We actually do irregular (they should be regular, but we've never found a schedule that works) reboots on work systems, just to make sure they come back up okay. Problems are rare, but do sometimes happen, and I'd much rather find out a system has reboot issues during scheduled downtime, rather than just after a power failure.
Kind of. There's tricks you can do, for example carousel, where you continously send the same file out again and again. So people can start listening at any point, receive to the end of the file in the current sending, then listen for the first half when it's broadcast again. Fcast (http://research.microsoft.com/barc/mbone/fcast.as px) is an example of this approach.
There are more complex tricks, but they're probably not worth going into.
Oh, or you could use swarmcasting (for example, BitTorrent). It's not as good, but tends to mean bandwidth usage is more localised (because clients will tend to connect to clients close to themselves in network terms).
Hell, as university _staff_ having a lab full of machines has been useful, when my desktop system has decided it just doesn't want to play nicely, and you don't have the time to get it working again. It's not a common thing, but it does happen.
Thought - has the president switched to using just a laptop? I don't mean, he has a laptop he brings into the office, and plugs into a monitor and keyboard, I mean that's all he uses? Maybe now's a good time to suggest that staff should be moving to laptops of the spec the students will be using, to work out any last minute bugs, then wait to see how long before one of them goes crazy.
People figure this out by looking at corner cases, and prodding stuff to see if it breaks. Most exploits are fairly simple though; we're finally getting away from buffer overflows, but they're easy to find by looking at where programs deal with a string, and seeing what happens if you put a much too large string in. Time consuming, but straight forward.
There are some genuinely skilled crackers out there, but they're fairly few and far between. I maintain a bunch of computers, and most of them deal with a cracking attempt a day. Let me give you a quick log extract:
Feb 21 03:22:56 <hostname> sshd[25243]: Invalid user firebird from <IP removed> Feb 21 03:22:57 <hostname> sshd[25245]: Invalid user art from <IP removed> Feb 21 03:22:59 <hostname> sshd[25247]: Invalid user manu from <IP removed> Feb 21 03:23:00 <hostname> sshd[25249]: Invalid user peru from <IP removed> Feb 21 03:23:02 <hostname> sshd[25251]: Invalid user contra from <IP removed> Feb 21 03:23:03 <hostname> sshd[25253]: Invalid user fbi from <IP removed> Feb 21 03:23:05 <hostname> sshd[25255]: Invalid user melanie from <IP removed>
That's just someone trying random username/password combinations and hoping. Eventually, they'll find somewhere with looser security, and get in, but that doesn't make them skilled, it makes them annoyingly persistant.
Don't get me wrong, this OS X exploit is actually fairly interesting, but most crackers have just enough knowledge to be dangerous, and not enough to use it wisely.
If you want impressive, have you considered the people securing these things? They don't have to find just one security hole, they have to find them all. They have to know every way someone might try breaking the system, and then some...
Hell yes! Four years after graduating, with continuous employment (finished my last exam on the Thursday, started work on the Tuesday), I find myself renting a 10'x10' room in a three bedroom ex-council property, because it's what I can afford. Rents around here have almost doubled in 4 years, it's terrible.
What's really frustrating is being told by people who bought their house for a fraction of its current value, that the housing market should be this high. Or that I should buy somewhere, and then I can start saving for somewhere bigger; if the $15,000 extra I'd need to buy somewhere around here suddenly dropped into my lap, and I bought a two bedroom place and rented out the second bedroom, I might be able to save some money. That sounds less like a plan and more like a recipe for disaster to me...
Let me get this straight; I'm not living in the most awkard way possible (the important part with food is that it is locally grown), and as such cannot critise you for massive waste of resources?
Errr?
I don't own a car, and I walk most places. On the infrequent occaisions I need items not sold locally, I take a bus to the nearest city. I share a three bedroom house with two other people, and insulate it as best I can given I'm merely renting it. I use energy saving bulbs, and turn lights off in rooms we're not using.
Sure, it's not perfect, but I think it gives me a point to say "I'm putting this much effort in, maybe you could at least try?"
Exactly what I was about to suggest. Best of all, no monthly fee, so you aren't shelling out however much per month for a game you'll have limited time to play...
