Actually, my point is that MS is more focused on B2B (business-to-business) relationships than business-to-person. 'Consumer' was the word I chose to represent individual customers as opposed to corporate customers. MS has totally lost sight of what it's like to design an efficient, intuitive interface for a task at hand. Instead, they seem more focused on packing in as many features as their sales reps get asked for by their "important" corporate clients, and don't realize that their individual clients don't want or need those same features, and prefer simplicity in design and workflow over the 'every feature under the sun' approach.
Wow, that's one of the more sensible things I've ever read by Dvorak. I can't say I disagree with any of his arguments, and MS truly is unexciting and completely lacking in vision and direction (except for the Xbox division). Then again, the Xbox division is pretty much the only one that makes products that target consumers instead of targetting corporate IT as their market. The major problem with MS software is that they've completely lost touch with consumers, and haven't the slightest idea of how to design software that consumers want to use. They add requested feature after feature, without any oversight on workflow or comprehensive interface design. In the end, we just get incredibly bloated software with functionality randomly scattered throughout the interface in inconsistent ways.
It seems they really need to refocus on individual consumer needs instead of what businesses need, and not be afraid to refactor their software with top-to-bottom interface redesigns when functionality and/or workflow changes significantly.
The problem I see is that you're actually paying for access to a network, and part of that network (the part run by the ISP) has the potential to deliver a lot more data, and the rest of it (outside connections) are limited by how much the ISP can afford/get away with. While it's true that ISPs could probably let the outside bandwidth languish as they build internal infrastructure, but I think we'd see an interesting new development.
As the larger ISPs grow their "first tier" networks, and license access to other ISPs, the bandwidth capacity of these internal ISP networks needs to be very high, and high-speed connections must be made to any sub-licensor ISP, and you effectively connect the two tiers from the different ISPs with a high-speed backbone. This will encourage huge investments in network infrastructure, and as more and more ISPs connect their first tier networks with better and better connections, the entire internet becomes a whole lot faster. It's kinda like the notion of a rising tide lifting all boats. Even smaller ISPs can leverage their geographic coverage to gain a market for local content distribution and ad placement. This just has the effect of rewarding the ISPs who invest in network infrastructure, and allow for localized high-bandwidth applications.
You can decide, by choosing the ISP you want to purchase service from. Letting end users modify their connection settings is the most ludicrous notion I've seen mentioned here in a while (from a network management perspective). All I'm saying is that connections that have to travel through different networks should be treated differently from connections over a single network; and it would be nice for ISPs to offer a service for content providers to guarantee access to hosted content at a minimum bandwidth to the ISP's subscribers.
Thanks for the informed reply. Yeah, I know it makes sense to do this, and ISPs do try to this as much as possible to reduce bandwidth costs, but I was just trying to make the argument that a tiered system has some obvious benefits, and we shouldn't try to make them illegal in the name of liberty. I think pushing the concept of tiers to network applications will really move us forward to take advantage of the way the internet is constructed, and maximize the bandwidth we could have at our disposal. The main thing ISPs could do to push development in this area would be to define the IP ranges of all the people in your tier, and allow faster connections to them; either that or make an extension to IPv4 that would track tiers as well as nodes, though that would be more difficult to get accepted.
If ISPs left the outside bandwidth unchanged, but added the ability for high-bandwidth first tier connections, it would really push a lot of power into the hands of the ISPs in several key ways. First, P2P would create a large demand for this type of service, pushing general adoption. Second, ISPs could upsell the first tier access as a type of Pro service (like digital cable TV being sold to cable internet subscribers). Third, as the first tier becomes popular and people become dependent on the level of bandwidth offered for local connections, ISPs could license access to the tier to other ISPs, as well as offering hosting directly to content providers.
While this means that ISPs would control a significant amount of the content on the network, it would really encourage the development of the ISP's network and storage infrastructure, and finally bring us some real high-bandwidth services in the only way I can really see this happening in a reasonable timescale. Maybe the bandwidth I mentioned (0.5-1Gbps) is a bit high, and we might be stuck with a few dozen Mbps for a while, that's still significantly better, and it's those asymmetrical upload speeds that totally kill most P2P apps, so that could be addressed with this new setup.
