Maybe it did when there were few BT clients running, but at this point, on a Cablevision OptimumOnline cable modem connection:
270KB/sec down, 35 KB/sec up.
Try it. If it goes slow, make sure port 6881 is open, although there are enough clients online now that it doesn't seem to matter. (I have it closed and it's working fine. I'd open it up except my crappy Belkin router's web UI and Lynx don't get along.)
I'm using the curses BT client under screen. Started at around 20-30 kB/sec, now I'm at approx. 270 kB/sec, and uploading 35K/sec even without opening up those listening ports.
(Note: I did use --ip to tell it my real "outside world" IP instead of giving the system my internal one.)
I'll leave it running until I reboot the machine (Sometime this evening.) That'll likely be 5-6 hours of serving up the files after the d/l completes.:)
I'm actually surprised you're not having problems with carts wearing out after 3-4 uses with the Lexmark as I was pretty sure they used a thermal inkjet technique just like HP's. (Essentially, a small portion of ink is vaporized and this drives the rest out the nozzle. HP and (I thought) Lexmark cartridges will start burning out those heating elements eventually. That's why HP and Lexmark carts have the nozzles built in to the cartridge.
Canon and Epson, on the other hand, use piezo-based technology. Except for potential clogging, their nozzle drivers will last forever. This is why Epson and Canon have the nozzles in the printer and use "dumb" ink tanks. (Which they can charge nearly as much as HP does for their cartridges, and make more profit because they're simpler.)
The only issue you might have with refilling Canon and Epson carts too often is that sometimes remnant ink can dry and that dried ink in the cart will alter the consistency of the refills.
Continuous ink systems remove this problem by ditching the sponge in the cartridge and replacing it by a tube going directly to the ink bottle.
The equation that has to be solved is floating-point.
The amount of data that has to be dealt with is just enough so that most consumer phones don't have enough CPU to perform the calculation. (This will change soon, already people are wondering if the pseudoranges on the upcoming Kyocera 7135 PalmOS phone will be somehow accessible to developers of applications on the Palm side of the phone.)
Qualcomm no longer makes handsets. Their former handset division is now owned by Kyocera. (In fact, many of the first Kyocera phones bore Qualcomm markings in places, as they began their design phase at Qualcomm.)
Lack of such info is usually not much of a problem if you're moving. Most GPSes (at least Garmin ones) will orient the map based on your direction of travel if you are moving.
More like 1 of the 1 "major" cell carriers in the US uses CDMA.
"major" being defined as "Actually covers 100% of the most densely populated state in the country" - That state being New Jersey.
AT&T used to be in this category, but their GSM network has nothing on their old D-AMPS (The TDMA variant they used) network coverage-wise.
Every other carrier can only pull off 50-66% coverage.
BTW, that one lone carrier that covers all of New Jersey and is the only one that provides service more than two miles outside of the city limits of Ithaca, NY is Verizon.
GSM is not an option in many parts of the United States because the GSM carriers all have horrendous coverage. (And GSM doesn't seem to handle multiple simultaneous providers very well - I've seen countless horror stories of T-Mobile users being unable to use their phones once AT&T and Cingular rolled out GSM in their area.)
" But CDMA is a waste of fucking money in that region because ALL their neighbours use GSM. Eventually the GSM people (i.e. Vodaphone et al.) will move in and people will change to GSM anyway just so they can use their phones when away from their own country. Even in the States you can use a GSM phone. Whilst CDMA is a better, newer technology than GSM it's losing out to GSM in the world, and GSM has an upgrade path to next genereation stuff." Who cares if the neighbors use GSM? 99% of the Iraqi population won't care because most of them won't ever need to use their phones in the country next door.
Iran? Fat chance, even after regime change there's too much bad blood between Iran and Iraq.
Turkey? Good luck getting through that border. Because of the situation with the Kurds, the Turkey/Iran border is going to be VERY tight.
