The deCSS source is out there. The current court arguments is it is primarly a pirate tool. Write your own program to read DVDs, with dvd as an integral part, and your tool is not a pirate tool it is a viewer. Courts are more likely to look at your program and say "WEll, reverse engineering is legal (THe PS emulator helps this case), and your program is obviously for viewing DVDs and doesn't provide a save function so it much be legal.
Been done for Years. In the late '70s a Company called network Systems released the first router, for the purpose letting mainframe devices sit in anouther location. CNT is their big complitor, which does a similear thing.
Of course the orginial routers didn't do what you want (and didn't do IP until much latter) but it was the direct ancestor of what companies use now what they want disk or tape drives located in a different state or country.
Interesing story, in 1985 this company sent a VP to california to examing and perhaps buy a small company that was just getting started then. AFter doing due dilligence they decided this company was going nowhere. The small company that wasn't going anywhere was Cisco. (Accually most of those close to Netowrk Systems today conclude that the only people who would win from that deal going through is 3com, but that is a different story)
I agree with you (I'm not a big investor, but I refuse to ignore P/E) However you seem to have messed things up a little. P/E is Price/Earnings. The P does not stand for profit, it stands for price. AFter paying all the emploiess, depts, and power, water, and other bills, Earnings is what you have left over. Price is what your paying for your share of the earnings. Divide Price by Earnings, and you have a simple you historicly useful measuire of the company.
I'll agree that PE isn't the whole story, but if this figgure gets high (The historical average is about 15) that means someone is paying extra for the profit, if it gets low, that means your getting a bargin. The critics point out (correctly) that a company that is growing fast will have a high PE, because looking to the future you can say expect the earnings to grow to make the PE more reasonable, while a company with a low PE is generally in trouble, and you may only own a share of those earnings for a few years before bankruptcy. Of course this is simplified, but you get the idea. If your going to invest money do some more research.
(ap) Sherwin-Winniams has announced a new paint that when electrified resists gravity.
We have known about this for 30 years in our paint labratories, but the early versions suffered from sudden loss of effect. Every april researchers would get this on each other shoes, and zap each other with static eleectrisity to make the other person float to the ceiling. However the victums would suddent crash to the floor with no warning several minutes latter.
So said cheif researcher, who prefered to go by bluGill. He went on to remark how similear this appeared at to quantum mechancis at first.
Our first tries in measuring how long until someone, or latter something when people got sick of sacrificing legs to science, fell were inconclusive. We discovered we could get a general idea quickly. Attempts to find out exactly turned out to be more difficult. It turned out that out measurements changed the amount of time by an unknown amount. This sounds like quantum mechanics, and we wasted 20 years and millions of dollars studiing that before realising that was the wrong answer.
Going back to the drawing board they decided to try something too obvious to start with: instead of a single static charge they applied a continious charge. Since the paint was always being activated it could not shut off without warning.
Of course that didn't solve the landing problem, but landing turned out to be simple engineering. Rather then making the entire surface anti-gravity we made many smaller parts anti-gravity. To take off we would make all parts anti-gravity, when landing we made some ant-grav, but turned others off. Any engineer can figgure terminal velocity and compare that to how much anti-gravity they have and make a soft a landing as they desire.
This paint would have been released years ago, but the first large test failed. They intended to levitate a sphere with humans in 100 feet off the gound (not exactly but they figgured they could get close), but the switching mechanism jamed when the stresses changed between grav and anti-grav areas.
Fortunatly with anti-grave so cheap we had put an entire hydrophonic garden onboard. Those onboard claim that they went as far as Mars, but our estimates of the speed they obtained suggest they only rose to the moon.
This paint is sure to bring out a new ear of space travel. Since mass is not an issue and the paint is cheep a backyard builder can build an inter-planity explorer in their backyard. The possibilities are endless.
Was I the only one who thought of Danny Dunn when I saw this story?
Sigh, replying to myself... some things that I just recalled
Things don't go away in the embedded world. One of our key products was designed in 1984. Several years ago the CPU supplier told us they were taling final orders. So we ordered a 15 year supply of CPUs. We had no choice, a 10 mhz cpu looks horridaly slow today, back then it was the best we could do, and the hardware depends on some timeing of when the CPU gets the bus to let other custom parts on the bus. We can't change CPUs (without spending $$$ on a product that is nearly obsete, there are no new cusotmers, but the old customers need more or replacements from time to time)
Anouther poster mentioned the Memory and CPU is generally constrained. We find this less true today then before. In fact we have found that we have to put more memory in than we need because that is all that is avaiable. Then we complain to the supplier that it uses to much power...
