The problem I have with this is that you're only getting half the story, if that. First, about twenty bloggers had to comment on the original graphs because they didn't include any version of Opera, which could only make a fairer test - FF vs IE is hardly a shocker. The Opera line on the graph instantly made a better competitor and blew most of their shocking claims about how well FF did compared to "other browsers" away - it sat dead in the narrow line between the two versions of Firefox. Next, the test is very specific about how it operates and requires lots of similar options to be enabled on all the tested browsers (to counteract memory caches and different cache-management algorithms). That all boils down to mean that in a certain, small area (*memory* management under a particular caching profile), the latest version of FF "wins". Not bad seeing as that's what the statistics were designed to show in the first place.
My main bugbears are that it doesn't take any account of CPU usage (I could decrease the memory usage of a program by half if CPU speed was no obstacle by just compressing everything I put into and take out of RAM), user-experience, cross-browser differences (so turning off some cache options present only in certain browsers is cheating because you still don't KNOW how they work or what they sacrifice for performance elsewhere. Default settings would have made more sense given that most people don't touch caching etc. options and that that's what a novice user would see. You would, of course, have to account for user-experience at the same time), the total amount of RAM installed (who's to say that Opera wouldn't adjust itself on a 256MB RAM machine to do less caching etc.?), total disk size (IE with several Gigs of disk cache isn't unusual nowadays but my Opera does just fine with a few hundred), "nice factor" to other programs (i.e. does it assume that you're ONLY running the browser, or is it pitched to let you use other programs alongside it well?) and a million and one other factors.
Given that I browse with Opera on a 600MHz laptop with 348Mb, I'm very happy with its performance. It can't do heavy-Java well but that's expected. It can do streaming video, dozens of simultaneous websites, Flash, all sorts without coming to a grinding halt. The startup speeds are reasonable and not due to disk-thrashing (it looks more CPU-bound to me). Firefox, any version, on the same machine slows to an absolute crawl after a few websites and thrashes like mad on startup.
I worry that programmers are looking at those graphs and congratulating themselves, without any real metric of what else has changed or stayed the same. Sure, memory allocations have dropped and there is more real RAM available to the machine but what does that mean if it's at the expense of other resources or, worse, the user experience?
Let's ignore all your points for a second and cut to the crux of the matter. The country you live in could legally enforce all of your suggestions absolutely perfectly. It wouldn't make a dent. You could do it in twenty, fifty countries. You still wouldn't make a dent. Law is not universal. In my continent you can't HAVE software patents, they actually do not exist. You aren't going to make that change any time soon no matter what your country does. Similarly for any legal resolution to spam, viruses, botnets etc. Even if 50% of the world's botnets are on American PC's (for example), by definition even the owner's don't want them or even know they are there. Nor do the ISP's, or the transport carriers, or anyone else along the line. But it's like suing people because they gave you a cold - they didn't want to catch the cold in the first place and, yes, although there are measures they can take to lessen their potential exposure to the virus, nothing is guaranteed.
1) "static IP's" - we can already trace where all the stuff comes from - there are complete trails back to the sending machines and from there back to the perpertrators. But most of it generally comes from computers abroad, or from people attacking computers from abroad, or via proxies, all of which are subject to different laws and untouchable. Even ASKING for the details belonging to a particular IP that resides in a foreign country is unbelievably difficult. And you won't get them, but your law enforcement might. And you think you can shut them off before they cause damage because you have their IP address? Nope. It's too late. By that time, the botnet's already moved on to take advantage of the next exploit. We have dynamically updating realtime, very expensive blocklists with dedicate people to add new machines as they are found - they don't stop that much, really.
2) "Laws that require people to assume some form of responsibility when they connect a computer to the net." - in every country in the world. With similar provisions. Quickly. Not going to happen. EVER. And then you're into why do you have to take responsibility and how do you ensure it? Your kid put a virus on your machine? I'll sue you, then. No? You caught a spyware toolbar which send me spam? I'll sue you, again. You'd either sue people literally off their computer seats, everything would get thrown out of court, or you've just helped the government introduce legislation to make them monitor everything you do at your computer, with fingerprint ID required to logon.
3) "Perhaps some form of compulsory insurance policy." - For owning a computer? No. If you could tax people for being stupid, the world would be split between the bankrupt and the filthy rich.
4) "Laws that require ISP's to disconnect spam bots and take some responsibility." - So now they're responsible for their users actions? They won't let you do it. If you do, they will shut themselves down and get out of the business. They ALREADY disconnect bots - it is in their interests. They ALREADY have to deny all responsibility for your actions. And they are ALREADY in deep legal grey areas because of the burden of proof of doing such things and the expense of a mistake (Sorry, Company X, I thought you sent a spam. I've just cut off your Internet by mistake. Bye-bye online business).
But the fact is that none of your measures are sensible or practical, some are even impossible, and all of them are in place in one way or another today. The fact is that every country in the world has a different idea. If we can't convince them all that death by execution or torture might be a bad idea, how the hell do you think you're going to get them to shut down botnets?
The comment about the computing power was not meant to demean, and certainly wasn't to be taken to represent this particular effort but "AI" efforts in general - it was meant to illustrate the sheer scale of the problem. The order of magnitude between a research student on a supercomputer and the actual computer power you have mentioned above is incomparable to all of the "knowledge" that you are trying to cram into the space (data storage, data diversity, computing time, etc.), as my comment was meant to show. It's just not going to work.
The most powerful supercomputers in the world can't get the weather right four days out of five (although that involves prediction, which is a different area entirely), they can *barely* beat a human expert at chess but only then mostly through *sheer brute force* with a couple of well-thought out tricks. And chess is an entirely logical, rule-based and *extremely* simple game in comparison to things such as Go, which is what my professors were studying trying to "beat" a human at when I was at university. Go is something like 10^19 times more complex but with vastly reduced rulesets compared to chess? I can't remember the exact number. And at the moment an "average" skilled human player will wipe the floor with basically any computer-based Go opponents, no matter what supercomputer is behind them. And that's in a perfectly controlled, 100% logical, rule-based environment, which is a different type of AI entirely.
We're not talking about a processing/storing "shortage" hindering our ability to really demonstrate even a basic, recognisable intelligence, we talking *orders of magnitude* out of our predicted range for even the next few decades to get a basic chatbot working enough to sound passable. It's like monkeys that can't add up trying to crack AES-512. As you rightly point out, there is some enormous brute force behind the chatbot (sorry, that's all it is) mentioned in the article, and still we get only quite a minor step forward, if some of the transcripts I've seen are anything to go by.
Hello chatbot. Decode this for me: "nrvsidr o fpm#t".
Serious answer: I do believe they can get close... but not for a few hundreds years or so at *absolute* minimum. But because EVERYONE is either going about it completely the wrong way (let's program what we know of 4-year-old's into a logical engine, or let's let something evolve that has about 100 "neurons" and hope it learns English) or the study that they need to do is far beyond our computational capability, the current techniques are next-to-useless.
How do intelligent things learn? Let's take a primitive example of an "intelligent" organism. They have *near-zero* historical data at startup - no expert databases, no pre-knowledge, nothing. Over the next ***four years***, they are constantly subjected to intense training, vast amounts of data and are heavily rewarded as and when they form connections. They don't have their internals poked and prodded and fine-tuned along the way, they just get on with it. They operate on scales and mechanisms that we can only dream of (super-computing wise), take in amounts of input that we could only just barely simulate and they are taught each item FROM SCRATCH. First primitive sounds, then associating those sounds with things, then the alphabet, then words, then grammar, then cleaning it up into a real language.... We don't do things the right way.
For example, in this article, grab a baby, deafen it and remove all other sensory perception except sight (and even that's much more info that this bot is getting), throw up some text for a few years and wait for it to hold an intelligent conversation by sending text back. It ain't gonna happen, even with a proven-intelligent being. There is no "meaning" to the words. You have to prime and train and exemplify and clarify and correct and all sorts. You can train programs to train programs but it's all the same in the end. It needs real feedback. It needs to know WHY what it said was wrong, not just get a "FALSE" entry in a database somewhere. It needs to learn that fire burns and holding your breath isn't wise to do for too long.
We could *train* an AI. I see that as a possibility - training some enormously complex system from scratch to recognise simple patterns and then build up to language slowly. But it takes a proven-successful organism at least a few years to get that far working 24 hours a day with a constantly-busy environment and enormous amounts of feedback that are automatically categorised into "good" (tastes nice, is fun, etc.) and "bad" (hurts) by the organism itself, not some pre-programmed list. And every time you want to improve the underlying algorithm, you're starting from birth all over again. We can train an AI, it'll take decades just for a basic one, but I very much doubt that we'll ever create one at all, certainly not within a few thousand years or so.
Oh come on, it's a chatbot not an AI. I did my Computing degree, I know how AI on computers is supposed to work and it's seriously laughable for anything more complex than extremely primitive, less-than-insect intelligence. You'll see AI "geese" flocking, you'll see AI "ants" making a good path, planning ahead and fending each other off but none of it is actually "AI".
AI at the moment consists of trying to cram millions of years of evolution, billions of pieces of information and decades of years of "actual learning/living time", from an organism capable of outpacing even the best supercomputers even when it's just a single-task (e.g. Kasparov vs Deeper Blue wasn't easy and I'd state that it was still a "win" for Kasparov in terms of the actual methods used) - let's not even mention a general-purpose AI - where just the data recorded by said organism in even quite a small experience or skillset is so phenomenally huge that we probably couldn't store it unless Google helped, into something that a research student can do in six months on a small mainframe. It's not going to work.
Computers work by doing what they are told, perfectly, quickly and repeatably. Now that is, in effect, how our bodies are constructed at the molecular/sub-molecular level. But as soon as you try to enforce your knowledge onto such a computer, you either create a database/expert system or a mess. It might even be a useful mess, sometimes, but it's still a mess and still not intelligence.
The only way I see so-called "intelligence" emerging artificially (let's say Turing-Test-passing but I'm probably talking post-Turing-Test intelligence as well) is if we were to run an absolutely unprecedented, enormous-scale genetic algorithm project for a few thousand years straight. That's the only way we've ever come across intelligence, from evolved lifeforms, which took millions of years to produce one fairly pathetic example that trampled over the rest of the species on the planet.