Oh, GSM doesn't have a true upgrade path to 3G. UMTS requires new frequency bands, new handsets, and new antennas. CDMA2000 requires none of the above. (Well, you need a new handset if you want the additional 3G features, but a cdmaOne phone will happily talk to a CDMA2000 tower, it's just restricted to the old circuit-switched 14.4 data. An old GSM phone, on the other hand, CANNOT communicate with a UMTS tower.)
"So fine, install a hugely expensive CDMA network. The GSM people will move in anyway, people will buy GSM phones and your precious US network will go bust. That's the price for FORCING a choice on someone." Oh, like Voicestream/T-Mobile? Or Cingular?
Yes, I'm sure the Iraqis will enjoy seeing their "Service Unavailable" messages. In every country where CDMA has been allowed to take on GSM in an open market, CDMA has won. The smallest CDMA provider in the US (Sprint) still has better coverage than Voicestream/T-Mobile or Cingular, the largest GSM providers in the US. AT&T may eventually become larger, but their GSM network can't compare in any way to their old TDMA network in coverage.
Or look at Japan. DoCoMo rolled out UMTS (3G GSM) while KDDI rolled out CDMA2000 1xEV-DO or EV-DV. Guess who has more subscribers for their 3G services? (Hint: Not DoCoMo).
If GSM is so much better than why can Verizon get away with charging so much more for their CDMA service? (Hint: You get what you pay for in terms of service quality.)
The only countries where GSM is dominant are ones where GSM was *forced on the customer by the government* - Exactly what you're arguing against.
If the Europeans want to build a GSM network in Iraq - Fine. Let them do so on their dime. But if it's on the USA's dime, it's going to be with the technology that won in the open market. GSM and CDMA can coexist in the same area, so I see no problem in the US installing CDMA and Europe installing GSM and seeing which one wins. If Europe wants to see GSM in Iraq - Pay up or shut up.
Or more appropriately, a variant of CDMA that no one has been able to build a handset for that had good battery life/didn't overheat. (Look at what happened to DoCoMo's name in Japan after they rolled out UMTS...)
It comes out to even less than what you state, and the quality rocks.
Also, with laser printers you don't have the option of continuous ink systems, archival inks, or specialized multi-greyscale inks. (People sell cartridges that replace the color inks in Epson with three shades of grey - Which makes for extremely high-quality greyscale printing.)
Laser printers that can print color are: a) Expensive b) Horrible at the job. I've never seen a color laser printer that printed a photo that wasn't completely ugly.
HP is much better than Lexmark, but they still suck.
I've used HP, Lexmark, and Epson printers. We had an HP Deskjet 500 that is STILL working well today. It's well over a decade old, and built like a tank. (Newer HPs might not be so hot.)
I bought a Lexmark 5700 my freshman year in college. The Windows drivers were crap and there was zero Linux support. My parents had a 5700 that broke 5 days after the warranty expired.
We all have Epsons now. They work beautifully under Linux, are built pretty well, and easy to refill. (Note: Epson also is starting to use cartridges with chips, but these have already been "cracked" and were "cracked" quite some time ago.)
The only CIS (Continuous Ink Systems) I've seen have been for Epsons. I don't have one but for someone that prints a LOT they look cool. (Essentially the carts are replaced with tubes going to large bottles of ink that you can buy at bulk prices.)
Modems have been doing compression for years. As others have pointed out, no compression technique is going to help you very much on precompressed files (Images, video, etc. All of the "big stuff" that people buy broadband to receive.) For text it will be a benefit, but not one already offered by modems. (As someone pointed out though, many modems have a limit at the serial port of 115k, so compression on the PC side of this would be of use. Note that the PPP standard has support for some sort of compression, most ISPs don't use it. If you have a machine on the "outside world" with a broadband connection, SSH tunneling to a remote HTTP proxy, with SSH compression turned on will give you a speed boost.)
Caching - Many ISPs already do this, although few ISPs do. It's built into almost all browsers, and if you want "smarter" caching, install Squid.
Persistent connections have been part of HTTP 1.1 for ages.
