We develop sales automation tools for handheld pcs. we couldn't do that on a PDA. there is no keyboard. how can a sales person lookup a customer / part, or enter an invoice / order with a stylus touch pad. answer, you can't you need a keyboard.
I disagree. Sales should have next to zero use for a keyboard. Product lines, customers, price lists... should all be selectable WITHOUT a keyboard. We're designing internal applications for our sales force using Palm PDAs. The sales force will have a PDA and a portable DVD player.
Product demo videos will be on the DVD players and they can give out CDs with documentation and other goodies. The Palm handlelds will have a few apps for quoting and the standard email, time tracking, contact management and datebook applications available.
We used to give them laptops which had to be upgraded every year or so, were prime targets for theft, crashed and were generally used to play games instead of sell. Now we're replacing them for less money, less space and 10x the effectiveness. Throw in a folding keyboard for writing[1] faxes/emails and a modem and they have everything they need.
[1] - yes, the odd time a keyboard is needed but it's very infrequent.
I agree that certain types of Bots will usually beat other types;
Yes, but that's not my complaint. My complaint is that if you build a bot capable of EMP it will always win, barring its own mechanical/electrical failure. You're no longer battling in the sense of "true" battle -- it's an unfair advantage. In fact, I believe that EMP armed bots are already disallowed.
(I also feel that the wedge-only bots are really lame. No offensive capability by anyones definition, no fight, no fun....)
I agree. This is the basis for my disapproval for the super-offensive bots: You pit bots of close calibre together for a real fight. It's no fun watching some $8000 superBattleBot go against a $400 robot built by a parent for his kids so they can get on the show.
but how could you realistically define multiple types of competition in the same weight class? Anyone could enter and win by default if they were the only stomp-bot with six alternating hammers, gas drive, CO2 flipping-arm-equipped, kevlar _AND_ lexan _AND_ titanium shielded wedge with battery backup.
My point exactly. And this is exactly what this battlebot is: A super-engineered robot killer. It should only be pitted against other super-engineered robot killers!
They aready have weight classes; why not damage classes within the weight classes, and then the winners from the various damage classes battle each other regardless of weight? I think that would be far more fair and (at least to me) far more entertaining. Yes it'd be more difficult to classify but then again that is why you have your weight classes and experience levels win martial arts (at least I believe this is so) -- the lighter, more experienced guy is still capable of trashing the heavy, inexperienced martial artist.
Those earlier vehicles with just two opposite rotating wheels that could only spin (and not laterally move) really sucked. They just sat there and waited for the other guy to come over and commit suicide.
You don't think this is just the same idea?
Personally these types of bots should be in a separate class -- there's no real point to trying to battle something with an impenetrable defense shield whirring around it waiting to hit anything with 10kf/p of torque. I'd classify this bot as purely defense. It can run around and defend. There's no real offensive tactic involved (spin up the rotor and attack? Please...)
I agree with some of the other posters -- the UK television show where the robots had to have good offense, defense and manouvre abilities was best. This show is simply "make it last and try to hurt the other guy" -- bo-ring.
Wow, talk about a technical response! Thank you! Please see my previous comment which talks about what I meant by being "in favour" of the technology.
There is only so much spectrum. The faster you change the signal, the broader the chunk of spectrum you use.
Yes of course. This is what I had meant. My thought process was "It doesn't use a specific chunk of the spectrum" rather than "It doesn't take up any space in the spectrum." I apologize for the confusion.
Want to send data a hundred miles? Imagine you were doing this with VISIBLE light. You're modulating an arc light at your transmitter, bright enough that the output of a solar cell a hundred miles away has more signal from your arc light than from all the other light sources (including any other arc lights) combined. Now imagine somebody on your block trying to read morse-code flashes from a distant colored lightbulb, using his own solar cell and a color filter.
This is actually a bad example as light is very directional and unless that arc light is pointed to your neighbour, he should be able to see his coloured light. (perhaps in a haze of white light.)
Run a protocol so the radio versions of your "arc lights" take turns and you can run a network. (Think of running "Ethernet" in the real aether.) But that's ALL you'll be able to run. Goodbye TV, goodbye AM and FM radio, goodbye aircraft band, police band, CB,...
What about using this technology for all communications. Pie in the sky, yes definately... but would it work any better?
Time-domain and frequency-domain signals don't play well together. And the time-domain kids got to the playground first. Do you want ANOTHER war with broadcast media? Remember, if they lose it's their death, so they'll fight REAL HARD.
You mean Frequency Domain, no?:-) But yes, I see what you are saying and this is the single biggest threat to UWB -- being shut down because of the interference.
Each one of those "arc lights" raises the noise floor - by a bunch. More noise means you can't measure the signal from the desired "arc light" as accurately - which means you get less bits-per-second from it.
As I'd stated in the message I posted at the start of this reply, I was only in favour for low-power applications: PANs basically. I don't seriously think you can get an antenna with enough dynamic range to effictively limit the splatter of a long-range signal. As you said, you can have speed, but it costs bandwidth. And wide-bandwidth antennae are hard to design.
You sacrifice a part of your bandwidth to send a pre-defined, typically repetitive, synchronization signal, to keep the receiver synchronized with the transmitter. Think of the start/stop bits on a serial line, the framing bits in T1, T3, or SONET, or the vertical and horizontal sync bars in a TV signal.
Ahh yes, this would work perfectly. I had not thought of that... (the most obvious answer of course.)
Thanks for that lengthy response. I see from my comment history that I did not make myself clear enough; my mistake but I am sort of glad I did. I got some really great responses.:-)
Let me get this straight. You "don't mind raising the noise floor for something like this," when it would be affecting GPS, public safety nets, air traffic, marine navigation and communications?
Don't jump to conclusions, you may hurt yourself.
