I'd like to program for KDE, just like I would have liked to program for Macs in 1992, just like I did program on the Atari 800 back before then... difference being that today I like to pay my mortgage, feed the family, etc. Qt is actually a pretty righteous choice, especially when your customer base contains 50% Windows XP, and 45% Windows Vista.
I'm surprised I had to scroll down this far to see Qt Creator mentioned... Following some of the imagery above, I'd sooner kill a buffalo and dress it with my bare hands than use most IDEs, and I've been trying them off and on for about 15 years now. Qt Creator actually makes my programming work easier without forcing me to learn a bunch of arcane switches hidden under six layers of GUI just to make simple things happen.
I'm not sure I'd use Qt Creator to work with libraries other than Qt, but if you are just a casual programmer, you could do worse than to start a project from one of the Qt examples.
download a copy of Visual [C++|C#|VB] and you can do all kinds of fun stuff.
I prefer the Qt SDK - yeah, it works on Windows (as well as Visual C++ does), and after you've invested 3000 hours learning the tool, you're not locked in.
I was going to mod you up, until you started going on about the scarcity of gas-power turbo cars (I happen to own one) and the small number of people who fool around tuning them - it depends on your perspective. There are probably low tens of thousands of turbo owner/tuners out there in the US. From the technical POV, it's one of the easier things to do to make significantly more power in a mostly normal car. Yes, that's a small number - but nowhere near zero.
Me, personally, not interested in tuning my car at the moment, been there, done that, got other things to mess with right now, I might get back to it in a decade or two. The tuners aren't all that hyped up about getting the most out of ethanol since you can get more out of pump gas. And... "boost" is not compression ratio, the actual ratio between maximum compression and where the exhaust/intake valves open/shut is pretty hard to change on the fly - you can get a little with variable valve timing, but VVT is usually used in other ways (such as to promote "better breathing" for low end torque and max power at high RPM.)
A variable compression ratio engine would be more expensive and prone to defects than the common fixed compression ratio engines - add to this the fact that using corn for auto fuel only makes sense to the corn farmers and I'm hoping that ethanol goes the way of the Edsel really soon...
I'm just hoping I'll live to see the day when the controllers are cheap, reliable, readily available, and worth bothering with.
Today you can buy a smart toaster, but it won't talk properly with your smart refrigerator or your PS3... It seems to have taken PCs about 30 years to reach some sort of convergence where you're not really "locked out" just because of the OS you run, or the hardware you run on... most software I care about today is available on the "big three" Win/Lin/Mac, and those three play relatively well with one another.
I was probably stupid to buy a PS3, but I thought that I would want a BluRay player to go with the new TV, turns out it really isn't such a big deal, I would probably have been better off with a Mac Mini or similar little PC instead of the PS3... it still is pretty obtuse in the playing nicely with others department (DLNA is crap compared to the "Power of Samba / NFS").
So, yeah, the embedded microcontrollers are around and they can do these things, but can they play nicely with the rest of the digital landscape yet? I don't really want to be rolling my own code to get the relevant data and functionality shared between these things - a full PC might seem like overkill, but it's going to be the quickest way to get something like a universal language between all the widgets, so that I can configure the night light to flash when then motion sensor outside the door is tripped, and more importantly, do unanticipated things like making the doorbell chime ring when outside temperature drops below 32 degrees or whatever...
Convenience is the key - X-10 had a single vendor solution for lots of these kinds of things 10, maybe 15 years ago, including a central microcontroller, I had one, it was crap (periodic lockups, unreliable hardware, etc.), and I couldn't do much about it because their components wouldn't interact with other vendors' stuff, and the others didn't have as many choices in functionality, etc. etc.
The Marvel guys really hit it on the head when they said these things should be inside everything else you buy... I'm thinking of:
Printers - should all be network attached, when is the last time you cared to lock out any user on your local network from accessing your printer? If so, this is smart enough to configure access controls on an http interface too.
Ditto scanners
Coffee makers... well, for a certain type of coffee drinker...
Cars, with WiFi:
As a "radio"
For a mobile copy of the service manual, that also records maintenance and diagnostic info
If you could trust Detroit to write secure, reliable software, the engine, transmission and ABS computers would be much cooler if they reported status via wireless http
Any appliance that has a user's manual that you actually need to read and maintenance requirements
The oven could record successful (and not so successful) temperature profiles and even take snapshots of how you arranged the baking trays.