Personally, while crippling connections to outside content is clearly unacceptable, I think a tiered internet would be necessary to really bring us to the next level of high-bandwidth services. If the ISP can host content on their servers, and offer very high-speed (maybe around 1Gbps peak, 0.5Gbps min) access to it for reasonable rates, that would offer a great alternative to current internet distribution options, even if it's limited to targetting the customers of specific ISPs. If a tiered internet means I can buy some hosting from Comcast, and get guaranteed high-bandwidth serving to all Comcast customers, it would really open the door to services that are just not possible with current net topology, serving to other ISPs that might not have access to the bandwidth required for the service.
An obvious example is on-demand delivery of high-definition video streams. Residential broadband connections are just not fast enough to enable those kinds of services. If I could buy hosting from the major ISPs, I would at least be able to target their customers without worrying about dropped packets and poor connections. It would require some minor changes to internet file access, only granting access to the stream if it's hosted on the requestor's ISP, but it's not really difficult to implement with current protocols.
This is really going to be like TV network affiliates for the new century. Major ISPs that cover large areas of the country could even offer to split up the access based on geography, allowing for services like geographically-targetted advertising and such. I don't really see a practical alternative to a tiered internet if we want to have the kinds of high-bandwidth services that need to happen.
While I realize that will create a market where bigger ISPs can offer more value than others, giving them a competitive advantage. This could be mitigated with some checks and balances, however. ISPs could be obliged to sublicense any hosted content to other ISPs for reasonable rates upon request, or maybe content providers could be forced to adopt a uniform licensing agreement and offer it to any ISP. Content affiliates could promote content and deal with mixing in advertising from local businesses to the customers they're targetting.
To just arbitrarily push for legislation that would make things like this illegal just seems like it's going against progress, and would do nothing but keep me from getting the proper high-bandwidth distribution options I want to see on the market. I can't think of any better way to encourage development of high-speed services, including high-speed MANs (Metropolitan Area Networks)...
The other tier would be the internet we know today, and that should continue to function in an unfiltered manner, as it does today. It's just that it makes sense, given the internet's infrastructure, to give ISPs the ability to serve content at guaranteed speeds, to meet a demand that couldn't be filled without limiting access on an ISP basis.
Also, realize that it could work both ways. If you can establish p2p connections to people on your tier, possibly with some auto-discovery, it could allow some very interesting possibilities for p2p apps. Since they would be first tier connections, it might be possible to access your computer hard drive remotely at close to full speed, or do some serious videoconferencing with local partners. Tiers would really adjust the internet's operation to reflect how it's actually put together, while still allowing outside connections as they do now... I bet ISPs would love to be able to shift all that p2p traffic onto their own network instead of having to pay for outside bandwidth...
If anything, Windows runtime should be a service on top of OSX (and maybe even BSD or Linux, if they're stable enough - API wise), not vice-versa. Structurally, Windows is a mess, but it has some good, mature APIs , and excellent developer tools (plus a huge library of software, obviously). Microsoft should build on their strengths and license Windows client, server, and developer runtimes for alternative operating systems, allowing them a slice of the pie no matter where the OS market share goes.
It's all about STM laptop bags, in particular, the Alley model. You can buy them for reasonable prices in the US from Radtech. They cost $45, and fit my MacBook Pro 15.5" lappy perfectly. These are definitely the most stylish bags I was able to find after an hour's worth of googling... You don't have to look like a tool just to carry your favorite tool with you...
Because Apple and Google are too tied to ensuring all the back catalog of the major studios are on board, which will never happen; at least not until the major labels start losing their shirts to new bands getting money without them. It is necessary to move forward with or without the major studios on board...
Nah. the cavemen are the chief executives of the major record labels, and the glacier is electronic distribution. Apple and probably Google will be supplanted by companies selling unencumbered digital downloads in standard formats for competitive prices, even if the music library from the major labels is unavailable. It'll take 10-15 years, but it's bound to happen, and major labels will be left wondering where they went wrong...