What I was wanting to say (and I am thankful for the number of technical responses) was that I am in favour of UWB for short-range communications. I do not like the idea of spewing out tens of watts (or even watts) of noise. I was thinking more along the lines of replacing the sub-watt communications systems used today for things like cordless phones, wireless mics, R/C cars, wireless keyboards/mice, 802.11[ab] and Bluetooth. I'm very much in favour of freeing up chunks of the spectrum and instead allocating them to better causes or for rather to communications links which can better exist in that spectrum. Yes UWB will raise the noise floor but again, in the miliwatt range and below I think this would be acceptable, but I have not done the studies to verify it.
As far as UWB being used for high speed long range links, I would imagine that the pilot of the Cessna would not be flying across the well-documented and regulated link, and that such a link would use super-directional antennas to avoid the amount of splatter.
Where do they get these guys? First he says that it doesn't use any spectrum...then he says that anything below 2 GHz will interfere with existing Nav and Comm systems. Gotta be one or the other. Can't be both.
Yes, it can be both
UWB works by sending single-cycle pulses. The information is carried by when the pulse is transmitted with respect to a reference.
Since there is no carrier, it doesn't affect a specific part of the spectrum. However, since there is no carrier, it affects all parts of the spectrum by adding to the noise floor. That is what the big problem with this technology is and why the FCC is looking so closely at it. The UWB Consortium has more information.
Personally I don't see a problem with raising the noise floor for this technology because, as I understand it, it raises the floor uniformly and, if I understand this correctly, the actual number of devices transmitting doesn't play into this.
I've been interested in this for a while. Time Domain (warning, flash-heavy site) is a company which has been playing with this for a long long time. I was rather skeptical of this when I first heard of it but my opinions on it are changing. Hell even EDN had an article on it.
The only thing I don't quite grok is how they can get two devices to have such rock-solid stable time references (we're talking sub-picosecond jitter) without secondary clock transmitters and keep them that way. If anyone out there can help shed some light on it I'd love to hear from you.
I too used to be one who ALWAYS clicked custom and always made sure programs went into Program Files, etc, etc. But some vendors still haven't caught on and I've run into too many troubles not allowing a program to install where it normally wants to.
I am one of those people too, but I find the problem is a lot LESS of a problem on Linux. Linux has softlinks and environment variables that the linker listens to, even if the program doesn't.
So with Linux, I've started a shift to letting stuff install where it wants to - I've found it reduces trouble down teh road.
Heh. I install everything that didn't come with my distro (Slackware) to/opt. That means VMWare is in/opt/vmware (I don't want it in the path). Acroread and CVS KDE are in/opt (where I have he paths modified so/opt/bin,/opt/lib and opt/man are part of the whole process). My MPEG tools are in/opt. Same with XMMS. Hell even my uClinux build tree is in/opt!
That way, I back up/opt and my home directories and I have very little to do if something destoroys my install.
Mind you, I also have a/data directory where all the tarballs go, and off of that directories for mp3s, avis and a documents directory for anything I work on. Backups and resotres are a snap. I found that Windows software tends to break the "My DOcuments" rule, sticking edited files in the Program Files directory and causing major trouble for backing up. Hell even Outlook Express does this (wtf, putting my amil under the windows subdirectory?!)
Re:looking forward to the russian response...
on
Sklyarov Indicted
·
· Score: 2
I seem to remeber a guy getting caned in China (or was it Hong Kong?) for breaking a law. I also seem to remeber the US population screaming and whining about the brutality and unfairness.
The little shit who did that committed the crime while in the country he was arrested in. Dmitry did nothing wrong while in the U.S.
Therefore I doubt Russia will raise a fuss, and I don't think Sklyarov should be set free. you don't break a country's laws (no matter how stupid), go to that country and expect not to get arrested.
See my comment to your first paragraph. If Dmitry did something wrong while in the U.S. (and I don't think giving a speech is illegal yet) then sure, he will have to be punished under U.S. Law. AFAIK, he did not write the eBook decryptor in the U.S.; he did that in Russia.
The strap provides good, continuous protection. The problem with holding your hand on the case is that 1) your hand -- especially the pads of your fingers and the palm -- is not a good conductor, and 2) it takes only a few milliseconds to build up enough charge to zap a component (just lifting your shoe off the carpet can build up a destructive charge). I consider a grounded strap essential when you've got your hands in the guts of a piece of equipment.
If you are having that kind of static buildup I would seriously reconsider the environment these things are in... That's insane. There's also the fact that these straps
aren't a direct ground for simple safety: they usually include anywhere from a 1M to 10Mohm resistor in series. Even if my fingertips were especially dry, say 100M or even 1000M, there is still a leakage path which static could flow from.
This is all inconsequential, of course, we're arguing after the fact.:-) I don't know exactly the type of cases you're talking about nor the environment. For all I know they could be ceramic cases in a wood kiln or something.
That doesn't help when you're holding the board by its edges, and maybe accidentally touching chips in the middle of the board.
I apologize for not making myself clear: The techs were pissed that they had to wear straps when connecting to ports on installed cards. That is what I was referring to.
The thing is that we don't know that the grounds are at equal potential, and in fact if the handheld is charged, we know that they are NOT at equal potential. Thus to prevent damage, we want the potential to change slowly -- something you accomplish with a lot of resistance.
So you're not suggesting that the resistance remain throughout the connection, only at the start. Ok. I had misread your original intent.
you did nothing to protect them... you're a pretty shitty designer,
err, read my post - I stated I didn't do the original design, it was inherited. and later in my post I speak the praises of using ESD and surge protection in a circuit design - but thanks for the description of my abilities all the same.
Thanks for chopping my comment in half. I did say "you're a shitty designer, or you had a shitty management (financial) decision laid down on you. Truth is that anyone who doesn't design with this in mind generally is a poor designer. Whether that be from inexperience or negligence doesn't matter.
