Round connectors are always easier when prototyping, because a drill is always easier and faster than a square punch. But on modern assembled, plastic-molded devices, this isn't necessarily the case.
Also: AFAICT, round connectors became standardized with the telephone switchboard, 40 or so years before the Germans wanted another war. (Some vendors still call a 1/4" plug a "Phone Connector," for this very reason.)
I can't be bothered to run around the house with a magnet and check, so take this with a grain of salt, but:
Cheap audio connectors are often (not always) steel, because steel is cheaper than brass. Expensive audio connectors are often (not always) brass, because brass conducts better but is more expensive than steel.
In either case: Various platings are applied for prettiness or corrosion resistance or both (with nickel being cheap and durable, and gold being expensive and pretty).
(Please note that I avoided inferring that any combination of these materials is in some way better than any other combination. My own single-data-point worth of experience with gold-plated steel connectors is that they can rust in normal use, while no other combination of materials ever does so.)
Read it? Of course. I even quoted it. Did you read it?
If LCD screens behaved in the manner that you say they do, the effect should be plainly visible. It isn't, as far as I can see.
That the effect is not visible causes me to cast doubt upon the your assertion (ie: it set off my bullshit detector). So, I asked for a citation (which you haven't provided).
But there will only be waste in large areas of black screen. In an image with a more even distribution of pixel states the light will bounce around until it finds a pixel that's on, and come out there.
Really? So when I have the backlight fixed at 100%, and display a screen of all black except for one white pixel, that pixel will be very, very bright? I don't see that effect on either of the LCDs in front of me (one TN, one IPS).
Or are you saying that the brightness of the pixels on one side of the screen affect the brightness of those on the other half? I don't see that, either.
Perhaps neighboring pixels? Nope - can't say I've seen that happening.
The point was not whether you could present me with an unrealistic objective, the point was that the power of the engine has little effect on how easy it is to drive in the snow. An F1 car can be difficult to get moving from a stop even on a dry road with the regular clutch arrangement they're fitted with.
Going out in a Porsche 911 in the snow is a bit silly simply on the basis that they tend to be expensive cars for most people.
A not-so-special 1995 Porsche 911 is specified to have 4.8 inches of ground clearance, weighs 3170 pounds, and produces 270HP @ 6100 RPM. It has ABS and traction control, and a limited-slip differential.
My 1995 BMW 325i has about the same ground clearance (I just went outside and measured it), and is specified as weighing 3175 pounds with 189HP @ 5900RPM. It also has ABS and traction control, but does not have LSD (which is a profound disadvantage from the 911 in snow).
Both are RWD, though one has the engine in the back and other has it in the front. They've both got fancy independent suspensions front and rear, and both can have similarly-sized wheels tires fitted. There's more things similar about the two cars than there is different.
The BMW is the best vehicle I've ever driven in snow. It just works. It grips so well with winter tires that it's almost boring to drive on all types of snow and ice (deep slush does get a bit more interesting). It goes, it stops, and it turns. I carry a heavy tow strap in it, just in case, but I've only ever used it to extract other cars from ditches (yeah, it does that).
The Porsche should not be profoundly different. Perhaps a little more tail-happy due to weight distribution, but that will still be under the control of the loud pedal.
And before you go back to saying "Oh, but the Porsche has 80 more HP!!" please realize those figures are at wide-open throttle with the engine at or near redline, and have nothing at all to do with how a sane person actually sets forth. (And if you actually believe that they do have an abundance of power available at low RPM with little throttle input, you really need to study up a bit more.)
Book value on the Porsche is close to 10x as as much as the BMW, which may make it sound inherently silly, but they're both just production cars put together on an assembly line: Neither of them are inherently unique. If someone wants to drive their Porsche in Alaska, who gives a shit? It's their money, and it'll work fine.
Yes, they've shown a lot of ability in that area. They've got more parallel CPU power than God.
Calibrating vehicles based on comparative performance results is interesting, as well.