I wrote an application like that once (I was young and foolish). But I'm better now, I swear (we spent most of summer removing that, and a few other screwups)...
Last week, I bought Kameo, a full price XBox 360 game, and Marble Blast Ultra, an XBox Live arcade downloadable game for a fifth the cost.
One of these games has myself and my flatmates addicted, the other I'm working slowly through mostly out of curiousity. I'm sure I don't have to tell you which one is which.
This is not to say that smaller, easier to play games are the way forward, or giant stunning masterpieces are a bad idea, it's just an example. However, I do strongly believe the only thing pushing up development costs is the developers themselves.
Let me go back to Kameo: the game is stunning. Fight scenes involving the main character and a few thousand trolls are incredible. The environments are richly detailed, and all have their own distinct character. However, I mostly find it over complex. There's nothing quite as frustrating as losing the camera behind a beautifully rendered tree while a couple of trolls beat you into the ground!
Developers need to focus on what's important. Make the game pretty, but not so much that it's over complex. Focus on making levels interesting, over attractive. Ensure the difficulty curve is effective, and you're most of the way there - you want to make sure the player is always challeneged, but never frustrated.
Ah, right, the problem is the article's support page is talking gibberish. It's out, both as a trial (downloadable, and on the PGR3 game disk - you can access it by walking around your garage and going over to the arcade machines) and as a full game...
IE still lacks XHTML support of any kind - I don't want to seem picky here, but it has been 6 years. Sure, I can have applications I work on spit out XHTML that's mostly like HTML 4, and send the appropriate MIME type based on the Accept header, but I'd really quite like to see IE support vaguely recent standards, y'know...
The issue I think for a lot of people is the difficulty in getting that experience. Let me go metaphorical for a second; ever been playing a computer game and had a bit that you just couldn't get past? I don't mean it was challenging, I mean hair-pulling out screaming at the devs hard.
Figuring out how to deal with people is like that, to me. Social situations are only fun if you can actually interact meaningfully with people; if on the other hand you're way out of your depth, missing half the conversation (as in, the half that's in body language), and have nothing to say, it's not exactly fun. Learning to deal with people, from that sort of starting point, is a multi-year task (I imagine you could do it in a year, maybe even less, with someone who knew what they were doing to coach you, but...).
So, yes, after years of throwing myself into social situations I had nearly no ability to deal meaningfully with, I can do most of this on intuition, but it's a hell of an effort for something most people seem to pick up instinctively.
Even worse, with things the way they are currently, it would be near-trivial to ruin someone's life just by e-mailing them kiddie-porn. Just include the words "Here are the files you requested", and make sure you tip off the police before your victim has time to find and delete the e-mail.
This has stopped becoming a search for justice, and rapidly becoming a witch-hunt...
There's a time and a place for security testing other people's systems. If you're playing around, and notice something that might be a security hole, correct solution is to find an admin that knows what they're talking about, and say "Hey, I see your system does , isn't that a problem?".
The real problem with "hobbyist" cracking is that, as a sys-admin, I have no way of telling that you're just playing around with a few things, as opposed to you've managed to rootkit my computer, and it's now running half a dozen freaky kernel modules that are hiding whatever you're doing to it, from me, unless I can prove that you didn't manage to get into anything deeper (for example, you might have managed to access a service you shouldn't, but I know that service couldn't give you root access).
Let me be more specific: If we believe a system has been compromised (and yes, we've done this to a system that was just doing "odd stuff" once), we back it up, reformat the drive, and re-install, change all the passwords, then rebuild it from the data in the backups. Depending on which system was compromised, this will generally be upwards of a days work, including time to fix problems that result from having a system down for reinstall.
This is not enough to ensure security is maintained, but is appropriate for the level of security we require. Theoretically, we'd have to do the same for all related systems (for example, we frequently ssh between servers, so a password could be sniffed).
Having said that, we're fairly easy going in terms of policies. For attempted cracking (port scans, HTTP exploits, attempts to brute force an SSH password), we send an e-mail to the person if we can, or to their ISP otherwise, saying there's been a problem with that IP address, and the person responsible should check it for viruses (probably not true in all cases, to I'd certainly expect it to be true for the majority).
