That $840 price they mentioned was the price of PCs there run around $840... not the Playstation. It said it will cost *less* than a PC, so you've got a max price of $840.
I've seen that general motors quote a few times, and quotes like it dozens of times. "Why would I give away something for free that costs millions to develop?"
Well in a lot of those cases, that's what they're doing. General Motors in a certain sense does give away their cars. When I spend $20k on a car, I'm paying a few hundred dollars for the labor of the people assembling the car, a big chunk of money for the cost of parts that are made by the company (frame, body, etc), and a significant chunk for parts made by outside part manufacturers.
Now I've bought my car, its completely open to me. State inspections ignored, I can take all the body panels off my car and tape cardboard spraypainted with fleckstone paint onto it. I can take the engine out and put another in, or rework how the engine works. I can do anything I want with it.
So what's the money paying for? Its for the parts, the time of the people assembling it and part of the profit goes to the company that developed it. If I really wanted to, I could buy the repair manual for the car, go out and buy one part after another and build it from the ground up. Happens all the time with hotrods and restoration cars. The design of the car is completely open. If I want a new cam for the car, I can buy one from the company that made it, but I can also take calipers and such, measure it, and make one myself if I've got the facilities for it or pay someone else to.
So compare that to open source. Sure I can go download all the parts of a Linux system and put it together myself, piece by piece. The manufacturing and materials cost of a piece of software is very low, so they're a lot cheaper than car parts, but those are essentially why they're cheaper. The first Linux system I used was built just that way. But now I tend to use RedHat linux. They build my "car" for me, and I can change it once I get it to how I'd like it to be.
And they sink a lot of money into developing new parts for the "car". I can buy a later model of the "car" if I want or I can aquire the parts and put them on myself.
My point is that the open source concept isn't *that* alien to business -- lots of them have been doing it all along, when you remove costs of materials and manufacturing from the equation. I don't need to go to Ford to build my own Mustang if I've got a good machine shop, skills, and time.
Come on people, lets make some effort to understand the issue before spouting off on it. Admittedly I didn't read the FCC release before posting my last comment on this, but what I said in it was essentially correct.
This has *nothing* to do with making ISP calls long distance, it just a simple ruling allowing telcos to choose if they desire to set up a reciprocal billing arrangement with ISPs (presumably larger ISPs) within their service area that they are providing lines to.
This doesn't mean you get billed per minute, it just means they have flexibility in how they choose to bill the line usage to the ISP.
They can, for example, class the incoming lines as inter-carrier lines, allowing them to be billed for usage rather than end-point charges.
And its not a requirement its just a clarification of their position on it. If I run fiber between two cities myself (ie, I own the fiber and I'm leasing the pole or line space underground) then I typically need those sort of reciprocal agreements with the carriers who are going to gateway traffic to my line in both terminating ends.
I don't think this is nearly as big of a deal as people have been making it out to be for the last six months that the rumor about this was going around. Some telcos may choose to gouge ISPs with these prices, but most probably won't charge anything. The release also made it seem like the actual situation would be based on the way existing reciprocal agreements within the state are handled.
I still wish I didn't need to use a telco though:)
I could be wrong but I think everyone may be misinterpreting this one. My understanding was this moves calls through your local telco to your ISP into the same classification as a call through your telco to a long distance provider, it does NOT make the calls long distance, it just lets telcos choose to bill ISP's for the traffic they cause, which they won't necessarily do. Any telco that has its own internet service would be shooting themselves in the foot since they'd have to bill the usage back to themselves at the same rate.
I can spend $25 a month and get unlimited calls on Sprint on the weekends, so the rates telcos are charging long distance companies for access through their CO's can't be that high.
I also think that billing in that manner to an ISP would mean that they can't charge for the line itself into the ISP but I could be wrong about that. If an ISP is paying $40 a month per line or circuit on a T1, then usage charges might not be much more per line except in unusual cases.
If that's the case, and this is really what the classification change means then I'd guess only very high usage people even have a chance of being affected. And someone who really has a valid reason to tie up their internet line 16 hours a day ought to have enough reason to pay $300 or $400 a month for a frame relay connection if cable, xDSL or other technology isn't available.
I noticed that company happens to sell a lot of VME products...
Does anyone know if any of the ports of Linux support VME-based systems? If any of them do, does the SMP support work when multiple VME boards with processors are on the VME bus or do the boards themselves have to be SMP specifically? (ie, can I pop two dual-processor VME boards onto the bus and have Linux use all four processors?) I've worked with 88110 and 68040 VME boards but don't really know how they actually worked from that standpoint...