$40 in 2 years isn't the really wild part - $4 in 10 years is where it starts getting crazy.
My old, underpowered (Celeron 800MHz) PC draws about 60 watts. My new(ish) QNap TS-109 NAS box draws something less than 20 and is a better file server... The new eeeBox desktop also draws less than 20.
Most of these could be USB accessories (many are already available as such), but they could be cheaper still if they were wired to an IO pin on the chipset, and a slicker product if they were molded into the wart instead of dangling off on a connector.
A spam filter worth having would require external (to the wart) storage, and I prefer larger organizations (Google, Yahoo, Corporate IT) to manage my spam filter for me, rather than trying to keep up with the battle myself... kinda defeats the purpose of a spam filter if you have to think about it.
First accessory I'd want for a smart wall-wart would be powerline networking (that works, unlike X-10 that seems to be about 99% reliable, which is crap when it's open loop.)
The next most obvious accessories include things like:
relay switching / dimming of the power line for lamp control, simple appliance control
An LED "night light" that could also convey information
A motion sensor
Microphone / speaker for intercom / VOIP functions
Temp sensors for room by room environmental data (and subsequent control of HVAC diverters / thermostat)
Battery power backup
Problem is that most of these functions would be happy without their own local CPU, but if the CPUs do drop below the $50 price point, it could be feasible to just have local CPUs anyway for data integrity, local signal processing, autonomous operation in the event of network failure, etc.
Did it strike anyone else as lame that the MIT dude said he'd have it run a spam filter?
Corporations, politicians, police, neighborhood association boards are all supposed to serve their constituents.
Of course, the corporation's constituents are the shareholders, not the customers... which is why there are laws against monopolies, price fixing, collusion, etc. Not that those laws have ever been effectively enforced, but the ideal has at least been established. What we need is real choice among auto manufacturers, as it is, they all screw the customer the same way, so if you want a car, you make your own, or buy from them.
When I looked into manufacturing my own OBDII to Palm Pilot interface cable (in about 1998), there were societies you had to be a member of to get access to the documentation, with membership fees of $25K per year and up.
Yeah, it's price fixing, collusion, etc. and it has been illegal in the US since the 1920s.
On the "openness" side, they are probably being encouraged by the feds to lock out access to Joe Sixpack as much as possible, otherwise he is liable to soup up his new ride and put out more emissions. New chips are almost all in-circuit reprogrammable, the interface is generally a couple of wires and should be able to be driven by any computer that the software is compatible with - so they make the software cryptic to discourage tinkering, and incidentally support a pretty big aftermarket filled with vendors who spend tons of time and money researching things that should have just been published in the first place.
We had a wheel speed sensor go out and give us an ABS malfunction light at 35,920 miles, when we were about 120 miles from home on a 36Kmi warranty.
The dealer took the vehicle for the whole day to make the swap out repair and charged us $200 for it.
ABS is overrated, but it is a nice thing to have on the rear wheels of an empty pickup truck. Of course, in my old (1977) pickup without ABS when someone would cut me off in traffic, I could just tap the brakes and make a big chattering tire smoking show for their rear-view... looks much more dangerous than it really is.
I've dealt with similar inspections in Florida (Dade county) and Texas (Houston), Dade county gave up after about 10 years of harassing their populace and having all the junkers registered in other counties to dodge the rules. Houston is still going strong AFAIK. I had to do about $400 worth of "repairs" to get past Houston emissions on my first try (tailpipe sniff test) - didn't really improve overall emissions, and it is such a crock in Houston to drive past a petrochemical plant that is belching literally tons of particulates into the air per minute and be dinged because your Miata is 3% over the unburned hydrocarbons limit.
There are hundreds, if not thousands of petrochemical plants there, and at least one day a month they run dirty enough to leave a sticky film on everything within 10 miles downwind of the plant.
I've jumped around watching a few TtSCC shows from hulu, and from that perspective it's just as full of the same high-school coming of age / romance angst as the rest of teen television (see: Buffy, Roswell, 90210, and a dozen others I'm sure...)