In short: Nothing you can't already do now for free, but the ISPs are making it easier.
Restricting underclocking (if it happens) would most likely be done to push the mobile editions of their various processors.
No "quick and dirty" FSB-based underclocking of a desktop CPU in a notebook - Nope, gotta buy the expensive mobile version and use SpeedStep.
I wouldn't mind this so much, since desktop CPUs should never have been allowed in notebooks in the first place, but the alternative (SpeedStep) utterly sucks. Transmeta does clock/voltage ramping much better than Intel does. (SpeedStep only has "full tilt" and "slow", while Transmeta CPUs have everything in between.)
Intel just needs to add an instruction that returns the "true" clockspeed rating to a register.
Then any software under any OS can read it, even DOS.
And since it's just one piddly instruction, it would be a simple matter to create a "clockspeed check" utility.
A protection system that makes any system on the planet that gets overclocked scream bloody murder isn't necessary. All that is needed is for a reliable way for the user to download a trusted application, fire it up, and see whether their PC is clocked properly or not.
Luxeon Star-based flashlights (as mentioned earlier) are *amazing*. 1 watt and 5 watt Luxeons are available (although 5W whites are rare, but getting more common.) The 1W units are used in a number of lights such as the Arc LS (Expensive, but excellent) and the CMG Reactor (Don't get it, worst LS light out there). Another person mentioned EL's Blaster series, which are good lights. Also there's plenty of info on the forums linked from the article (candlepowerforums) on converting a 2D or 3D Maglight into a resistored LS light.
For small and relatively cheap, register on the forums and investigate the BadBoy and MadMax drop-ins for MiniMags. These "pills" replace the bulb in your minimag and include a 1W Luxeon on a DC/DC step-up converter. The MadMax is constant-voltage, the BB constant-current (the latter is the best way to drive an LED if you want consistent brightness.) They are both BRIGHT. My BadBoy 400 can tangle with a 2D Maglight and win, with a much whiter and smoother beam.
If you order a BadBoy, I suggest trying the McFlood optics, also from dat2zip's site. They give a nice smooth flood beam. For an intermediate beam, I have a 22.5mm diameter DCX lens from www.anchoroptical.com mounted in a MiniMag accessory holder in front of the McFlood, for a "tight flood" beam.
Nice thing about these "pills" is that if the body gets scuffed up, replacement bodies are only $7-8 each at any local store.
Unfortunately McLuxes are no longer in production - At $120 or so per light, these were quite expensive but they were VERY solidly built with an excellent design, and they were also one of the smallest LS lights out there.
"trickle charging" the laptop itself will not likely be possible - Almost all laptops expect full current for charging. (Even when turned off, my dad's Dell Inspiron 8000 would overload and shut down our old inverter, which was too cheap and too small.)
Trickling the batteries - Be VERY careful. Li-Ions cannot be constantly trickle charged. You SHOULD be safe giving the battery a low-current charge if you use a regulator set to 4.1 or 4.2v/cell (Depends on the exact battery chemistry, the type of electrode changes the nominal voltage slightly.) with a current-limiting resistor. But don't leave the battery in the charger, this will be bad.
The way Li-Ions are supposed to be charged is a current-limited charge, but with the charge voltage never exceeding 4.1 or 4.2 volts per cell. At this threshold, the charger must go into constant voltage mode. Once the current goes below a certain threshold, the charger should start a timer and then shut down at the end of 30-60 minutes. Trickle-charging it beyond that 30-60 minutes is BAD. NiMHs and NiCds can be trickle charged at low currents (C/16 or less, the lower the better for the battery's health), and lead-acids (gel cells are a variant of lead-acid) like having a "float" charge of a constant voltage (Around 13-14 volts for a 12v lead-acid battery I think) applied constantly or regularly. (Car alternators typically provide this charge in automotives.)
Note that even "deep cycle" lead-acids or gel cells should not be cycled deeply on a regular basis. lead-acids like to be topped off, as do those rechargable alkalines.