I've seen true-blue SysVR4 running on VME systems that way, it'd be cool if Linux did too.
Hey I bet this would have the horsepower to play Gran Tourismo and simulate car damage. (A key feature in any good two player driving game where my friends are concerned...)
I remember reading that AC article back in 1996 about this. Seemed like a really cool technology, but if you notice the commercials doing it now and compare it to the ones on the site, the ones using that system have the worst quality in terms of image stability while the "movement" is happening, color continuity, etc, etc, etc...
That credit card commercial in particular I thought looked like crap. Any of the ones involving moves through a larger amount of space, because they seem to spread the units out too far and double or triple up on frames like older cartoons did. It makes for very choppy movement.
The commercials where its done very clearly are typically shot from fewer vantage points and digitally morphed, like the Gap ad. (Which still tracks poorly, but looks a lot better anyway...)
I'm more impressed with the CGI in those chrysler ads where the car peels away. Very slick, very clean. Anyone know which agency/effects house did those?
The x86 line can't run virtual 386+ machines in hardware. If their product really did what they say it does, then ANY os would work on it, not just WinNT and Linux.
My bets on them having recoded the lowest layer of the Linux kernel to abstract it a step from the hardware... Three years ago when I was bored out of my mind at work and was being forced to use NT, I'd thought about the possibility of having a core microkernel sitting directly on the hardware, and having customized kernels for Linux and NT that mapped their memory management, task switching, etc all through the microkernel.
I wonder if they're claiming a patent on it... I know I've got a whole notebook of notes sitting around on the idea.:)
Sure, a flat-panel 50 inch screen is nice, they say its HD upgradable, but for $25k I'd expect an HD upgradable to be able to actually do HD resolution. 1280x768 isn't.
Doesn't mean I won't accept it if someone was to give me one.:)
Before everyone gets into a knock down dirty flame war about Linux vs Windows and what's more secure, remember the problem is people having access in front of the computer. On windows, you go and look in the temp directory. If you think the permissions mean anything under Linux I may be able to just hit control-alt-delete on your system, catch it at lilo, hit tab to see what your linux boot profile is called (lets say its "linux")
LILO> linux init=/bin/sh
Guess what, your computer just dropped me into a shell without asking for your root password. If you let people in front of your computer, Linux isn't much more secure than Windows. If you don't, at least someone can't telnet into Windows. I'd be more concerned about this being an issue under Linux.
I've had wierd problems like that with MySQL... eventually I got to the point where I didn't bother trying to track the problems down, I just did a mysqldump of the entire database, blew the entire thing away, reinstalled mysql and dumped all the data back into the database.
Worked like a charm. Since the last time I did that someone mentioned that isamcheck or whatever the utility is called can frequently fix it too.
This technique for 3-D visualization is OLD. Sure, using a hologram is new, but from the sound of it (and some experience with holograms years ago) it seems like a step back.
Systems I've seen in the past that do this use those ribbed plastic screens they glue on the front of baseball cards and the like to make them 3-D. The 3-D effect drops out and reforms itself as you move to the side, but there's not just one "sweet spot", there's a number of them moving in either direction from the screen.
When I was young I played with this idea using one of those plastic sheets and an Atari 800 hooked up to a TV.
I saw a demo a few years back that used a high-refresh monitor and a rotating mirror to project 3-d images that not only could be viewed from the side, but your perspective on what you were looking at would be different from the side than the front. You could move around the display and see the side of things...
Even casual references to a website on slashdot generates a noticeable jump in hits on a server. On Friday I mentioned the autoLinux project I was working on, and had 2,500 visitors within two or three hours of the post, and it wasn't even a featured story, just a url I mentioned in the comment.
I don't recall seeing things on any of those papers mentioned though, so I'm wondering what kind of a load a server gets in the case of the stories that really catch people's attention. 100 hits per minute is pretty low, even if you're going to a site doing dynamic content. I'm wondering what sort of server hardware/software is being used on sites that have real problems as a result of the slashdot effect.
100 hits per minute of files that take more than a dozen seconds or so to download could bring servers down that run Apache and have limits on the number of children they can spawn. (150 default?)
My guess is its large numbers of modem users not large numbers of hits that really causes the/. effect.
