My own idea was to build the map of the network into the address space itself. Imagine a unch of routers, each of which is only connected to 4 others, in a regular tiled square grid. Each router could be said to have an XY coordinate. Simple then to just assign subnets such that each router got 10.x.y.0/24.
Mind you, it's not very efficient, but if you assign subnets on non-byte boundaries, there are a few higher-dimensional geometries where the connectedness of any two hosts is no more than about 40 hops (worst case scenario, the vast majority would still be 20 hops or less).
At that point, you'd be able to route a packet based simply on its destination IP. Might still want to send metrics (congestion at this point on the network, etc), but you could tell BGP to go to hell.
The funny thing was, no one could understand how you'd assign subnets, they kept arguing that it needed to be centralized. "What if I connect to XY 4,63 and I decide I want XY 5, 209?". Gee idiot, if you connect to XY 4,63, then you have only a few choices such as 3,63 5,63 4,62 and 4,63. Oh well. I'd even figured out how to get such routing to coexist with the traditional orthodox routing table (stick an entry in it to route all 10.x packets to some virtual device, and have it juggle packets out to the vpn tunnels).
This story reminded me of a project of my own that died not so many months ago. May it rest in peace (Aug 2003 - Aug 2005).
At it's heyday, we had maybe 50 hosts, BGP routing, working DNS with altenate top level domains and reverse dns, websites, email, IRC, a few wikis and so forth. All links were openvpn tunnels, in a decentralized setup where even if one link was severed many still had connectivity. People had no clues who they were communicating with past the first hop, and all first hops connected to someone in a nation legally antagonistic to your own.
But most people don't want that. They just want to be anonymous on the "big internet". They certainly don't want something that isn't some lame java program, and they don't want anything that actually has human readable URLs. They'd rather reinvent every single fucking protocol over and over (freenet's file trading and usenet-like frost, anyone?).
They also apparently don't like a system that means they have to decide who to trust, rather than trusting some software they don't understand to pick and choose who they connect to. Most seemed unwilling or unable to understand the simplicity of the concept, but were more than enthusiastic about DHT algorithms that they couldn't comprehend. More than one seemed downright manic about wifi networks, even though it's generally pretty trivial to use radio equipment to triangulate just who is sending out those packets.
So, to answer the question, you'd have to find enough non-idiots with the spare time and networking skills to have some sort of critical mass. I don't think it will happen myself.
If you're asking me if it's in theory possible to make annoying SVG ads, then yes, I suppose it is. But tell me, how much SVG have you seen on the web, period? Me, there's a guy who links an SVG diagram on his k5 sig, and of course we have the croczilla examples. That's it.
If we're judging each by its deeds, then yeh, for now SVG is nicer.
(I won't even point out that it's open, and others are free to see how I do things and build on it, and with notepad no less...)
Yes. When we invent magical technology that is actually energy dense enough, or if we ever decide to live in some neopaleolithic society, where the kids get to watch an hour of television once a year, then solar is the way to go.
Nuclear works, it's safe, and we've got several hundred years worth of fuel left if we start acting smart about it. That's enough time to get fusion working.
Fusion is the perfect energy source, anyone that says different is a retard. Whether or not it's possible is debatable, I'll give. But your comment smells like tripe.
The cost of masks is related to the process itself. So a 65nm cutting edge process will be several million dollars, but there's plenty of old lithography equipment out there from the 80s, which is good enough to mask what he wants. I have no idea if it costs $50,000 for a set of masks like that, but it doesn't seem like you're too far off...
Unless you're making 1 million of them, it just doesn't make sense. FPGA's and CPLDs aren't just for prototyping anymore, many small-run products use them.
Hell, depending on the simplicity, are you sure you can't get away with a pic microcontroller? That's what the OTPs are for, after all.
Is learning how to break up your programs into multiple files. In C, learning to do this is tricky... from there it's not far from calling an external function from some library you've linked in.
In javascript or PHP, there is syntax for loading external files. Play with it, do dumb little experiments. Make up a file.js that just runs and alert(), and then call it from another, and see how it pops up a message box just as if you had called alert() directly.
From there, it's only a matter of thinking up functions that (with a little tweaking) could be used over and over, by different programs. You'll start to discover that naming things properly is important, that you can't name something so that it only makes sense in the context of where it is now, that cantThinkOfAGoodNameForThis() might work well when you're hacking out little one-off scripts, but that when you want to reuse it, it really confuses things. You'll discover that "blah" is not a good variable name.