But if it gets to that point, wouldn't it be simply easier to measure their actual speed through the turns? No calibrated accelerometers needed.
Suppose they automatically tag every posted speed sign on a curve based on Streetview photographic data, and record the average speed of vehicles that traverse that curve. They then have a baseline correction for posted-worst-case vs. actual-every-day speeds, based on DOT specification and The Real World.
This will allow them to extrapolate what a given turn should be posted an unmarked curve, while maintaining consistency with how the rest of the turns on that section of road are already posted
Bonus points for further extrapolating that data into a prediction of what speed a specific car should be able to navigate the curve at, but I don't like the liability aspects of that myself...
Would you like to file the patent on this yourself, or shall we apply jointly?
You're likely totaled by the first moose that hits you no matter what sort of normal road-going vehicle you're in. There's just something about a couple of tons of car intersecting a ton of meat that tends to ruin one's day.
For car-vs-moose games, we either drive heavy trucks, or something else. And while I'll be the first to admit that I've always wanted a proper dump truck just to tool around in, I find myself driving something else instead.
That's a pretty extreme example, but I'll bite: If you find me an F1 car with suitable tires and with friction material in the clutch that doesn't engage like a lightswitch, I'll show you myself how to drive one in the snow (again, and again, and again, as long ground clearance isn't a problem).
(Yeah, I added a condition. But regular street cars, in including Porsche 911s, come equipped with much more reasonable clutch arrangements than F1 cars.)
Power is controlled by the loud pedal (the one on the right), and can be further modulated with one or more other pedals on the left. Driver skill makes up for an abundance of power.
Old man story time: I used to drive a 1996 Firebird in the snow. It had bad weight distribution, was grossly overpowered for the conditions, had a high-ratio diff, and had impressively wide tires with an impressively unsuitable tread pattern and composition, and was quite low to the ground. On the face of it, there was nothing good about the car for winter driving.
But it did have positraction and a button to tell the automatic transmission to start out in second gear instead of first. Posi helped a lot, obviously. The 2nd gear start button helped a lot too (by causing the power transfer at low speeds to be completely gummified by the torque converter), but only at first.
At first, it was a terrifying thing. I'm no stranger to driving in snow, but this car wouldn't fucking go, or do anything else for that matter. Eventually, the car forced me to learn it better, and to better modulate the throttle. After that, like a kid taking off the training wheels on his bike, I just left the tranny in its default mode, and didn't have any issues.
Subsequently, I could drive that car on anything, as long as ground clearance wasn't a problem and I wasn't in a hurry. Stopping and turning were still laughable, so I was unable to safely drive the car very fast, but it always worked and never surprised me, despite being the antithesis of a good winter vehicle.
Except the Cyanogenmod-using crowd consists of those who are most likely to actually fix their own phones when broken.
I replaced the digitizer and the LCD screen in my Motorola Droid. I'd never have paid someone to do it for me, but as a tinkerer geek it just seemed like the right thing to do. The digitizer was free from someone else's water damaged phone, while the LCD screen is a cheap Ebay part that looks and works just like the original.
Between fixing things myself and Cyanogenmod, I'm not really interested in a new phone. On the other hand, when it comes time to get something different, I'll surely get something that is both easily hacked and already has good support from CM.
Even for the 2 weeks of the year you can drive it.
Spoken as someone who has never seriously looked into the concept of "winter tires".
As long as the tires (and, optionally, chains) are appropriate, and ground clearance is not an issue, any car works fine in the frozen north -- especially if the driven axle has the majority of the vehicle's weight on it. And this is, obviously, the case of a rear-engined, RWD 911.
Compared in particular to a typical front-heavy RWD pickup truck, I'd suspect that the Porsche would do quite well at going, stopping, and turning.