Actually, the point of moving over to Intel was access to more powerful chips that use less power. IBM chips are actually pretty cheap, relatively, and it's not surprising they had to bump up the price of the mini with Intel Core and Core Duo chips, as those parts are pretty expensive...
S. Jobs would never allow sub-par 3rd party software in a default install of OS X. In fact, Apple as a whole has never allowed sub-par 3rd party software on the Mac. Anytime there is bundled 3rd party software, it is always best-of-class. Jobs, and mostly the company's ethic as a whole, just has too much taste and sense to ruin the user experience with crappy 3rd party software. That would be like a high-end Mercedes shipping with Firestone tires...
Yeah, great idea... let's make everyone's tax forms available publicly... can't see any potential for privacy abuse there, nooooooooo... I can't believe a government could be so stupid!
There's already a solution to that provided by the law; it's called a civil suit. You find someone who's violating copyright by redistributing copyrighted content without a license, and sue them for lots of money. It happens all the time. Copyright violations are a civil matter between the violator and the copyright holder. You're not going to go out and dull every knife in the marketplace to prevent stabbings; same thing here, you're not going to lock out everyone from using the content they purchase just to prevent a bit of piracy. Part of living in a free society is that you have to give up total control, and use market forces such as pricing, availability, marketing and lawsuits to guide your business dealings. We have a word for total control, and it's called totalitarianism. Whether it's coming from a government or a corporation, people are perfectly justified in revolting against such behavior...
Check out the TV series Firefly, now available on DVD. It's by the same guy who did Serenity (many of the same actors as well), but much, much more interesting and fun to watch (for the most part). I hadn't even heard of it when it aired, and only found out about it when Serenity came out on DVD and people were talking about it in the blogosphere... Definitely worth checking out...
Forget Blockbuster, Apple already has a network of retail stores covering the majority of the US population (as well as significant parts of other markets), already with high-speed net access, computers, etc, all they would need is an XServe RAID full of movies, and they could be loading up iPods (or DVD-R) all over the place. You could even queue up some content on the Apple website, and have it burned to DVD-R or ready to load on the iPod when you get to the store. There is definitely a market for this type of service.
Yeah, right... trust the Chinese government to uphold our privacy rights. Anyone who runs Red Flag Linux voluntarily should have their head examined. I think Gentoo might be a safe bet...
Agreed; this has been the bane of Mac laptops and other models that use slim DVD drives for a while. If only Apple would switch to Pioneer low profile drives... As a part of the MacTheRipper project (DVD ripper for OS X), these Matshita drives have been a huge problem, with the only solution being to attempt a drive replacement with a slim Pioneer drive, or to get an external FireWire enclosure and drive for DVD ripping from any region.
Repairing the permissions on your OSX drive using Disk Utility never hurts. This should be done about once a month, and before and after all system updates. If problems persist, try creating a new user and see if that solves it. If so, move your settings over to the new account and start using that one. For major OS upgrades (10.3.x -> 10.4.x, etc.), always choose the Archive and Install update option, and never a simple update. You might just try an Archive and Install installation of your current OS version again, and see if that solves the problem. If not, you can do a clean install and manually update the critical settings you want from your old install.
First, I want to apologize for my messed up formatting in that last post... Having said that, is the lag necessarily true for local connections? I thought Mesh was slow because it relies on sharing regular ISP TCP/IP connections, but maybe if it prioritized local direct connections (routes without going through ISPs), maybe those would get good performance. I guess the bandwidth and the router's data cache would have to be very high for this to be scalable in any way just for the routing overhead, and routers would have to be pretty smart in a coordinated way to figure out optimal routes on the fly, but I think there might be hope for low latency decentralized networks, especially if most of the traffic is within a couple mile radius.
I guess it would be best to set up super-nodes with long distance range and high bandwidth, and then have most end users on lower range and lower bandwidth nodes. The only problem is providing incentive for super-nodes to happen in a decentralized fashion. Maybe the routers could be capable of being either a super or a regular node, and would configure themselves as the network requires. That seems like it would go a long way towards solving any latency problems, and keep the decentralized unregulated network topology. Now we just have to wait for the technology for long range high bandwidth hops to happen in a manner that doesn't violate FCC regulations.