Naming conventions is a science unto itself, so don't feel intimidated if it's not obvious right away. Apparently those who've written the compsci textbooks for 40 years still argue over all of it, so it's nothing you or I will ever "master".
I pretty much suspected as much. Doesn't matter, I'll continue doing it. The attitude that there's no point in bothering is the evil here, and when they fuck up the planet beyond repair, I'll have earned a right to bitch.
I've not seen any arguments for them, either. You want to cover the entire desktop with a big window that does nothing, other than serve as a desktop replacement for the little windows? Do you somehow cover all of the MDI parent window with them? And if so, why bother having the parent there in the first place?
And if you don't have the child windows covering everything, why waste the space painting up a gray background? You could stick gaim in one of the unused portions, and follow a conversation, or whatever in the hell you need there. MDI is a waste of screen real estate.
Photoshop: No substitute is available. Even if we filter out all the whiny bullshit that some of the graphic artist weenies expect (I want all 4000 commercial photoshop plugins too!), we're still left with things that really matter that Gimp can't do. CYMK is the killer feature. And it's apparently nothing that can be hacked in so easily. There are still some usability issues that need to be addressed (though again, some of the weenies will never be happy unless it matched pixel for pixel). There are undoubtedly major issues that a non-photoshop user like myself aren't even aware of. For now we have Gimp, but it is no substitute.
Autocad: No substitute is available. Again, it's a case of all the commercial plugins... if they really make photoshop worthwhile, well, then they basically *ARE* autocad. They make all the difference. This is going to be a tough act to follow, and worse, there are 100 graphic artist wannabees in open source for every engineer wannabe. I'm not familiar with any of those suggested by the article, but I expect they are pretty much to Autocad what Gimp is to photoshop. No real substitutes available.
Dreamweaver: Nvu. It's pretty damn close. It could be Dreamweaver with not an incredible amount of work. But I hope that we don't do that. Mozilla/Firefox aren't just IE, they're better than it is. That's what Nvu should be, or some branch off of it (know it's Mozilla Composer at its core, but is it OSS or proprietary? I never really checked it out). The best part is, that it shares some heritage with Firefox and Thunderbird, and that means in theory, writing plugins for it should be possible. I think that could be really useful in an application like that.
iTunes: Didn't we just see an article about Songbird here recently? The screenshots look pretty slick. Again, based off of mozilla code, I think this could end up being a replacement, even if it isn't yet. Though nothing would ever satisfy the mac weenies, I suspect.
Flash: Inkscape. It's not there yet, animation isn't ready. They're actually trying to design the interface correctly, rather than just imitate all the other animation software we've seen over the years. Also, they do seem to sort of be waiting for software that can view it (for most purposes, this means browsers that support SVG/SMIL). This will probably be every bit as powerful as Flash... there will be those who disagree of course, but who wouldn't have laughed if you'd suggested that mozilla would be the superior of IE in the beginning?
I've not seen a monkeypunching banner in god, years now.
SVG is nice, I'm using it to do stuff I'd only have been able to do in flash or java before. I've got several interactive diagrams for a webapp of mine, even as you change the parameters, so does the image (there are too many combinations to just switch out one image for another). It's pretty neat. I'm planning on doing even cooler things, including a few 3d applets.
And the best part about it, I do not need a windows machine to run the macrodobe software, nor the $x00 for the software itself.
Laugh it up, but I recycle everything that leaves my apartment. I have one grocery-sized bag worth of garbage per week, tops. Sometimes alot less.
I fill 90 gallons worth of garbage can full of recyclables though. Won't do a damn bit of good, but I do it anyway.
The surprising thing is, if I still lived in a house with even a small yard, I'd have a composter again, which would reduce my landfilled waste by yet half. Most people are just too damn lazy.
Fission if it's all we've got, fusion if we can figure it out. I don't see any other viable alternative myself, unless we stumble onto physics even more exotic than those two.
It *is* an energy problem. If energy were 100% free, or so cheap as that it might as well be so, making hydrocarbons is no big deal at all. Any chemical that doesn't rely on rare elements becomes feasible. And as far as I know, no one is claiming that we'll soon reach peak gold...
Hell, if we did solve energy, even elements might be feasible.
Really? They're one of the few countries that murdered their king and kicked the ass of every aristocrat they could find. On second thought, I'm not shocked at all by your disdain for the french...
Yes, and Brazil's total gasoline consumption is comparable to the US. Haha.