We can't even define intelligence properly, we've never been able to simulate it or emulate it, let alone "create" it. We have one fairly pathetic example to work from with a myriad of lesser forms, none of which we've ever been able to surpass - we might be able to build "ant-like" things but we've never made anything as intelligent as an ant. That doesn't mean we should stop but we should seriously think about exactly how we think "intelligence" will just jump out at us if we get the software right.
You can't "write" an AI. It's silly to try unless you have very limited targets in mind. But one day we might be able to let one evolve and then we could copy it and "train" it to do different things.
And every chatbot I ever tried has serious problems - They can't reason gobbledegook properly because they can't spot patterns. That's the bit that shows a sign of real intelligence, being able to spot patterns in most things. If you started talking in Swedish to an English-only chatbot, it blows its mind. If you started talking in Swedish to an English person, they'd be trying to work out what you said, using context and history of the conversation to attempt to learn your language, try to re-start the conversation from a known base ("I'm sorry, could you start again please?" or even just "Hello?" or "English?"), or give up and ignore you. The bots can't get close to that sort of reasoning.
I live on the trailing edge. It's easier, more comfortable, cheaper (how's that HD-DVD purchase looking now?) and makes much more sense. You don't end up buying into much that disappears quickly. You don't end up with teething problems. You don't end up wasting money on vast amounts of testing because other people do that for you. You don't end up worrying about that next big upgrade which is "vital" because one bit of software demands it. You just wait until EVERYTHING demands it or, more likely, nothing demands it and the project dies a death.
I do network management for schools and this is the only sensible way to go about things. You cannot upgrade every single bit of software to the latest version as soon as it's released. Too much breaks. Even Windows Update throws default printers, etc. onto a clean machine and that's the easy stuff to detect and clean up.
For instance, last year one school went from Office 2000 to Office 2003 (Yep, that was our upgrade last year - and that was forced on us by the "people above" because 2000 is such an old number, basically. Nothing to do with practicality, compatibility or anything else and we'd already installed the filters that let you open newer Word docs etc. so there wasn't a simple complaint when other schools switched to XP/2003/2007 and started sending us their files). That simple upgrade broke everything - printer allocations, all our Office GPO settings, basically every PowerPoint file in the school, some software called Mindjet Manager that just so happened to be on the desktop of all the top-dogs in the schools, you name it.
The Mindjet problem wasn't even documented, we found it out the hard way (some sort of DLL conflict to do with the order of installation that just crashed the program without error - sysinternals utilities saved the day again!). Upgrading breaks more than it fixes about 75% of the time.
The only reason to upgrade is if you get something out of it. That's how companies sell upgrades - "look what you can do NOW". If you don't need that thing, or can already do it somehow, then why upgrade? Just keep plodding along with your perfectly working software with all the latest security updates that does everything you could possibly want.
Even Becta (the UK schools IT bods) recognise this - they specifically recommend not to move to Vista or the new Office until EVERYTHING has been checked out thoroughly, i.e. give it a few years yet. Better the devil you know and all that. This is why MS is having a hard time selling Vista - it's only the "we must have it because it's got a shiny new version number" crowd that are the ones actually pushing for it and unfortunately sometimes they are in charge.
What happens in a modern office that is totally impossible with a old office suite? Not a lot. And most of it is power-user features that most places won't even touch. Even I never use 5% of the functionality of Excel, Word, etc. when I'm doing budgets with pretty graphs in and the kids are basically doing DTP using Word (against my advice, but hey, Word works and we have it so why spend money on something else?) and they aren't using 5% of it anyway. Where's the incentive?
And you don't get bitten anywhere near as often. Schools go through fads - we had interactive whiteboards and flatscreens and the new ones are "Virtual Learning Environments", electronic signage (a big telly with an advert on it) and wireless classroom trolleys filled with laptops. Every single fad starts up, fails miserably in a number of schools (usually because they have copied it straight from the commercial sector and are thinking along the business-consumer lines rather than "THIS IS A SCHOOL. It might not work like a business does.") and THEN the suppliers learn what the schools want/need and bring out the really useful gear that the following schools can then actually use - for instance:
Soft whiteboards got vandalised by the older kids. So they changed to hard. Hard whiteboards weren't tactile enough for
I'm really not a Microsoft advocate but I work in Windows-only schools... I do completely understand the frustration but off the top of my head:
1) Group Policy integration
I can set browser settings (every single one that you see in IE's Internet options etc.) for a group of computers and fine-tune each setting to particular users. You don't want kids changing their proxy settings, you don't want them playing with trusted zones, you don't want them getting overwhelmed with dialogs when they log in for the first time. You don't want to have to set it up per-user.
2) Content that is IE only.
Everything is web-based nowadays. For instance, schools are pulling lesson plans, educational games, etc. off websites or "cached content" (i.e. stick a Linux box full of some companies flash-heavy website in your school and every month or so they'll update it for you... this is an actual product from several very expensive companies). Most of it works only in IE, although I have seen a few that will work on Firefox too. ActiveX is seriously over-used in the educational sector for example. And therefore if you want support (which schools in particular see as vital), you have to use IE.
3) The little blue E
Seriously, I have dozens of users (and I get more each year) who when they want the Internet go for the little blue E shortcut. I have actually replaced it with an identical-looking one that ran Firefox and only the "power-users" on the network actually noticed anything wrong. 450 kids just carried on working as if there was nothing different.
As soon as everybody gets off their backsides and gives me non-IE support on all the content that the schools carry, I'd move every school over. The next hurdle after that is making it easy to customise every option on a per-user/per-computer basis properly. The third is easily overcome even if it is selling your soul to put a decent browser behind the IE icon...
In most of the above points the presumption seems to be that because the resolution difference exists, the average person can distinguish it from a given distance. This is clearly not true.
We can have the same old arguments again if you want. Barring artifacts caused by mains flicker, enormous screen size, peripheral vision, strobing of ambient lighting etc. - the average person can't really see a difference between 50 and 60Hz refresh rates, or 60 and 100Hz. They can't really see the difference between 16-bit (properly done) and 24-bit. They can't hear the difference between MP3 at certain bitrates and the raw 48-bit audio data at 96KHz. They can't really see the difference between interlaced and progressive (hence why TV over in Britain has been 576 lines, 25fps, interlaced, for the last few decades and nobody really "noticed" anything out of place).
At reasonable distances, the average person just cannot discern, in a fair test, the difference between these formats. There will be a minority (much more of a minority than those who "claim" they can see it, as there are so many other factors to take account of) who can detect some of them, of course, (e.g. photosensitive epileptics and, for example, people who can "see" the flicker in flourescent lighting) but most of the difference I see is down to different colour mappings producing more vivid colours and greater contrast, wider screens, closer distance. I also spot deinterlacing artifacts, poor SDTV sources being compared to clean HDTV sources, TV's that deliberately make SDTV look worse through poor upscaling, etc. and they stand out more to me than one pixel becoming four.
Nobody EVER sells me HDTV based on those things above (e.g. that HDTV's have newer substances in their displays that can give me a better colour range etc.) They all sell it to me on the fact that there are more dots. Dots that, once I take a step backwards, I can't distinguish one dot from the next. Seriously, on a smooth, moving video (not a black pixel on a white screen), pick out a single pixel and tell me what colour it is and track it as it changes. Yes, I can spot a stuck green pixel on a standard LCD screen at twenty yards and wouldn't be able to watch a LCD TV with a stuck pixel. But if you were to play some video and switch between 1024x768 and 800x600 on a good, fully-working monitor at a reasonable distance without affecting "actual" size of the video or colour ranges, I (and most of the population) wouldn't be able to tell. In the average use, on a *fair* HDTV you just will not notice any difference between SDTV and HDTV unless you buy an enormous HDTV or sit with your face in the screen. I don't want to do either. Six to ten feet is a good distance for TV viewing for me.
I have done these tests countless times in shops, friends houses etc. I'm itching for an excuse to buy a good gadget, but HDTV and its associated technology (e.g. HDTV sources like HDTV channels and Blu-Ray etc.) are just not worth ANYTHING to me. People take me round their house and show me their HDTV and it's just like watching TV. I've even gone to the point of dragging out an old SDTV and running the same signal through it because my friend's model of HDTV seemed to "cheat" when displaying SDTV and it actually was WORSE through the HDTV than through a ten-year-old television of comparable size. That sounds like sucker-sales-tactics to me ("hah, when they switch to an SD signal it'll look horrible and they'll think it was like that all along!") and it's obvious that my friend was one of the people that fell for it, and I doubt he's alone.
Oh, and my old 4:3 screen actually showed a larger visible image than his widescreen TV could, even with the black borders and the fact that my TV, when bought brand new ten years ago, cost less than half of his did today. Hence I don't have any widescreen TV's either because, for many years, a bigger 4:3 TV was actually superior up to the size of most affordable widescreens.
Back in my day, Windows was a WIMP GUI. Now they call it an operating system.
But yes, the old-style viruses tend to have lost out the past few years. I can remember quaking in fear when I read about a virus that was polymorphic, stealth, boot-sector infecting, "hold your partition table to ransom", able to transfer to floppies, hard disks and even CD's (WOW!), plus across IPX networks, randomising data destruction etc.
I don't have vast experience of Vista because I decided against deploying it on the last few networks I managed. However, it seems that it must still be incredibly easy to access the MBR even, as you point out, as a non-administrator user.
However I assume that "Boot sector protection" as available in most modern BIOS's should stop this stone dead (I know that I implement it but I doubt everyone does). It's like 1989 all over again...
Granted, the virus is easily cleaned, although it's potential effects may not be (identity theft etc.).
Maybe it's just me remembering the good old days of DOS viruses but none of this actually seems "new", except for "calling the NDIS API directly, using a 'code pullout' technique to load just the parts of ndis.sys that it needs", which is just a shortcut to direct hardware access... it's basically loading another copy of a library to use it seperately from the OS (and therefore, presumably, OS security) to access the network card. It's a nice way to access a network card from "below" the OS in a hardware-independent way (i.e. let's pinch the Windows driver rather than try to work out what card we are using and create 1000 drivers for each different brand of card.). But on the whole, this is hardly "new" or "shocking".