But we also provided shoe straps (slip it on in the morning - take off when you leave) which wern't too bad.
This is a good solution too. What is the difference between a static strap and touching the grounded case of the unit you're about to work on? I can't see any, and from personal experience I'm far more apt to touch the case than I am to make sure I have the strap on and that the strap is grounded.
And the bottom line is - tough shit. Your fooling around with millions of dollars of corporate hardware and you work for the company - deal with it or take a walk/find new job.
I agree with you that static protection is necessary -- don't get me wrong. Where I was digging in, however, was with the choice of static protection you had offered. Now of course, the ionizing air filters and so on weren't always around and neither were the shoe straps / antistatic shoes, but I was more or less taking the side of the techs who found the straps a pain in the ass. I feel I have a better solution (touching the case) and it seems to work for me. It doesn't work for everyone and it likely won't fly on an ISO sheet. You and I both know how much ISO compliance is worth though, so long as it appears that you're following it.:-)
I can't tell you how nervous my HW techs got holding processor cards worth hundreds of thousands of dollars - its scary!
I can imagine... but really why weren't these super-duper-hyper-expensive cards built with some form of ESD protection (hell for that price it could have been DAMN GOOD protection) on the external ports?
I'm still amazed that Mobo makers don't toss in dispoable straps with their retail boxed boards - hell they only cost 50 cents or so.
Pure economics. $0.50 on a strap is $0.50 less in profit per mobo. It's far cheaper to have the warning in the book. (I don't agree that it's right, just that that's the way it is.)
got news for you, NONE of these chips have ESD hardening.
If you used which which weren't able to withstand ESD and you did nothing to protect them... you're a pretty shitty designer, or you had a shitty management (financial) decision laid down on you. It's not hard to put a 10 ohm resistor in series with the input pins, throw on a bit of a filter and finish it off with some good fast 1W transzorbs. Takes care of ESD and EMI/RFI.
As for static straps - its amazing how people hate them so. I managed a 10k sq ft data center with almost 700 servers, from small $5000 machines to monster Auspex boxes costing millions.
Static straps and other such protective measures are a pain in the ass. Get used to it. It probably would have been better for everyone if you just used ionizing air filters and kept the air relatively humid.
The worst were the Sysadmins - they figured since they didn't touch teh cards themselves it was OK, yet they were plugging serial cables into exposed serial ports to hook up root terminals (before we had a networked root term setup) It was amazing the resistance I encountered for such a simple thing.
Whenever I have to touch a computer, I make special note to touch the case often. It may not drain every last volt of potential difference from the case and I, but I've never had troubles. I'd bitch and complain if I were a sysadmin too. Or rather I would touch the case and the connector before plugging anything in.
The real problem here is that nearly every integrated mobo has the serial port contained in the northbridge/southbridge chipset, so discharge to the port means discharge to a critical IC in the computer as well.
I doubt that it fried anything bigger than a TTL<-->RS232 converter IC. The reason? The designs of the large ICs don't work well when designing conversion circuitry. Sure, the UART will be part of the chipset but the signals on the I/O will be TTL or CMOS level outputs.
Those signals will then hit an IC like the MAX232 (An RS232 converter IC from Maxim Electronics) which contains the charge pumps and converts that +5/GND (or +3/GND) signal to +12/-12V required to meet the RS232 spec and back. Chipmakers like Maxim also make static-protected versions as well. (In Maxim's case, they usually designate ESD-protected devices with an E suffix.) These chips are good for a 11kV direct zap using the human body model.
No, I don't work for Maxim. Burr-Brown (Now part of TI), National Semiconductor (now spun off to Fairchild) and a host of others make these chips. I'm just most familliar with Maxim's.
If the motherboard fried, they used substandard (IMO) converter ICs. I've hit my laptop and several PCs very hard with ESD and I've yet to have a problem. The biggest problem is that ESD is a slow killer. Rarely does it fry something outright. Usually it just weakens the oxide layer on the semiconductor FETs and causes early death and spurious operation.
Palm should be no more liable for this than every company which manufactures serial, parallel, USB, FireWire and really ANY external device. If the guy bought a shit motherboard, deal with it. It's not Palm's fault.
Careful mechanical design to further reduce the problem - arrange that the "grounds" always touch first - preferably through a few hundred thousand ohms of resistance.
Ummm... how the hell is ground to be effective with that kind of impedance? Ground potential should be equal.
Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof;
The unclear part is how to apply this when the each of these two adjacent sentence clauses (no establishment/no prohibition) contradict each other.
It is unclear whether providing overhead (electricity, land, janitorial services etc) support for people egaged in religious activity constitutes a move, however tiny, towards the establishment of religion. Neither is it perfectly clear that forbidding equally all such support to all such groups effectively prohibits their right to excercise freedom of religion.
This is the old "reducing to absurdity" logical fallacy. You could argue that the use of a state park or parking lot constitues a move, however tiny, to the establishment of religion.
There are times to nitpick and times not to. This is (to me) clearly a time not to do so. You take the law in the context it was meant to be in, instead of breaking it apart to this level. I believe there is a judicial term for this but I haven't got a clue what it is.
Their claims were that they could put their units in the power distribution stations and then everyone would be able to buy little boxes to plug into their outlets that would be able to recieve and transmit data. They claimed data transmission rates in the gigabit range.
Yeah that's them. They claimed to be able to "tunnel" through the magnetic field caused my electrical current flow with a maser. I downloaded their patent but couldn't make sense of it.
Why not turn bridging on in the kernel and bridge to an internally connected 802.11b card.