US (1999): 132 million cars Brazil: 13.5 million (1998 estimate of 80 cars per 1000 people, times population for that year).
So we have 10 times as many right there. If we dedicate as much cropland as they do, we can hit 2.5%.
As for the quality of fuel that is ethanol, consider that we had it for far longer than we had refined petroleum products, and yet the internal combustion engine was only invented after petroleum started to be exploited. I'm not a chemist, but how easily it combusts, and how much energy that produces, for a given amount counts for a whole lot.
Do the math. Even if every square inch of the ground was converted to ethanol production, we could hit what, 15% of the current demand for gasoline?
Ethanol is actually a pretty shitty fuel, when it comes down to it. We'd probably do better to synthesize gasoline, but that's only possible when we come up with a real power source that doesn't involve letting plants grow out in the sun for 4 months. Hopefully though, by then, we can switch completely to electric or something...
That's even assuming that tokamaks are the correct way to do fusion. I'm not quite convinced yet. There are other possibilities that produce pretty much zero radioactive waste.
It's kind of funny how no one understands energy storage vs. energy production, isn't it?
That's pretty dumb. One day, the earth will be consumed in the fires of our sun going nova... so yeh, dire things will eventually happen. If we run out of oil next year, are you still going to be sarcastic, because they were off by 7 years in the prediction?
We need to find better ways of doing things. But the rich are too comfortable to continue making money with the old ways, and probably aren't clever enough to find new ways.
With current consumption, we'd have to farm an entire continent basically, just to make up for how much oil we use. So, do we tell the people of South America or the people of Africa, that they have to move so we can build our giant ethanol farm?
Nuclear is the only way, and we really need fusion. We should have spent $400 billion on a crash fusion program, not on the Iraq folly.
And when they start selling 747s to consumers, it will still be an aircraft design problem? It's funny, no one would believe that they can just hop in the cockpit and start flying... and they certainly wouldn't call you up during a dangerous descent oh my god, what's wrong with this stupid airplane, I want to land it and I'm pushing buttons all over the place but nothing's happening!.
It's not a design problem. It's a people wanting to do cool things on a magic box without a fucking clue how any of it works.
Slashdot Mirror
My own idea was to build the map of the network into the address space itself. Imagine a unch of routers, each of which is only connected to 4 others, in a regular tiled square grid. Each router could be said to have an XY coordinate. Simple then to just assign subnets such that each router got 10.x.y.0/24.
Mind you, it's not very efficient, but if you assign subnets on non-byte boundaries, there are a few higher-dimensional geometries where the connectedness of any two hosts is no more than about 40 hops (worst case scenario, the vast majority would still be 20 hops or less).
At that point, you'd be able to route a packet based simply on its destination IP. Might still want to send metrics (congestion at this point on the network, etc), but you could tell BGP to go to hell.
The funny thing was, no one could understand how you'd assign subnets, they kept arguing that it needed to be centralized. "What if I connect to XY 4,63 and I decide I want XY 5, 209?". Gee idiot, if you connect to XY 4,63, then you have only a few choices such as 3,63 5,63 4,62 and 4,63. Oh well. I'd even figured out how to get such routing to coexist with the traditional orthodox routing table (stick an entry in it to route all 10.x packets to some virtual device, and have it juggle packets out to the vpn tunnels).
This story reminded me of a project of my own that died not so many months ago. May it rest in peace (Aug 2003 - Aug 2005).
At it's heyday, we had maybe 50 hosts, BGP routing, working DNS with altenate top level domains and reverse dns, websites, email, IRC, a few wikis and so forth. All links were openvpn tunnels, in a decentralized setup where even if one link was severed many still had connectivity. People had no clues who they were communicating with past the first hop, and all first hops connected to someone in a nation legally antagonistic to your own.
But most people don't want that. They just want to be anonymous on the "big internet". They certainly don't want something that isn't some lame java program, and they don't want anything that actually has human readable URLs. They'd rather reinvent every single fucking protocol over and over (freenet's file trading and usenet-like frost, anyone?).
They also apparently don't like a system that means they have to decide who to trust, rather than trusting some software they don't understand to pick and choose who they connect to. Most seemed unwilling or unable to understand the simplicity of the concept, but were more than enthusiastic about DHT algorithms that they couldn't comprehend. More than one seemed downright manic about wifi networks, even though it's generally pretty trivial to use radio equipment to triangulate just who is sending out those packets.
So, to answer the question, you'd have to find enough non-idiots with the spare time and networking skills to have some sort of critical mass. I don't think it will happen myself.