Maybe a slashdot poll asking what speeds people view slashdot at?
(the URL for autoLinux is http://www.bangsplat.org/autolinux for anyone interested who missed it before...)
Well I was getting ready to announce this anyway, so now's as good a time as any...
I've been working on developing hardware and software combinations for about six months now for using Linux in a car. I've recently been given a copy of the TIGER/Line database and am working on a format for storing navigation mapping data in a compact enough format to use in a car for GPS automotive applications under Linux.
There's a mailing list for the project, even though its really in its infancy. The mailing list is meant for discussion about automotive uses of Linux, as well as hardware and software issues with using it in that environment.
I have one Winblows machine in my house -- its one of the ones that the Linux system can turn on. I use it for only one thing, music production. Linux is a tad weak when it comes to music production software, that I've found. If your Microsoft phone works for you, that's all fine and dandy. It costs a bundle. It relies on Windows (and as such isn't reliable...) You have no control over how it does what it does. Hopefully you'll get a lot of use out of it, Microsoft's consumer hardware isn't bad, their joysticks are nice and I use their mice exclusively. But I have to replace them once or twice a year. Hopefully your phone will have better reliability.
Gotta remember something though, most of this kind of stuff is just plain stupid -- particularly that house. What's the point if all you're doing is plugging in someone else's closed-box system, plugging in an expensive piece of hardware, and turning it all on? None of those features are very useful IMHO, and you get nothing personally out of it.
I set my stuff up because I could. I use the Internet access once in a while when I forget files at home, and I fairly frequently wonder during the day if I turned off the coffee pot and being able to turn it off from remote is kinda useful.
Oughta make ya proud being able to put "Microsoft Inside" on your house... lmao
I'm not too impressed. Windows 95? Sheesh... run Linux. I've got my linux router/automation server set up to handle my voicemail using vgetty. I can punch in a code and turn on or off most of the lights and major appliances in my apartment, set my thermostats back, program away cycles for my lights, as well as check my voicemail. Enter another code and it will hang up and dial my house network up to the Internet (no high speed access around here unfortunately) and then page me with its IP-address. It'll even let me turn on any of the other PC's on the network in case I need a file off one of them and I'm not at home.
Of course all of that can be done through the web server as well, with the added ability to read received faxes and listen to the voice messages online. Lets see Winblows do all that. For free.:)
I have a Linux system that handles my voicemail, firewalling, and home automation at home that's running in worse shape than it sounds your computer arrived in. Motherboard and powersupply sitting on a shelf, no case. Harddrive on its back next to it, piece of paper keeping things from shorting and a floppy drive taped to the top of it. Couldn't find the connectors for the serial ports so the modem shucked its case and is hardwired to the serial port through a ribbon cable.
Looks sort of like something out of that movie Pi.
The only practical way to make money doing that, unless you're only doing it on a very small scale and aren't paying for bandwidth, is get investment money, sink millions into marketing, get several million users and either IPO or sell it. Unless you're on an IPO or buyout track you've got to make a profit and I've yet to see that happen doing that sort of a business.
Coincidentally I've been working on precisely this for several months now. My intention was to release a stripped down subset of the Tiger/Line database optimised for street mapping. (It by default for anyone who hasn't seen it has a *ton* of data that isn't really all that useful).
I've got Perl code that parses the database and have been experimenting with various ways to index and organize the data to rapidly determine which data segments fall within a given visual area without doing too much real-time sorting. No display stuff yet, but I can take coordinates for a given location and very quickly retrieve the data necessary for displaying the map.
The code I'm working on is for a module for an SVGALIB based framework for automotive applications under Linux that's going in my car. E-mail, GPS, mapping, ICQ and other such features. I was planning on getting a more formal project going in another month or two when my workload at the job goes down a bit, and I'm certainly interested in talking to anyone else working on similar things.
This resold Tiger/Line database is great -- the damn thing is free (for anyone who wasn't aware of that) but you've got to find someone who has it... I only have a single CD-ROM, not the full country.
If anyone wants to discuss these sort of automotive applications for Linux, please feel free to drop me a line at georgeh-nospam@blowtorch.com. Get rid of -nospam (obviously...)
Slashdot Mirror
Go back and reread the article.
That $840 price they mentioned was the price of PCs there run around $840... not the Playstation. It said it will cost *less* than a PC, so you've got a max price of $840.
I'm guessing $349 to start...