The real question is: What's Summer Glau's next psycho chick series role going to be?
If your Solus Pro comes with (reasonably priced) updates that cover a wide range of manufacturers, then things are improving. I looked at getting into OBDII code reading when the Palm Pilot was all the rage (1998ish), at that time the only complete code reading systems on the market were running $50K+, and they were for single manufacturer lines.
Right now, you need a laptop to get at those codes - if ASUS can deliver complete netbooks for $300, it shouldn't cost much more than that to integrate a 7" touch screen display into the dash of your $30,000 new vehicle.
The implications of High cyl head temp could be, in laymen's terms, something to the effect of:
"Engine is hot, continued driving may cause engine failure and damage requiring expensive repair or replacement of the engine. Stop the vehicle in a safe location and keep the engine running."
When in park,
"Cylinder head temperature is dropping, do not turn off your engine, unless there is a safety issue. The cylinder head temperature gauge is measuring 477 degrees. Normal operating range for cylinder head temperature in this vehicle is 180 to 350 degrees. Operation with cylinder head temperatures above 450 degrees for more than a few minutes can lead to head gasket failure, lubrication failure, and will eventually lead to complete engine failure. Possible causes for elevated cylinder head temperature include cooling system failure, lean fuel mixture, and operation at high engine load for extended time in a hot environment. As long as the cylinder head temperature continues to drop quickly, the engine should be kept running to allow the cooling system to operate. This vehicle has experienced high cylinder head temperatures 3 times in the past year, indicating a possible intermittent problem with the engine, or operation in extreme conditions. Consultation with a qualified mechanic is recommended."
Later,
Cylinder head temperature is 342 degrees, this is within normal operating range (180-350). All other engine status monitors are within normal range. If cylinder head temperatures continue to run above normal your vehicle should be serviced by a qualified mechanic as soon as possible.
Most of this information is already codified in a reasonable form for the service techs, it wouldn't take as much effort or expense to put it into owner consumable form as it does to formulate a seasonal marketing strategy for a splinter brand (Pontiac, for example.)
One diagnostic that I could have used would have read something like:
The engine coolant temperature gauge is measuring 877 degrees. This indicates a lack of coolant in the system, or possibly a faulty temperature sensor. All other engine status monitors are within normal range. Recommend service with a qualified technician as soon as possible, the engine monitoring system cannot function properly without accurate engine coolant temperature information.
Instead, I just got a pegged temperature gauge and a check engine light. This, coincidentally, happened just after I hooked up a rather heavy trailer to a new-ish pickup truck. Freaked me out, but it was just a bad sensor, we were lucky to find a mechanic in a strange town who would fix it on a Saturday afternoon. If the light had come on 3 hours later, it would have been a nerve wracking Sunday driving 400 miles with the temp needle spiked, check engine light on, and nothing really wrong.
In 1977, I can understand that blinking lights were state of the art, and recorded messages informing you that your door is a jelly container (jar) had their limits.
In 2009, there's no reason they couldn't have a 3rd grade reading level explanation, with pictures and videos, telling that your gas cap needs to be tightened. I believe most cars still send you to the dealer with a check engine light for that one.
If I can buy an MP3 player that has a 2" video screen for $50 - the auto manufacturers have no excuse for not having a user-friendly display (better than cryptic flashing lights) built in with the OBDII interface. It should (in the US) use plain English to describe exactly what is wrong and what the implications are - no reference manual required - hell, the reference manual should be available on an on-board http server with a WiFi network that both serves the info to the owner's notebook PC, and downloads updates and tech bulletins (automatically, for free) when driven onto a dealer's lot.
Everything I have described above costs less than one air-bag, and should be standard equipment on all but the most basic models, and provided as an "at cost" option for any car it doesn't come standard on.
Should... in a fantasy world where the corporations are actually serving their customers.
you already can. Go get an OBD2 reader or have a shop pull the codes. Interpreting them is a bit harder - I thought there was legislation requiring manufacturers to divulge the codes, but I'm not sure.
Point of the article is that the standard OBD2 readers aren't cutting it anymore, they're giving the legally required (smog related) codes and nothing else. If you want a reader like the dealer uses, prepare to fork out more than you paid for the car... This is why the independent mechanics are feeling screwed.