Because I bought Orinoco Wavelan cards and the bridging firmware is sent to the card by the access point at every boot up.:-(
Mind you, I've got a logic analyzer with enough lines to handle a 16-bit PCMCIA bus, I just have to beg/borrow/steal an access point and spend the next few weeks untangling the init code to see what they send so I can tell Linux to do the same.:-)
Are all cards that use the same chipset as the WaveLan cards (Hermes 2 I think?) incapable of bridging without the firmware? If not, I'll just buy a cheap one and use it instead, or see if I can't do a dump of the firmware and muck with sending it to the Wavelan.:-)
They don't get around the problem of high frequencies being attenuated by the pole transformers, they simply don't worry about it because they have their CuPlus units sitting up there too.
Deployment of this technology will be far slower than DSL or cable, simply because every neighbourhood (or fraction of a neighbourhood) requires a CuPlus unit in order for this to work. That's one CuPlus unit for every pole transformer!
In my area, I see a pole transformer service at most two dozen households. The deployment for this technology will be insanely long.
The other question I have which isn't answered is how the RuPlus or NtPlus units can be plugged into any outlet and work when the split-phase wiring in North America effectively isolates the two 120VAC circuits from each other? Your pole transformer usually takes a 1200-13kV line and neutral and then gives you a center-tapped, 240VAC output (3 wires) to your house. Do they have the equivalent of a hub and two modems to distribute the incoming (to your house) signal across both phases and take the possibly distinct return signals, figure them out and put them back on to the main communications bus?
Whatever happened to those guys who claimed to be using a maser to modulate the magnetic field, thereby defeating Maxwell's equations and getting high frequency data through the 60Hz transformers? I'd love to see a working demo.
Get your stuff from Digikey [digikey.com] for quality (and inexpensive) electronics.
<cough> did you say cheap and digikey in the same sentence?!!
DigiKey is known as the Radio Shack of the professional electronics industry. Sure, they can get you pretty much anything and fast, but they are not cheap by any means, even when you look over on the table to the bulk prices.
You're far better to go to Future-Active, Arrow or even Farnell for your parts. Digikey is quick and fast, and their quality is the same as anyone else. But they are NOT inexpensive for anything. You're paying a premium to have them ship 16 identical catalogs to your shop every quarter.
We use the megabit modems extensively for our business DSL deployment (no DSL offered from the telco, so we do it the hard way). When we started we couldn't justify buying a DSLAM so we just hooked them up back-to-back as described in Cringely's article and as long as you have the rate set the same on both ends, they just work. No, they're not RADSL and personally, I prefer that.
We've had zero trouble with these units, having installed about a dozen or so over the past 3 years. Great for businesses who KNOW they want on the 'net at high speed, but for a personal connection or as a trial they're a bit pricey. That's why I've been working on some alternatives.
I've just purchased a pair of Efficient Networks 5250 SDSL bridges. They don't specifically state that they'll work back-to-back but after some research and initial legwork I think they'll work just fine as a cheap alternative. They can be had for USD$50 from Ebay.
Pairgain stuff has the longest "reach" of all the DSL equipment we've investigated but they are also one of the more expensive ones out there. I suppose you get what you pay for.:-)
Cough %$Bullshit$% a DIVx rip from dvd under the mystical 700mb cd limit or thereabouts for a 2 hour flick has more artifacts than King Tut's Funeral Chamber.
Without postprocessing, you're 100% correct. I use mplayer with the opendivx libraries. On a dual celeron 466 and no postprocessing, it is what I would consider the digital equivalent to an EP recording of a copy. However with the postprocessing set to its highest level (4 for DIVx) it is wonderful. Of course, my system isn't good enough to handle this so the audio gets out of sync with the video very quickly.:-)
System: Abit BP6 (Dual Celeron 466), 256M RAM, GeForce2MX-400 (32M). Yes this is in fullscreen under X 4.0.3 with the latest nVidia drivers (1251 I believe). The movie: Varsity Blues. Filesize: 629441536.
Who will pay pay slightly less and $500 for a TV out card to get a shithouse quality movie(paticualry after going though a consumer level TVOUT card) on your 4foot plasma?
You're targetting the wrong demographic.
Think any TV-out card (nVidia, ATI, etc.) going to a standard TV. The type of people who buy a $500 TV-out card and have a plasma television are also the type of people who use component video and $500+ DVD players.
I think the type of people to download a 500M movie are the same type of people who download DivX movies from eDonkey and burn them to CD anyway... quality should be about the same and if you have a ballsy-enough computer to do all the postprocessing it looks like a good VHS tape on a normal TV. Hell it looks like a DVD to me on my 17" monitor but I end up dropping frames since my computer isn't ballsy enough.:-(
The reality is that the presence of RI helps sophisticated query rewrite engines to do some very nifty transformations. So having RI actually helps you get better performing plans for read queries and hurts performance a little for writes.
I totally agree. If the backend knows that it can't compromise integrity though optimization (i.e. an optimization trial fails to maintain RI, etc.) then it also knows that when such tests don't fail, it has an optimized access method which kicks ass.
(ie: Linux SMP box hanging under high network load (which makes backups a real bitch), forcing me to power cycle : flawed APIC handling for the 3c905 ethernet card),
Out of curiousity, did you find a fix for this? I think that may be explaining the odd lockup I get on my system that I haven't bene able to pin down...
Of course, MySQL has other drawbacks, namely that it doesn't support triggers or table inheritance or some of the more complex nuances of standard SQL, but the 95% of stuff it does have is very fast, and the other 5% can be handled in code.
The instant you decide to move data and/or referential integrity from the DB into 'code', you've lost the battle. I can't believe someone would even suggest this. Sure, stored procedures are a plus and most times they aren't necessary, but you simply cannot have your integrity checks outside the DB. That's the whole god-damned point of an RDBMS!
If your integrity depends on unusual interactions in the data store, stored procedures are often the only way out. And you can't have stored procs that work to enforce integrity without triggers.
If you haven't got the integrity, you may as well be using a hashed filesystem to store your data for all the difference it would make. Hell, the hashed fs would probably be faster since it isn't pretending to be an RDBMS.
