Forgot your password?
typodupeerror

Comment: Re:Cold Fusion isn't like Perpetual Motion (Score 1) 975

No, constructing something impossible, like a perpetual motion machine, is impossible. Science quite often says "definitely" or "definitely not".

Constructing something highly unlikely but not provably impossible, like cold fusion, is highly unlikely, especially if you're doing stuff by accident instead of actually understanding theory, but what science says about cold fusion is "everybody assumed it wasn't possible, but somebody did it, and then we showed that what they did was bogus but interesting, so we're back to assuming it's probably not possible so you'd better do a really good job of explaining what you're doing if you want us to spend our time looking at it again."

Comment: Re:2G-wireless GPSs (re: rant.) (Score 1) 156

Oh, right, I never owned a 90s car :-) My wife's 1985 car lasted until 2001, my 1987 van lasted until 2012 (with one engine replacement), and I never played with the digital busses on either my 2012 car or my wife's 2001 car (which IIRC only had the dumber version of OBD, not the current CAN bus.) I suppose I should try that some time. Both of those cars have the electronic speedometer with analog readout you refer to.

Comment: Network Transparency Use Cases I use often (Score 1) 224

by billstewart (#48174325) Attached to: Lead Mir Developer: 'Mir More Relevant Than Wayland In Two Years'

Many of the remote management applications I used to do with X are being replaced by HTTP interfaces, optionally with AJAX, or sometimes REST APIs, but there are still a number that aren't.

  • - Remote logins to run Unix shell - xterm really rocks for this (these days usually with X over SSH), as long as I'm using a Linux desktop, and it's a bit lame but workable from a Windows desktop (usually using Xming.)
  • - Windows Remote Desktop Protocol remote logins to Windows servers - RDP mostly works, though running any sort of video over it is painful, and audio's worse, even over Ethernets. VNC would be similar, if I were talking to a Unix machine.
  • - VMware console access to Unix or sometimes Windows VMs - the biggest problem here is that too many different applications are carving away chunks of the screen, so there's only 800x600 or maybe 1024x768 left; and I'm usually accessing the VMware server from a web browser on a Windows machine accessed by RDP through a firewall, so there's no real way to compare video speeds or get audio.

Comment: Supporting Mobile Devices on X or NeWS (Score 1) 224

by billstewart (#48174255) Attached to: Lead Mir Developer: 'Mir More Relevant Than Wayland In Two Years'

We definitely have to dumb down protocols to run on mobile phones with 1024x768 screens and only 1 GHz CPUs and 1GB RAM, because they can't possibly run anywhere near as fast as they did on 1152x900 screens with 10 MHz CPUs and 4 MB RAM running X10, or 640x480 screens with 25-33 MHz 386 CPUs and (I forget how much, but not enough) RAM and X11 with Motif.

And yes, there were really good reasons for running NeWS instead of X, because some changes in which work you did on which end of the wire could make a huge difference in responsiveness and speed, and using Postscript meant that what you saw was really what you'd get, and it let you deal with problems like mouse tracking latency a lot better.

Comment: Re:Auto deals worried? Luxury Cars, biz models (Score 1) 156

No, besides disliking competition in general, auto dealers and car makers have two big reasons to try to block Tesla sales

  • - Tesla sells high-priced cars, competing with the other high-priced high-profit-margin cars that dealers like to sell. They wouldn't be as worried about threats to the low-profit-margin econobox sales.
  • - They also threaten the whole business model that US car dealers have, affecting who gets what cut of the car buyer's money. This may bother the manufacturers less than the dealer, but it still upsets the whole value chain, especially if those evil Tesla car buyers then resell their old cars on Craigslist or some other Internet site instead of trading it in at the dealer or at least selling it to a used car lot.

Comment: Re:Too many discrete components (Score 1) 156

I'd expect most of them are sensors for the various battery and motor things, or components to connect the sensors safely to the other electronics (opto-isolators, etc. to keep potentially high random voltages and currents from frying the whole system.) Once you've turned the analog data into bits, even with small-volume production it'd be fairly easy to use an FPGA or programmable microcontrollers to do the rest, rather than building lots of custom discrete parts.

Comment: 2G-wireless GPSs (re: rant.) (Score 1) 156

My Garmin Nuvi had some cool features that depended on 2G, like using Google search instead of just built-in, and also checking movies, weather, etc. It also used that to get traffic data, instead of whatever other traffic data services are available. Now the 2G wireless is going away, since the carrier won't renew the contract, so there's no more traffic data :-( But at least it's a separate GPS, so I could replace it if I wanted to. (Instead, I use the AM radio you dislike to listen to Traffic Every 10 Minutes Radio.)