Ingenuity, huh? Well, that ingenuity has finally come back to bite Joe and Detroit in the ass. Today, we finally start seriously saving fuel, or else.
Or else what? There's no laws, rules, or even social norms to discourage me from driving a 10 liter 3 ton pickup truck everywhere I go, in point of fact, we own a 5.9 liter 1/2 ton truck that could be used as our daily driver if we wanted, we choose to leave it parked unless needed, but my wife likes to drive it when the weather is foul because it makes her feel safer.
The real waste in our two driver household is actually the third vehicle - we probably emitted more carbon footprint in the purchase of our "lightweight" around town car than we will save in fuel consumption difference over its lifetime. I didn't do a carbon analysis, but dollar-wise, assuming the (purchased at 1 year old) $13K commuter lasts 8 years and resells for $1K, that's $1500 per year in capital costs, plus about $500/yr in additional insurance / maintenance, so we'd have to save 800 gallons a year (at $2.50/gallon) to make the choice truly economical. Say the truck gets 12mpg and the car gets 36 (to be really generous, our car gets more like 27 around town)... we'd need to transfer about 14,400 miles per year from the truck to the car to "break even" on fuel consumption dollars. Considering that the car has only been driven about 12,000 miles a year, it's not really saving us money. What it is doing is giving us a small, easy to park "right sized" vehicle to serve our around town driving. It looks more economical than going everywhere in the truck, but it isn't.
We have legitimate reasons for using the truck, about 10 times a year. Renting might be slightly more economical, but it completely destroys the convenience and power of owning your own vehicle, ready at a moment's notice. There's also the convenience of the redundant backup, the "third" vehicle is 19 years old - well maintained, but about once a year it needs some repairs and it's nice to be able to park it and do the repairs at leisure rather than having the pressure of "needing" the vehicle. The truck is 10 years old, so it's going to start falling into that periodic repair category soon too.
Legislating increased fuel economy in new vehicles isn't biting anyone in the ass. It's about time, just like in the late '60s / early '70s, our engine technology is producing more power than is really useful for getting from A to B. It's about time to turn that technology away from making overpowered vehicles into making them more efficient, just like they did with the initial CAFE standards. The free market clearly values "fun" over efficiency, and why not? Life is short. It won't seriously hurt anyone's happiness for CAFE to rise by 30%, electronic engine management systems can pull off that and more, but not without additional incentives outside the free market.
The first alpha of 9.10 was released a couple days ago with new kernel, new gcc, lots of new libraries... you should not be surprised things don't work well yet. Jaunty seems pretty stable to me. Minor issues with my intel video card, but works fine for all my daily work.
Yes, but you must not care to hear Biff bark...
The summary complains about sound, and the datapoints I have on Linux sound are this:
1998 - the LinuxSoundHOWTO makes derisive statements to the effect of "well, if you must have sound, these are the hoops you jump through:...", implying that real free beer swilling penguin huggers don't need sound, period.
2006 - Debian Stable with KDE - Turning on desktop sounds completely hosed one user account, never worked quite right and eventually crashed and burned to the point that it was easier to wipe the account and start over without sounds than to unravel the damage, login to the account would lock up the whole machine.
2009 - Kubuntu 9.04 - sound still stutters and stalls in some circumstances - yes, I can play Pandora flawlessly through Firefox, but TuxType still stutters - shun on TuxType all you like, it's my 5 year old's favorite interactive app...
2009 - Fedora (recent, not sure exact release, not my machine) - ported a Qt/portaudio app from Vista, compiled and linked with remarkably little trouble, ran flawlessly the first time, then got hosed up and would only play 2 seconds of sound before sound hung up, at least the rest of the app soldiered on. Restarting the app would give another 2 seconds of sound then cut out again. Running a music player in-between would clear out whatever the problem is and the ported app would work fine again, unless there was any kind of unclean exit (not going through the PA close, and going through a PA close after things were off-rails doesn't help). Sure, the app should behave, that's what Win3.1 said.
I have had similar problems with motherboard based ethernet ports, except they get off the rails and never return - plug in an older $5 ethernet card on the bus and everything in hunky-dory again, but it still shows that the drivers aren't up to snuff in handling problem states in some cases. I knew a serious Debian guru who resorted to the same hardware based fix just because that's how things are, easier to spend the $5 than try to unravel the driver issue.