I've seen that general motors quote a few times, and quotes like it dozens of times. "Why would I give away something for free that costs millions to develop?"
Well in a lot of those cases, that's what they're doing. General Motors in a certain sense does give away their cars. When I spend $20k on a car, I'm paying a few hundred dollars for the labor of the people assembling the car, a big chunk of money for the cost of parts that are made by the company (frame, body, etc), and a significant chunk for parts made by outside part manufacturers.
Now I've bought my car, its completely open to me. State inspections ignored, I can take all the body panels off my car and tape cardboard spraypainted with fleckstone paint onto it. I can take the engine out and put another in, or rework how the engine works. I can do anything I want with it.
So what's the money paying for? Its for the parts, the time of the people assembling it and part of the profit goes to the company that developed it. If I really wanted to, I could buy the repair manual for the car, go out and buy one part after another and build it from the ground up. Happens all the time with hotrods and restoration cars. The design of the car is completely open. If I want a new cam for the car, I can buy one from the company that made it, but I can also take calipers and such, measure it, and make one myself if I've got the facilities for it or pay someone else to.
So compare that to open source. Sure I can go download all the parts of a Linux system and put it together myself, piece by piece. The manufacturing and materials cost of a piece of software is very low, so they're a lot cheaper than car parts, but those are essentially why they're cheaper. The first Linux system I used was built just that way. But now I tend to use RedHat linux. They build my "car" for me, and I can change it once I get it to how I'd like it to be.
And they sink a lot of money into developing new parts for the "car". I can buy a later model of the "car" if I want or I can aquire the parts and put them on myself.
My point is that the open source concept isn't *that* alien to business -- lots of them have been doing it all along, when you remove costs of materials and manufacturing from the equation. I don't need to go to Ford to build my own Mustang if I've got a good machine shop, skills, and time.
Come on people, lets make some effort to understand the issue before spouting off on it. Admittedly I didn't read the FCC release before posting my last comment on this, but what I said in it was essentially correct.
:)
This has *nothing* to do with making ISP calls long distance, it just a simple ruling allowing telcos to choose if they desire to set up a reciprocal billing arrangement with ISPs (presumably larger ISPs) within their service area that they are providing lines to.
This doesn't mean you get billed per minute, it just means they have flexibility in how they choose to bill the line usage to the ISP.
They can, for example, class the incoming lines as inter-carrier lines, allowing them to be billed for usage rather than end-point charges.
And its not a requirement its just a clarification of their position on it. If I run fiber between two cities myself (ie, I own the fiber and I'm leasing the pole or line space underground) then I typically need those sort of reciprocal agreements with the carriers who are going to gateway traffic to my line in both terminating ends.
I don't think this is nearly as big of a deal as people have been making it out to be for the last six months that the rumor about this was going around. Some telcos may choose to gouge ISPs with these prices, but most probably won't charge anything. The release also made it seem like the actual situation would be based on the way existing reciprocal agreements within the state are handled.
I still wish I didn't need to use a telco though
I could be wrong but I think everyone may be misinterpreting this one. My understanding was this moves calls through your local telco to your ISP into the same classification as a call through your telco to a long distance provider, it does NOT make the calls long distance, it just lets telcos choose to bill ISP's for the traffic they cause, which they won't necessarily do. Any telco that has its own internet service would be shooting themselves in the foot since they'd have to bill the usage back to themselves at the same rate.
I can spend $25 a month and get unlimited calls on Sprint on the weekends, so the rates telcos are charging long distance companies for access through their CO's can't be that high.
I also think that billing in that manner to an ISP would mean that they can't charge for the line itself into the ISP but I could be wrong about that. If an ISP is paying $40 a month per line or circuit on a T1, then usage charges might not be much more per line except in unusual cases.
If that's the case, and this is really what the classification change means then I'd guess only very high usage people even have a chance of being affected. And someone who really has a valid reason to tie up their internet line 16 hours a day ought to have enough reason to pay $300 or $400 a month for a frame relay connection if cable, xDSL or other technology isn't available.
Anyone know anything more specific about this?
Actually, I do have a VME backplane, but no cards for it. :)
I noticed that company happens to sell a lot of VME products...
Does anyone know if any of the ports of Linux support VME-based systems? If any of them do, does the SMP support work when multiple VME boards with processors are on the VME bus or do the boards themselves have to be SMP specifically? (ie, can I pop two dual-processor VME boards onto the bus and have Linux use all four processors?) I've worked with 88110 and 68040 VME boards but don't really know how they actually worked from that standpoint...