The obvious final outcome here is that people are going to start forming grassroots wireless networks on a metropolitan area level, and interconnecting these networks through encrypted tunnels through standard ISPs. It'll start by people getting access points to access metropolitan area services at high bandwidth, something like Mesh would be used to provide the network infrastructure, and eventually, this will end up as an ad-hoc wireless internet. Obviously, the software is going to have to evolve to be very scalable and have acceptable performance, but what we will be left with is a truly free network by construction (decentralization), even if we can't include everyone in the middle of nowhere.
People will first use it for gaming and sharing media, but businesses will jump at the opportunity to get access to a targeted metropolitan network, and TV affiliates might find it an attractive method of reaching local viewers. Even if the connections outside the MAN are poor, most cities and towns would surely find good uses for high speed local networks. As these networks grow, they will become more and more interconnected, and hopefully replace the centralized internet we have today for all but the longest hops.
The killer (commercial) application for this is obviously going to be video and local services. To get TV affiliates interested, though, there would need to be a proven audience, and p2p filesharing and low-lag gaming is going to be the driving force for this adoption. Once the infrastructure is in place and companies can get free high-speed access to customers in their area, a huge new market for broadband services will open up, along with the advertising possibilities targeted at local customers.
The rest is the hardware; current wireless routers have insufficient range and bandwidth to make starting such a network on a large scale viable. We would need a cheap, high bandwidth, and long(er) range solution to get this party really started. Maybe in 5-10 years that kind of technology will be available, and then this ad-hoc network will almost create itself. I really find it hard to believe that anyone is really capable of preventing this from happening at this point. We have the will, we have the communication capability, we have the ingenious coders and open source community, I'd say it's pretty much a done deal. All we need is the obvious hardware solution to make it all a reality.
Slashdot Mirror
Actually, my point is that MS is more focused on B2B (business-to-business) relationships than business-to-person. 'Consumer' was the word I chose to represent individual customers as opposed to corporate customers. MS has totally lost sight of what it's like to design an efficient, intuitive interface for a task at hand. Instead, they seem more focused on packing in as many features as their sales reps get asked for by their "important" corporate clients, and don't realize that their individual clients don't want or need those same features, and prefer simplicity in design and workflow over the 'every feature under the sun' approach.
Wow, that's one of the more sensible things I've ever read by Dvorak. I can't say I disagree with any of his arguments, and MS truly is unexciting and completely lacking in vision and direction (except for the Xbox division). Then again, the Xbox division is pretty much the only one that makes products that target consumers instead of targetting corporate IT as their market. The major problem with MS software is that they've completely lost touch with consumers, and haven't the slightest idea of how to design software that consumers want to use. They add requested feature after feature, without any oversight on workflow or comprehensive interface design. In the end, we just get incredibly bloated software with functionality randomly scattered throughout the interface in inconsistent ways.
It seems they really need to refocus on individual consumer needs instead of what businesses need, and not be afraid to refactor their software with top-to-bottom interface redesigns when functionality and/or workflow changes significantly.
The problem I see is that you're actually paying for access to a network, and part of that network (the part run by the ISP) has the potential to deliver a lot more data, and the rest of it (outside connections) are limited by how much the ISP can afford/get away with. While it's true that ISPs could probably let the outside bandwidth languish as they build internal infrastructure, but I think we'd see an interesting new development.
As the larger ISPs grow their "first tier" networks, and license access to other ISPs, the bandwidth capacity of these internal ISP networks needs to be very high, and high-speed connections must be made to any sub-licensor ISP, and you effectively connect the two tiers from the different ISPs with a high-speed backbone. This will encourage huge investments in network infrastructure, and as more and more ISPs connect their first tier networks with better and better connections, the entire internet becomes a whole lot faster. It's kinda like the notion of a rising tide lifting all boats. Even smaller ISPs can leverage their geographic coverage to gain a market for local content distribution and ad placement. This just has the effect of rewarding the ISPs who invest in network infrastructure, and allow for localized high-bandwidth applications.