For the two times anyone got through our security, the time someone came to us and said "Err, there's a giant security hole here", we basically said "Yeah, we know, could you not tell anyone for the next week while we finish the patch we're working on?". The other time, someone broke into one of our servers from off-site, and last I heard central IT services were tracking them down...
One day? One day??? They installed them at the new building we've just moved in to; every 15 minutes I have to lean to the left and wave my arm frantically to get the lights to work!
*twitch*
Don't start me on the fact that we can't turn the lights off, so they're blazing away throughout summer. Although we kinda need them, because they made the windows tiny "to save energy".
Graphics: Nothing I've seen gives me any indication of how powerful this will actually be, only that it's designed by ATi, and designed as a completely different architecture to everything else. Having said that, I think we're at the point where it's fairly trivial to stuff as many polygons into a 480i/p display as you want, and unless it's missing some funky technology such as the lighting/shader stuff the XBox 360 and PS3 (will) have, there shouldn't be any real difference.
By which I mean, I think you're right. Pity it's not R5xx based though, it looked like XBox 360/Revolution dual format releases were going to be more or less trivial...
Memory: Yikes, that's not much at all. I'm suprised; memory is fairly cheap stuff, putting 256mb in would probably have made life a lot easier on devs, but nevermind...
The more I read about the supposed specs of the Revolution the more it sounds like it will, outside of higher resolution, easily outpace the 360 in performance. The dual 970ish CPU that is in the Revolution will easily outperform the 360 CPU - which looks like it has realworld performance around a 2 to 2.5GHz dual 970 system. And the custom ATI graphics system in the Revolution sounds like it will be much more advanced than the essentially current gen pc GPU in the 360 - outside of resolution once again.
Err, huh? Most of the rumours I'm seeing put the architecture of the XBox 360 and Revolution to be identical, with the following differences:
Triple core 3.2 GHz CPU in XBox 360, dual core 2.8Ghz CPU in Revolution 500Mhz R500 core for the XBox 360, 600Mhz R520 core for the Revolution (this seems to more be the order they were designed in, rather than an indication of features) 48 shader pipelines for XBox 360, 32 for Revolution 512MB shared RAM on XBox 360, 512MB system and 256MB graphics for Revolution
So... the Revolution will probably have more memory, be slower in CPU terms, and a close match in graphics. It will probably be cheaper, but I'd imagine Microsoft will be looking at a price drop not long after the Revolution is released.
> Several are going into the Playstation 3, so will this require a fan? You're aware the PS2 had a fan, right? I think it may just be behind the PSU, but still, there's one in there...
My XBox 360, with it's triple-core G5-varient processor, is not what I'd call quiet (yes, I could drown it out by putting the TV volume up, but it's definitely audible at the volume I like to play games). I'd expect the Cell to have similar heat output too...
"Sex is not a mere role-playing concession in Sociolotron; it is a fully-baked and detailed gameplay mechanic."
Does this sound like a really bad idea to anyone else? Reminds me worryingly of F.A.T.A.L. (which is an abomination, IMHO).
Consider, why does the game need a gameplay mechanic for sex? Answer, because the outcome is important ("the fluids that are exchanged and spilled during sexual encounters are critical components of some potions") - at that point, I think the game is moving beyond not avoiding sex, and into becoming conceringly focused on it.
I'm sure this has a market, and they're welcome to it, but I wouldn't consider this "Sex and the Modern MMOG", I'd consider it "Sex-orientated MMOG".
> Even worse is if its a google cache. Does anyone use Google cache except where the page they're after is unavaiable (server down/overloaded, or page moved)? Certainly it's all I use it for...
> It's full of trickery and scams, people who would kill you for just coming into their asteroid belt, PC characters more evil than any of the NPCs, spying, scheming, infiltration, betrayal, and distrust...
Thing is, I see that as different, not necessarily better. Certainly, it's not something I want out of a MMORPG - I play these things to relax, I want to not spend my entire time watching my back, y'know? I tend to avoid PvP for the same reason - it's good on occaision, but not what I want out of most of my game time...
Slashdot Mirror
We actually do irregular (they should be regular, but we've never found a schedule that works) reboots on work systems, just to make sure they come back up okay. Problems are rare, but do sometimes happen, and I'd much rather find out a system has reboot issues during scheduled downtime, rather than just after a power failure.
