On the other hand, we will be seeing sites with a lot of completely useless content that make you search around all day to find what you're looking for.
Sure, sessions will work for sites like forums. However, is there going to be anything shown by session length that won't be shown by page views in that case? What about pages that you can really spend days to weeks at a time staring at, such as the glibc manual or the Coyotos microkernel specification? If the user never refreshes the page before the end of the session, information-packed sites aren't going to be measured at all.
"I have a very, very hard time coming up with reasons why I should recommend someone go out and buy a new GNU/Linux PC, because it seems to make so much more sense to recommend they go out and buy a new Mac."
It's even easier to convince them not to buy a new computer, because Linux will run like new on their old machine.
All animals impact their environment, and humans are not exempt from that. We're beginning to realise that there are some things we can change. Allowing our farms to be overrun by plants that kill off our crops and get in the way is not a viable solution for large farms, but using our smarts to remove weeds instead of herbicides is a step in the cleaner direction. And it's a step in a more enlightened direction, too, if it means we spend less effort farming.
The code is pretty well commented, and it's usually worth having a good look at the sections you're interested in. It's well organised and clean [in the interesting parts]. It has to be, it's modified by a lot of people. Seriously, you won't be sorry. Above this, Google is your best friend. IBM's developer works has good holistic info, and Linux HQ has lots of links to great information. Kernel hacking on your own time is pretty easy, I don't think it needs to be taught. If you want to work on THE Linux kernel, the official archives and documentation are a must.
I'm not saying it's that easy, what I said wasn't even a first order approximation on how to share state in networked games. But the moment you realise that you're designing a network game, you need to start assuming the worst. Assuming the client understands all the information you've sent it, and can send anything it likes. Attempting to lock down hardware that you don't control anyway isn't the answer- it's thinking about the value of the information you're sending.
Of course, what you're saying is right- you can't hard clip without a margin to account for latency [and partially visible characters]. Especially when the server must calculate this, because then you're talking about the latency of a round trip. So you're always going to have SOME players possibly seeing through walls. But, planning for security often makes a significant difference.
I've written for similar scenarios, and the correct way is probably not to send the position state of the other players unless visible. That might sound like a lot of computing power is required and a lot of bandwidth would be used, but cutting planes can be tested on regions in three multiplications, two additions, and a compare, per player. Actual per-player state that would need to be sent would likely be what weapon they were holding, the angle of their legs, their position [a three vector] and their orientation [at most a three vector]. The set of vertices of the actual character model do not need to be sent, but can be held in arrays pending the location of the player.
Of course, since this was not the mindset that the developers worked with, it's not what happened. I haven't seen the source myself, but something tells me it could have been done. With a little meditation on the subject, security like this isn't always that difficult or that expensive- in fact, it's usually the Right Way.
The question is, will this hardware stop cheating? It just makes me wonder if the developers ran the idea past someone with a degree in information security first and didn't like the response, or didn't bother.
All you can ever hope for is checking that the packets sent by the machine are those sent by the machine. You might be able to control how the software behaves on the machine, but you will never stop a user without your hardware protection from sending the exact same packets.
You're completely right, but see the [astroturfing] parent of my post for context. They were [supposedly] using kernel and all, but weren't distributing anything, so it was a moot point, but they felt it had to be included in their troll, hence the clarification.
To be honest, posts like theirs scare the crap out of me. If, for example, Microsoft can convince big business to pay patent royalties without Microsoft informing the people who wrote the software in question, than posts like the one I'd replied to may convince small business owners and such to stay away from anything GPL'd.:(
Sorry, you seem to be mistaken. You're only required to publish changes if you/distribute/ code, which is not what is happening at all. And, you don't need to GPL things compiled by gcc, you need to GPL things compiled with glibc, ie, when you actually include GPL'd code in yours. I think. For all I know, glibc could be LGPL'd [now that I think about it, it probably is]. The reason being, you don't actually distribute any part of gcc with your software.
That would be like having being told what you can and can't do with things written in Microsoft Notepad.