Having said that, it's nice to see something where people have actually invested time and skill into creating a program that bypasses the OS in such a way, rather than just another re-written script with a couple of variables changed.
Lines like:
"This malware is very professionally written and produced. Which of course means it's not written for fun."
might annoy some, though. The old DOS viruses NEVER really acheieved anything useful (even with blackmail attempts while holding your boot sector to ransom) etc. and were written "just because" by teenagers. That didn't stop them from appearing professionally written and breaking genuinely new ground for the time. Just because people are now using such malware for financial gain, doesn't mean that it's ALWAYS the case. And Linux zealots are sure to jump on the above quote with all their hearts.:-)
And then you have the obvious - why is the OS allowing you to modify the MBR without appropriate rights and/or why are users running as users with the rights necessary to do this? This is STILL a problem harking back to the DOS days - everyone as administrator. With a new twist - the average user hasn't needed to BE administrator for quite a long time now.
I use Linux BECAUSE of its backwards compatibility. I recycle old machines into computers for schools, etc. and if Linux weren't backwards compatible none of them would even boot. As it is, you can still run Linux on a 386 with ISA cards (is MCA still supported, I can't remember?), the same can't be said about Windows which blatantly removes old architectures in order to force upgrades (ISA support disappeared in XP, even though you can still hack it a bit to make it work perfectly). I'm not saying there's not a justification for not supporting ancient systems but the fact is that Linux does it for as long as ANYBODY with coding experience cares, not just the Microsoft pencil-pushers.
Not only is Linux backwards-compatible, but it supports things I've never even heard of, and things I absolutely no need for, everything from ten-year-old watchdog timers to the latest and greatest datacentre interconnects. It's also sideways-compatible in that, if you're feeling sadistic, you can run Windows NDIS drivers on it without modification.
Linux has always been backwards compatible in virtually all senses of the word. I can still load ext2 filesystems, a filesystem over fourteen years old, and that's not even the weirdest or oldest FS that I can use! I can still use ancient hardware for which drivers don't exist past DOS or Windows 3.1. I can still use programs that were written for Linux 1.0, so long as I have the right libraries in place.
Now OVERALL compatibility may be slightly worse than Windows (I don't believe it, given the things that Linux supports that Windows couldn't do in a million years) in that it can't use EVERY scanner, printer, etc. because of rubbishy, propreitry protocols from companies where nobody even remembers how they work any more, let alone document them or write a driver for them. But to say Linux isn't backwards compatible is like saying that your average Windows driver from a £10 scanner is stable.
Although the statements are all well and good, you have to make a lot of assumptions for the maths to work - namely that people are ordered, logical and will know where they are going.
Many moons ago, my A-Level project consisted of something quite similar. I had to do a traffic simulation and work out what the "ideal" interval was for traffic-light changes. I spent months on the project, got top marks because there were absolutely no mathematical errors in it and it had quite a lot of higher-level maths in there.
The results of my formulae were applied to a simple crossroads, with cars accelerating and deccelerating perfectly at fixed points of the traffic light cycle, depending on their speed, the position and interval of the lights, etc. Cars entered the system at a uniform rate from all junctions and several phases of the traffic lights were simulated.
Unfortunately, the formulae took all this information and the answers (for traffic light change intervals) were given as solutions to a large polynomial. Despite the mathematics being verified several times by several people, and the assumptions being called "reasonable", the optimal answers given for this simplest of simulations was to have traffic lights change at intervals of -5 seconds and 0.02 seconds. Neither of which gave the cars time to accelerate across the junction (or even be able to SEE the lights change) completely.
Factoring in some more terms for things like "being able to cross the junction before the lights change", "being able to physically react to the light changing", etc., even more bizarre answers were arrived at until a bit of "fudge factoring" in the initial assumptions made it give sensible answers.
I quite believe that the methods used currently are possibly the worst *mathematically* but I doubt they're the worst for airports and humans to handle. If it was, it would have changed already to get people on/off and get the planes turned around faster.
Still to do this day, I wish the school had had the resources at the time for me to do a proper computer-controlled visual simulation. Alas, they were still teaching 18-year-olds BBC BASIC at the time.
Yes. I have. Several friends have them. My comment still stands. I've purchased several for my job - we use large ordinary HDTV's as "digital signage", i.e. things showing looped Powerpoint presentations. I admit that you CAN tell the difference, close up, with a nice, high-res signal (HDTV or even just a decent PC resolution) than the same on a "normal" SDTV, the same way you can between a monitor and a SDTV showing the same signal... of course you can - but that's not the point. The point is, in normal use, from a reasonable distance, you really don't notice the difference. I've also looked at them seriously for purchasing for personal use and because I'm *that* guy that everyone comes to asking what the best model of *vaguely-electronic-equipment* is.
I've seriously spent hours in shops and looked at every single HDTV on offer. I try them all out, one by one (first thing I look for is what connectors they have on the back). I then make sure that any display TV's are displaying actual HDTV content. I carefully measure the distance back so that I'm at my usual viewing position, and I can't see picture difference - except I do notice that, with a *few* brands, you can side-by-side compare a similarly priced large SDTV and HDTV, while both showing SD picture, and *sometimes* the HDTV makes the SD signal look ever so slightly "worse" (I can't define it, I'm sorry) than the SDTV showing the same SD signal. A ploy to make you think that you HDTV is actually an improvement when you see a "real" HD signal? Some technical reason to do with the interpolation etc? I have no idea.
The only exception is if the screen is MUCH larger than one I can a) afford or b) fit on my wall / in my living room comfortably and then the difference is mainly that it's BIGGER. Don't get me wrong, in a few years you won't be able to get an SDTV, in the same way that you won't be able to get a CRT monitor except for specialist uses, or a non-digital-reception TV etc. Maybe my eyes are just crap. But I don't see the difference. I've even been found in shops by my wife, squinting from a few inches at a HDTV trying to see what the difference is. As soon as you get to a reasonable viewing distance, my eyes just can't discern one pixel from another and it's no better than SD shown at a similar size on a similarly priced "normal" TV. Call me mad, call me a liar but **I** am not paying for anything that **I** can't notice no matter how many people tell me how much better it is, or how much knowledge I have of exactly how many pixels the thing has got.
It's the same old story, to a point. The performance required to do a relatively simple job (play fullscreen video) in a new way (using HD content and a new storage medium) means that it becomes impractical without upgrades. I can remember having to tweak computers to be powerful enough to play MP3's without skipping, but there at least you had the advantage that the storage space saved compared to even the best-compressed formats of the time was phenomenal.
I freely admit that I absolutely do not "get" the HD fuss. It's the same thing we've had for years, with more pixels, that you can't reasonably see on a fair test past a certain distance (although I would say that on a high-res laptop you are more likely to spot the difference because of the unusually close eye-screen distance), with new storage formats, new compression, new software, new DRM and new performance characteristics... which are killing battery life. And, yes, eventually they'll start making "blu-ray acceleration cards" just like MPEG-acceleration, 3D-acceleration, etc., although in this day and age they're called "software on the GPU". But at the end of the day, you've gained little (a higher res that you might not be able to distinguish) for enormous performance increases.
Where's the advantage in it when a "Blu-ray" PC can still play the DVD's of previous years but at much, much less expense... if you can play a blu-ray for two hours or you can play MPEG-2 for six (while compiling stuff in the background without jerkiness) on the same machine, what are you going to end up using if you watch a lot of video on your laptop?
When I go away and know that I might want to view movies on my laptop (e.g. long trip staying in cheap hotels, stay over at a friends house etc), I take either DVD's, or I have a bunch of MPG's/AVI's/VOB's etc. on the laptop itself or on DVD-R's ahead of time. Quality isn't really the factor there and the advantage to having everything in a simple format that everyone can read easily and which doesn't tax the laptop is key.
It's another case of "laptop = general purpose computer, so let's turn it into a media centre and make it do everything". It's nice that it's CAPABLE of everything but you can't expect a portable device to do it all AND give you good performance at everything. Laptops are not even desktop-substitutes for most work (the times I have to explain this to people... it costs pounds to repair a broken desktop, hundreds to repair a broken laptop).
Let the early adopters waste their money. Even if Blu-Ray becomes the de-facto standard, I'd much rather just decrypt-to-disk and convert to a format that's easily readable, with extremely cheap media, that plays the video "good enough" for most things if I'm intending to carry it around with me. Much better 1 x DVD-R with a couple of full movies on it that I can watch one-after-the-other and make a backup copy for pennies than 1 x Blu-Ray that I can't give my friends with only a single movie on it that kills my batteries just watching it.
There was a time when I did exactly the same with DVD vs VCD - it's actually trivial to just copy several DVD's worth of movie/tv show to a DVD-R or even a CD-R and not worry about the quality. You're travelling - who cares whether it's HD or VCD-quality so long as you can tell what's going on without eyestrain?
Erm... You might have been technically correct fifty years ago when everybody still did Latin but not any longer. It's standard practice in every company I've ever worked, for the last 20-odd years and probably a lot longer than that. CV = personal details + list of qualifications + (summarised) work history + references + (possibly) brief statements about important project you've done in the past (unless you're under 18 when it seems you're taught to list "interests" like "watching TV"... I've actually seen it on a CV). You usually cram it all onto 1 or maybe 2 A4 pages. Possibly a personal statement (1 page of A4) if requested stating briefly why you'd be good at the job, clarify anything listed on the CV. Cover letter to wrap it all up (two lines and some decent letter-writing layout).
In the UK, no job agency would distinguish between CV and resume, unless they only deal with very pretentious clients, but I've never heard of anyone distinguishing over such a minor detail even in the City. Job adverts include the line "Send CV to:". Never any mention of anything else, except possibly "CV and Personal Statement". We don't even CALL it, or any part of it, a resume. We might have done once (I don't know the etymology, but it sounds French) but we haven't for years.