Kind of. There's tricks you can do, for example carousel, where you continously send the same file out again and again. So people can start listening at any point, receive to the end of the file in the current sending, then listen for the first half when it's broadcast again. Fcast (http://research.microsoft.com/barc/mbone/fcast.as px) is an example of this approach.
There are more complex tricks, but they're probably not worth going into.
Oh, or you could use swarmcasting (for example, BitTorrent). It's not as good, but tends to mean bandwidth usage is more localised (because clients will tend to connect to clients close to themselves in network terms).
Thirded.
Hell, as university _staff_ having a lab full of machines has been useful, when my desktop system has decided it just doesn't want to play nicely, and you don't have the time to get it working again. It's not a common thing, but it does happen.
Thought - has the president switched to using just a laptop? I don't mean, he has a laptop he brings into the office, and plugs into a monitor and keyboard, I mean that's all he uses? Maybe now's a good time to suggest that staff should be moving to laptops of the spec the students will be using, to work out any last minute bugs, then wait to see how long before one of them goes crazy.
People figure this out by looking at corner cases, and prodding stuff to see if it breaks. Most exploits are fairly simple though; we're finally getting away from buffer overflows, but they're easy to find by looking at where programs deal with a string, and seeing what happens if you put a much too large string in. Time consuming, but straight forward.
There are some genuinely skilled crackers out there, but they're fairly few and far between. I maintain a bunch of computers, and most of them deal with a cracking attempt a day. Let me give you a quick log extract:
Feb 21 03:22:56 <hostname> sshd[25243]: Invalid user firebird from <IP removed>
Feb 21 03:22:57 <hostname> sshd[25245]: Invalid user art from <IP removed>
Feb 21 03:22:59 <hostname> sshd[25247]: Invalid user manu from <IP removed>
Feb 21 03:23:00 <hostname> sshd[25249]: Invalid user peru from <IP removed>
Feb 21 03:23:02 <hostname> sshd[25251]: Invalid user contra from <IP removed>
Feb 21 03:23:03 <hostname> sshd[25253]: Invalid user fbi from <IP removed>
Feb 21 03:23:05 <hostname> sshd[25255]: Invalid user melanie from <IP removed>
That's just someone trying random username/password combinations and hoping. Eventually, they'll find somewhere with looser security, and get in, but that doesn't make them skilled, it makes them annoyingly persistant.
Don't get me wrong, this OS X exploit is actually fairly interesting, but most crackers have just enough knowledge to be dangerous, and not enough to use it wisely.
If you want impressive, have you considered the people securing these things? They don't have to find just one security hole, they have to find them all. They have to know every way someone might try breaking the system, and then some...
I've got an idea. Buy a PowerMac, get a friend to take it to pieces, and then you can have the fun of putting it together? :D
Hell yes! Four years after graduating, with continuous employment (finished my last exam on the Thursday, started work on the Tuesday), I find myself renting a 10'x10' room in a three bedroom ex-council property, because it's what I can afford. Rents around here have almost doubled in 4 years, it's terrible.
What's really frustrating is being told by people who bought their house for a fraction of its current value, that the housing market should be this high. Or that I should buy somewhere, and then I can start saving for somewhere bigger; if the $15,000 extra I'd need to buy somewhere around here suddenly dropped into my lap, and I bought a two bedroom place and rented out the second bedroom, I might be able to save some money. That sounds less like a plan and more like a recipe for disaster to me...
Let me get this straight; I'm not living in the most awkard way possible (the important part with food is that it is locally grown), and as such cannot critise you for massive waste of resources?
Errr?
I don't own a car, and I walk most places. On the infrequent occaisions I need items not sold locally, I take a bus to the nearest city. I share a three bedroom house with two other people, and insulate it as best I can given I'm merely renting it. I use energy saving bulbs, and turn lights off in rooms we're not using.
Sure, it's not perfect, but I think it gives me a point to say "I'm putting this much effort in, maybe you could at least try?"
Exactly what I was about to suggest. Best of all, no monthly fee, so you aren't shelling out however much per month for a game you'll have limited time to play...
I wrote an application like that once (I was young and foolish). But I'm better now, I swear (we spent most of summer removing that, and a few other screwups)...