If you're asking me if it's in theory possible to make annoying SVG ads, then yes, I suppose it is. But tell me, how much SVG have you seen on the web, period? Me, there's a guy who links an SVG diagram on his k5 sig, and of course we have the croczilla examples. That's it.
If we're judging each by its deeds, then yeh, for now SVG is nicer.
(I won't even point out that it's open, and others are free to see how I do things and build on it, and with notepad no less...)
Yes. When we invent magical technology that is actually energy dense enough, or if we ever decide to live in some neopaleolithic society, where the kids get to watch an hour of television once a year, then solar is the way to go.
Nuclear works, it's safe, and we've got several hundred years worth of fuel left if we start acting smart about it. That's enough time to get fusion working.
Fusion is the perfect energy source, anyone that says different is a retard. Whether or not it's possible is debatable, I'll give. But your comment smells like tripe.
The cost of masks is related to the process itself. So a 65nm cutting edge process will be several million dollars, but there's plenty of old lithography equipment out there from the 80s, which is good enough to mask what he wants. I have no idea if it costs $50,000 for a set of masks like that, but it doesn't seem like you're too far off...
Unless you're making 1 million of them, it just doesn't make sense. FPGA's and CPLDs aren't just for prototyping anymore, many small-run products use them.
Hell, depending on the simplicity, are you sure you can't get away with a pic microcontroller? That's what the OTPs are for, after all.
Is learning how to break up your programs into multiple files. In C, learning to do this is tricky... from there it's not far from calling an external function from some library you've linked in.
In javascript or PHP, there is syntax for loading external files. Play with it, do dumb little experiments. Make up a file.js that just runs and alert(), and then call it from another, and see how it pops up a message box just as if you had called alert() directly.
From there, it's only a matter of thinking up functions that (with a little tweaking) could be used over and over, by different programs. You'll start to discover that naming things properly is important, that you can't name something so that it only makes sense in the context of where it is now, that cantThinkOfAGoodNameForThis() might work well when you're hacking out little one-off scripts, but that when you want to reuse it, it really confuses things. You'll discover that "blah" is not a good variable name.
Naming conventions is a science unto itself, so don't feel intimidated if it's not obvious right away. Apparently those who've written the compsci textbooks for 40 years still argue over all of it, so it's nothing you or I will ever "master".
After careful consideration, no.
I pretty much suspected as much. Doesn't matter, I'll continue doing it. The attitude that there's no point in bothering is the evil here, and when they fuck up the planet beyond repair, I'll have earned a right to bitch.
I've not seen any arguments for them, either. You want to cover the entire desktop with a big window that does nothing, other than serve as a desktop replacement for the little windows? Do you somehow cover all of the MDI parent window with them? And if so, why bother having the parent there in the first place?
And if you don't have the child windows covering everything, why waste the space painting up a gray background? You could stick gaim in one of the unused portions, and follow a conversation, or whatever in the hell you need there. MDI is a waste of screen real estate.
Photoshop: No substitute is available. Even if we filter out all the whiny bullshit that some of the graphic artist weenies expect (I want all 4000 commercial photoshop plugins too!), we're still left with things that really matter that Gimp can't do. CYMK is the killer feature. And it's apparently nothing that can be hacked in so easily. There are still some usability issues that need to be addressed (though again, some of the weenies will never be happy unless it matched pixel for pixel). There are undoubtedly major issues that a non-photoshop user like myself aren't even aware of. For now we have Gimp, but it is no substitute.
Autocad: No substitute is available. Again, it's a case of all the commercial plugins... if they really make photoshop worthwhile, well, then they basically *ARE* autocad. They make all the difference. This is going to be a tough act to follow, and worse, there are 100 graphic artist wannabees in open source for every engineer wannabe. I'm not familiar with any of those suggested by the article, but I expect they are pretty much to Autocad what Gimp is to photoshop. No real substitutes available.
Dreamweaver: Nvu. It's pretty damn close. It could be Dreamweaver with not an incredible amount of work. But I hope that we don't do that. Mozilla/Firefox aren't just IE, they're better than it is. That's what Nvu should be, or some branch off of it (know it's Mozilla Composer at its core, but is it OSS or proprietary? I never really checked it out). The best part is, that it shares some heritage with Firefox and Thunderbird, and that means in theory, writing plugins for it should be possible. I think that could be really useful in an application like that.
iTunes: Didn't we just see an article about Songbird here recently? The screenshots look pretty slick. Again, based off of mozilla code, I think this could end up being a replacement, even if it isn't yet. Though nothing would ever satisfy the mac weenies, I suspect.