My compile/test/fix cycle is up to nearly an hour now. That is After fixing a big and recompiling (which doesn't take any longer) I now have to download software, burn proms, and reboot. Each reboot runs 20 minutes of hardware diagnostics (We are looking at getting that down, but we don't want to do a less then complete coverage) before I can simulate the situation I need. I'm doing some unusal endcases that take time to setup after that.
Still, I like embedded systems. It is a refreshing change form the microsoft crashes often, or linux it can crash though it normally doesn't world. We can crash, but only if we have something redundant to take over.
I'm aware of (but don't use) programs that would help the situation. My point was that the human contact is lacking at times. Those proposing a technical solution are sticking a finger in the dike and ignoreing the water pouring over the top.
When you can't be there technical solutions are better then noting, but the lack of human interaction of not being there is a price that you will have to pay no matter what.
I'd like to agree. If that Palm can go wireless he is set, just put a PCMCIA ethernet card in one (linux) laptop, and let the laptop route to wireless when you need to connect to someone else's machine/network. Just get two cables, one crossover, one normal and your set.
Many buisnesses nowadays have waveLan or similear installed (I think waveLan is biggest, but I'm not sure) which means with DHCP you can connect directly to their network as needed.
Again, the palm might be a problem, but AFAIK you can't connect ethernet to that either, so you should be able to go wireless. If you don't go wireless you will kick yoruself the first time one of your ethernet cables breaks just before a presentation and you can't do what you want. (Although I suppose some cusomters may have so much wireless congestion or RF noise that this isn't perfect either)
Where I work we are considering Linux for our systems. We attempting to get in on the embedded linux consortum (Not sure of the name) becuase there is only one other non-vender company in on that. We want some influence if we are to use it.
Embedded systems can be anything from a PC monitorying some hardware (I think IBM's mainframe printer uses OS/2 for that) Then there are embedded systems where the comptuer makes the decisions, like an ATM. And CNC (Computer control of metal working equipment) is different.
The product my company is working on is somewhat like a router (I can't give it away), and we have special hardware to get the packet speed up, so the OS only needs to monitor tempatures and change configurations. This is different from CNC where the comptuer has to change which windings on the moter power is going through at the right speed, which depends on how fast the work is spinning, amoung other things. Then there is medical equipment which needs to monitor the health of someone, and adjust drug doses, and call a doctor/nurse if things get out of hand.
Now look at requirements. Marketing assures me that in our case the customer wants relability before all else. Our MTBCF [if you don't know what this means you don't belong in this industry] is 500,000 hour. Some of our componants (harddrives in particular) are not rated for that long, even before you get into statistical analysis. Since our machine breaking could cause every computer in the company to go down we have to meet those numbers. The CNC machine is expected to be reliable, but if it breaks unexpectedly it only costs one person's days work, which isn't nearly as big a deal. The medical device is different. It cannot break in operation, whatsoever. Any problem could cause someone to die, and that is not acceptable. (I know that doctors are not perfect, and for that matter the machine isn't either, but that doesn't change the fact that no breakage is acceptable). That means the device must be capaiable of detecting all faults imeadiatly and sounding an alarm, and preventing any possibility of overdose. (That is if a medicina valves breaks open there better be some other way to shut off the flow) Fortunatly in most cases drug flow can be cut off for a few minutes (meaning that you can sound an alarm if a problem occures and the staff has time to replace the machine)
I could go on, but you quikcly realise that embedded systems have differing goals. I would say that if you want to get into them, write some error recovery. Get involved in detecting memory errors or something, and continuing on. A sig 11 in linux normally means bad memory, (often realted to overclocking) can you find a way around doing a core dump? Of course there is also the functionality. If you want to be involved in the medical mahcine, then I expect you to have some training in medacine, The CNC programer should spend some time in a metal shop. Of course when your linking computers you only need comptuer expirence, but this is a very small portion of the embedded systems market.
This guy has a doctorate, so I dont want to call him an idiot, but nothing in this interview gave me the idea that poplar trees were really that much better then say oak trees, wheet, or algee [in a pond]. He did in fact say they needed data. I looks to me though like he likes trees, so he is calling trees the solution. Now maybe they are, but I don't see any evidence.
Even if he manages to clean up polution, how is it better to have polution in the cells of trees that your burn and release into the air? Of course organics don't have this problem (much), but other chemicals just get absorbed into the plant and then what? A question I immeadiatly have that isn't answered.
Like I said, the guy has a doctorate in this, he must have done some studing, but I don't see anything that makes me belive he did something.