Problem is that the pinouts and connectors for almost all laptop batteries are nonstandard, and some expect various control lines to be connected while charging. These connections are never documented.
Nothing says that you have to inject the water after the fuel, nor do you have to inject the water in the same manner the fuel is injected.
Note that some of the older fuel injection designs, instead of having multiport injection (one injector/cylinder), they used throttle body injection. (One injector at the throttle body). You could just add a water injector at the throttle body. One guy did it on his Dodge Shadow (http://www.karlsnet.com/mopar/) - There's a lot of info not on that page, he was an active poster on the Allpar (http://www.allpar.com/) forums until he changed jobs and no longer had much time to experiment or post on the forums. Too bad, the guy had some great ideas and was very good at documenting his results, and open to suggestions/constructive criticism. (He had a few errors in some formulas and said, "oops, you're right" and subsequently made corrections on his page.)
Just one note: Do NOT put ethanol into your gas tank unless you know for a fact that your car is an FFV.
Ethanol is VERY corrosive. While the engine management systems of FFVs are often unchanged, anything that comes into contact with the fuel must often be redesigned with additional corrosion resistance. (Chrysler's FFV engines had some fancy corrosion-resistant coatings on their valves, for instance. They also had special fuel lines.)
Slashdot Mirror
Maybe it did when there were few BT clients running, but at this point, on a Cablevision OptimumOnline cable modem connection:
270KB/sec down, 35 KB/sec up.
Try it. If it goes slow, make sure port 6881 is open, although there are enough clients online now that it doesn't seem to matter. (I have it closed and it's working fine. I'd open it up except my crappy Belkin router's web UI and Lynx don't get along.)
On OptimumOnline cable - 270/35 here.
That's INCREDIBLY fast for something that just got released today!
Their site is sparse on images and the BT client is small, plus there is no dynamic content. They seem to be handling the load fine.
And the RedHat download - This is the fastest RedHat download I've EVER done... BT rocks.
Use the headless or curses versions.
:)
I'm using the curses BT client under screen. Started at around 20-30 kB/sec, now I'm at approx. 270 kB/sec, and uploading 35K/sec even without opening up those listening ports.
(Note: I did use --ip to tell it my real "outside world" IP instead of giving the system my internal one.)
I'll leave it running until I reboot the machine (Sometime this evening.) That'll likely be 5-6 hours of serving up the files after the d/l completes.
BitTorrent rocks!
I'm actually surprised you're not having problems with carts wearing out after 3-4 uses with the Lexmark as I was pretty sure they used a thermal inkjet technique just like HP's. (Essentially, a small portion of ink is vaporized and this drives the rest out the nozzle. HP and (I thought) Lexmark cartridges will start burning out those heating elements eventually. That's why HP and Lexmark carts have the nozzles built in to the cartridge.
Canon and Epson, on the other hand, use piezo-based technology. Except for potential clogging, their nozzle drivers will last forever. This is why Epson and Canon have the nozzles in the printer and use "dumb" ink tanks. (Which they can charge nearly as much as HP does for their cartridges, and make more profit because they're simpler.)
The only issue you might have with refilling Canon and Epson carts too often is that sometimes remnant ink can dry and that dried ink in the cart will alter the consistency of the refills.
Continuous ink systems remove this problem by ditching the sponge in the cartridge and replacing it by a tube going directly to the ink bottle.
The equation that has to be solved is floating-point.
The amount of data that has to be dealt with is just enough so that most consumer phones don't have enough CPU to perform the calculation. (This will change soon, already people are wondering if the pseudoranges on the upcoming Kyocera 7135 PalmOS phone will be somehow accessible to developers of applications on the Palm side of the phone.)
You will not ever be able to replace your current cell phone if you have one.
New phones are required to have GPS capability as part of E911 so that your location can be pinpointed by 911 operators.
Note: According to a few other posters, transmission of GPS information can be disabled for non-911 calls if the user desires.