You can decide, by choosing the ISP you want to purchase service from. Letting end users modify their connection settings is the most ludicrous notion I've seen mentioned here in a while (from a network management perspective). All I'm saying is that connections that have to travel through different networks should be treated differently from connections over a single network; and it would be nice for ISPs to offer a service for content providers to guarantee access to hosted content at a minimum bandwidth to the ISP's subscribers.
Thanks for the informed reply. Yeah, I know it makes sense to do this, and ISPs do try to this as much as possible to reduce bandwidth costs, but I was just trying to make the argument that a tiered system has some obvious benefits, and we shouldn't try to make them illegal in the name of liberty. I think pushing the concept of tiers to network applications will really move us forward to take advantage of the way the internet is constructed, and maximize the bandwidth we could have at our disposal. The main thing ISPs could do to push development in this area would be to define the IP ranges of all the people in your tier, and allow faster connections to them; either that or make an extension to IPv4 that would track tiers as well as nodes, though that would be more difficult to get accepted.
If ISPs left the outside bandwidth unchanged, but added the ability for high-bandwidth first tier connections, it would really push a lot of power into the hands of the ISPs in several key ways. First, P2P would create a large demand for this type of service, pushing general adoption. Second, ISPs could upsell the first tier access as a type of Pro service (like digital cable TV being sold to cable internet subscribers). Third, as the first tier becomes popular and people become dependent on the level of bandwidth offered for local connections, ISPs could license access to the tier to other ISPs, as well as offering hosting directly to content providers.
While this means that ISPs would control a significant amount of the content on the network, it would really encourage the development of the ISP's network and storage infrastructure, and finally bring us some real high-bandwidth services in the only way I can really see this happening in a reasonable timescale. Maybe the bandwidth I mentioned (0.5-1Gbps) is a bit high, and we might be stuck with a few dozen Mbps for a while, that's still significantly better, and it's those asymmetrical upload speeds that totally kill most P2P apps, so that could be addressed with this new setup.
Personally, while crippling connections to outside content is clearly unacceptable, I think a tiered internet would be necessary to really bring us to the next level of high-bandwidth services. If the ISP can host content on their servers, and offer very high-speed (maybe around 1Gbps peak, 0.5Gbps min) access to it for reasonable rates, that would offer a great alternative to current internet distribution options, even if it's limited to targetting the customers of specific ISPs. If a tiered internet means I can buy some hosting from Comcast, and get guaranteed high-bandwidth serving to all Comcast customers, it would really open the door to services that are just not possible with current net topology, serving to other ISPs that might not have access to the bandwidth required for the service.
An obvious example is on-demand delivery of high-definition video streams. Residential broadband connections are just not fast enough to enable those kinds of services. If I could buy hosting from the major ISPs, I would at least be able to target their customers without worrying about dropped packets and poor connections. It would require some minor changes to internet file access, only granting access to the stream if it's hosted on the requestor's ISP, but it's not really difficult to implement with current protocols.
This is really going to be like TV network affiliates for the new century. Major ISPs that cover large areas of the country could even offer to split up the access based on geography, allowing for services like geographically-targetted advertising and such. I don't really see a practical alternative to a tiered internet if we want to have the kinds of high-bandwidth services that need to happen.
While I realize that will create a market where bigger ISPs can offer more value than others, giving them a competitive advantage. This could be mitigated with some checks and balances, however. ISPs could be obliged to sublicense any hosted content to other ISPs for reasonable rates upon request, or maybe content providers could be forced to adopt a uniform licensing agreement and offer it to any ISP. Content affiliates could promote content and deal with mixing in advertising from local businesses to the customers they're targetting.
To just arbitrarily push for legislation that would make things like this illegal just seems like it's going against progress, and would do nothing but keep me from getting the proper high-bandwidth distribution options I want to see on the market. I can't think of any better way to encourage development of high-speed services, including high-speed MANs (Metropolitan Area Networks)...
The other tier would be the internet we know today, and that should continue to function in an unfiltered manner, as it does today. It's just that it makes sense, given the internet's infrastructure, to give ISPs the ability to serve content at guaranteed speeds, to meet a demand that couldn't be filled without limiting access on an ISP basis.