-Adolf (not planning on switching my 50/50 weight distribution, Blizzak-equipped RWD BMW for anything else in snow or ice, ever, though I might entertain the notion of something with a bit more rearward weight bias just for particularly ugly winter days) Osborne
My own forays into accelerometer data have been meh, at best: At the rates that even my 325i with fairly soft street tires corners at, the results are horribly clouded with body roll and weight transfer. I think it's a bit bold to say that accelerometer data, by itself, is useful to these ends without heavily massaging the data to take the individual car into account.
(My GMC Safari is way, way worse -- the lateral acceleration figures I get from [random app] are mostly approximate of sheer lies, as the thing pans and tilts its way down a curvy road.)
You can't accurately map the G load from an accelerometer if you're under the influence of gravity and also suffering from body roll (and, as you know, -all- cars with not-completely-rigid suspensions (ie: air-filled tires) roll to some extent).
Or at least you can't with my Motorola Droid. Or my first-gen iPod Touch.
The iPhone 4 allegedly has gyroscopes as well as accelerometers, which should make an easier time of that, but that's a relatively recent development in the whole Google Streetview world of things, and I doubt that they currently have such data on very many roads.
Should they? Yes. Will they make use of it if they do have it? Doubtful. Will proper signage make a bigger improvement in safety? I think so, but if the signage is reviewed and found to be good, then the resultant data should also be good (vis-a-vis my last post).
Where did you find information that the FCC let the ISPs know which subscribers are participating? I also took part in the study and I was expecting that sort of thing to happen, but I never saw anything to confirm it.
I'd like to know this as well.
(The story has scrolled down far enough that neither of us are likely to get a useful response, but I figure it's worth a shot.)
When I last had Road Runner, I had dropouts regularly whenever things were wet. I called, complained, and 2 minutes later had an appointment for a tech to show up and fix things during the next day.
The tech apparently did show up (he never rang the doorbell), and apparently did fix things, because after that everything was golden.
Protip: Your ISP won't fix your connection problems if you don't tell them that you have them. (Whether they should fix them on their own, without prompting, is another discussion.)
I can watch a couple of Netflix streams on different devices, load up on torrents, goof off with Youtube, SSH into a remote *nix box, and web browse at the same time as my wife plays WoW, regardless of time of day.
All on a single consumer-grade VDSL loop that is about 300 feet beyond spec and consists partly of lead-sheathed cable that dates back to the infancy of the PSTN. And I wasn't even part of the study!
That folks seem to think this is an unusual level of performance is pretty disturbing.
From my side: The aftermarket world was a strange soup back then. We didn't have ASUS and eVGA and Gigabyte to blame for our woes: We just bought whatever board we could get our hands on (whether at a swap-meet or from the back pages of Computer Shopper) and were happy with that (or not).
PC Chips-based boards were always a curse. I'm glad to have not run across them recently.
That said: In 1985 I was still cutting my teeth on a combination of TRS-80 Model I and VIC-20. It wasn't until '88 that I got my fingers into an 8088. I'll get off your lawn now...
Perhaps we just have a different perspective on history.
Every properly-configured 386SX/DX board I ran across worked fine, whether big-name OEM or aftermarket...except for one strange AMD 386SX/40 board I had which had issues with DMA.
Which is just one example of slight (easy to work around under DOS) badness out of a whole sea of others that worked without episode.
(Meanwhile: Of course jumpers made a difference. That's why they were there. (Did you expect that they'd installed jumper headers onto a board with the intention of having them do nothing at all? FWIW, I can tweak a single RAM setting on my Q6600 box and the machine will play dead, too. BIOS settings from the past decade == the jumper settings of the prior decade.))
That sounds like a lot of work: Even GPS units which supposedly show overall speed limits on long stretches of road seem to get it wrong much of the time, and those are at least possible to research in a not-totally-insane fashion through datamining various public DOT info and local regs.
Individual turns, though. That just sounds hairy. Is this data published anywhere?