"Kyocera, Samsung, Qualcomm"
Qualcomm no longer makes handsets. Their former handset division is now owned by Kyocera. (In fact, many of the first Kyocera phones bore Qualcomm markings in places, as they began their design phase at Qualcomm.)
The GPS included in many of these phones is not a complete GPS system.
Specifically, the phone doesn't have enough CPU to obtain a location fix on its own. All AGPS (Assiste GPS) location processing is done by the towers.
The advantage: It costs MUCH less to implement, since the phone doesn't need all that extra CPU.
Disadvantage: No user-accessible location fix.
Lack of such info is usually not much of a problem if you're moving. Most GPSes (at least Garmin ones) will orient the map based on your direction of travel if you are moving.
More like 1 of the 1 "major" cell carriers in the US uses CDMA.
"major" being defined as "Actually covers 100% of the most densely populated state in the country" - That state being New Jersey.
AT&T used to be in this category, but their GSM network has nothing on their old D-AMPS (The TDMA variant they used) network coverage-wise.
Every other carrier can only pull off 50-66% coverage.
BTW, that one lone carrier that covers all of New Jersey and is the only one that provides service more than two miles outside of the city limits of Ithaca, NY is Verizon.
GSM is not an option in many parts of the United States because the GSM carriers all have horrendous coverage. (And GSM doesn't seem to handle multiple simultaneous providers very well - I've seen countless horror stories of T-Mobile users being unable to use their phones once AT&T and Cingular rolled out GSM in their area.)
" But CDMA is a waste of fucking money in that region because ALL their neighbours use GSM. Eventually the GSM people (i.e. Vodaphone et al.) will move in and people will change to GSM anyway just so they can use their phones when away from their own country. Even in the States you can use a GSM phone. Whilst CDMA is a better, newer technology than GSM it's losing out to GSM in the world, and GSM has an upgrade path to next genereation stuff."
Who cares if the neighbors use GSM? 99% of the Iraqi population won't care because most of them won't ever need to use their phones in the country next door.
Iran? Fat chance, even after regime change there's too much bad blood between Iran and Iraq.
Turkey? Good luck getting through that border. Because of the situation with the Kurds, the Turkey/Iran border is going to be VERY tight.
Oh, GSM doesn't have a true upgrade path to 3G. UMTS requires new frequency bands, new handsets, and new antennas. CDMA2000 requires none of the above. (Well, you need a new handset if you want the additional 3G features, but a cdmaOne phone will happily talk to a CDMA2000 tower, it's just restricted to the old circuit-switched 14.4 data. An old GSM phone, on the other hand, CANNOT communicate with a UMTS tower.)
"So fine, install a hugely expensive CDMA network. The GSM people will move in anyway, people will buy GSM phones and your precious US network will go bust. That's the price for FORCING a choice on someone."
Oh, like Voicestream/T-Mobile? Or Cingular?
Yes, I'm sure the Iraqis will enjoy seeing their "Service Unavailable" messages. In every country where CDMA has been allowed to take on GSM in an open market, CDMA has won. The smallest CDMA provider in the US (Sprint) still has better coverage than Voicestream/T-Mobile or Cingular, the largest GSM providers in the US. AT&T may eventually become larger, but their GSM network can't compare in any way to their old TDMA network in coverage.
Or look at Japan. DoCoMo rolled out UMTS (3G GSM) while KDDI rolled out CDMA2000 1xEV-DO or EV-DV. Guess who has more subscribers for their 3G services? (Hint: Not DoCoMo).
If GSM is so much better than why can Verizon get away with charging so much more for their CDMA service? (Hint: You get what you pay for in terms of service quality.)
The only countries where GSM is dominant are ones where GSM was *forced on the customer by the government* - Exactly what you're arguing against.
If the Europeans want to build a GSM network in Iraq - Fine. Let them do so on their dime. But if it's on the USA's dime, it's going to be with the technology that won in the open market. GSM and CDMA can coexist in the same area, so I see no problem in the US installing CDMA and Europe installing GSM and seeing which one wins. If Europe wants to see GSM in Iraq - Pay up or shut up.