Slashdot Mirror
We develop sales automation tools for handheld pcs. we couldn't do that on a PDA. there is no keyboard. how can a sales person lookup a customer / part, or enter an invoice / order with a stylus touch pad. answer, you can't you need a keyboard.
I disagree. Sales should have next to zero use for a keyboard. Product lines, customers, price lists... should all be selectable WITHOUT a keyboard. We're designing internal applications for our sales force using Palm PDAs. The sales force will have a PDA and a portable DVD player.
Product demo videos will be on the DVD players and they can give out CDs with documentation and other goodies. The Palm handlelds will have a few apps for quoting and the standard email, time tracking, contact management and datebook applications available.
We used to give them laptops which had to be upgraded every year or so, were prime targets for theft, crashed and were generally used to play games instead of sell. Now we're replacing them for less money, less space and 10x the effectiveness. Throw in a folding keyboard for writing[1] faxes/emails and a modem and they have everything they need.
[1] - yes, the odd time a keyboard is needed but it's very infrequent.
I agree that certain types of Bots will usually beat other types;
Yes, but that's not my complaint. My complaint is that if you build a bot capable of EMP it will always win, barring its own mechanical/electrical failure. You're no longer battling in the sense of "true" battle -- it's an unfair advantage. In fact, I believe that EMP armed bots are already disallowed.
(I also feel that the wedge-only bots are really lame. No offensive capability by anyones definition, no fight, no fun....)
I agree. This is the basis for my disapproval for the super-offensive bots: You pit bots of close calibre together for a real fight. It's no fun watching some $8000 superBattleBot go against a $400 robot built by a parent for his kids so they can get on the show.
but how could you realistically define multiple types of competition in the same weight class? Anyone could enter and win by default if they were the only stomp-bot with six alternating hammers, gas drive, CO2 flipping-arm-equipped, kevlar _AND_ lexan _AND_ titanium shielded wedge with battery backup.
My point exactly. And this is exactly what this battlebot is: A super-engineered robot killer. It should only be pitted against other super-engineered robot killers!
They aready have weight classes; why not damage classes within the weight classes, and then the winners from the various damage classes battle each other regardless of weight? I think that would be far more fair and (at least to me) far more entertaining. Yes it'd be more difficult to classify but then again that is why you have your weight classes and experience levels win martial arts (at least I believe this is so) -- the lighter, more experienced guy is still capable of trashing the heavy, inexperienced martial artist.
Those earlier vehicles with just two opposite rotating wheels that could only spin (and not laterally move) really sucked. They just sat there and waited for the other guy to come over and commit suicide.
You don't think this is just the same idea?
Personally these types of bots should be in a separate class -- there's no real point to trying to battle something with an impenetrable defense shield whirring around it waiting to hit anything with 10kf/p of torque. I'd classify this bot as purely defense. It can run around and defend. There's no real offensive tactic involved (spin up the rotor and attack? Please...)
I agree with some of the other posters -- the UK television show where the robots had to have good offense, defense and manouvre abilities was best. This show is simply "make it last and try to hurt the other guy" -- bo-ring.
Wow, talk about a technical response! Thank you! Please see my previous comment which talks about what I meant by being "in favour" of the technology.
There is only so much spectrum. The faster you change the signal, the broader the chunk of spectrum you use.
Yes of course. This is what I had meant. My thought process was "It doesn't use a specific chunk of the spectrum" rather than "It doesn't take up any space in the spectrum." I apologize for the confusion.
Want to send data a hundred miles? Imagine you were doing this with VISIBLE light. You're modulating an arc light at your transmitter, bright enough that the output of a solar cell a hundred miles away has more signal from your arc light than from all the other light sources (including any other arc lights) combined. Now imagine somebody on your block trying to read morse-code flashes from a distant colored lightbulb, using his own solar cell and a color filter.
This is actually a bad example as light is very directional and unless that arc light is pointed to your neighbour, he should be able to see his coloured light. (perhaps in a haze of white light.)
Run a protocol so the radio versions of your "arc lights" take turns and you can run a network. (Think of running "Ethernet" in the real aether.) But that's ALL you'll be able to run. Goodbye TV, goodbye AM and FM radio, goodbye aircraft band, police band, CB, ...
What about using this technology for all communications. Pie in the sky, yes definately... but would it work any better?
Time-domain and frequency-domain signals don't play well together. And the time-domain kids got to the playground first. Do you want ANOTHER war with broadcast media? Remember, if they lose it's their death, so they'll fight REAL HARD.
You mean Frequency Domain, no? :-) But yes, I see what you are saying and this is the single biggest threat to UWB -- being shut down because of the interference.
Each one of those "arc lights" raises the noise floor - by a bunch. More noise means you can't measure the signal from the desired "arc light" as accurately - which means you get less bits-per-second from it.
As I'd stated in the message I posted at the start of this reply, I was only in favour for low-power applications: PANs basically. I don't seriously think you can get an antenna with enough dynamic range to effictively limit the splatter of a long-range signal. As you said, you can have speed, but it costs bandwidth. And wide-bandwidth antennae are hard to design.
You sacrifice a part of your bandwidth to send a pre-defined, typically repetitive, synchronization signal, to keep the receiver synchronized with the transmitter. Think of the start/stop bits on a serial line, the framing bits in T1, T3, or SONET, or the vertical and horizontal sync bars in a TV signal.
Ahh yes, this would work perfectly. I had not thought of that... (the most obvious answer of course.)
Thanks for that lengthy response. I see from my comment history that I did not make myself clear enough; my mistake but I am sort of glad I did. I got some really great responses. :-)
Let me get this straight. You "don't mind raising the noise floor for something like this," when it would be affecting GPS, public safety nets, air traffic, marine navigation and communications?