Last week, I bought Kameo, a full price XBox 360 game, and Marble Blast Ultra, an XBox Live arcade downloadable game for a fifth the cost.
One of these games has myself and my flatmates addicted, the other I'm working slowly through mostly out of curiousity. I'm sure I don't have to tell you which one is which.
This is not to say that smaller, easier to play games are the way forward, or giant stunning masterpieces are a bad idea, it's just an example. However, I do strongly believe the only thing pushing up development costs is the developers themselves.
Let me go back to Kameo: the game is stunning. Fight scenes involving the main character and a few thousand trolls are incredible. The environments are richly detailed, and all have their own distinct character. However, I mostly find it over complex. There's nothing quite as frustrating as losing the camera behind a beautifully rendered tree while a couple of trolls beat you into the ground!
Developers need to focus on what's important. Make the game pretty, but not so much that it's over complex. Focus on making levels interesting, over attractive. Ensure the difficulty curve is effective, and you're most of the way there - you want to make sure the player is always challeneged, but never frustrated.
Ah, right, the problem is the article's support page is talking gibberish. It's out, both as a trial (downloadable, and on the PGR3 game disk - you can access it by walking around your garage and going over to the arcade machines) and as a full game...
IE still lacks XHTML support of any kind - I don't want to seem picky here, but it has been 6 years. Sure, I can have applications I work on spit out XHTML that's mostly like HTML 4, and send the appropriate MIME type based on the Accept header, but I'd really quite like to see IE support vaguely recent standards, y'know...
Yes.
The issue I think for a lot of people is the difficulty in getting that experience. Let me go metaphorical for a second; ever been playing a computer game and had a bit that you just couldn't get past? I don't mean it was challenging, I mean hair-pulling out screaming at the devs hard.
Figuring out how to deal with people is like that, to me. Social situations are only fun if you can actually interact meaningfully with people; if on the other hand you're way out of your depth, missing half the conversation (as in, the half that's in body language), and have nothing to say, it's not exactly fun. Learning to deal with people, from that sort of starting point, is a multi-year task (I imagine you could do it in a year, maybe even less, with someone who knew what they were doing to coach you, but...).
So, yes, after years of throwing myself into social situations I had nearly no ability to deal meaningfully with, I can do most of this on intuition, but it's a hell of an effort for something most people seem to pick up instinctively.
Even worse, with things the way they are currently, it would be near-trivial to ruin someone's life just by e-mailing them kiddie-porn. Just include the words "Here are the files you requested", and make sure you tip off the police before your victim has time to find and delete the e-mail.
This has stopped becoming a search for justice, and rapidly becoming a witch-hunt...
There's a time and a place for security testing other people's systems. If you're playing around, and notice something that might be a security hole, correct solution is to find an admin that knows what they're talking about, and say "Hey, I see your system does , isn't that a problem?".
The real problem with "hobbyist" cracking is that, as a sys-admin, I have no way of telling that you're just playing around with a few things, as opposed to you've managed to rootkit my computer, and it's now running half a dozen freaky kernel modules that are hiding whatever you're doing to it, from me, unless I can prove that you didn't manage to get into anything deeper (for example, you might have managed to access a service you shouldn't, but I know that service couldn't give you root access).
Let me be more specific: If we believe a system has been compromised (and yes, we've done this to a system that was just doing "odd stuff" once), we back it up, reformat the drive, and re-install, change all the passwords, then rebuild it from the data in the backups. Depending on which system was compromised, this will generally be upwards of a days work, including time to fix problems that result from having a system down for reinstall.
This is not enough to ensure security is maintained, but is appropriate for the level of security we require. Theoretically, we'd have to do the same for all related systems (for example, we frequently ssh between servers, so a password could be sniffed).
Having said that, we're fairly easy going in terms of policies. For attempted cracking (port scans, HTTP exploits, attempts to brute force an SSH password), we send an e-mail to the person if we can, or to their ISP otherwise, saying there's been a problem with that IP address, and the person responsible should check it for viruses (probably not true in all cases, to I'd certainly expect it to be true for the majority).
For the two times anyone got through our security, the time someone came to us and said "Err, there's a giant security hole here", we basically said "Yeah, we know, could you not tell anyone for the next week while we finish the patch we're working on?". The other time, someone broke into one of our servers from off-site, and last I heard central IT services were tracking them down...