Or more appropriately, a variant of CDMA that no one has been able to build a handset for that had good battery life/didn't overheat. (Look at what happened to DoCoMo's name in Japan after they rolled out UMTS...)
I buy ink in bulk and refill the carts.
It comes out to even less than what you state, and the quality rocks.
Also, with laser printers you don't have the option of continuous ink systems, archival inks, or specialized multi-greyscale inks. (People sell cartridges that replace the color inks in Epson with three shades of grey - Which makes for extremely high-quality greyscale printing.)
Laser printers that can print color are:
a) Expensive
b) Horrible at the job. I've never seen a color laser printer that printed a photo that wasn't completely ugly.
HP is much better than Lexmark, but they still suck.
I've used HP, Lexmark, and Epson printers. We had an HP Deskjet 500 that is STILL working well today. It's well over a decade old, and built like a tank. (Newer HPs might not be so hot.)
I bought a Lexmark 5700 my freshman year in college. The Windows drivers were crap and there was zero Linux support. My parents had a 5700 that broke 5 days after the warranty expired.
We all have Epsons now. They work beautifully under Linux, are built pretty well, and easy to refill. (Note: Epson also is starting to use cartridges with chips, but these have already been "cracked" and were "cracked" quite some time ago.)
The only CIS (Continuous Ink Systems) I've seen have been for Epsons. I don't have one but for someone that prints a LOT they look cool. (Essentially the carts are replaced with tubes going to large bottles of ink that you can buy at bulk prices.)
Modems have been doing compression for years. As others have pointed out, no compression technique is going to help you very much on precompressed files (Images, video, etc. All of the "big stuff" that people buy broadband to receive.) For text it will be a benefit, but not one already offered by modems. (As someone pointed out though, many modems have a limit at the serial port of 115k, so compression on the PC side of this would be of use. Note that the PPP standard has support for some sort of compression, most ISPs don't use it. If you have a machine on the "outside world" with a broadband connection, SSH tunneling to a remote HTTP proxy, with SSH compression turned on will give you a speed boost.)
Caching - Many ISPs already do this, although few ISPs do. It's built into almost all browsers, and if you want "smarter" caching, install Squid.
Persistent connections have been part of HTTP 1.1 for ages.
In short: Nothing you can't already do now for free, but the ISPs are making it easier.
Restricting underclocking (if it happens) would most likely be done to push the mobile editions of their various processors.
No "quick and dirty" FSB-based underclocking of a desktop CPU in a notebook - Nope, gotta buy the expensive mobile version and use SpeedStep.
I wouldn't mind this so much, since desktop CPUs should never have been allowed in notebooks in the first place, but the alternative (SpeedStep) utterly sucks. Transmeta does clock/voltage ramping much better than Intel does. (SpeedStep only has "full tilt" and "slow", while Transmeta CPUs have everything in between.)
I had no problems with an AT&T 486SX-25 OCed to 33.
Intel just needs to add an instruction that returns the "true" clockspeed rating to a register.
Then any software under any OS can read it, even DOS.
And since it's just one piddly instruction, it would be a simple matter to create a "clockspeed check" utility.
A protection system that makes any system on the planet that gets overclocked scream bloody murder isn't necessary. All that is needed is for a reliable way for the user to download a trusted application, fire it up, and see whether their PC is clocked properly or not.
Damn this article... I just replayed RCR about a year ago, now I'm having urges to pull my Nintendo out of the basement again.
:)
River City Ransom - One of the best games EVER.
Yes, I agree with the article's author, we need a River City Ransom 2.
I can't live without my Acro Circus... That was one kickass skill.
Where's the "I forgot something" button.
The Arc AAA is an excellent light for those looking for something smaller (and less bright) than the Luxeon.