Flash: Inkscape. It's not there yet, animation isn't ready. They're actually trying to design the interface correctly, rather than just imitate all the other animation software we've seen over the years. Also, they do seem to sort of be waiting for software that can view it (for most purposes, this means browsers that support SVG/SMIL). This will probably be every bit as powerful as Flash... there will be those who disagree of course, but who wouldn't have laughed if you'd suggested that mozilla would be the superior of IE in the beginning?
Flashblock.
I've not seen a monkeypunching banner in god, years now.
SVG is nice, I'm using it to do stuff I'd only have been able to do in flash or java before. I've got several interactive diagrams for a webapp of mine, even as you change the parameters, so does the image (there are too many combinations to just switch out one image for another). It's pretty neat. I'm planning on doing even cooler things, including a few 3d applets.
And the best part about it, I do not need a windows machine to run the macrodobe software, nor the $x00 for the software itself.
Laugh it up, but I recycle everything that leaves my apartment. I have one grocery-sized bag worth of garbage per week, tops. Sometimes alot less.
I fill 90 gallons worth of garbage can full of recyclables though. Won't do a damn bit of good, but I do it anyway.
The surprising thing is, if I still lived in a house with even a small yard, I'd have a composter again, which would reduce my landfilled waste by yet half. Most people are just too damn lazy.
Atoms.
Fission if it's all we've got, fusion if we can figure it out. I don't see any other viable alternative myself, unless we stumble onto physics even more exotic than those two.
It *is* an energy problem. If energy were 100% free, or so cheap as that it might as well be so, making hydrocarbons is no big deal at all. Any chemical that doesn't rely on rare elements becomes feasible. And as far as I know, no one is claiming that we'll soon reach peak gold...
Hell, if we did solve energy, even elements might be feasible.
Really? They're one of the few countries that murdered their king and kicked the ass of every aristocrat they could find. On second thought, I'm not shocked at all by your disdain for the french...
Really? I spend all my disposable income on buying tanks of CO2 and venting it into the air...
Yes, and Brazil's total gasoline consumption is comparable to the US. Haha.
US (1999): 132 million cars
Brazil: 13.5 million (1998 estimate of 80 cars per 1000 people, times population for that year).
So we have 10 times as many right there. If we dedicate as much cropland as they do, we can hit 2.5%.
As for the quality of fuel that is ethanol, consider that we had it for far longer than we had refined petroleum products, and yet the internal combustion engine was only invented after petroleum started to be exploited. I'm not a chemist, but how easily it combusts, and how much energy that produces, for a given amount counts for a whole lot.
Do the math. Even if every square inch of the ground was converted to ethanol production, we could hit what, 15% of the current demand for gasoline?
Ethanol is actually a pretty shitty fuel, when it comes down to it. We'd probably do better to synthesize gasoline, but that's only possible when we come up with a real power source that doesn't involve letting plants grow out in the sun for 4 months. Hopefully though, by then, we can switch completely to electric or something...
Thanks for the link, hope you're not implying I don't understand the issues.
That's even assuming that tokamaks are the correct way to do fusion. I'm not quite convinced yet. There are other possibilities that produce pretty much zero radioactive waste.
It's kind of funny how no one understands energy storage vs. energy production, isn't it?
That's pretty dumb. One day, the earth will be consumed in the fires of our sun going nova... so yeh, dire things will eventually happen. If we run out of oil next year, are you still going to be sarcastic, because they were off by 7 years in the prediction?
We need to find better ways of doing things. But the rich are too comfortable to continue making money with the old ways, and probably aren't clever enough to find new ways.
It's not that simple.
With current consumption, we'd have to farm an entire continent basically, just to make up for how much oil we use. So, do we tell the people of South America or the people of Africa, that they have to move so we can build our giant ethanol farm?
Nuclear is the only way, and we really need fusion. We should have spent $400 billion on a crash fusion program, not on the Iraq folly.
And when they start selling 747s to consumers, it will still be an aircraft design problem? It's funny, no one would believe that they can just hop in the cockpit and start flying... and they certainly wouldn't call you up during a dangerous descent oh my god, what's wrong with this stupid airplane, I want to land it and I'm pushing buttons all over the place but nothing's happening!.
It's not a design problem. It's a people wanting to do cool things on a magic box without a fucking clue how any of it works.
So stealing my laptop will allow anyone to go to websites and impersonate me?