Getting caught by the boss would probably mean just having to fix half a dozen more sensors, rather than getting into trouble :)
One day? One day??? They installed them at the new building we've just moved in to; every 15 minutes I have to lean to the left and wave my arm frantically to get the lights to work!
*twitch*
Don't start me on the fact that we can't turn the lights off, so they're blazing away throughout summer. Although we kinda need them, because they made the windows tiny "to save energy".
Arrgh!
Ah, so what you're saying is the title should have been "Asynchronous Requests for XML with Javascript and Ajax"? :)
It's really useful if you reference your sources, BTW, so here's some I found for you:
m e-new-info-on-the-nintendo-revolution/t ml2 491
http://www.engadget.com/2005/12/06/ign-digs-up-so
http://revolution.ign.com/articles/673/673578p1.h
http://www.eurogamer.net/article.php?article_id=6
Graphics: Nothing I've seen gives me any indication of how powerful this will actually be, only that it's designed by ATi, and designed as a completely different architecture to everything else. Having said that, I think we're at the point where it's fairly trivial to stuff as many polygons into a 480i/p display as you want, and unless it's missing some funky technology such as the lighting/shader stuff the XBox 360 and PS3 (will) have, there shouldn't be any real difference.
By which I mean, I think you're right. Pity it's not R5xx based though, it looked like XBox 360/Revolution dual format releases were going to be more or less trivial...
Memory: Yikes, that's not much at all. I'm suprised; memory is fairly cheap stuff, putting 256mb in would probably have made life a lot easier on devs, but nevermind...
CPU: Yup, agree.
Err, huh? Most of the rumours I'm seeing put the architecture of the XBox 360 and Revolution to be identical, with the following differences:
Triple core 3.2 GHz CPU in XBox 360, dual core 2.8Ghz CPU in Revolution
500Mhz R500 core for the XBox 360, 600Mhz R520 core for the Revolution (this seems to more be the order they were designed in, rather than an indication of features)
48 shader pipelines for XBox 360, 32 for Revolution
512MB shared RAM on XBox 360, 512MB system and 256MB graphics for Revolution
So... the Revolution will probably have more memory, be slower in CPU terms, and a close match in graphics. It will probably be cheaper, but I'd imagine Microsoft will be looking at a price drop not long after the Revolution is released.
Revolution rumours mostly taken from:
http://arstechnica.com/news.ars/post/20050923-534
> Several are going into the Playstation 3, so will this require a fan?
You're aware the PS2 had a fan, right? I think it may just be behind the PSU, but still, there's one in there...
My XBox 360, with it's triple-core G5-varient processor, is not what I'd call quiet (yes, I could drown it out by putting the TV volume up, but it's definitely audible at the volume I like to play games). I'd expect the Cell to have similar heat output too...
"Sex is not a mere role-playing concession in Sociolotron; it is a fully-baked and detailed gameplay mechanic."
Does this sound like a really bad idea to anyone else? Reminds me worryingly of F.A.T.A.L. (which is an abomination, IMHO).
Consider, why does the game need a gameplay mechanic for sex? Answer, because the outcome is important ("the fluids that are exchanged and spilled during sexual encounters are critical components of some potions") - at that point, I think the game is moving beyond not avoiding sex, and into becoming conceringly focused on it.
I'm sure this has a market, and they're welcome to it, but I wouldn't consider this "Sex and the Modern MMOG", I'd consider it "Sex-orientated MMOG".
> Even worse is if its a google cache.
Does anyone use Google cache except where the page they're after is unavaiable (server down/overloaded, or page moved)? Certainly it's all I use it for...
> It's full of trickery and scams, people who would kill you for just coming into their asteroid belt, PC characters more evil than any of the NPCs, spying, scheming, infiltration, betrayal, and distrust...
Thing is, I see that as different, not necessarily better. Certainly, it's not something I want out of a MMORPG - I play these things to relax, I want to not spend my entire time watching my back, y'know? I tend to avoid PvP for the same reason - it's good on occaision, but not what I want out of most of my game time...
Or a sys-admin, maybe? I work with computer systems that need to be kept reasonably in sync, time wise, and NTP is a good way of doing that...