What you miss is that NOW, you can turn it off, all by removing a file. Before, attributes were all part of the HTML. If it looked like crap on your machine, you were stuck with it.
With all due respect, you're talking about twenty years of separation. Not to mention that the Cray was a 64-bit design, with enough registers to put any modern processor to shame.
I've poked around their website several times, and I can't work out how to order a MS-free laptop. Is it one of the enterprisey features? I'm looking to buy a new laptop soon, and I'd love to stick with Lenovo, but I'll go elsewhere if I'd be paying a Microsoft tax.
The full name of the OS is in fact "Plan 9 From Bell Labs". There's also a port of the API to a more popular standard called "Plan 9 From User Space", which is cute.
Same here. The music was brilliant, and while the graphics leave something to be desired today, they were pretty good for the time. Not to mention, there were plenty of courses, and four completely different modes of play. It's got pride of place among my other Miyamoto games. I'd wager the SRFX gameplay actually lasted longer than Zelda did.
That silly Sunburnt! Next thing you know, e'll be suggesting that the USA have weapons of mass destruction!
On topic, I think that maybe looking at this from another angle would be better- that of the reader. Anonymous comments may be completely worthless, but we still feel we have some right to sue whoever said them. If you're crying every time some fourteen year old mods you down and calls you fat, you need a psychiatrist, not a lawyer.
Maybe in the past it was different, when people were new to the internet, and this sort of thing was new. But now, it's time to realise that some dork posting comments about you is going to be seen for what e is, and it's not going to reflect on you, and reacting strongly is probably the worst thing you can do. It's time for the internet to grow up.
Because they are flag-burning commies who don't respond to "bend over"?
I understand that there is more to it than that, but the Australian government bends over quite readily, and it has been treated quite well [for smallish values of 'well'] of late.
Slashdot Mirror
On the other hand, we will be seeing sites with a lot of completely useless content that make you search around all day to find what you're looking for.
Sure, sessions will work for sites like forums. However, is there going to be anything shown by session length that won't be shown by page views in that case? What about pages that you can really spend days to weeks at a time staring at, such as the glibc manual or the Coyotos microkernel specification? If the user never refreshes the page before the end of the session, information-packed sites aren't going to be measured at all.
you can't measure that...
What it NEEDS is for wine to be reasonably complete! Time to get hacking..
Btw, lotus 123 didn't run on Windows. It didn't need an operating system at all. It was the hardware that was important.
"I have a very, very hard time coming up with reasons why I should recommend someone go out and buy a new GNU/Linux PC, because it seems to make so much more sense to recommend they go out and buy a new Mac."
It's even easier to convince them not to buy a new computer, because Linux will run like new on their old machine.
An electric razor does the job just fine?
All animals impact their environment, and humans are not exempt from that. We're beginning to realise that there are some things we can change. Allowing our farms to be overrun by plants that kill off our crops and get in the way is not a viable solution for large farms, but using our smarts to remove weeds instead of herbicides is a step in the cleaner direction. And it's a step in a more enlightened direction, too, if it means we spend less effort farming.
Ssh! Next thing you know, M$ will have their legal department designing the file formats. Imagine a document format in legalese!
The code is pretty well commented, and it's usually worth having a good look at the sections you're interested in. It's well organised and clean [in the interesting parts]. It has to be, it's modified by a lot of people. Seriously, you won't be sorry. Above this, Google is your best friend. IBM's developer works has good holistic info, and Linux HQ has lots of links to great information. Kernel hacking on your own time is pretty easy, I don't think it needs to be taught. If you want to work on THE Linux kernel, the official archives and documentation are a must.
I'm not saying it's that easy, what I said wasn't even a first order approximation on how to share state in networked games. But the moment you realise that you're designing a network game, you need to start assuming the worst. Assuming the client understands all the information you've sent it, and can send anything it likes. Attempting to lock down hardware that you don't control anyway isn't the answer- it's thinking about the value of the information you're sending.
Of course, what you're saying is right- you can't hard clip without a margin to account for latency [and partially visible characters]. Especially when the server must calculate this, because then you're talking about the latency of a round trip. So you're always going to have SOME players possibly seeing through walls. But, planning for security often makes a significant difference.