And, working in schools, I can tell you that this is EXACTLY how children are taught to apply for jobs and have been for about the last two decades (I was, when I was at school/university). And it's exactly how every adult job application I've ever seen has arrived. You send a CV which incorporates all your qualifications and work experience, wrapped in a two-line cover letter (Dear Sir/Madam, My CV, Thanks). If asked, you provide a "personal statement" which is an A4 diatribe about your work life and what you can bring to a company.
Maybe you're just nit-picking on the origins of the word but in the UK, you have a CV that tells you EVERYTHING you need on one page. You normally head it "Curriculum Vitae" (which lets you instantly discard those who can't be bothered to look up "Vitae").
There are hundreds of obsolete skills. But there is one that NEVER goes out of fashion, never gets obsoleted, never stops coming in handy... the ability to learn quickly, and to remember what you learn. Seriously. I make a point of putting it on my CV (resume for you Americans). Learning quickly means you can adapt to ANY environment quickly. Remembering what you've learned means you can draw parallels and keep "generic" knowledge going. Bung me in front of a particular UNIX server and I might not have any idea how to do much but login. Give me ten minutes to acclimatise and I can be doing ANYTHING on it. TCP/IP is TCP/IP... the places where you change the settings may differ but knowledge of protocols, routing, etc. is the same whether it's a Commodore 64 on the web or a network of virtualised Vista PC's.
I don't have MCSE, CCNA or anything else because the sheer fact is that by the time you've passed the course and been using it for a year, its content is out of date. Not all of it, but quite a bit of it. Especially on those courses designed for particular bits of software. And they are nothing but memory tests. That's not learning.
I've done assembly, I've done BASIC and everything in between. My University tried to teach me Java until I stopped attending the lectures for that part and was instead "hired out" to other students as the person to ask about the Java coursework. I'd only ever dabbled in it but having programmed in a lot of other languages it was no more than a curiousity to flick through a Java book and pick up the syntax. I did the coursework myself at home, taught many others to pass the course, and passed myself (good grades for that course) with barely a sweat. I'd dabbled in Java before but it was merely a matter of flicking through a half-decent book on the subject, applying everything else you already know and making sure you have a list of function-method-procedure (call them whatever you like, OO is just a shortcut that saves you typing so much functional-programming code) name changes handy. KMP search algorithms are the same in any language, it's just a matter of learning or merely memorising (which is NOT learning) the differences between languages.
Similarly, my primary job is being hired by schools to manage their networks. First one was 98-standalones with Ethernet cables basically used for display.:-) The next was a full Windows 98 network with custom management software and an NT server. The next was a 2000 Server and XP network with custom management software. The next was plain 2000 + XP with Active Directory. The next was similar but with some other custom management software bodged to perform some of the more tedious tasks. The next was Server 2003 + XP + Vista. And so on. The last one I had was another "design me a network from scratch" for a school, and so they got Server 2003, XP and Linux for some tasks (it was just easier, made more sense and cheaper).
Formal training in any of the above OS, network management, network management software or application software? Zilch. Number of networks exploded? Zilch. Number of networks more productive once I had finished with them? 100%. Number of schools chasing me for further employment to work on their next big network, next OS, next suite of applications? I lose count. And these are critical networks - they run everything from the canteen to the staff wages to the legally required paperwork to the student desktops to the fire and security systems. You have no idea how crippled a school is nowadays if its servers go down... lessons stop, systems go haywire and the students get sent home. And they literally fight over getting an imbecile like me in to manage their systems, or even just clean them up so that they can employ a "normal" technician next year.
If you can learn, you can run any OS, of any age, at any time, in any combination without a problem. If you can't then you're stuck memorising "Windows Vista for Dummies" until the next OS comes out a
Er... well... I've personally come across several brands of USB drive that automatically ran programs on stock Windows installs from manufacturers (maybe Dell or somebody turned the options on, I don't know, the point it is shouldn't even be an option). If I remember, you can even specify "actions" for USB drives to present in the "what do you want to do" dialog that appears when you pop it into a slot, quite easily. Usually USB drives have a setup.exe for encryption etc. but a lot of them are also bootable disks with a boot partition, or have a hidden partition to run setup programs from. I'm not saying that you're not right, that's it not the default, but I've seen it happen on "new" machines.
Additionally, I was only positing a theory... the fact is that to catch a "virus" you have to execute it. This means either these users double-clicked an unknown executable (not inconceivable for stupid people) or it somehow executed itself on insertion. Maybe they were pre-SP2 installs of XP, I don't know.
- Run an OS that does not automatically try to mount devices, without user interaction. - Run an OS that does not execute programs on devices once mounted, without user interaction but preferably not at all. (Autorun, I'm looking at you)
Although what doesn't seem to mentioned specifically is if the viruses are contained on the memory of the frames themselves (i.e. just like any other removeable drive) or whether they are on some sort of driver/bundle CD. It does seem to hint that it means the device itself, which begs the question how is it getting executed? Is there a setup.exe that autoruns like on certain brands of USB drive (DUMB IDEA OF THE CENTURY)? Are there infected data files like JPEG's that just so happen to allow execution of their code on certain OS's? Is there an actual executable that isn't supposed to be on there at all that autoruns or waits for the user to double-click it?
Either way, it's hardly a brilliant way to spread and only a dozen or so people seem to have been affected out of whichever country it's talking about (presumably the US). That sounds more like they had the virus already and it made its way onto their digital photo frames when they first connected them. Yes, it's a worry that malicious code could make its way onto a consumer device at the factory, but more at fault here are the OS and the user practices - we had all this back in the 80's/90's... don't take floppies off people you don't trust without scanning them first. Have we seriously come full-circle to the same dumb, preventable "problem"?
It's interesting, yes, but it all seems to be making excuses for a poor SMB implementation.
The true test, obviously, is if copying files over SMB using Linux/Samba is the comparable to Vista in terms of speed, CPU usage, RAM usage etc. My guess is that Vista still appears to suck, but I haven't done that test. It's a file copy. You can come up with all the fancy explanations and function-flow diagrams that you want... the fact is that it takes TOO LONG to copy a relatively small amount of files over a network using Vista. It worked fine in XP. It works fine in other OS's using compatibility layers to talk the same protocols. There really is no excuse. I don't care if you're using 64kb packets or 60kb, I don't care what order you're doing things in so long as the data appears on the other end in a reasonable amount of time.
It worries me more when they can't fix the simple problems... what the hell do they do when they find a BIG problem like data corruption?
Seriously, you're telling me that a version number jump in the kernel during a Service Pack is somehow news? And not only that but *unconfirmed* reports of that. *With screenshots*. Wow.
And what does it do. What does the new 0.0.1 add to Windows? Dunno. There isn't a word about it in the article, just some screenshots of version numbers.
How the bloody hell does this make the front page?
"You DO know you can change the login dialog for Vista back to the old dialog that lets to select the Domain, just like in XP, don't you? You can even set this as a Group Policy."
Try it. Really. Without having to download third-party software. It's *close* but it's not what you think. It really doesn't do what I need it to. I need drop-down Domain/Local.
"Please tell me which imaging system allows you to push down a 4GB image over PIIX boot in less than 5 minutes."
Certainly. Ghost. Security Suite 2 (Ghost 10 or 11, I lose count). GhostCast running via some boot-disk-emulation over a PXE boot menu. Average deployment time of an image with XP + Office + all the usual (Flash, Quicktime, Shockwave, Realplayer, Adobe Reader, an unzip util etc.) in an admittedly highly-compressed image - 3:30. You should really get yourself a new network. Unicast is much faster for what I'm talking about (single machine rebuild) but multicast does more machines, so mass deployments are best done on a closed all-gigabit network without other traffic but I get these same times over our actual production network (which isn't anything fancy to be absolutely honest). Maybe another minute or two for things like SID-regenerations etc. but then you have a working desktop.
"I've never used Ghost for Vista images (because you're not supposed to)."
See below, it's supported by Ghost. But... My point exactly. Why can't I? They've just made another reason for me to not abandon my already long-established machine build/rebuild procedure for little to no reason.
"Ghost is a shitty way to deploy Windows. Period."
It works for us. It works faster than you could get it to work yourself. It's simple, clean, efficient, and ties in perfeclty to the way the network operates. Users can do it THEMSELVES if you give them the password - boot, press F12 before Windows comes up, select from the nice netboot menu, type in password, the machine rebuilds in about 5 minutes (depends on the machine obviously, because it's mostly disk writing). And each time you do it, you KNOW the computer will end up in a perfect working state exactly how you intended. And the fact is that it works for all OS that we'll ever want to use, so long as the PC in question can PXE boot and we can get a disk image of the disk.
"If you have to have lots of multi-boot images you're doing something weird, like hardware testing."
Or, say, having more than one OS. Or even having "unusual" partitions to perform utility/rebuild functions as above. It's really not that difficult or unusual and to be honest, it's really not that important. My point is that I ALREADY possess a superior solution to the suggested built-in-to-Windows-and-crippled equivalent, as do most people who do it for a living.
"BTW, Ghost was discontinued years ago. Symantec Ghost is actually rebranded ImageCast and, unsurprisingly, it's not compatible with old Ghost images."
Quote: Support for Microsoft Windows Vista and 64-Bit OS's. Purchasing Options: This product is available for purchase online in the Symantec Store as well as through resellers
I bought 50 additional licenses from an large reseller just the other day and media was still available if I wanted it. How is that discontinued?
Discontinued nope. Maybe not perfectly up-to-date but then I don't really care because, as is the point of this thread, I don't care about Vista or 64-bit, like 99% of places that deploy PC's. Plus all it does is bit-copy disks to a file from a DOS shell so it's hardly a complex program to support.
"You simply don't know what you're talking about. Do you understand how these tools are even used? You do realize that you NEVER have to make custom images again?"