Don't jump to conclusions, you may hurt yourself.
What I was wanting to say (and I am thankful for the number of technical responses) was that I am in favour of UWB for short-range communications. I do not like the idea of spewing out tens of watts (or even watts) of noise. I was thinking more along the lines of replacing the sub-watt communications systems used today for things like cordless phones, wireless mics, R/C cars, wireless keyboards/mice, 802.11[ab] and Bluetooth. I'm very much in favour of freeing up chunks of the spectrum and instead allocating them to better causes or for rather to communications links which can better exist in that spectrum. Yes UWB will raise the noise floor but again, in the miliwatt range and below I think this would be acceptable, but I have not done the studies to verify it.
As far as UWB being used for high speed long range links, I would imagine that the pilot of the Cessna would not be flying across the well-documented and regulated link, and that such a link would use super-directional antennas to avoid the amount of splatter.
Where do they get these guys? First he says that it doesn't use any spectrum...then he says that anything below 2 GHz will interfere with existing Nav and Comm systems. Gotta be one or the other. Can't be both.
Yes, it can be both
UWB works by sending single-cycle pulses. The information is carried by when the pulse is transmitted with respect to a reference.
Since there is no carrier, it doesn't affect a specific part of the spectrum. However, since there is no carrier, it affects all parts of the spectrum by adding to the noise floor. That is what the big problem with this technology is and why the FCC is looking so closely at it. The UWB Consortium has more information.
Personally I don't see a problem with raising the noise floor for this technology because, as I understand it, it raises the floor uniformly and, if I understand this correctly, the actual number of devices transmitting doesn't play into this.
I've been interested in this for a while. Time Domain (warning, flash-heavy site) is a company which has been playing with this for a long long time. I was rather skeptical of this when I first heard of it but my opinions on it are changing. Hell even EDN had an article on it.
The only thing I don't quite grok is how they can get two devices to have such rock-solid stable time references (we're talking sub-picosecond jitter) without secondary clock transmitters and keep them that way. If anyone out there can help shed some light on it I'd love to hear from you.
I too used to be one who ALWAYS clicked custom and always made sure programs went into Program Files, etc, etc. But some vendors still haven't caught on and I've run into too many troubles not allowing a program to install where it normally wants to.
I am one of those people too, but I find the problem is a lot LESS of a problem on Linux. Linux has softlinks and environment variables that the linker listens to, even if the program doesn't.
So with Linux, I've started a shift to letting stuff install where it wants to - I've found it reduces trouble down teh road.
Heh. I install everything that didn't come with my distro (Slackware) to /opt. That means VMWare is in /opt/vmware (I don't want it in the path). Acroread and CVS KDE are in /opt (where I have he paths modified so /opt/bin, /opt/lib and opt/man are part of the whole process). My MPEG tools are in /opt. Same with XMMS. Hell even my uClinux build tree is in /opt!
That way, I back up /opt and my home directories and I have very little to do if something destoroys my install.
Mind you, I also have a /data directory where all the tarballs go, and off of that directories for mp3s, avis and a documents directory for anything I work on. Backups and resotres are a snap. I found that Windows software tends to break the "My DOcuments" rule, sticking edited files in the Program Files directory and causing major trouble for backing up. Hell even Outlook Express does this (wtf, putting my amil under the windows subdirectory?!)
I seem to remeber a guy getting caned in China (or was it Hong Kong?) for breaking a law. I also seem to remeber the US population screaming and whining about the brutality and unfairness.
The little shit who did that committed the crime while in the country he was arrested in. Dmitry did nothing wrong while in the U.S.
Therefore I doubt Russia will raise a fuss, and I don't think Sklyarov should be set free. you don't break a country's laws (no matter how stupid), go to that country and expect not to get arrested.
See my comment to your first paragraph. If Dmitry did something wrong while in the U.S. (and I don't think giving a speech is illegal yet) then sure, he will have to be punished under U.S. Law. AFAIK, he did not write the eBook decryptor in the U.S.; he did that in Russia.
The strap provides good, continuous protection. The problem with holding your hand on the case is that 1) your hand -- especially the pads of your fingers and the palm -- is not a good conductor, and 2) it takes only a few milliseconds to build up enough charge to zap a component (just lifting your shoe off the carpet can build up a destructive charge). I consider a grounded strap essential when you've got your hands in the guts of a piece of equipment.
If you are having that kind of static buildup I would seriously reconsider the environment these things are in... That's insane. There's also the fact that these straps
aren't a direct ground for simple safety: they usually include anywhere from a 1M to 10Mohm resistor in series. Even if my fingertips were especially dry, say 100M or even 1000M, there is still a leakage path which static could flow from.
This is all inconsequential, of course, we're arguing after the fact. :-) I don't know exactly the type of cases you're talking about nor the environment. For all I know they could be ceramic cases in a wood kiln or something.
That doesn't help when you're holding the board by its edges, and maybe accidentally touching chips in the middle of the board.
I apologize for not making myself clear: The techs were pissed that they had to wear straps when connecting to ports on installed cards. That is what I was referring to.
The thing is that we don't know that the grounds are at equal potential, and in fact if the handheld is charged, we know that they are NOT at equal potential. Thus to prevent damage, we want the potential to change slowly -- something you accomplish with a lot of resistance.
So you're not suggesting that the resistance remain throughout the connection, only at the start. Ok. I had misread your original intent.
err, read my post - I stated I didn't do the original design, it was inherited. and later in my post I speak the praises of using ESD and surge protection in a circuit design - but thanks for the description of my abilities all the same.
Thanks for chopping my comment in half. I did say "you're a shitty designer, or you had a shitty management (financial) decision laid down on you. Truth is that anyone who doesn't design with this in mind generally is a poor designer. Whether that be from inexperience or negligence doesn't matter.