Not to discount the usefulness. As a safety measure, knowing the that the turn ahead is a blind, decreasing-radius whore of a thing with opposite camber that even a 911 GT3 with soft tires would take issue with would be far more useful than in terms of safety than knowing that the McVillage that you're driving through has lowered the speed limit on their main drag to 25MPH.
Here's how I think it should be done: First, I think, the signs need fixed. This will be expensive, but it ought to be within the charter of whatever organization it is that is supposed to take care of the road anyway. A top-to-bottom review would be best, but crowd-sourcing (by soliciting complaints and sending a qualified crew out to have a look when they're in the vicinity) would be a hell of a good start toward figuring out how to post signage.
Second, the details need to be made available in a public fashion using some sort of format which is easily absorbed into a GPS's database. And since there's going to be a lot of review (and new data) happening anyway, this part should be just about free to implement (on the grand scale of things).
It doesn't do anything for me, since I live in flat country where the roads are straight, and the occasional bend on a main road is always very well marked. But it'd be a big help when I'm elsewhere on unfamiliar roads...
I have an AT&T/NCR 386SL laptop running a (quite old) version of Slackware. There were no nightmares involved with the installation that had anything to do with the type of CPU. It's as stable as a rock, although it does do slightly more stuff than most rocks...
If there are problems with the design that affect software, I'd like to say that they were sorted quite a long time ago.
Where I'm from, all 2-lane-ish rural roads (except at the outskirts of cities, and even then only for a mile or two) are 55MPH. It doesn't matter if they're US routes, or state routes, or county roads, or ruddy little township roads barely big enough for a single compact car: They're all the same.
All 4-lane roads are 65MPH, with rare exception, but those are already routed around with "Avoid highways."
If I set up a TomTom to avoid roads posted over 50MPH, the only sane result would be that it would instruct me to stay home.
Slashdot Mirror
(Playing dumb) Is this $120 hardware kit able to connect directly to my TV?
[of course not]
(/dumb) How much does the rest of the PC cost?
A PC with 3D Blu-Ray playback and a Six-Axxis controller for less than $249?
Where?
Cheers. Hope the scotch is working.
Round connectors are always easier when prototyping, because a drill is always easier and faster than a square punch. But on modern assembled, plastic-molded devices, this isn't necessarily the case.
Also: AFAICT, round connectors became standardized with the telephone switchboard, 40 or so years before the Germans wanted another war. (Some vendors still call a 1/4" plug a "Phone Connector," for this very reason.)
I can't be bothered to run around the house with a magnet and check, so take this with a grain of salt, but:
Cheap audio connectors are often (not always) steel, because steel is cheaper than brass. Expensive audio connectors are often (not always) brass, because brass conducts better but is more expensive than steel.
In either case: Various platings are applied for prettiness or corrosion resistance or both (with nickel being cheap and durable, and gold being expensive and pretty).
(Please note that I avoided inferring that any combination of these materials is in some way better than any other combination. My own single-data-point worth of experience with gold-plated steel connectors is that they can rust in normal use, while no other combination of materials ever does so.)
I have one word for you - hinge. Use a hinge to have it move out of the way.
The cover over the airbag already does this. Adding a lightweight LCD+touchscreen wouldn't change that much.
Read it? Of course. I even quoted it. Did you read it?
If LCD screens behaved in the manner that you say they do, the effect should be plainly visible. It isn't, as far as I can see.
That the effect is not visible causes me to cast doubt upon the your assertion (ie: it set off my bullshit detector). So, I asked for a citation (which you haven't provided).
Really? So when I have the backlight fixed at 100%, and display a screen of all black except for one white pixel, that pixel will be very, very bright? I don't see that effect on either of the LCDs in front of me (one TN, one IPS).
Or are you saying that the brightness of the pixels on one side of the screen affect the brightness of those on the other half? I don't see that, either.
Perhaps neighboring pixels? Nope - can't say I've seen that happening.
So. [citation needed], if you don't mind.