Yep. But the build of one/two custom images means that I save more time than it takes to even set up the systems to deploy the
Slashdot Mirror
The problem I have with this is that you're only getting half the story, if that. First, about twenty bloggers had to comment on the original graphs because they didn't include any version of Opera, which could only make a fairer test - FF vs IE is hardly a shocker. The Opera line on the graph instantly made a better competitor and blew most of their shocking claims about how well FF did compared to "other browsers" away - it sat dead in the narrow line between the two versions of Firefox. Next, the test is very specific about how it operates and requires lots of similar options to be enabled on all the tested browsers (to counteract memory caches and different cache-management algorithms). That all boils down to mean that in a certain, small area (*memory* management under a particular caching profile), the latest version of FF "wins". Not bad seeing as that's what the statistics were designed to show in the first place.
My main bugbears are that it doesn't take any account of CPU usage (I could decrease the memory usage of a program by half if CPU speed was no obstacle by just compressing everything I put into and take out of RAM), user-experience, cross-browser differences (so turning off some cache options present only in certain browsers is cheating because you still don't KNOW how they work or what they sacrifice for performance elsewhere. Default settings would have made more sense given that most people don't touch caching etc. options and that that's what a novice user would see. You would, of course, have to account for user-experience at the same time), the total amount of RAM installed (who's to say that Opera wouldn't adjust itself on a 256MB RAM machine to do less caching etc.?), total disk size (IE with several Gigs of disk cache isn't unusual nowadays but my Opera does just fine with a few hundred), "nice factor" to other programs (i.e. does it assume that you're ONLY running the browser, or is it pitched to let you use other programs alongside it well?) and a million and one other factors.
Given that I browse with Opera on a 600MHz laptop with 348Mb, I'm very happy with its performance. It can't do heavy-Java well but that's expected. It can do streaming video, dozens of simultaneous websites, Flash, all sorts without coming to a grinding halt. The startup speeds are reasonable and not due to disk-thrashing (it looks more CPU-bound to me). Firefox, any version, on the same machine slows to an absolute crawl after a few websites and thrashes like mad on startup.
I worry that programmers are looking at those graphs and congratulating themselves, without any real metric of what else has changed or stayed the same. Sure, memory allocations have dropped and there is more real RAM available to the machine but what does that mean if it's at the expense of other resources or, worse, the user experience?
Let's ignore all your points for a second and cut to the crux of the matter. The country you live in could legally enforce all of your suggestions absolutely perfectly. It wouldn't make a dent. You could do it in twenty, fifty countries. You still wouldn't make a dent. Law is not universal. In my continent you can't HAVE software patents, they actually do not exist. You aren't going to make that change any time soon no matter what your country does. Similarly for any legal resolution to spam, viruses, botnets etc. Even if 50% of the world's botnets are on American PC's (for example), by definition even the owner's don't want them or even know they are there. Nor do the ISP's, or the transport carriers, or anyone else along the line. But it's like suing people because they gave you a cold - they didn't want to catch the cold in the first place and, yes, although there are measures they can take to lessen their potential exposure to the virus, nothing is guaranteed.
1) "static IP's" - we can already trace where all the stuff comes from - there are complete trails back to the sending machines and from there back to the perpertrators. But most of it generally comes from computers abroad, or from people attacking computers from abroad, or via proxies, all of which are subject to different laws and untouchable. Even ASKING for the details belonging to a particular IP that resides in a foreign country is unbelievably difficult. And you won't get them, but your law enforcement might. And you think you can shut them off before they cause damage because you have their IP address? Nope. It's too late. By that time, the botnet's already moved on to take advantage of the next exploit. We have dynamically updating realtime, very expensive blocklists with dedicate people to add new machines as they are found - they don't stop that much, really.
2) "Laws that require people to assume some form of responsibility when they connect a computer to the net." - in every country in the world. With similar provisions. Quickly. Not going to happen. EVER. And then you're into why do you have to take responsibility and how do you ensure it? Your kid put a virus on your machine? I'll sue you, then. No? You caught a spyware toolbar which send me spam? I'll sue you, again. You'd either sue people literally off their computer seats, everything would get thrown out of court, or you've just helped the government introduce legislation to make them monitor everything you do at your computer, with fingerprint ID required to logon.
3) "Perhaps some form of compulsory insurance policy." - For owning a computer? No. If you could tax people for being stupid, the world would be split between the bankrupt and the filthy rich.
4) "Laws that require ISP's to disconnect spam bots and take some responsibility." - So now they're responsible for their users actions? They won't let you do it. If you do, they will shut themselves down and get out of the business. They ALREADY disconnect bots - it is in their interests. They ALREADY have to deny all responsibility for your actions. And they are ALREADY in deep legal grey areas because of the burden of proof of doing such things and the expense of a mistake (Sorry, Company X, I thought you sent a spam. I've just cut off your Internet by mistake. Bye-bye online business).
But the fact is that none of your measures are sensible or practical, some are even impossible, and all of them are in place in one way or another today. The fact is that every country in the world has a different idea. If we can't convince them all that death by execution or torture might be a bad idea, how the hell do you think you're going to get them to shut down botnets?
The comment about the computing power was not meant to demean, and certainly wasn't to be taken to represent this particular effort but "AI" efforts in general - it was meant to illustrate the sheer scale of the problem. The order of magnitude between a research student on a supercomputer and the actual computer power you have mentioned above is incomparable to all of the "knowledge" that you are trying to cram into the space (data storage, data diversity, computing time, etc.), as my comment was meant to show. It's just not going to work.
The most powerful supercomputers in the world can't get the weather right four days out of five (although that involves prediction, which is a different area entirely), they can *barely* beat a human expert at chess but only then mostly through *sheer brute force* with a couple of well-thought out tricks. And chess is an entirely logical, rule-based and *extremely* simple game in comparison to things such as Go, which is what my professors were studying trying to "beat" a human at when I was at university. Go is something like 10^19 times more complex but with vastly reduced rulesets compared to chess? I can't remember the exact number. And at the moment an "average" skilled human player will wipe the floor with basically any computer-based Go opponents, no matter what supercomputer is behind them. And that's in a perfectly controlled, 100% logical, rule-based environment, which is a different type of AI entirely.
We're not talking about a processing/storing "shortage" hindering our ability to really demonstrate even a basic, recognisable intelligence, we talking *orders of magnitude* out of our predicted range for even the next few decades to get a basic chatbot working enough to sound passable. It's like monkeys that can't add up trying to crack AES-512. As you rightly point out, there is some enormous brute force behind the chatbot (sorry, that's all it is) mentioned in the article, and still we get only quite a minor step forward, if some of the transcripts I've seen are anything to go by.
Hello chatbot. Decode this for me: "nrvsidr o fpm#t".
Serious answer: I do believe they can get close... but not for a few hundreds years or so at *absolute* minimum. But because EVERYONE is either going about it completely the wrong way (let's program what we know of 4-year-old's into a logical engine, or let's let something evolve that has about 100 "neurons" and hope it learns English) or the study that they need to do is far beyond our computational capability, the current techniques are next-to-useless.
How do intelligent things learn? Let's take a primitive example of an "intelligent" organism. They have *near-zero* historical data at startup - no expert databases, no pre-knowledge, nothing. Over the next ***four years***, they are constantly subjected to intense training, vast amounts of data and are heavily rewarded as and when they form connections. They don't have their internals poked and prodded and fine-tuned along the way, they just get on with it. They operate on scales and mechanisms that we can only dream of (super-computing wise), take in amounts of input that we could only just barely simulate and they are taught each item FROM SCRATCH. First primitive sounds, then associating those sounds with things, then the alphabet, then words, then grammar, then cleaning it up into a real language.... We don't do things the right way.
For example, in this article, grab a baby, deafen it and remove all other sensory perception except sight (and even that's much more info that this bot is getting), throw up some text for a few years and wait for it to hold an intelligent conversation by sending text back. It ain't gonna happen, even with a proven-intelligent being. There is no "meaning" to the words. You have to prime and train and exemplify and clarify and correct and all sorts. You can train programs to train programs but it's all the same in the end. It needs real feedback. It needs to know WHY what it said was wrong, not just get a "FALSE" entry in a database somewhere. It needs to learn that fire burns and holding your breath isn't wise to do for too long.
We could *train* an AI. I see that as a possibility - training some enormously complex system from scratch to recognise simple patterns and then build up to language slowly. But it takes a proven-successful organism at least a few years to get that far working 24 hours a day with a constantly-busy environment and enormous amounts of feedback that are automatically categorised into "good" (tastes nice, is fun, etc.) and "bad" (hurts) by the organism itself, not some pre-programmed list. And every time you want to improve the underlying algorithm, you're starting from birth all over again. We can train an AI, it'll take decades just for a basic one, but I very much doubt that we'll ever create one at all, certainly not within a few thousand years or so.
Oh come on, it's a chatbot not an AI. I did my Computing degree, I know how AI on computers is supposed to work and it's seriously laughable for anything more complex than extremely primitive, less-than-insect intelligence. You'll see AI "geese" flocking, you'll see AI "ants" making a good path, planning ahead and fending each other off but none of it is actually "AI".
AI at the moment consists of trying to cram millions of years of evolution, billions of pieces of information and decades of years of "actual learning/living time", from an organism capable of outpacing even the best supercomputers even when it's just a single-task (e.g. Kasparov vs Deeper Blue wasn't easy and I'd state that it was still a "win" for Kasparov in terms of the actual methods used) - let's not even mention a general-purpose AI - where just the data recorded by said organism in even quite a small experience or skillset is so phenomenally huge that we probably couldn't store it unless Google helped, into something that a research student can do in six months on a small mainframe. It's not going to work.
Computers work by doing what they are told, perfectly, quickly and repeatably. Now that is, in effect, how our bodies are constructed at the molecular/sub-molecular level. But as soon as you try to enforce your knowledge onto such a computer, you either create a database/expert system or a mess. It might even be a useful mess, sometimes, but it's still a mess and still not intelligence.
The only way I see so-called "intelligence" emerging artificially (let's say Turing-Test-passing but I'm probably talking post-Turing-Test intelligence as well) is if we were to run an absolutely unprecedented, enormous-scale genetic algorithm project for a few thousand years straight. That's the only way we've ever come across intelligence, from evolved lifeforms, which took millions of years to produce one fairly pathetic example that trampled over the rest of the species on the planet.