The latest desktops are mightily impressive, but I still bought an XP machine for my wife because: Eudora is _still_ more stable than Thunderbird, these little quirks are just un-necessary for her to deal with, and I don't want to be bothered trying to get Reader Rabbit to Run under Wine, especially with the sound issues.
Slashdot Mirror
As much as it might pain the free software crowd, Microsoft has done a good job with Visual Studio Express.
It doesn't really pain them much at all, since VSE can't build a single thing that runs in *nix.
I'd like to program for KDE, just like I would have liked to program for Macs in 1992, just like I did program on the Atari 800 back before then... difference being that today I like to pay my mortgage, feed the family, etc. Qt is actually a pretty righteous choice, especially when your customer base contains 50% Windows XP, and 45% Windows Vista.
I'm surprised I had to scroll down this far to see Qt Creator mentioned... Following some of the imagery above, I'd sooner kill a buffalo and dress it with my bare hands than use most IDEs, and I've been trying them off and on for about 15 years now. Qt Creator actually makes my programming work easier without forcing me to learn a bunch of arcane switches hidden under six layers of GUI just to make simple things happen.
I'm not sure I'd use Qt Creator to work with libraries other than Qt, but if you are just a casual programmer, you could do worse than to start a project from one of the Qt examples.
download a copy of Visual [C++|C#|VB] and you can do all kinds of fun stuff.
I prefer the Qt SDK - yeah, it works on Windows (as well as Visual C++ does), and after you've invested 3000 hours learning the tool, you're not locked in.
I was going to mod you up, until you started going on about the scarcity of gas-power turbo cars (I happen to own one) and the small number of people who fool around tuning them - it depends on your perspective. There are probably low tens of thousands of turbo owner/tuners out there in the US. From the technical POV, it's one of the easier things to do to make significantly more power in a mostly normal car. Yes, that's a small number - but nowhere near zero.
Me, personally, not interested in tuning my car at the moment, been there, done that, got other things to mess with right now, I might get back to it in a decade or two. The tuners aren't all that hyped up about getting the most out of ethanol since you can get more out of pump gas. And... "boost" is not compression ratio, the actual ratio between maximum compression and where the exhaust/intake valves open/shut is pretty hard to change on the fly - you can get a little with variable valve timing, but VVT is usually used in other ways (such as to promote "better breathing" for low end torque and max power at high RPM.)
A variable compression ratio engine would be more expensive and prone to defects than the common fixed compression ratio engines - add to this the fact that using corn for auto fuel only makes sense to the corn farmers and I'm hoping that ethanol goes the way of the Edsel really soon...
I'm just hoping I'll live to see the day when the controllers are cheap, reliable, readily available, and worth bothering with.
Today you can buy a smart toaster, but it won't talk properly with your smart refrigerator or your PS3... It seems to have taken PCs about 30 years to reach some sort of convergence where you're not really "locked out" just because of the OS you run, or the hardware you run on... most software I care about today is available on the "big three" Win/Lin/Mac, and those three play relatively well with one another.
I was probably stupid to buy a PS3, but I thought that I would want a BluRay player to go with the new TV, turns out it really isn't such a big deal, I would probably have been better off with a Mac Mini or similar little PC instead of the PS3... it still is pretty obtuse in the playing nicely with others department (DLNA is crap compared to the "Power of Samba / NFS").
So, yeah, the embedded microcontrollers are around and they can do these things, but can they play nicely with the rest of the digital landscape yet? I don't really want to be rolling my own code to get the relevant data and functionality shared between these things - a full PC might seem like overkill, but it's going to be the quickest way to get something like a universal language between all the widgets, so that I can configure the night light to flash when then motion sensor outside the door is tripped, and more importantly, do unanticipated things like making the doorbell chime ring when outside temperature drops below 32 degrees or whatever...
Convenience is the key - X-10 had a single vendor solution for lots of these kinds of things 10, maybe 15 years ago, including a central microcontroller, I had one, it was crap (periodic lockups, unreliable hardware, etc.), and I couldn't do much about it because their components wouldn't interact with other vendors' stuff, and the others didn't have as many choices in functionality, etc. etc.