Sure. It won't be as fast as charging it from a power strip that is plugged into itself, but it should do fine when out and about.
The point was not whether you could present me with an unrealistic objective, the point was that the power of the engine has little effect on how easy it is to drive in the snow. An F1 car can be difficult to get moving from a stop even on a dry road with the regular clutch arrangement they're fitted with.
Going out in a Porsche 911 in the snow is a bit silly simply on the basis that they tend to be expensive cars for most people.
A not-so-special 1995 Porsche 911 is specified to have 4.8 inches of ground clearance, weighs 3170 pounds, and produces 270HP @ 6100 RPM. It has ABS and traction control, and a limited-slip differential.
My 1995 BMW 325i has about the same ground clearance (I just went outside and measured it), and is specified as weighing 3175 pounds with 189HP @ 5900RPM. It also has ABS and traction control, but does not have LSD (which is a profound disadvantage from the 911 in snow).
Both are RWD, though one has the engine in the back and other has it in the front. They've both got fancy independent suspensions front and rear, and both can have similarly-sized wheels tires fitted. There's more things similar about the two cars than there is different.
The BMW is the best vehicle I've ever driven in snow. It just works. It grips so well with winter tires that it's almost boring to drive on all types of snow and ice (deep slush does get a bit more interesting). It goes, it stops, and it turns. I carry a heavy tow strap in it, just in case, but I've only ever used it to extract other cars from ditches (yeah, it does that).
The Porsche should not be profoundly different. Perhaps a little more tail-happy due to weight distribution, but that will still be under the control of the loud pedal.
And before you go back to saying "Oh, but the Porsche has 80 more HP!!" please realize those figures are at wide-open throttle with the engine at or near redline, and have nothing at all to do with how a sane person actually sets forth. (And if you actually believe that they do have an abundance of power available at low RPM with little throttle input, you really need to study up a bit more.)
Book value on the Porsche is close to 10x as as much as the BMW, which may make it sound inherently silly, but they're both just production cars put together on an assembly line: Neither of them are inherently unique. If someone wants to drive their Porsche in Alaska, who gives a shit? It's their money, and it'll work fine.
Yes, they've shown a lot of ability in that area. They've got more parallel CPU power than God.
Calibrating vehicles based on comparative performance results is interesting, as well.
But if it gets to that point, wouldn't it be simply easier to measure their actual speed through the turns? No calibrated accelerometers needed.
Suppose they automatically tag every posted speed sign on a curve based on Streetview photographic data, and record the average speed of vehicles that traverse that curve. They then have a baseline correction for posted-worst-case vs. actual-every-day speeds, based on DOT specification and The Real World.
This will allow them to extrapolate what a given turn should be posted an unmarked curve, while maintaining consistency with how the rest of the turns on that section of road are already posted
Bonus points for further extrapolating that data into a prediction of what speed a specific car should be able to navigate the curve at, but I don't like the liability aspects of that myself...
Would you like to file the patent on this yourself, or shall we apply jointly?
You're likely totaled by the first moose that hits you no matter what sort of normal road-going vehicle you're in. There's just something about a couple of tons of car intersecting a ton of meat that tends to ruin one's day.
For car-vs-moose games, we either drive heavy trucks, or something else. And while I'll be the first to admit that I've always wanted a proper dump truck just to tool around in, I find myself driving something else instead.
That's a pretty extreme example, but I'll bite: If you find me an F1 car with suitable tires and with friction material in the clutch that doesn't engage like a lightswitch, I'll show you myself how to drive one in the snow (again, and again, and again, as long ground clearance isn't a problem).
(Yeah, I added a condition. But regular street cars, in including Porsche 911s, come equipped with much more reasonable clutch arrangements than F1 cars.)
Power is controlled by the loud pedal (the one on the right), and can be further modulated with one or more other pedals on the left. Driver skill makes up for an abundance of power.