I work at on Office in MN, and can in theory tellecommute anytime. (THere are dial in servers and the like to get around the firewall. As a Unix orianted enviroment we have it easier) There is one guy who lives in Arizona (For those who don't know US geography, he is about 1700 KM away) He gets his work done, I get my work done. However when we have to work togather it is difficult. I cannot show him what is on my screen, he cannot see what is on mine. (not well anyway) I cannot point to a line of code and ask what that does. He cannot take control of my screen for a moment to write a line of code that he understands better then me. I need to look over his shoulder when he does because I understand the pig picture of that function better then him, but he knows the function I want to call better...
OTOH, many of the guys around here will send a message once or twice a month "Working from home this afternoon". We find that once you know how a piece of code works, how it interacts with everything else, it is much more productive to go home where there is nobody around (Note, see the children comment someone else made) to get a lot of code written and tested quickly. Turn that around though, if you don't understand howyour part of the program works, or how it interacts with everything else, you are better off at the office where you can wonder the hallways: "Hey Dave, soon as your out of the restroom there, mike and I have a question for you." Meaning that if you need someone your better off face to face, maybe with a white board to sketch things on.
Telecommuting can work. Sometimes it works better, other times it is worse. My boss encourages us to tellecommute once in a while because he knows that once in a while you are more productive. (Besides, I'm in Minnesota, once in a while the snow gets bad enough that you cannot get to work, much less safely get to work. Tellecommuting doesn't cost him near as much.)
I find that the majority of my time in X is spent with something onscreen not running on the local machine. Even if we call the programs running on the same machine as my window manager this is true.
Maybe your desktop machine is fast enough, but not mine. I used to have an ultra sparc on my desktop. I hated it, genuine noisy fan, poor keyboard (You think Sun would know the controll key belongs next to the A. xmodmap is for more difficult tasks than that) And when I kicked the power cord all my compiles stoped. Now the server is in a back room, I have a NCD (A dumb terminal that can run X) on my desk.
As a devolper I find that my programs have to run on a machine in the lab. Pop an X window from the lab machine to my desktop and its like working in the lab. THen I can go home and pop an X window from the lab to my home FreeBSD machine. (I don't normally work from home, but if I can check my compile over night and fix the one typo after supper I save time at work the next day)
I used to have a sun3 in one room, and my main machine in anouther. In theory I could run programs on the sun3, but since the main machine is 400 mhz, and the sun is 16 (20? doesn't matter as m68k and x86 of different generations don't compare well by looking at mhz, but you can get the idea) who wants to? X still has life. Xfree86 is very fast on a local machine, and works just fine on a remote one.
Lemma: If csh doesn't have it, then it is evil. Proff: This old foggy started when csh was the best avaiable and can't be bothered to switch csh doesn't have tab completion. Tab completion is evil. QED excercise 1: Use lemma to prove tcsh is evil. exercise 2: prove all sh dervitives are evil. Exercise 3 (for advanced students) Prove that sh is evil but not as evil as any other shell
Seriously though, you can't assume that your favorite shell is the one I use. For that matter your favorite shell may not be avaiable on the computers I use, which include version of UNIX from when Unix was direct decendant of Bell Labs code, and not standard that could be applied to any OS.
Note that the converse is not true, as csh is very evil.
Besides the above, the unix way is not 8.3 filenames, but smaller ones. Unix would have only two letter filenames allowed, but the limit to uniqe two letter filenames is smaller then a typical/ur/lc/bn, and the early test to make sure the new longer filename filesystem would work changes that path to/usr/local/bin.
P.S. I have this nice swam^h^h^h^h parcel of land in Florida if anyone is interested.
This is fine if you're just playing around sampling some stuff. But for recording a live performance you need an operating system that has a high level of reliability. That's something that is not currently available for Windows.
Bingo! I know many musicians who have gone from $1000 computers to $20,000 dedicated machines with less functionality. You see when you are recording with a group you get 70 tries. ON the 70th try the magic is suddenly there and the song moves right. This is not an explainable thing since the other 69 tries are note perfect, but the 70th somehow has a level of energy that didn't exists. The last thing you want is to get the magic (whatever it is) and then find out the recording device didn't capture it.