$40 in 2 years isn't the really wild part - $4 in 10 years is where it starts getting crazy.
My old, underpowered (Celeron 800MHz) PC draws about 60 watts. My new(ish) QNap TS-109 NAS box draws something less than 20 and is a better file server... The new eeeBox desktop also draws less than 20.
Most of these could be USB accessories (many are already available as such), but they could be cheaper still if they were wired to an IO pin on the chipset, and a slicker product if they were molded into the wart instead of dangling off on a connector.
A spam filter worth having would require external (to the wart) storage, and I prefer larger organizations (Google, Yahoo, Corporate IT) to manage my spam filter for me, rather than trying to keep up with the battle myself ... kinda defeats the purpose of a spam filter if you have to think about it.
Problem is that most of these functions would be happy without their own local CPU, but if the CPUs do drop below the $50 price point, it could be feasible to just have local CPUs anyway for data integrity, local signal processing, autonomous operation in the event of network failure, etc.
Did it strike anyone else as lame that the MIT dude said he'd have it run a spam filter?
Corporations, politicians, police, neighborhood association boards are all supposed to serve their constituents.
Of course, the corporation's constituents are the shareholders, not the customers... which is why there are laws against monopolies, price fixing, collusion, etc. Not that those laws have ever been effectively enforced, but the ideal has at least been established. What we need is real choice among auto manufacturers, as it is, they all screw the customer the same way, so if you want a car, you make your own, or buy from them.
When I looked into manufacturing my own OBDII to Palm Pilot interface cable (in about 1998), there were societies you had to be a member of to get access to the documentation, with membership fees of $25K per year and up.
Yeah, it's price fixing, collusion, etc. and it has been illegal in the US since the 1920s.
On the "openness" side, they are probably being encouraged by the feds to lock out access to Joe Sixpack as much as possible, otherwise he is liable to soup up his new ride and put out more emissions. New chips are almost all in-circuit reprogrammable, the interface is generally a couple of wires and should be able to be driven by any computer that the software is compatible with - so they make the software cryptic to discourage tinkering, and incidentally support a pretty big aftermarket filled with vendors who spend tons of time and money researching things that should have just been published in the first place.
We had a wheel speed sensor go out and give us an ABS malfunction light at 35,920 miles, when we were about 120 miles from home on a 36Kmi warranty.
The dealer took the vehicle for the whole day to make the swap out repair and charged us $200 for it.
ABS is overrated, but it is a nice thing to have on the rear wheels of an empty pickup truck. Of course, in my old (1977) pickup without ABS when someone would cut me off in traffic, I could just tap the brakes and make a big chattering tire smoking show for their rear-view... looks much more dangerous than it really is.
I've dealt with similar inspections in Florida (Dade county) and Texas (Houston), Dade county gave up after about 10 years of harassing their populace and having all the junkers registered in other counties to dodge the rules. Houston is still going strong AFAIK. I had to do about $400 worth of "repairs" to get past Houston emissions on my first try (tailpipe sniff test) - didn't really improve overall emissions, and it is such a crock in Houston to drive past a petrochemical plant that is belching literally tons of particulates into the air per minute and be dinged because your Miata is 3% over the unburned hydrocarbons limit.
There are hundreds, if not thousands of petrochemical plants there, and at least one day a month they run dirty enough to leave a sticky film on everything within 10 miles downwind of the plant.
I've jumped around watching a few TtSCC shows from hulu, and from that perspective it's just as full of the same high-school coming of age / romance angst as the rest of teen television (see: Buffy, Roswell, 90210, and a dozen others I'm sure...)
The real question is: What's Summer Glau's next psycho chick series role going to be?
If your Solus Pro comes with (reasonably priced) updates that cover a wide range of manufacturers, then things are improving. I looked at getting into OBDII code reading when the Palm Pilot was all the rage (1998ish), at that time the only complete code reading systems on the market were running $50K+, and they were for single manufacturer lines.
Right now, you need a laptop to get at those codes - if ASUS can deliver complete netbooks for $300, it shouldn't cost much more than that to integrate a 7" touch screen display into the dash of your $30,000 new vehicle.
