figure out my mistake, and correct it within 1 or 2 tries - something I never had the opportunity to do in classes I took earlier in my academic career. I truly believe that having those additional chances to correct my own mistakes helped me to learn the material better.
This is one example of something that is not easy to do without computers, but quite easy to do with computers.
"It comes from the (frankly quite bizarre) idea that it is somehow unethical to randomise educational methods."
But, isn't that exactly what is happening here? We have this IT director from a UK school, asking /. about ideas for how to setup computer systems to somehow improve the educational experience of their students. It sounds like the educational methods are being randomized (to the extent that they aren't necessarily in uniformity to the methods in place at other UK schools) in this situation.
I don't think this is, already, that much different, in practice, from what I've described, except to the extent that I'm suggesting that instead of setting things up in a particular way just because someone thinks it's a good idea, that it be done as part of a controlled experiment designed to validate or invalidate a hypothesis regarding educational outcomes. That it be done in such a way that some meaningful metrics be obtained regarding the effectiveness (or lack thereof), for the benefit of other schools in the future.
You know, this is one of those areas where I wish educational systems would take a more scientific approach to these types of problems. It's pretty much the same way here in the U.S.
There seems to be little 'method' to the ways we try to figure out the 'best' ways to integrate IT into education.
It seems to me that in situations like this, schools could benefit from systematically applying the scientific method - Observations, Hypothesis, Prediction, Experiment, Analysis. (Repeat as necessary.)
Start building a *theory* of education and IT, and then make your school IT decisions and budgets based upon the body of theory thus developed.
So, this means that you gather lots of ideas from all sorts of people (everyone from Education Ph.Ds, down to teachers and IT staff at the schools, even to interested members of the public who have ideas) about how IT could be better implemented in education, and start using a small number of schools as experimental test beds (and other schools as 'control data' for the experiments).
These experiments should be, first, submitted to and approved by some national 'school board', or at least something like a group of professors at a University education department, who are tasked with tracking and eventually reporting on the results. From the results, this 'school board', or university task force, or whoever is responsible, can start creating recommendations and best practices.
Oh, oh, you forgot. . .
. . .a beowulf cluster running this. . .
Seriously, not being dependent on foreign companies for critical national technological infrastructure is in the strategic national interests of every nation on earth. If you are a foreign nation, how do you know that the OS you are getting from $OS_Vendor doesn't have 'wiretaps', back-doors, remote kill switches, or other secrets in the software which $OS_Vendor, or the nation to which $OS_Vendor is based out of, can use to cripple you? Another problem is, that $OS_Vendor could simply stop providing you with necessary patches to update known problems and vulnerabilities in the OS.
One possible solution would be, if you are using a closed-source vendor, to require that vendor to provide the government with buildable source code, which could be reviewed by your own Computer Scientists, then built by your government, and distributed throughout the nation. This also allows your developers to provide your nation with patches and support if you are cut off from support from $OS_Vendor. That is not true Open-Source, but that is still, effectively, a "National Operating System". Open Source is one step better though, because you have, potentially, a lot larger base of people that are reviewing the code. That whole Eric Raymond thing to the effect that with sufficiently many eyes, all bugs are shallow.
Just saying that some foreign leader that is not well liked has something in common with another leader is sort of mis-leading, because there will often be many things in common between good leaders and bad leaders - what's important often isn't the similarities, but the differences.
I think I might be about ready to give up on PC gaming completely, lol. When I bought the computer, 2 years ago, 1 Gig of ram was quite a lot (though there were some systems selling with 2G, that was about the top).
PC Gaming is the only consumer market that thinks it's reasonable that something you buy today, will be almost completely unable to use content produced only 1 year after you buy the device, without constant upgrading. That'd be like an iPhone that couldn't play some of the songs on iTMS after a year, or a TV, which couldn't display some TV shows after 1 year.
At least with consoles, you get about 3 years before a system gets replaced with a 'new generation', and when it does, consoles are (comparatively) cheap.
That said, I'll probably at least spend the money to upgrade to 2 or 4 Gigs of RAM, as that is pretty cheap, but when I reach the day that this computer can no longer play most of the new games coming out, I think I'll just go the console route, like most other gamers.
Yes, it was about 2 years ago, I guess. I have, actually, been thinking about upgrading the RAM. The thing is, most games play OK for me, even on that system. Like I said, I know it's not top of the line, but games need to run on systems that are not top of the line, too. Historically, Valve games have run pretty well on reasonably older hardware. Half-Life 2, the Episodes, and Portal seemed to all do OK on my laptop (after I turned down the graphics some), just not TF2.
As for the nVidia, the GPU claims to have 256MB of it's own RAM. It should not be shared memory, so I should really have 1 Gig (of course, Windows XP and background services seem to use a good 400M+ when I'm running nothing else, so that is cutting me down to about 500M of useable RAM for games).