I've seen true-blue SysVR4 running on VME systems that way, it'd be cool if Linux did too.
Hey I bet this would have the horsepower to play Gran Tourismo and simulate car damage. (A key feature in any good two player driving game where my friends are concerned...)
I remember reading that AC article back in 1996 about this. Seemed like a really cool technology, but if you notice the commercials doing it now and compare it to the ones on the site, the ones using that system have the worst quality in terms of image stability while the "movement" is happening, color continuity, etc, etc, etc...
That credit card commercial in particular I thought looked like crap. Any of the ones involving moves through a larger amount of space, because they seem to spread the units out too far and double or triple up on frames like older cartoons did. It makes for very choppy movement.
The commercials where its done very clearly are typically shot from fewer vantage points and digitally morphed, like the Gap ad. (Which still tracks poorly, but looks a lot better anyway...)
I'm more impressed with the CGI in those chrysler ads where the car peels away. Very slick, very clean. Anyone know which agency/effects house did those?
The x86 line can't run virtual 386+ machines in hardware. If their product really did what they say it does, then ANY os would work on it, not just WinNT and Linux.
:)
My bets on them having recoded the lowest layer of the Linux kernel to abstract it a step from the hardware... Three years ago when I was bored out of my mind at work and was being forced to use NT, I'd thought about the possibility of having a core microkernel sitting directly on the hardware, and having customized kernels for Linux and NT that mapped their memory management, task switching, etc all through the microkernel.
I wonder if they're claiming a patent on it... I know I've got a whole notebook of notes sitting around on the idea.
Sure, a flat-panel 50 inch screen is nice, they say its HD upgradable, but for $25k I'd expect an HD upgradable to be able to actually do HD resolution. 1280x768 isn't.
:)
Doesn't mean I won't accept it if someone was to give me one.
Before everyone gets into a knock down dirty flame war about Linux vs Windows and what's more secure, remember the problem is people having access in front of the computer. On windows, you go and look in the temp directory. If you think the permissions mean anything under Linux I may be able to just hit control-alt-delete on your system, catch it at lilo, hit tab to see what your linux boot profile is called (lets say its "linux")
LILO> linux init=/bin/sh
Guess what, your computer just dropped me into a shell without asking for your root password. If you let people in front of your computer, Linux isn't much more secure than Windows. If you don't, at least someone can't telnet into Windows. I'd be more concerned about this being an issue under Linux.
I've had wierd problems like that with MySQL... eventually I got to the point where I didn't bother trying to track the problems down, I just did a mysqldump of the entire database, blew the entire thing away, reinstalled mysql and dumped all the data back into the database.
Worked like a charm. Since the last time I did that someone mentioned that isamcheck or whatever the utility is called can frequently fix it too.
*shrug* maybe it would work for Slashdot.
This technique for 3-D visualization is OLD. Sure, using a hologram is new, but from the sound of it (and some experience with holograms years ago) it seems like a step back.
Systems I've seen in the past that do this use those ribbed plastic screens they glue on the front of baseball cards and the like to make them 3-D. The 3-D effect drops out and reforms itself as you move to the side, but there's not just one "sweet spot", there's a number of them moving in either direction from the screen.
When I was young I played with this idea using one of those plastic sheets and an Atari 800 hooked up to a TV.
I saw a demo a few years back that used a high-refresh monitor and a rotating mirror to project 3-d images that not only could be viewed from the side, but your perspective on what you were looking at would be different from the side than the front. You could move around the display and see the side of things...
Damn, where's this guy live? $.99 for a seven layer burrito? They're a buck thirty nine around here...
Hmph... no cable modems, no DSL, expensive taco bell... why do I live here?
Even casual references to a website on slashdot generates a noticeable jump in hits on a server. On Friday I mentioned the autoLinux project I was working on, and had 2,500 visitors within two or three hours of the post, and it wasn't even a featured story, just a url I mentioned in the comment.
/. effect.
I don't recall seeing things on any of those papers mentioned though, so I'm wondering what kind of a load a server gets in the case of the stories that really catch people's attention. 100 hits per minute is pretty low, even if you're going to a site doing dynamic content. I'm wondering what sort of server hardware/software is being used on sites that have real problems as a result of the slashdot effect.