The implications of High cyl head temp could be, in laymen's terms, something to the effect of:
"Engine is hot, continued driving may cause engine failure and damage requiring expensive repair or replacement of the engine. Stop the vehicle in a safe location and keep the engine running."
When in park,
"Cylinder head temperature is dropping, do not turn off your engine, unless there is a safety issue. The cylinder head temperature gauge is measuring 477 degrees. Normal operating range for cylinder head temperature in this vehicle is 180 to 350 degrees. Operation with cylinder head temperatures above 450 degrees for more than a few minutes can lead to head gasket failure, lubrication failure, and will eventually lead to complete engine failure. Possible causes for elevated cylinder head temperature include cooling system failure, lean fuel mixture, and operation at high engine load for extended time in a hot environment. As long as the cylinder head temperature continues to drop quickly, the engine should be kept running to allow the cooling system to operate. This vehicle has experienced high cylinder head temperatures 3 times in the past year, indicating a possible intermittent problem with the engine, or operation in extreme conditions. Consultation with a qualified mechanic is recommended."
Later,
Cylinder head temperature is 342 degrees, this is within normal operating range (180-350). All other engine status monitors are within normal range. If cylinder head temperatures continue to run above normal your vehicle should be serviced by a qualified mechanic as soon as possible.
Most of this information is already codified in a reasonable form for the service techs, it wouldn't take as much effort or expense to put it into owner consumable form as it does to formulate a seasonal marketing strategy for a splinter brand (Pontiac, for example.)
One diagnostic that I could have used would have read something like:
The engine coolant temperature gauge is measuring 877 degrees. This indicates a lack of coolant in the system, or possibly a faulty temperature sensor. All other engine status monitors are within normal range. Recommend service with a qualified technician as soon as possible, the engine monitoring system cannot function properly without accurate engine coolant temperature information.
Instead, I just got a pegged temperature gauge and a check engine light. This, coincidentally, happened just after I hooked up a rather heavy trailer to a new-ish pickup truck. Freaked me out, but it was just a bad sensor, we were lucky to find a mechanic in a strange town who would fix it on a Saturday afternoon. If the light had come on 3 hours later, it would have been a nerve wracking Sunday driving 400 miles with the temp needle spiked, check engine light on, and nothing really wrong.
In 1977, I can understand that blinking lights were state of the art, and recorded messages informing you that your door is a jelly container (jar) had their limits.
In 2009, there's no reason they couldn't have a 3rd grade reading level explanation, with pictures and videos, telling that your gas cap needs to be tightened. I believe most cars still send you to the dealer with a check engine light for that one.
If I can buy an MP3 player that has a 2" video screen for $50 - the auto manufacturers have no excuse for not having a user-friendly display (better than cryptic flashing lights) built in with the OBDII interface. It should (in the US) use plain English to describe exactly what is wrong and what the implications are - no reference manual required - hell, the reference manual should be available on an on-board http server with a WiFi network that both serves the info to the owner's notebook PC, and downloads updates and tech bulletins (automatically, for free) when driven onto a dealer's lot.
Everything I have described above costs less than one air-bag, and should be standard equipment on all but the most basic models, and provided as an "at cost" option for any car it doesn't come standard on.
Should... in a fantasy world where the corporations are actually serving their customers.
you already can. Go get an OBD2 reader or have a shop pull the codes. Interpreting them is a bit harder - I thought there was legislation requiring manufacturers to divulge the codes, but I'm not sure.
Point of the article is that the standard OBD2 readers aren't cutting it anymore, they're giving the legally required (smog related) codes and nothing else. If you want a reader like the dealer uses, prepare to fork out more than you paid for the car... This is why the independent mechanics are feeling screwed.
Ingenuity, huh? Well, that ingenuity has finally come back to bite Joe and Detroit in the ass. Today, we finally start seriously saving fuel, or else.
Or else what? There's no laws, rules, or even social norms to discourage me from driving a 10 liter 3 ton pickup truck everywhere I go, in point of fact, we own a 5.9 liter 1/2 ton truck that could be used as our daily driver if we wanted, we choose to leave it parked unless needed, but my wife likes to drive it when the weather is foul because it makes her feel safer.