But we also provided shoe straps (slip it on in the morning - take off when you leave) which wern't too bad.
This is a good solution too. What is the difference between a static strap and touching the grounded case of the unit you're about to work on? I can't see any, and from personal experience I'm far more apt to touch the case than I am to make sure I have the strap on and that the strap is grounded.
And the bottom line is - tough shit. Your fooling around with millions of dollars of corporate hardware and you work for the company - deal with it or take a walk/find new job.
I agree with you that static protection is necessary -- don't get me wrong. Where I was digging in, however, was with the choice of static protection you had offered. Now of course, the ionizing air filters and so on weren't always around and neither were the shoe straps / antistatic shoes, but I was more or less taking the side of the techs who found the straps a pain in the ass. I feel I have a better solution (touching the case) and it seems to work for me. It doesn't work for everyone and it likely won't fly on an ISO sheet. You and I both know how much ISO compliance is worth though, so long as it appears that you're following it. :-)
I can't tell you how nervous my HW techs got holding processor cards worth hundreds of thousands of dollars - its scary!
I can imagine... but really why weren't these super-duper-hyper-expensive cards built with some form of ESD protection (hell for that price it could have been DAMN GOOD protection) on the external ports?
I'm still amazed that Mobo makers don't toss in dispoable straps with their retail boxed boards - hell they only cost 50 cents or so.
Pure economics. $0.50 on a strap is $0.50 less in profit per mobo. It's far cheaper to have the warning in the book. (I don't agree that it's right, just that that's the way it is.)
got news for you, NONE of these chips have ESD hardening.
If you used which which weren't able to withstand ESD and you did nothing to protect them... you're a pretty shitty designer, or you had a shitty management (financial) decision laid down on you. It's not hard to put a 10 ohm resistor in series with the input pins, throw on a bit of a filter and finish it off with some good fast 1W transzorbs. Takes care of ESD and EMI/RFI.
As for static straps - its amazing how people hate them so. I managed a 10k sq ft data center with almost 700 servers, from small $5000 machines to monster Auspex boxes costing millions.
Static straps and other such protective measures are a pain in the ass. Get used to it. It probably would have been better for everyone if you just used ionizing air filters and kept the air relatively humid.
The worst were the Sysadmins - they figured since they didn't touch teh cards themselves it was OK, yet they were plugging serial cables into exposed serial ports to hook up root terminals (before we had a networked root term setup) It was amazing the resistance I encountered for such a simple thing.
Whenever I have to touch a computer, I make special note to touch the case often. It may not drain every last volt of potential difference from the case and I, but I've never had troubles. I'd bitch and complain if I were a sysadmin too. Or rather I would touch the case and the connector before plugging anything in.
The real problem here is that nearly every integrated mobo has the serial port contained in the northbridge/southbridge chipset, so discharge to the port means discharge to a critical IC in the computer as well.
I doubt that it fried anything bigger than a TTL<-->RS232 converter IC. The reason? The designs of the large ICs don't work well when designing conversion circuitry. Sure, the UART will be part of the chipset but the signals on the I/O will be TTL or CMOS level outputs.
Those signals will then hit an IC like the MAX232 (An RS232 converter IC from Maxim Electronics) which contains the charge pumps and converts that +5/GND (or +3/GND) signal to +12/-12V required to meet the RS232 spec and back. Chipmakers like Maxim also make static-protected versions as well. (In Maxim's case, they usually designate ESD-protected devices with an E suffix.) These chips are good for a 11kV direct zap using the human body model.
No, I don't work for Maxim. Burr-Brown (Now part of TI), National Semiconductor (now spun off to Fairchild) and a host of others make these chips. I'm just most familliar with Maxim's.
If the motherboard fried, they used substandard (IMO) converter ICs. I've hit my laptop and several PCs very hard with ESD and I've yet to have a problem. The biggest problem is that ESD is a slow killer. Rarely does it fry something outright. Usually it just weakens the oxide layer on the semiconductor FETs and causes early death and spurious operation.
Palm should be no more liable for this than every company which manufactures serial, parallel, USB, FireWire and really ANY external device. If the guy bought a shit motherboard, deal with it. It's not Palm's fault.
Careful mechanical design to further reduce the problem - arrange that the "grounds" always touch first - preferably through a few hundred thousand ohms of resistance.
Ummm... how the hell is ground to be effective with that kind of impedance? Ground potential should be equal.
The unclear part is how to apply this when the each of these two adjacent sentence clauses (no establishment/no prohibition) contradict each other.
It is unclear whether providing overhead (electricity, land, janitorial services etc) support for people egaged in religious activity constitutes a move, however tiny, towards the establishment of religion. Neither is it perfectly clear that forbidding equally all such support to all such groups effectively prohibits their right to excercise freedom of religion.
This is the old "reducing to absurdity" logical fallacy. You could argue that the use of a state park or parking lot constitues a move, however tiny, to the establishment of religion.
There are times to nitpick and times not to. This is (to me) clearly a time not to do so. You take the law in the context it was meant to be in, instead of breaking it apart to this level. I believe there is a judicial term for this but I haven't got a clue what it is.
Their claims were that they could put their units in the power distribution stations and then everyone would be able to buy little boxes to plug into their outlets that would be able to recieve and transmit data. They claimed data transmission rates in the gigabit range.
Yeah that's them. They claimed to be able to "tunnel" through the magnetic field caused my electrical current flow with a maser. I downloaded their patent but couldn't make sense of it.
Why not turn bridging on in the kernel and bridge to an internally connected 802.11b card.