100 hits per minute of files that take more than a dozen seconds or so to download could bring servers down that run Apache and have limits on the number of children they can spawn. (150 default?)
My guess is its large numbers of modem users not large numbers of hits that really causes the
Maybe a slashdot poll asking what speeds people view slashdot at?
(the URL for autoLinux is http://www.bangsplat.org/autolinux for anyone interested who missed it before...)
Well I was getting ready to announce this anyway, so now's as good a time as any...
I've been working on developing hardware and software combinations for about six months now for using Linux in a car. I've recently been given a copy of the TIGER/Line database and am working on a format for storing navigation mapping data in a compact enough format to use in a car for GPS automotive applications under Linux.
There's a mailing list for the project, even though its really in its infancy. The mailing list is meant for discussion about automotive uses of Linux, as well as hardware and software issues with using it in that environment.
Anyone interested should check out:
http://www.bangsplat.org/autolinux
I have one Winblows machine in my house -- its one of the ones that the Linux system can turn on. I use it for only one thing, music production. Linux is a tad weak when it comes to music production software, that I've found. If your Microsoft phone works for you, that's all fine and dandy. It costs a bundle. It relies on Windows (and as such isn't reliable...) You have no control over how it does what it does. Hopefully you'll get a lot of use out of it, Microsoft's consumer hardware isn't bad, their joysticks are nice and I use their mice exclusively. But I have to replace them once or twice a year. Hopefully your phone will have better reliability.
Gotta remember something though, most of this kind of stuff is just plain stupid -- particularly that house. What's the point if all you're doing is plugging in someone else's closed-box system, plugging in an expensive piece of hardware, and turning it all on? None of those features are very useful IMHO, and you get nothing personally out of it.
I set my stuff up because I could. I use the Internet access once in a while when I forget files at home, and I fairly frequently wonder during the day if I turned off the coffee pot and being able to turn it off from remote is kinda useful.
Oughta make ya proud being able to put "Microsoft Inside" on your house... lmao
I'm not too impressed. Windows 95? Sheesh... run Linux. I've got my linux router/automation server set up to handle my voicemail using vgetty. I can punch in a code and turn on or off most of the lights and major appliances in my apartment, set my thermostats back, program away cycles for my lights, as well as check my voicemail. Enter another code and it will hang up and dial my house network up to the Internet (no high speed access around here unfortunately) and then page me with its IP-address. It'll even let me turn on any of the other PC's on the network in case I need a file off one of them and I'm not at home.
:)
Of course all of that can be done through the web server as well, with the added ability to read received faxes and listen to the voice messages online. Lets see Winblows do all that. For free.
I have a Linux system that handles my voicemail, firewalling, and home automation at home that's running in worse shape than it sounds your computer arrived in. Motherboard and powersupply sitting on a shelf, no case. Harddrive on its back next to it, piece of paper keeping things from shorting and a floppy drive taped to the top of it. Couldn't find the connectors for the serial ports so the modem shucked its case and is hardwired to the serial port through a ribbon cable.
Looks sort of like something out of that movie Pi.
:)
The only practical way to make money doing that, unless you're only doing it on a very small scale and aren't paying for bandwidth, is get investment money, sink millions into marketing, get several million users and either IPO or sell it. Unless you're on an IPO or buyout track you've got to make a profit and I've yet to see that happen doing that sort of a business.
Coincidentally I've been working on precisely this for several months now. My intention was to release a stripped down subset of the Tiger/Line database optimised for street mapping. (It by default for anyone who hasn't seen it has a *ton* of data that isn't really all that useful).
I've got Perl code that parses the database and have been experimenting with various ways to index and organize the data to rapidly determine which data segments fall within a given visual area without doing too much real-time sorting. No display stuff yet, but I can take coordinates for a given location and very quickly retrieve the data necessary for displaying the map.
The code I'm working on is for a module for an SVGALIB based framework for automotive applications under Linux that's going in my car. E-mail, GPS, mapping, ICQ and other such features. I was planning on getting a more formal project going in another month or two when my workload at the job goes down a bit, and I'm certainly interested in talking to anyone else working on similar things.
This resold Tiger/Line database is great -- the damn thing is free (for anyone who wasn't aware of that) but you've got to find someone who has it... I only have a single CD-ROM, not the full country.
If anyone wants to discuss these sort of automotive applications for Linux, please feel free to drop me a line at georgeh-nospam@blowtorch.com. Get rid of -nospam (obviously...)