Old man story time: I used to drive a 1996 Firebird in the snow. It had bad weight distribution, was grossly overpowered for the conditions, had a high-ratio diff, and had impressively wide tires with an impressively unsuitable tread pattern and composition, and was quite low to the ground. On the face of it, there was nothing good about the car for winter driving.
But it did have positraction and a button to tell the automatic transmission to start out in second gear instead of first. Posi helped a lot, obviously. The 2nd gear start button helped a lot too (by causing the power transfer at low speeds to be completely gummified by the torque converter), but only at first.
At first, it was a terrifying thing. I'm no stranger to driving in snow, but this car wouldn't fucking go, or do anything else for that matter. Eventually, the car forced me to learn it better, and to better modulate the throttle. After that, like a kid taking off the training wheels on his bike, I just left the tranny in its default mode, and didn't have any issues.
Subsequently, I could drive that car on anything, as long as ground clearance wasn't a problem and I wasn't in a hurry. Stopping and turning were still laughable, so I was unable to safely drive the car very fast, but it always worked and never surprised me, despite being the antithesis of a good winter vehicle.
Except the Cyanogenmod-using crowd consists of those who are most likely to actually fix their own phones when broken.
I replaced the digitizer and the LCD screen in my Motorola Droid. I'd never have paid someone to do it for me, but as a tinkerer geek it just seemed like the right thing to do. The digitizer was free from someone else's water damaged phone, while the LCD screen is a cheap Ebay part that looks and works just like the original.
Between fixing things myself and Cyanogenmod, I'm not really interested in a new phone. On the other hand, when it comes time to get something different, I'll surely get something that is both easily hacked and already has good support from CM.
Spoken as someone who has never seriously looked into the concept of "winter tires".
As long as the tires (and, optionally, chains) are appropriate, and ground clearance is not an issue, any car works fine in the frozen north -- especially if the driven axle has the majority of the vehicle's weight on it. And this is, obviously, the case of a rear-engined, RWD 911.
Compared in particular to a typical front-heavy RWD pickup truck, I'd suspect that the Porsche would do quite well at going, stopping, and turning.
-Adolf (not planning on switching my 50/50 weight distribution, Blizzak-equipped RWD BMW for anything else in snow or ice, ever, though I might entertain the notion of something with a bit more rearward weight bias just for particularly ugly winter days) Osborne
Interesting idea.
My own forays into accelerometer data have been meh, at best: At the rates that even my 325i with fairly soft street tires corners at, the results are horribly clouded with body roll and weight transfer. I think it's a bit bold to say that accelerometer data, by itself, is useful to these ends without heavily massaging the data to take the individual car into account.
(My GMC Safari is way, way worse -- the lateral acceleration figures I get from [random app] are mostly approximate of sheer lies, as the thing pans and tilts its way down a curvy road.)
You can't accurately map the G load from an accelerometer if you're under the influence of gravity and also suffering from body roll (and, as you know, -all- cars with not-completely-rigid suspensions (ie: air-filled tires) roll to some extent).
Or at least you can't with my Motorola Droid. Or my first-gen iPod Touch.
The iPhone 4 allegedly has gyroscopes as well as accelerometers, which should make an easier time of that, but that's a relatively recent development in the whole Google Streetview world of things, and I doubt that they currently have such data on very many roads.
Should they? Yes. Will they make use of it if they do have it? Doubtful. Will proper signage make a bigger improvement in safety? I think so, but if the signage is reviewed and found to be good, then the resultant data should also be good (vis-a-vis my last post).
IMHO, of course.
I'd like to know this as well.
(The story has scrolled down far enough that neither of us are likely to get a useful response, but I figure it's worth a shot.)
Have you even bothered to complain?
When I last had Road Runner, I had dropouts regularly whenever things were wet. I called, complained, and 2 minutes later had an appointment for a tech to show up and fix things during the next day.
The tech apparently did show up (he never rang the doorbell), and apparently did fix things, because after that everything was golden.