Because I bought Orinoco Wavelan cards and the bridging firmware is sent to the card by the access point at every boot up. :-(
Mind you, I've got a logic analyzer with enough lines to handle a 16-bit PCMCIA bus, I just have to beg/borrow/steal an access point and spend the next few weeks untangling the init code to see what they send so I can tell Linux to do the same. :-)
Are all cards that use the same chipset as the WaveLan cards (Hermes 2 I think?) incapable of bridging without the firmware? If not, I'll just buy a cheap one and use it instead, or see if I can't do a dump of the firmware and muck with sending it to the Wavelan. :-)
They don't get around the problem of high frequencies being attenuated by the pole transformers, they simply don't worry about it because they have their CuPlus units sitting up there too.
Deployment of this technology will be far slower than DSL or cable, simply because every neighbourhood (or fraction of a neighbourhood) requires a CuPlus unit in order for this to work. That's one CuPlus unit for every pole transformer!
In my area, I see a pole transformer service at most two dozen households. The deployment for this technology will be insanely long.
The other question I have which isn't answered is how the RuPlus or NtPlus units can be plugged into any outlet and work when the split-phase wiring in North America effectively isolates the two 120VAC circuits from each other? Your pole transformer usually takes a 1200-13kV line and neutral and then gives you a center-tapped, 240VAC output (3 wires) to your house. Do they have the equivalent of a hub and two modems to distribute the incoming (to your house) signal across both phases and take the possibly distinct return signals, figure them out and put them back on to the main communications bus?
Whatever happened to those guys who claimed to be using a maser to modulate the magnetic field, thereby defeating Maxwell's equations and getting high frequency data through the 60Hz transformers? I'd love to see a working demo.
Get your stuff from Digikey [digikey.com] for quality (and inexpensive) electronics.
<cough> did you say cheap and digikey in the same sentence?!!
DigiKey is known as the Radio Shack of the professional electronics industry. Sure, they can get you pretty much anything and fast, but they are not cheap by any means, even when you look over on the table to the bulk prices.
You're far better to go to Future-Active, Arrow or even Farnell for your parts. Digikey is quick and fast, and their quality is the same as anyone else. But they are NOT inexpensive for anything. You're paying a premium to have them ship 16 identical catalogs to your shop every quarter.
We use the megabit modems extensively for our business DSL deployment (no DSL offered from the telco, so we do it the hard way). When we started we couldn't justify buying a DSLAM so we just hooked them up back-to-back as described in Cringely's article and as long as you have the rate set the same on both ends, they just work. No, they're not RADSL and personally, I prefer that.
We've had zero trouble with these units, having installed about a dozen or so over the past 3 years. Great for businesses who KNOW they want on the 'net at high speed, but for a personal connection or as a trial they're a bit pricey. That's why I've been working on some alternatives.
I've just purchased a pair of Efficient Networks 5250 SDSL bridges. They don't specifically state that they'll work back-to-back but after some research and initial legwork I think they'll work just fine as a cheap alternative. They can be had for USD$50 from Ebay.
Pairgain stuff has the longest "reach" of all the DSL equipment we've investigated but they are also one of the more expensive ones out there. I suppose you get what you pay for. :-)
Cough %$Bullshit$% a DIVx rip from dvd under the mystical 700mb cd limit or thereabouts for a 2 hour flick has more artifacts than King Tut's Funeral Chamber.
Without postprocessing, you're 100% correct. I use mplayer with the opendivx libraries. On a dual celeron 466 and no postprocessing, it is what I would consider the digital equivalent to an EP recording of a copy. However with the postprocessing set to its highest level (4 for DIVx) it is wonderful. Of course, my system isn't good enough to handle this so the audio gets out of sync with the video very quickly. :-)
System: Abit BP6 (Dual Celeron 466), 256M RAM, GeForce2MX-400 (32M). Yes this is in fullscreen under X 4.0.3 with the latest nVidia drivers (1251 I believe). The movie: Varsity Blues. Filesize: 629441536.
Who will pay pay slightly less and $500 for a TV out card to get a shithouse quality movie(paticualry after going though a consumer level TVOUT card) on your 4foot plasma?
You're targetting the wrong demographic.
Think any TV-out card (nVidia, ATI, etc.) going to a standard TV. The type of people who buy a $500 TV-out card and have a plasma television are also the type of people who use component video and $500+ DVD players.
I think the type of people to download a 500M movie are the same type of people who download DivX movies from eDonkey and burn them to CD anyway... quality should be about the same and if you have a ballsy-enough computer to do all the postprocessing it looks like a good VHS tape on a normal TV. Hell it looks like a DVD to me on my 17" monitor but I end up dropping frames since my computer isn't ballsy enough. :-(
The reality is that the presence of RI helps sophisticated query rewrite engines to do some very nifty transformations. So having RI actually helps you get better performing plans for read queries and hurts performance a little for writes.
I totally agree. If the backend knows that it can't compromise integrity though optimization (i.e. an optimization trial fails to maintain RI, etc.) then it also knows that when such tests don't fail, it has an optimized access method which kicks ass.
(ie: Linux SMP box hanging under high network load (which makes backups a real bitch), forcing me to power cycle : flawed APIC handling for the 3c905 ethernet card),
Out of curiousity, did you find a fix for this? I think that may be explaining the odd lockup I get on my system that I haven't bene able to pin down...
Of course, MySQL has other drawbacks, namely that it doesn't support triggers or table inheritance or some of the more complex nuances of standard SQL, but the 95% of stuff it does have is very fast, and the other 5% can be handled in code.
The instant you decide to move data and/or referential integrity from the DB into 'code', you've lost the battle. I can't believe someone would even suggest this. Sure, stored procedures are a plus and most times they aren't necessary, but you simply cannot have your integrity checks outside the DB. That's the whole god-damned point of an RDBMS! If your integrity depends on unusual interactions in the data store, stored procedures are often the only way out. And you can't have stored procs that work to enforce integrity without triggers.
If you haven't got the integrity, you may as well be using a hashed filesystem to store your data for all the difference it would make. Hell, the hashed fs would probably be faster since it isn't pretending to be an RDBMS.