The real waste in our two driver household is actually the third vehicle - we probably emitted more carbon footprint in the purchase of our "lightweight" around town car than we will save in fuel consumption difference over its lifetime. I didn't do a carbon analysis, but dollar-wise, assuming the (purchased at 1 year old) $13K commuter lasts 8 years and resells for $1K, that's $1500 per year in capital costs, plus about $500/yr in additional insurance / maintenance, so we'd have to save 800 gallons a year (at $2.50/gallon) to make the choice truly economical. Say the truck gets 12mpg and the car gets 36 (to be really generous, our car gets more like 27 around town)... we'd need to transfer about 14,400 miles per year from the truck to the car to "break even" on fuel consumption dollars. Considering that the car has only been driven about 12,000 miles a year, it's not really saving us money. What it is doing is giving us a small, easy to park "right sized" vehicle to serve our around town driving. It looks more economical than going everywhere in the truck, but it isn't.
We have legitimate reasons for using the truck, about 10 times a year. Renting might be slightly more economical, but it completely destroys the convenience and power of owning your own vehicle, ready at a moment's notice. There's also the convenience of the redundant backup, the "third" vehicle is 19 years old - well maintained, but about once a year it needs some repairs and it's nice to be able to park it and do the repairs at leisure rather than having the pressure of "needing" the vehicle. The truck is 10 years old, so it's going to start falling into that periodic repair category soon too.
Legislating increased fuel economy in new vehicles isn't biting anyone in the ass. It's about time, just like in the late '60s / early '70s, our engine technology is producing more power than is really useful for getting from A to B. It's about time to turn that technology away from making overpowered vehicles into making them more efficient, just like they did with the initial CAFE standards. The free market clearly values "fun" over efficiency, and why not? Life is short. It won't seriously hurt anyone's happiness for CAFE to rise by 30%, electronic engine management systems can pull off that and more, but not without additional incentives outside the free market.
Run that "correlation is not causation" past the target audience (government bureaucrats) and see if you get anything other than furrowed brows...
The first alpha of 9.10 was released a couple days ago with new kernel, new gcc, lots of new libraries... you should not be surprised things don't work well yet. Jaunty seems pretty stable to me. Minor issues with my intel video card, but works fine for all my daily work.
Yes, but you must not care to hear Biff bark...
The summary complains about sound, and the datapoints I have on Linux sound are this:
1998 - the LinuxSoundHOWTO makes derisive statements to the effect of "well, if you must have sound, these are the hoops you jump through:...", implying that real free beer swilling penguin huggers don't need sound, period.
2006 - Debian Stable with KDE - Turning on desktop sounds completely hosed one user account, never worked quite right and eventually crashed and burned to the point that it was easier to wipe the account and start over without sounds than to unravel the damage, login to the account would lock up the whole machine.
2009 - Kubuntu 9.04 - sound still stutters and stalls in some circumstances - yes, I can play Pandora flawlessly through Firefox, but TuxType still stutters - shun on TuxType all you like, it's my 5 year old's favorite interactive app...
2009 - Fedora (recent, not sure exact release, not my machine) - ported a Qt/portaudio app from Vista, compiled and linked with remarkably little trouble, ran flawlessly the first time, then got hosed up and would only play 2 seconds of sound before sound hung up, at least the rest of the app soldiered on. Restarting the app would give another 2 seconds of sound then cut out again. Running a music player in-between would clear out whatever the problem is and the ported app would work fine again, unless there was any kind of unclean exit (not going through the PA close, and going through a PA close after things were off-rails doesn't help). Sure, the app should behave, that's what Win3.1 said.
I have had similar problems with motherboard based ethernet ports, except they get off the rails and never return - plug in an older $5 ethernet card on the bus and everything in hunky-dory again, but it still shows that the drivers aren't up to snuff in handling problem states in some cases. I knew a serious Debian guru who resorted to the same hardware based fix just because that's how things are, easier to spend the $5 than try to unravel the driver issue.
The latest desktops are mightily impressive, but I still bought an XP machine for my wife because: Eudora is _still_ more stable than Thunderbird, these little quirks are just un-necessary for her to deal with, and I don't want to be bothered trying to get Reader Rabbit to Run under Wine, especially with the sound issues.