Also, realize that it could work both ways. If you can establish p2p connections to people on your tier, possibly with some auto-discovery, it could allow some very interesting possibilities for p2p apps. Since they would be first tier connections, it might be possible to access your computer hard drive remotely at close to full speed, or do some serious videoconferencing with local partners. Tiers would really adjust the internet's operation to reflect how it's actually put together, while still allowing outside connections as they do now... I bet ISPs would love to be able to shift all that p2p traffic onto their own network instead of having to pay for outside bandwidth...
Meant to reply to the parent of the post I replied to...
If anything, Windows runtime should be a service on top of OSX (and maybe even BSD or Linux, if they're stable enough - API wise), not vice-versa. Structurally, Windows is a mess, but it has some good, mature APIs , and excellent developer tools (plus a huge library of software, obviously). Microsoft should build on their strengths and license Windows client, server, and developer runtimes for alternative operating systems, allowing them a slice of the pie no matter where the OS market share goes.
It's all about STM laptop bags, in particular, the Alley model. You can buy them for reasonable prices in the US from Radtech. They cost $45, and fit my MacBook Pro 15.5" lappy perfectly. These are definitely the most stylish bags I was able to find after an hour's worth of googling... You don't have to look like a tool just to carry your favorite tool with you...
Because Apple and Google are too tied to ensuring all the back catalog of the major studios are on board, which will never happen; at least not until the major labels start losing their shirts to new bands getting money without them. It is necessary to move forward with or without the major studios on board...
No one has mod points anymore 'cuz they're all on digg...
Nah. the cavemen are the chief executives of the major record labels, and the glacier is electronic distribution. Apple and probably Google will be supplanted by companies selling unencumbered digital downloads in standard formats for competitive prices, even if the music library from the major labels is unavailable. It'll take 10-15 years, but it's bound to happen, and major labels will be left wondering where they went wrong...
Actually, the point of moving over to Intel was access to more powerful chips that use less power. IBM chips are actually pretty cheap, relatively, and it's not surprising they had to bump up the price of the mini with Intel Core and Core Duo chips, as those parts are pretty expensive...
S. Jobs would never allow sub-par 3rd party software in a default install of OS X. In fact, Apple as a whole has never allowed sub-par 3rd party software on the Mac. Anytime there is bundled 3rd party software, it is always best-of-class. Jobs, and mostly the company's ethic as a whole, just has too much taste and sense to ruin the user experience with crappy 3rd party software. That would be like a high-end Mercedes shipping with Firestone tires...
George W. Bush, is that you? Hard to tell... could be almost anyone in his cabinet or extensive list of consultants and advisors, really...
Yeah, great idea... let's make everyone's tax forms available publicly... can't see any potential for privacy abuse there, nooooooooo... I can't believe a government could be so stupid!
There's already a solution to that provided by the law; it's called a civil suit. You find someone who's violating copyright by redistributing copyrighted content without a license, and sue them for lots of money. It happens all the time. Copyright violations are a civil matter between the violator and the copyright holder. You're not going to go out and dull every knife in the marketplace to prevent stabbings; same thing here, you're not going to lock out everyone from using the content they purchase just to prevent a bit of piracy. Part of living in a free society is that you have to give up total control, and use market forces such as pricing, availability, marketing and lawsuits to guide your business dealings. We have a word for total control, and it's called totalitarianism. Whether it's coming from a government or a corporation, people are perfectly justified in revolting against such behavior...
Check out the TV series Firefly, now available on DVD. It's by the same guy who did Serenity (many of the same actors as well), but much, much more interesting and fun to watch (for the most part). I hadn't even heard of it when it aired, and only found out about it when Serenity came out on DVD and people were talking about it in the blogosphere... Definitely worth checking out...
Forget Blockbuster, Apple already has a network of retail stores covering the majority of the US population (as well as significant parts of other markets), already with high-speed net access, computers, etc, all they would need is an XServe RAID full of movies, and they could be loading up iPods (or DVD-R) all over the place. You could even queue up some content on the Apple website, and have it burned to DVD-R or ready to load on the iPod when you get to the store. There is definitely a market for this type of service.