We can't even define intelligence properly, we've never been able to simulate it or emulate it, let alone "create" it. We have one fairly pathetic example to work from with a myriad of lesser forms, none of which we've ever been able to surpass - we might be able to build "ant-like" things but we've never made anything as intelligent as an ant. That doesn't mean we should stop but we should seriously think about exactly how we think "intelligence" will just jump out at us if we get the software right.
You can't "write" an AI. It's silly to try unless you have very limited targets in mind. But one day we might be able to let one evolve and then we could copy it and "train" it to do different things.
And every chatbot I ever tried has serious problems - They can't reason gobbledegook properly because they can't spot patterns. That's the bit that shows a sign of real intelligence, being able to spot patterns in most things. If you started talking in Swedish to an English-only chatbot, it blows its mind. If you started talking in Swedish to an English person, they'd be trying to work out what you said, using context and history of the conversation to attempt to learn your language, try to re-start the conversation from a known base ("I'm sorry, could you start again please?" or even just "Hello?" or "English?"), or give up and ignore you. The bots can't get close to that sort of reasoning.
I live on the trailing edge. It's easier, more comfortable, cheaper (how's that HD-DVD purchase looking now?) and makes much more sense. You don't end up buying into much that disappears quickly. You don't end up with teething problems. You don't end up wasting money on vast amounts of testing because other people do that for you. You don't end up worrying about that next big upgrade which is "vital" because one bit of software demands it. You just wait until EVERYTHING demands it or, more likely, nothing demands it and the project dies a death.
I do network management for schools and this is the only sensible way to go about things. You cannot upgrade every single bit of software to the latest version as soon as it's released. Too much breaks. Even Windows Update throws default printers, etc. onto a clean machine and that's the easy stuff to detect and clean up.
For instance, last year one school went from Office 2000 to Office 2003 (Yep, that was our upgrade last year - and that was forced on us by the "people above" because 2000 is such an old number, basically. Nothing to do with practicality, compatibility or anything else and we'd already installed the filters that let you open newer Word docs etc. so there wasn't a simple complaint when other schools switched to XP/2003/2007 and started sending us their files). That simple upgrade broke everything - printer allocations, all our Office GPO settings, basically every PowerPoint file in the school, some software called Mindjet Manager that just so happened to be on the desktop of all the top-dogs in the schools, you name it.
The Mindjet problem wasn't even documented, we found it out the hard way (some sort of DLL conflict to do with the order of installation that just crashed the program without error - sysinternals utilities saved the day again!). Upgrading breaks more than it fixes about 75% of the time.
The only reason to upgrade is if you get something out of it. That's how companies sell upgrades - "look what you can do NOW". If you don't need that thing, or can already do it somehow, then why upgrade? Just keep plodding along with your perfectly working software with all the latest security updates that does everything you could possibly want.
Even Becta (the UK schools IT bods) recognise this - they specifically recommend not to move to Vista or the new Office until EVERYTHING has been checked out thoroughly, i.e. give it a few years yet. Better the devil you know and all that. This is why MS is having a hard time selling Vista - it's only the "we must have it because it's got a shiny new version number" crowd that are the ones actually pushing for it and unfortunately sometimes they are in charge.
What happens in a modern office that is totally impossible with a old office suite? Not a lot. And most of it is power-user features that most places won't even touch. Even I never use 5% of the functionality of Excel, Word, etc. when I'm doing budgets with pretty graphs in and the kids are basically doing DTP using Word (against my advice, but hey, Word works and we have it so why spend money on something else?) and they aren't using 5% of it anyway. Where's the incentive?
And you don't get bitten anywhere near as often. Schools go through fads - we had interactive whiteboards and flatscreens and the new ones are "Virtual Learning Environments", electronic signage (a big telly with an advert on it) and wireless classroom trolleys filled with laptops. Every single fad starts up, fails miserably in a number of schools (usually because they have copied it straight from the commercial sector and are thinking along the business-consumer lines rather than "THIS IS A SCHOOL. It might not work like a business does.") and THEN the suppliers learn what the schools want/need and bring out the really useful gear that the following schools can then actually use - for instance:
Soft whiteboards got vandalised by the older kids. So they changed to hard.
Hard whiteboards weren't tactile enough for
I'm really not a Microsoft advocate but I work in Windows-only schools... I do completely understand the frustration but off the top of my head:
1) Group Policy integration
I can set browser settings (every single one that you see in IE's Internet options etc.) for a group of computers and fine-tune each setting to particular users. You don't want kids changing their proxy settings, you don't want them playing with trusted zones, you don't want them getting overwhelmed with dialogs when they log in for the first time. You don't want to have to set it up per-user.
2) Content that is IE only.
Everything is web-based nowadays. For instance, schools are pulling lesson plans, educational games, etc. off websites or "cached content" (i.e. stick a Linux box full of some companies flash-heavy website in your school and every month or so they'll update it for you... this is an actual product from several very expensive companies). Most of it works only in IE, although I have seen a few that will work on Firefox too. ActiveX is seriously over-used in the educational sector for example. And therefore if you want support (which schools in particular see as vital), you have to use IE.
3) The little blue E
Seriously, I have dozens of users (and I get more each year) who when they want the Internet go for the little blue E shortcut. I have actually replaced it with an identical-looking one that ran Firefox and only the "power-users" on the network actually noticed anything wrong. 450 kids just carried on working as if there was nothing different.
As soon as everybody gets off their backsides and gives me non-IE support on all the content that the schools carry, I'd move every school over. The next hurdle after that is making it easy to customise every option on a per-user/per-computer basis properly. The third is easily overcome even if it is selling your soul to put a decent browser behind the IE icon...
In most of the above points the presumption seems to be that because the resolution difference exists, the average person can distinguish it from a given distance. This is clearly not true.
We can have the same old arguments again if you want. Barring artifacts caused by mains flicker, enormous screen size, peripheral vision, strobing of ambient lighting etc. - the average person can't really see a difference between 50 and 60Hz refresh rates, or 60 and 100Hz. They can't really see the difference between 16-bit (properly done) and 24-bit. They can't hear the difference between MP3 at certain bitrates and the raw 48-bit audio data at 96KHz. They can't really see the difference between interlaced and progressive (hence why TV over in Britain has been 576 lines, 25fps, interlaced, for the last few decades and nobody really "noticed" anything out of place).
At reasonable distances, the average person just cannot discern, in a fair test, the difference between these formats. There will be a minority (much more of a minority than those who "claim" they can see it, as there are so many other factors to take account of) who can detect some of them, of course, (e.g. photosensitive epileptics and, for example, people who can "see" the flicker in flourescent lighting) but most of the difference I see is down to different colour mappings producing more vivid colours and greater contrast, wider screens, closer distance. I also spot deinterlacing artifacts, poor SDTV sources being compared to clean HDTV sources, TV's that deliberately make SDTV look worse through poor upscaling, etc. and they stand out more to me than one pixel becoming four.
Nobody EVER sells me HDTV based on those things above (e.g. that HDTV's have newer substances in their displays that can give me a better colour range etc.) They all sell it to me on the fact that there are more dots. Dots that, once I take a step backwards, I can't distinguish one dot from the next. Seriously, on a smooth, moving video (not a black pixel on a white screen), pick out a single pixel and tell me what colour it is and track it as it changes. Yes, I can spot a stuck green pixel on a standard LCD screen at twenty yards and wouldn't be able to watch a LCD TV with a stuck pixel. But if you were to play some video and switch between 1024x768 and 800x600 on a good, fully-working monitor at a reasonable distance without affecting "actual" size of the video or colour ranges, I (and most of the population) wouldn't be able to tell. In the average use, on a *fair* HDTV you just will not notice any difference between SDTV and HDTV unless you buy an enormous HDTV or sit with your face in the screen. I don't want to do either. Six to ten feet is a good distance for TV viewing for me.
I have done these tests countless times in shops, friends houses etc. I'm itching for an excuse to buy a good gadget, but HDTV and its associated technology (e.g. HDTV sources like HDTV channels and Blu-Ray etc.) are just not worth ANYTHING to me. People take me round their house and show me their HDTV and it's just like watching TV. I've even gone to the point of dragging out an old SDTV and running the same signal through it because my friend's model of HDTV seemed to "cheat" when displaying SDTV and it actually was WORSE through the HDTV than through a ten-year-old television of comparable size. That sounds like sucker-sales-tactics to me ("hah, when they switch to an SD signal it'll look horrible and they'll think it was like that all along!") and it's obvious that my friend was one of the people that fell for it, and I doubt he's alone.
Oh, and my old 4:3 screen actually showed a larger visible image than his widescreen TV could, even with the black borders and the fact that my TV, when bought brand new ten years ago, cost less than half of his did today. Hence I don't have any widescreen TV's either because, for many years, a bigger 4:3 TV was actually superior up to the size of most affordable widescreens.
Next you'll be telling me that Bose Wave S
Back in my day, Windows was a WIMP GUI. Now they call it an operating system.
But yes, the old-style viruses tend to have lost out the past few years. I can remember quaking in fear when I read about a virus that was polymorphic, stealth, boot-sector infecting, "hold your partition table to ransom", able to transfer to floppies, hard disks and even CD's (WOW!), plus across IPX networks, randomising data destruction etc.
Now THAT was a virus to be scared of.
I don't have vast experience of Vista because I decided against deploying it on the last few networks I managed. However, it seems that it must still be incredibly easy to access the MBR even, as you point out, as a non-administrator user.
However I assume that "Boot sector protection" as available in most modern BIOS's should stop this stone dead (I know that I implement it but I doubt everyone does). It's like 1989 all over again...
Granted, the virus is easily cleaned, although it's potential effects may not be (identity theft etc.).
Maybe it's just me remembering the good old days of DOS viruses but none of this actually seems "new", except for "calling the NDIS API directly, using a 'code pullout' technique to load just the parts of ndis.sys that it needs", which is just a shortcut to direct hardware access... it's basically loading another copy of a library to use it seperately from the OS (and therefore, presumably, OS security) to access the network card. It's a nice way to access a network card from "below" the OS in a hardware-independent way (i.e. let's pinch the Windows driver rather than try to work out what card we are using and create 1000 drivers for each different brand of card.). But on the whole, this is hardly "new" or "shocking".