Protip: Your ISP won't fix your connection problems if you don't tell them that you have them. (Whether they should fix them on their own, without prompting, is another discussion.)
Another single data point:
I can watch a couple of Netflix streams on different devices, load up on torrents, goof off with Youtube, SSH into a remote *nix box, and web browse at the same time as my wife plays WoW, regardless of time of day.
All on a single consumer-grade VDSL loop that is about 300 feet beyond spec and consists partly of lead-sheathed cable that dates back to the infancy of the PSTN. And I wasn't even part of the study!
That folks seem to think this is an unusual level of performance is pretty disturbing.
From my side: The aftermarket world was a strange soup back then. We didn't have ASUS and eVGA and Gigabyte to blame for our woes: We just bought whatever board we could get our hands on (whether at a swap-meet or from the back pages of Computer Shopper) and were happy with that (or not).
PC Chips-based boards were always a curse. I'm glad to have not run across them recently.
That said: In 1985 I was still cutting my teeth on a combination of TRS-80 Model I and VIC-20. It wasn't until '88 that I got my fingers into an 8088. I'll get off your lawn now...
I haven't run across that. Perhaps my neighborhood is simply not contentious enough.
Perhaps we just have a different perspective on history.
Every properly-configured 386SX/DX board I ran across worked fine, whether big-name OEM or aftermarket...except for one strange AMD 386SX/40 board I had which had issues with DMA.
Which is just one example of slight (easy to work around under DOS) badness out of a whole sea of others that worked without episode.
(Meanwhile: Of course jumpers made a difference. That's why they were there. (Did you expect that they'd installed jumper headers onto a board with the intention of having them do nothing at all? FWIW, I can tweak a single RAM setting on my Q6600 box and the machine will play dead, too. BIOS settings from the past decade == the jumper settings of the prior decade.))
That sounds like a lot of work: Even GPS units which supposedly show overall speed limits on long stretches of road seem to get it wrong much of the time, and those are at least possible to research in a not-totally-insane fashion through datamining various public DOT info and local regs.
Individual turns, though. That just sounds hairy. Is this data published anywhere?
Not to discount the usefulness. As a safety measure, knowing the that the turn ahead is a blind, decreasing-radius whore of a thing with opposite camber that even a 911 GT3 with soft tires would take issue with would be far more useful than in terms of safety than knowing that the McVillage that you're driving through has lowered the speed limit on their main drag to 25MPH.
Here's how I think it should be done: First, I think, the signs need fixed. This will be expensive, but it ought to be within the charter of whatever organization it is that is supposed to take care of the road anyway. A top-to-bottom review would be best, but crowd-sourcing (by soliciting complaints and sending a qualified crew out to have a look when they're in the vicinity) would be a hell of a good start toward figuring out how to post signage.
Second, the details need to be made available in a public fashion using some sort of format which is easily absorbed into a GPS's database. And since there's going to be a lot of review (and new data) happening anyway, this part should be just about free to implement (on the grand scale of things).
It doesn't do anything for me, since I live in flat country where the roads are straight, and the occasional bend on a main road is always very well marked. But it'd be a big help when I'm elsewhere on unfamiliar roads...
For the record:
I have an AT&T/NCR 386SL laptop running a (quite old) version of Slackware. There were no nightmares involved with the installation that had anything to do with the type of CPU. It's as stable as a rock, although it does do slightly more stuff than most rocks...
If there are problems with the design that affect software, I'd like to say that they were sorted quite a long time ago.
Where I'm from, all 2-lane-ish rural roads (except at the outskirts of cities, and even then only for a mile or two) are 55MPH. It doesn't matter if they're US routes, or state routes, or county roads, or ruddy little township roads barely big enough for a single compact car: They're all the same.
All 4-lane roads are 65MPH, with rare exception, but those are already routed around with "Avoid highways."
If I set up a TomTom to avoid roads posted over 50MPH, the only sane result would be that it would instruct me to stay home.