Yeah, right... trust the Chinese government to uphold our privacy rights. Anyone who runs Red Flag Linux voluntarily should have their head examined. I think Gentoo might be a safe bet...
Or you could use WINE and just run the Windows software you need right in OS X... That, and it looks like Garmin has announced Mac support...
Agreed; this has been the bane of Mac laptops and other models that use slim DVD drives for a while. If only Apple would switch to Pioneer low profile drives... As a part of the MacTheRipper project (DVD ripper for OS X), these Matshita drives have been a huge problem, with the only solution being to attempt a drive replacement with a slim Pioneer drive, or to get an external FireWire enclosure and drive for DVD ripping from any region.
Repairing the permissions on your OSX drive using Disk Utility never hurts. This should be done about once a month, and before and after all system updates. If problems persist, try creating a new user and see if that solves it. If so, move your settings over to the new account and start using that one. For major OS upgrades (10.3.x -> 10.4.x, etc.), always choose the Archive and Install update option, and never a simple update. You might just try an Archive and Install installation of your current OS version again, and see if that solves the problem. If not, you can do a clean install and manually update the critical settings you want from your old install.
First, I want to apologize for my messed up formatting in that last post... Having said that, is the lag necessarily true for local connections? I thought Mesh was slow because it relies on sharing regular ISP TCP/IP connections, but maybe if it prioritized local direct connections (routes without going through ISPs), maybe those would get good performance. I guess the bandwidth and the router's data cache would have to be very high for this to be scalable in any way just for the routing overhead, and routers would have to be pretty smart in a coordinated way to figure out optimal routes on the fly, but I think there might be hope for low latency decentralized networks, especially if most of the traffic is within a couple mile radius.
I guess it would be best to set up super-nodes with long distance range and high bandwidth, and then have most end users on lower range and lower bandwidth nodes. The only problem is providing incentive for super-nodes to happen in a decentralized fashion. Maybe the routers could be capable of being either a super or a regular node, and would configure themselves as the network requires. That seems like it would go a long way towards solving any latency problems, and keep the decentralized unregulated network topology. Now we just have to wait for the technology for long range high bandwidth hops to happen in a manner that doesn't violate FCC regulations.
The obvious final outcome here is that people are going to start forming grassroots wireless networks on a metropolitan area level, and interconnecting these networks through encrypted tunnels through standard ISPs. It'll start by people getting access points to access metropolitan area services at high bandwidth, something like Mesh would be used to provide the network infrastructure, and eventually, this will end up as an ad-hoc wireless internet. Obviously, the software is going to have to evolve to be very scalable and have acceptable performance, but what we will be left with is a truly free network by construction (decentralization), even if we can't include everyone in the middle of nowhere. People will first use it for gaming and sharing media, but businesses will jump at the opportunity to get access to a targeted metropolitan network, and TV affiliates might find it an attractive method of reaching local viewers. Even if the connections outside the MAN are poor, most cities and towns would surely find good uses for high speed local networks. As these networks grow, they will become more and more interconnected, and hopefully replace the centralized internet we have today for all but the longest hops. The killer (commercial) application for this is obviously going to be video and local services. To get TV affiliates interested, though, there would need to be a proven audience, and p2p filesharing and low-lag gaming is going to be the driving force for this adoption. Once the infrastructure is in place and companies can get free high-speed access to customers in their area, a huge new market for broadband services will open up, along with the advertising possibilities targeted at local customers. The rest is the hardware; current wireless routers have insufficient range and bandwidth to make starting such a network on a large scale viable. We would need a cheap, high bandwidth, and long(er) range solution to get this party really started. Maybe in 5-10 years that kind of technology will be available, and then this ad-hoc network will almost create itself. I really find it hard to believe that anyone is really capable of preventing this from happening at this point. We have the will, we have the communication capability, we have the ingenious coders and open source community, I'd say it's pretty much a done deal. All we need is the obvious hardware solution to make it all a reality.