When the satellite XM radio free-with-new-car subscription on my current car ran out, no problem, that just meant there was one button on the dashboard that was no longer useful; the most likely interface to become obsolete is the Bluetooth cellphone support. There'd be a lot more risk of obsolescence if I'd gotten the hopelessly-overpriced navigation/radio/etc. console only that came with the fancy trim package (which also had the bigger engine that I didn't want, and the spare tire I really did need, and pushbutton combination door lock I'd also have liked.) While I like having a remote-control door lock, which is probably already insecure, it's built in to the keys, which means I have to carry a big clunky not-waterproof key system with me instead of a probably-waterproof slightly clunky RFID key like my wife's car has or a simple key like older cars - really annoying when I'm going surfing.

Digital speedometers might be lying; analog speedometers also might be lying, especially if there's a problem with the cable, or you've put on different sized tires.

Comment: Santa Ana's losing war (Score 1) 323

by billstewart (#48147555) Attached to: How English Beat German As the Language of Science

Mexican history considers Santa Ana to be one of their least competent generals over the years. But after he lost, the US rejected his initial peace proposals, which would not only have given the invaders the land they got, but also the states of Sonora and Chihuahua. While they were officially Mexican territory at the time, the local Native Americans had other opinions about whether they were interested in being run by the Spanish or Mexican colonialists, and groups like the Apaches and Comanches wouldn't have been much more cooperative to the US than they were to the Mexicans.

Comment: Re:German illegal? (Score 2) 323

by billstewart (#48147543) Attached to: How English Beat German As the Language of Science

Any argument that only has two sides is a boring and overly limited view of reality. Texas was part of Mexico at the time, and there were a bunch of illegal immigrant gringos who came in and wanted to be able to own slaves.

But separately from that, during the various wars in Europe in the early-mid 1800s, there were a lot of German immigrants who moved to Texas and the rest of Northern Mexico. The Texas German dialect is dying out, but you'll still see a lot of German culture in places like New Braunfels, some of the cooking in San Antonio, and a lot of mariachi music is strongly German oompah-band stuff.

Comment: Cold Fusion isn't like Perpetual Motion (Score 1) 975

Perpetual Motion violates the laws of physics - can't be done, so any patent application is bogus, either wrong or fraudulent, not worth wasting time on.*

Cold Fusion might or might not be possible - the scientific community at large hasn't seen a valid description of the physics or chemistry, and without somebody understanding the science, it's extremely unlikely that they'll engineer a successful implementation by tinkering around, and unlikely that somebody who's keeping the science a "trade secret" has actually done real science, as opposed to waving their hands around in ways that seem pleasing to their scientifically untrained eye, and the mere fact that they haven't blown themselves up isn't proof that it works.

* ("Free energy" is a different case - it usually refers to quackery, but sometimes is used to refer to things like taking advantage of heat differences in the ocean or earth or other things that you might be able to engineer usefully into a long-term economically viable power source, but probably can't.)

Comment: RTFA, and articles like it (Score 1) 717

by billstewart (#48126187) Attached to: Why the Trolls Will Always Win

I've read too many articles like this recently to keep track of who said what, but one of them pointed out that women especially get attacked by trolls when they're starting to become well-known and people are listening to their opinions. Kathy Sierra, for instance, started the Head First line of programming books, which I found useful, and got enough sexist trolling that she left the business. It's happened to other authors I know as well. And of course there are the trolls who hate having women in gaming.

Comment: Presbyopia (Score 1) 155

by billstewart (#48126137) Attached to: Liking Analog Meters Doesn't Make You a Luddite (Video)

As a guy in my 50s who now needs reading glasses, I'm finding digital displays increasingly frustrating, especially small cheap LCDs that aren't very distinct unless you're looking at them from straight on. I much prefer digital clocks, but if you can't tell a 3 from an 8 or 0 or a 1 from a 7, they're not very useful, and it's almost always easy to tell the big hand from the little hand. Similarly with digital meters, big numbers with fat segments are still easy to read, small skinny ones aren't.

"We are on the verge: Today our program proved Fermat's next-to-last theorem." -- Epigrams in Programming, ACM SIGPLAN Sept. 1982

Working...