Luxeon Star-based flashlights (as mentioned earlier) are *amazing*. 1 watt and 5 watt Luxeons are available (although 5W whites are rare, but getting more common.) The 1W units are used in a number of lights such as the Arc LS (Expensive, but excellent) and the CMG Reactor (Don't get it, worst LS light out there). Another person mentioned EL's Blaster series, which are good lights. Also there's plenty of info on the forums linked from the article (candlepowerforums) on converting a 2D or 3D Maglight into a resistored LS light.
For small and relatively cheap, register on the forums and investigate the BadBoy and MadMax drop-ins for MiniMags. These "pills" replace the bulb in your minimag and include a 1W Luxeon on a DC/DC step-up converter. The MadMax is constant-voltage, the BB constant-current (the latter is the best way to drive an LED if you want consistent brightness.) They are both BRIGHT. My BadBoy 400 can tangle with a 2D Maglight and win, with a much whiter and smoother beam.
If you order a BadBoy, I suggest trying the McFlood optics, also from dat2zip's site. They give a nice smooth flood beam. For an intermediate beam, I have a 22.5mm diameter DCX lens from www.anchoroptical.com mounted in a MiniMag accessory holder in front of the McFlood, for a "tight flood" beam.
Nice thing about these "pills" is that if the body gets scuffed up, replacement bodies are only $7-8 each at any local store.
Unfortunately McLuxes are no longer in production - At $120 or so per light, these were quite expensive but they were VERY solidly built with an excellent design, and they were also one of the smallest LS lights out there.
"trickle charging" the laptop itself will not likely be possible - Almost all laptops expect full current for charging. (Even when turned off, my dad's Dell Inspiron 8000 would overload and shut down our old inverter, which was too cheap and too small.)
Trickling the batteries - Be VERY careful. Li-Ions cannot be constantly trickle charged. You SHOULD be safe giving the battery a low-current charge if you use a regulator set to 4.1 or 4.2v/cell (Depends on the exact battery chemistry, the type of electrode changes the nominal voltage slightly.) with a current-limiting resistor. But don't leave the battery in the charger, this will be bad.
The way Li-Ions are supposed to be charged is a current-limited charge, but with the charge voltage never exceeding 4.1 or 4.2 volts per cell. At this threshold, the charger must go into constant voltage mode. Once the current goes below a certain threshold, the charger should start a timer and then shut down at the end of 30-60 minutes. Trickle-charging it beyond that 30-60 minutes is BAD. NiMHs and NiCds can be trickle charged at low currents (C/16 or less, the lower the better for the battery's health), and lead-acids (gel cells are a variant of lead-acid) like having a "float" charge of a constant voltage (Around 13-14 volts for a 12v lead-acid battery I think) applied constantly or regularly. (Car alternators typically provide this charge in automotives.)
Note that even "deep cycle" lead-acids or gel cells should not be cycled deeply on a regular basis. lead-acids like to be topped off, as do those rechargable alkalines.
Problem is that the pinouts and connectors for almost all laptop batteries are nonstandard, and some expect various control lines to be connected while charging. These connections are never documented.
Nothing says that you have to inject the water after the fuel, nor do you have to inject the water in the same manner the fuel is injected.
Note that some of the older fuel injection designs, instead of having multiport injection (one injector/cylinder), they used throttle body injection. (One injector at the throttle body). You could just add a water injector at the throttle body. One guy did it on his Dodge Shadow (http://www.karlsnet.com/mopar/) - There's a lot of info not on that page, he was an active poster on the Allpar (http://www.allpar.com/) forums until he changed jobs and no longer had much time to experiment or post on the forums. Too bad, the guy had some great ideas and was very good at documenting his results, and open to suggestions/constructive criticism. (He had a few errors in some formulas and said, "oops, you're right" and subsequently made corrections on his page.)
Just one note: Do NOT put ethanol into your gas tank unless you know for a fact that your car is an FFV.
Ethanol is VERY corrosive. While the engine management systems of FFVs are often unchanged, anything that comes into contact with the fuel must often be redesigned with additional corrosion resistance. (Chrysler's FFV engines had some fancy corrosion-resistant coatings on their valves, for instance. They also had special fuel lines.)