I've written for similar scenarios, and the correct way is probably not to send the position state of the other players unless visible. That might sound like a lot of computing power is required and a lot of bandwidth would be used, but cutting planes can be tested on regions in three multiplications, two additions, and a compare, per player. Actual per-player state that would need to be sent would likely be what weapon they were holding, the angle of their legs, their position [a three vector] and their orientation [at most a three vector]. The set of vertices of the actual character model do not need to be sent, but can be held in arrays pending the location of the player. Of course, since this was not the mindset that the developers worked with, it's not what happened. I haven't seen the source myself, but something tells me it could have been done. With a little meditation on the subject, security like this isn't always that difficult or that expensive- in fact, it's usually the Right Way. The question is, will this hardware stop cheating? It just makes me wonder if the developers ran the idea past someone with a degree in information security first and didn't like the response, or didn't bother.
All you can ever hope for is checking that the packets sent by the machine are those sent by the machine. You might be able to control how the software behaves on the machine, but you will never stop a user without your hardware protection from sending the exact same packets.
The Quake fiasco has already taught us plenty about this: don't trust the user.
You're completely right, but see the [astroturfing] parent of my post for context. They were [supposedly] using kernel and all, but weren't distributing anything, so it was a moot point, but they felt it had to be included in their troll, hence the clarification.
:(
To be honest, posts like theirs scare the crap out of me. If, for example, Microsoft can convince big business to pay patent royalties without Microsoft informing the people who wrote the software in question, than posts like the one I'd replied to may convince small business owners and such to stay away from anything GPL'd.
Sorry, you seem to be mistaken. You're only required to publish changes if you /distribute/ code, which is not what is happening at all. And, you don't need to GPL things compiled by gcc, you need to GPL things compiled with glibc, ie, when you actually include GPL'd code in yours. I think. For all I know, glibc could be LGPL'd [now that I think about it, it probably is]. The reason being, you don't actually distribute any part of gcc with your software.
That would be like having being told what you can and can't do with things written in Microsoft Notepad.
What you miss is that NOW, you can turn it off, all by removing a file. Before, attributes were all part of the HTML. If it looked like crap on your machine, you were stuck with it.
With all due respect, you're talking about twenty years of separation. Not to mention that the Cray was a 64-bit design, with enough registers to put any modern processor to shame.
No doubt they will get some machines too, and now, instead of working their orchards, there will be extra long siestas!
I've poked around their website several times, and I can't work out how to order a MS-free laptop. Is it one of the enterprisey features? I'm looking to buy a new laptop soon, and I'd love to stick with Lenovo, but I'll go elsewhere if I'd be paying a Microsoft tax.
It's the only way really when you've got tax or child support problems.
What do you think this is, 2092? Who has a colour screen anyway?
The full name of the OS is in fact "Plan 9 From Bell Labs". There's also a port of the API to a more popular standard called "Plan 9 From User Space", which is cute.
Same here. The music was brilliant, and while the graphics leave something to be desired today, they were pretty good for the time. Not to mention, there were plenty of courses, and four completely different modes of play. It's got pride of place among my other Miyamoto games. I'd wager the SRFX gameplay actually lasted longer than Zelda did.
That silly Sunburnt! Next thing you know, e'll be suggesting that the USA have weapons of mass destruction!
On topic, I think that maybe looking at this from another angle would be better- that of the reader. Anonymous comments may be completely worthless, but we still feel we have some right to sue whoever said them. If you're crying every time some fourteen year old mods you down and calls you fat, you need a psychiatrist, not a lawyer.
Maybe in the past it was different, when people were new to the internet, and this sort of thing was new. But now, it's time to realise that some dork posting comments about you is going to be seen for what e is, and it's not going to reflect on you, and reacting strongly is probably the worst thing you can do. It's time for the internet to grow up.
Because they are flag-burning commies who don't respond to "bend over"?
I understand that there is more to it than that, but the Australian government bends over quite readily, and it has been treated quite well [for smallish values of 'well'] of late.