:-)
Having said that, it's nice to see something where people have actually invested time and skill into creating a program that bypasses the OS in such a way, rather than just another re-written script with a couple of variables changed.
Lines like:
"This malware is very professionally written and produced. Which of course means it's not written for fun."
might annoy some, though. The old DOS viruses NEVER really acheieved anything useful (even with blackmail attempts while holding your boot sector to ransom) etc. and were written "just because" by teenagers. That didn't stop them from appearing professionally written and breaking genuinely new ground for the time. Just because people are now using such malware for financial gain, doesn't mean that it's ALWAYS the case. And Linux zealots are sure to jump on the above quote with all their hearts.
And then you have the obvious - why is the OS allowing you to modify the MBR without appropriate rights and/or why are users running as users with the rights necessary to do this? This is STILL a problem harking back to the DOS days - everyone as administrator. With a new twist - the average user hasn't needed to BE administrator for quite a long time now.
This is just blatantly not true.
I use Linux BECAUSE of its backwards compatibility. I recycle old machines into computers for schools, etc. and if Linux weren't backwards compatible none of them would even boot. As it is, you can still run Linux on a 386 with ISA cards (is MCA still supported, I can't remember?), the same can't be said about Windows which blatantly removes old architectures in order to force upgrades (ISA support disappeared in XP, even though you can still hack it a bit to make it work perfectly). I'm not saying there's not a justification for not supporting ancient systems but the fact is that Linux does it for as long as ANYBODY with coding experience cares, not just the Microsoft pencil-pushers.
Not only is Linux backwards-compatible, but it supports things I've never even heard of, and things I absolutely no need for, everything from ten-year-old watchdog timers to the latest and greatest datacentre interconnects. It's also sideways-compatible in that, if you're feeling sadistic, you can run Windows NDIS drivers on it without modification.
Linux has always been backwards compatible in virtually all senses of the word. I can still load ext2 filesystems, a filesystem over fourteen years old, and that's not even the weirdest or oldest FS that I can use! I can still use ancient hardware for which drivers don't exist past DOS or Windows 3.1. I can still use programs that were written for Linux 1.0, so long as I have the right libraries in place.
Now OVERALL compatibility may be slightly worse than Windows (I don't believe it, given the things that Linux supports that Windows couldn't do in a million years) in that it can't use EVERY scanner, printer, etc. because of rubbishy, propreitry protocols from companies where nobody even remembers how they work any more, let alone document them or write a driver for them. But to say Linux isn't backwards compatible is like saying that your average Windows driver from a £10 scanner is stable.
Although the statements are all well and good, you have to make a lot of assumptions for the maths to work - namely that people are ordered, logical and will know where they are going.
Many moons ago, my A-Level project consisted of something quite similar. I had to do a traffic simulation and work out what the "ideal" interval was for traffic-light changes. I spent months on the project, got top marks because there were absolutely no mathematical errors in it and it had quite a lot of higher-level maths in there.
The results of my formulae were applied to a simple crossroads, with cars accelerating and deccelerating perfectly at fixed points of the traffic light cycle, depending on their speed, the position and interval of the lights, etc. Cars entered the system at a uniform rate from all junctions and several phases of the traffic lights were simulated.
Unfortunately, the formulae took all this information and the answers (for traffic light change intervals) were given as solutions to a large polynomial. Despite the mathematics being verified several times by several people, and the assumptions being called "reasonable", the optimal answers given for this simplest of simulations was to have traffic lights change at intervals of -5 seconds and 0.02 seconds. Neither of which gave the cars time to accelerate across the junction (or even be able to SEE the lights change) completely.
Factoring in some more terms for things like "being able to cross the junction before the lights change", "being able to physically react to the light changing", etc., even more bizarre answers were arrived at until a bit of "fudge factoring" in the initial assumptions made it give sensible answers.
I quite believe that the methods used currently are possibly the worst *mathematically* but I doubt they're the worst for airports and humans to handle. If it was, it would have changed already to get people on/off and get the planes turned around faster.
Still to do this day, I wish the school had had the resources at the time for me to do a proper computer-controlled visual simulation. Alas, they were still teaching 18-year-olds BBC BASIC at the time.
Yes. I have. Several friends have them. My comment still stands. I've purchased several for my job - we use large ordinary HDTV's as "digital signage", i.e. things showing looped Powerpoint presentations. I admit that you CAN tell the difference, close up, with a nice, high-res signal (HDTV or even just a decent PC resolution) than the same on a "normal" SDTV, the same way you can between a monitor and a SDTV showing the same signal... of course you can - but that's not the point. The point is, in normal use, from a reasonable distance, you really don't notice the difference. I've also looked at them seriously for purchasing for personal use and because I'm *that* guy that everyone comes to asking what the best model of *vaguely-electronic-equipment* is.
I've seriously spent hours in shops and looked at every single HDTV on offer. I try them all out, one by one (first thing I look for is what connectors they have on the back). I then make sure that any display TV's are displaying actual HDTV content. I carefully measure the distance back so that I'm at my usual viewing position, and I can't see picture difference - except I do notice that, with a *few* brands, you can side-by-side compare a similarly priced large SDTV and HDTV, while both showing SD picture, and *sometimes* the HDTV makes the SD signal look ever so slightly "worse" (I can't define it, I'm sorry) than the SDTV showing the same SD signal. A ploy to make you think that you HDTV is actually an improvement when you see a "real" HD signal? Some technical reason to do with the interpolation etc? I have no idea.
The only exception is if the screen is MUCH larger than one I can a) afford or b) fit on my wall / in my living room comfortably and then the difference is mainly that it's BIGGER. Don't get me wrong, in a few years you won't be able to get an SDTV, in the same way that you won't be able to get a CRT monitor except for specialist uses, or a non-digital-reception TV etc. Maybe my eyes are just crap. But I don't see the difference. I've even been found in shops by my wife, squinting from a few inches at a HDTV trying to see what the difference is. As soon as you get to a reasonable viewing distance, my eyes just can't discern one pixel from another and it's no better than SD shown at a similar size on a similarly priced "normal" TV. Call me mad, call me a liar but **I** am not paying for anything that **I** can't notice no matter how many people tell me how much better it is, or how much knowledge I have of exactly how many pixels the thing has got.
It's the same old story, to a point. The performance required to do a relatively simple job (play fullscreen video) in a new way (using HD content and a new storage medium) means that it becomes impractical without upgrades. I can remember having to tweak computers to be powerful enough to play MP3's without skipping, but there at least you had the advantage that the storage space saved compared to even the best-compressed formats of the time was phenomenal.
I freely admit that I absolutely do not "get" the HD fuss. It's the same thing we've had for years, with more pixels, that you can't reasonably see on a fair test past a certain distance (although I would say that on a high-res laptop you are more likely to spot the difference because of the unusually close eye-screen distance), with new storage formats, new compression, new software, new DRM and new performance characteristics... which are killing battery life. And, yes, eventually they'll start making "blu-ray acceleration cards" just like MPEG-acceleration, 3D-acceleration, etc., although in this day and age they're called "software on the GPU". But at the end of the day, you've gained little (a higher res that you might not be able to distinguish) for enormous performance increases.
Where's the advantage in it when a "Blu-ray" PC can still play the DVD's of previous years but at much, much less expense... if you can play a blu-ray for two hours or you can play MPEG-2 for six (while compiling stuff in the background without jerkiness) on the same machine, what are you going to end up using if you watch a lot of video on your laptop?
When I go away and know that I might want to view movies on my laptop (e.g. long trip staying in cheap hotels, stay over at a friends house etc), I take either DVD's, or I have a bunch of MPG's/AVI's/VOB's etc. on the laptop itself or on DVD-R's ahead of time. Quality isn't really the factor there and the advantage to having everything in a simple format that everyone can read easily and which doesn't tax the laptop is key.
It's another case of "laptop = general purpose computer, so let's turn it into a media centre and make it do everything". It's nice that it's CAPABLE of everything but you can't expect a portable device to do it all AND give you good performance at everything. Laptops are not even desktop-substitutes for most work (the times I have to explain this to people... it costs pounds to repair a broken desktop, hundreds to repair a broken laptop).
Let the early adopters waste their money. Even if Blu-Ray becomes the de-facto standard, I'd much rather just decrypt-to-disk and convert to a format that's easily readable, with extremely cheap media, that plays the video "good enough" for most things if I'm intending to carry it around with me. Much better 1 x DVD-R with a couple of full movies on it that I can watch one-after-the-other and make a backup copy for pennies than 1 x Blu-Ray that I can't give my friends with only a single movie on it that kills my batteries just watching it.
There was a time when I did exactly the same with DVD vs VCD - it's actually trivial to just copy several DVD's worth of movie/tv show to a DVD-R or even a CD-R and not worry about the quality. You're travelling - who cares whether it's HD or VCD-quality so long as you can tell what's going on without eyestrain?
their
Fake. 'Nuff said.
Speaking for the UK:
Erm... You might have been technically correct fifty years ago when everybody still did Latin but not any longer. It's standard practice in every company I've ever worked, for the last 20-odd years and probably a lot longer than that. CV = personal details + list of qualifications + (summarised) work history + references + (possibly) brief statements about important project you've done in the past (unless you're under 18 when it seems you're taught to list "interests" like "watching TV"... I've actually seen it on a CV). You usually cram it all onto 1 or maybe 2 A4 pages. Possibly a personal statement (1 page of A4) if requested stating briefly why you'd be good at the job, clarify anything listed on the CV. Cover letter to wrap it all up (two lines and some decent letter-writing layout).
In the UK, no job agency would distinguish between CV and resume, unless they only deal with very pretentious clients, but I've never heard of anyone distinguishing over such a minor detail even in the City. Job adverts include the line "Send CV to:". Never any mention of anything else, except possibly "CV and Personal Statement". We don't even CALL it, or any part of it, a resume. We might have done once (I don't know the etymology, but it sounds French) but we haven't for years.
And, working in schools, I can tell you that this is EXACTLY how children are taught to apply for jobs and have been for about the last two decades (I was, when I was at school/university). And it's exactly how every adult job application I've ever seen has arrived. You send a CV which incorporates all your qualifications and work experience, wrapped in a two-line cover letter (Dear Sir/Madam, My CV, Thanks). If asked, you provide a "personal statement" which is an A4 diatribe about your work life and what you can bring to a company.
Maybe you're just nit-picking on the origins of the word but in the UK, you have a CV that tells you EVERYTHING you need on one page. You normally head it "Curriculum Vitae" (which lets you instantly discard those who can't be bothered to look up "Vitae").
There are hundreds of obsolete skills. But there is one that NEVER goes out of fashion, never gets obsoleted, never stops coming in handy... the ability to learn quickly, and to remember what you learn. Seriously. I make a point of putting it on my CV (resume for you Americans). Learning quickly means you can adapt to ANY environment quickly. Remembering what you've learned means you can draw parallels and keep "generic" knowledge going. Bung me in front of a particular UNIX server and I might not have any idea how to do much but login. Give me ten minutes to acclimatise and I can be doing ANYTHING on it. TCP/IP is TCP/IP... the places where you change the settings may differ but knowledge of protocols, routing, etc. is the same whether it's a Commodore 64 on the web or a network of virtualised Vista PC's.
:-) The next was a full Windows 98 network with custom management software and an NT server. The next was a 2000 Server and XP network with custom management software. The next was plain 2000 + XP with Active Directory. The next was similar but with some other custom management software bodged to perform some of the more tedious tasks. The next was Server 2003 + XP + Vista. And so on. The last one I had was another "design me a network from scratch" for a school, and so they got Server 2003, XP and Linux for some tasks (it was just easier, made more sense and cheaper).
I don't have MCSE, CCNA or anything else because the sheer fact is that by the time you've passed the course and been using it for a year, its content is out of date. Not all of it, but quite a bit of it. Especially on those courses designed for particular bits of software. And they are nothing but memory tests. That's not learning.
I've done assembly, I've done BASIC and everything in between. My University tried to teach me Java until I stopped attending the lectures for that part and was instead "hired out" to other students as the person to ask about the Java coursework. I'd only ever dabbled in it but having programmed in a lot of other languages it was no more than a curiousity to flick through a Java book and pick up the syntax. I did the coursework myself at home, taught many others to pass the course, and passed myself (good grades for that course) with barely a sweat. I'd dabbled in Java before but it was merely a matter of flicking through a half-decent book on the subject, applying everything else you already know and making sure you have a list of function-method-procedure (call them whatever you like, OO is just a shortcut that saves you typing so much functional-programming code) name changes handy. KMP search algorithms are the same in any language, it's just a matter of learning or merely memorising (which is NOT learning) the differences between languages.
Similarly, my primary job is being hired by schools to manage their networks. First one was 98-standalones with Ethernet cables basically used for display.
Formal training in any of the above OS, network management, network management software or application software? Zilch. Number of networks exploded? Zilch. Number of networks more productive once I had finished with them? 100%. Number of schools chasing me for further employment to work on their next big network, next OS, next suite of applications? I lose count. And these are critical networks - they run everything from the canteen to the staff wages to the legally required paperwork to the student desktops to the fire and security systems. You have no idea how crippled a school is nowadays if its servers go down... lessons stop, systems go haywire and the students get sent home. And they literally fight over getting an imbecile like me in to manage their systems, or even just clean them up so that they can employ a "normal" technician next year.
If you can learn, you can run any OS, of any age, at any time, in any combination without a problem. If you can't then you're stuck memorising "Windows Vista for Dummies" until the next OS comes out a
Yeah, they would have been much more accepted if they had pronounced it "Heidi DVD". :-)
I always think the funniest acronym is PXE UNDI - it sounds like fairy knicker to me.
Er... well... I've personally come across several brands of USB drive that automatically ran programs on stock Windows installs from manufacturers (maybe Dell or somebody turned the options on, I don't know, the point it is shouldn't even be an option). If I remember, you can even specify "actions" for USB drives to present in the "what do you want to do" dialog that appears when you pop it into a slot, quite easily. Usually USB drives have a setup.exe for encryption etc. but a lot of them are also bootable disks with a boot partition, or have a hidden partition to run setup programs from. I'm not saying that you're not right, that's it not the default, but I've seen it happen on "new" machines.
Additionally, I was only positing a theory... the fact is that to catch a "virus" you have to execute it. This means either these users double-clicked an unknown executable (not inconceivable for stupid people) or it somehow executed itself on insertion. Maybe they were pre-SP2 installs of XP, I don't know.
- Run an OS that does not automatically try to mount devices, without user interaction.
- Run an OS that does not execute programs on devices once mounted, without user interaction but preferably not at all. (Autorun, I'm looking at you)
Although what doesn't seem to mentioned specifically is if the viruses are contained on the memory of the frames themselves (i.e. just like any other removeable drive) or whether they are on some sort of driver/bundle CD. It does seem to hint that it means the device itself, which begs the question how is it getting executed? Is there a setup.exe that autoruns like on certain brands of USB drive (DUMB IDEA OF THE CENTURY)? Are there infected data files like JPEG's that just so happen to allow execution of their code on certain OS's? Is there an actual executable that isn't supposed to be on there at all that autoruns or waits for the user to double-click it?
Either way, it's hardly a brilliant way to spread and only a dozen or so people seem to have been affected out of whichever country it's talking about (presumably the US). That sounds more like they had the virus already and it made its way onto their digital photo frames when they first connected them. Yes, it's a worry that malicious code could make its way onto a consumer device at the factory, but more at fault here are the OS and the user practices - we had all this back in the 80's/90's... don't take floppies off people you don't trust without scanning them first. Have we seriously come full-circle to the same dumb, preventable "problem"?
It's interesting, yes, but it all seems to be making excuses for a poor SMB implementation.
The true test, obviously, is if copying files over SMB using Linux/Samba is the comparable to Vista in terms of speed, CPU usage, RAM usage etc. My guess is that Vista still appears to suck, but I haven't done that test. It's a file copy. You can come up with all the fancy explanations and function-flow diagrams that you want... the fact is that it takes TOO LONG to copy a relatively small amount of files over a network using Vista. It worked fine in XP. It works fine in other OS's using compatibility layers to talk the same protocols. There really is no excuse. I don't care if you're using 64kb packets or 60kb, I don't care what order you're doing things in so long as the data appears on the other end in a reasonable amount of time.
It worries me more when they can't fix the simple problems... what the hell do they do when they find a BIG problem like data corruption?
Seriously, you're telling me that a version number jump in the kernel during a Service Pack is somehow news? And not only that but *unconfirmed* reports of that. *With screenshots*. Wow.
And what does it do. What does the new 0.0.1 add to Windows? Dunno. There isn't a word about it in the article, just some screenshots of version numbers.
How the bloody hell does this make the front page?
"You DO know you can change the login dialog for Vista back to the old dialog that lets to select the Domain, just like in XP, don't you? You can even set this as a Group Policy."
Try it. Really. Without having to download third-party software. It's *close* but it's not what you think. It really doesn't do what I need it to. I need drop-down Domain/Local.
"Please tell me which imaging system allows you to push down a 4GB image over PIIX boot in less than 5 minutes."
Certainly. Ghost. Security Suite 2 (Ghost 10 or 11, I lose count). GhostCast running via some boot-disk-emulation over a PXE boot menu. Average deployment time of an image with XP + Office + all the usual (Flash, Quicktime, Shockwave, Realplayer, Adobe Reader, an unzip util etc.) in an admittedly highly-compressed image - 3:30. You should really get yourself a new network. Unicast is much faster for what I'm talking about (single machine rebuild) but multicast does more machines, so mass deployments are best done on a closed all-gigabit network without other traffic but I get these same times over our actual production network (which isn't anything fancy to be absolutely honest). Maybe another minute or two for things like SID-regenerations etc. but then you have a working desktop.
"I've never used Ghost for Vista images (because you're not supposed to)."
See below, it's supported by Ghost. But... My point exactly. Why can't I? They've just made another reason for me to not abandon my already long-established machine build/rebuild procedure for little to no reason.
"Ghost is a shitty way to deploy Windows. Period."
It works for us. It works faster than you could get it to work yourself. It's simple, clean, efficient, and ties in perfeclty to the way the network operates. Users can do it THEMSELVES if you give them the password - boot, press F12 before Windows comes up, select from the nice netboot menu, type in password, the machine rebuilds in about 5 minutes (depends on the machine obviously, because it's mostly disk writing). And each time you do it, you KNOW the computer will end up in a perfect working state exactly how you intended. And the fact is that it works for all OS that we'll ever want to use, so long as the PC in question can PXE boot and we can get a disk image of the disk.
"If you have to have lots of multi-boot images you're doing something weird, like hardware testing."
Or, say, having more than one OS. Or even having "unusual" partitions to perform utility/rebuild functions as above. It's really not that difficult or unusual and to be honest, it's really not that important. My point is that I ALREADY possess a superior solution to the suggested built-in-to-Windows-and-crippled equivalent, as do most people who do it for a living.
"BTW, Ghost was discontinued years ago. Symantec Ghost is actually rebranded ImageCast and, unsurprisingly, it's not compatible with old Ghost images."
http://www.symantec.com/business"/products/overview.jsp?pcid=2260&pvid=865_1
Quote: Support for Microsoft Windows Vista and 64-Bit OS's.
Purchasing Options: This product is available for purchase online in the Symantec Store as well as through resellers
I bought 50 additional licenses from an large reseller just the other day and media was still available if I wanted it. How is that discontinued?
Discontinued nope. Maybe not perfectly up-to-date but then I don't really care because, as is the point of this thread, I don't care about Vista or 64-bit, like 99% of places that deploy PC's. Plus all it does is bit-copy disks to a file from a DOS shell so it's hardly a complex program to support.
"You simply don't know what you're talking about. Do you understand how these tools are even used? You do realize that you NEVER have to make custom images again?"
Yep. But the build of one/two custom images means that I save more time than it takes to even set up the systems to deploy the