Of course musicans have also discovered that for all the promise of digital, the orginal masters (running at 15 ips on 1/2 inch tape) can be recorded and edited in analog without anymore (noticable) hiss then then digital and analog editing has advantages. (mind you digital has advantages too, unless your are an engineer with expirence I wouldn't trust your judgement on which is better)
Security is all nice, but you have to first consider how it will be used
So you get a ticket to the next linux convention out of town. (Nice of the boss to declair it buisness related) you live your comptuer on with all this monitoring equipment. Local theif who has been casing your home for a while notices that you didn't come home last night, your a geek, and there is a linux conf... late the next night breaks in and takes your comptuers. You come back to find a broken window, but the computer that you stored the images from the camera is gone. (And the theif has probably raped your computer with an illegal copy of windows on the machine)
Now lets go the other way. Your arrive home with your security system on. Three buddies with their comptuers in hand are behind you for a network party. In your excitement you forget to turn the alarm off. Just as you get the comptuers booting 120db blasts in your ears. Opps, forget to turn the alarm off. After a few times you quit turning on the alarm, or the neighbors get used to it going off at all hours when you forget to turn it off. (Remember break-ins are rare)
What I'm getting at is two points: The alarm system must be useful in either preventing crime, or capturing the criminal after words. If the alarm is not on or the neighbors are traned to ignore it the theif is home free. If the theif in the later case steals the VCR/comptuer recorder with his picture on you have nothing to give the police and you are free.
Seriously consider paying a big company to monitor your alarm (which they will insist on installing). It costs money, but if your really going to deal with the hasstle you may as well have them call the police. (Far as I know it is illegal for an alarm system to call the police, but the alarm system can call the monitoring station, and then they call the police) And the big advantage is by connecting the fire alarms to the alarm system you can accually get the fire dept out when that neighbor kid figgured he can't break in so he burns the house down.
Lets assume (for discussion) that Solaris is more accurate then linux. That does not make solaris accerate enough.
I did poorly in my numerical anaysis class, but I do remember a couple things: rounding errors (which floating point almost always has) increase greatly. Often you can get a result from a calculation that after analysis has zero significant digets. That is you could get your result from a random number generator and have just as much confidence in the answer. This of course depends on the calculations involved.
Anouther place the professor pointed out that adding more data to your real world sample increases the error out of all bounds. (This was related to finding a polynomial for some line, and basicly if the line is right, there may be an equation, but not a polynomial. In that case they can prove there does not exist a function that can tell you how far off you are when calculating a data point)
In summery, differences between what solaris calculates and what linux comes up with should not be your first concern. First you should understand all the errors involved, and most of them are fundamental to a computer with a finite number of bits.
(In addition to what the AC said about the quality of old glass) When they made those old cathedrals they put the widest edge down.
So, yes glass flows (I think anouther person estimated it at 200 microns for some large quanity of time), but the quality of glass back then was not that great therefore any measurments made to old glass are inconclusive.
See Wine Hq. The license on Wine just (is in the process of?) changed from GPL to BSD to fit the goals of Wine better.
First of all, all old GPL code is still GPLed. They simply contacted all the authors they could find (and in fact gave them some time decide) Results were about 95% in favor, 2% opposed, and the rest unfindable. The 2% of coders were deemed insignificant in that it wouldn't be too difficult to re-code their contributions. The other will probably end up re-coded eventially if the authoers cannot be found.
There are pros and cons to all licenses, please do not get into an argument of should Wine have changed. They did, and unless you contribute code you have no voice in the change. The topic is can it be done, and obviously they have proven it can. If you want to do the same things read their archives, there is no point in repeating mistakes they made.
I live in Minnesota, and drive 20 miles to work. (Lets assume that works provides a recharging station so I have a full charge when I leave) Take a cold winter morning, -25, with a coule traffic accidents. It now takes not 30 minutes to get to work, but 75. Since the tempature is so cold I expect the heater to be running the entire time. The electric car does not have the range I need already, just running the heater on full.
I can write with either hand (not well, but I can) as an expiriment I signed all my credit cards with my right hand, and then all the sales slips with my left hand. Onle ONE clerk has noticed that my signature didn't match the one on the card and demanded further identification. (Once noting my drivers license had the same name, and signature and my picture on it he accepted the sale)
Somehow the above tidbit fits into this topic, and it forms some arguement in here. I don't know what though.
Really there are two problems, and nobody admits they have to be addressed seperatly.
First there is commercial tradmarks and the area the trademark covers. For instance: McDonalds.com - the fast food restaruant we all know McDonalds.com - the orginal restaruant that Ray Kroc bought the idea from (Are they still around, in any case the rights of that deal should handle this) McDonalds.com - a local electrical company in some small city run by some brothers McDonald. McDonalds.com Same as the last one, but diffierenty city. Maybe these guys are plumbers, whaterver. McDonalds.com Again a small local buisness. Maybe a hairstylists that takes appointments on the web
You get the idea, a major recignised worldwide name, but it cannot do anything about the small reginal companies that are not compittion. Somehow the needs of all these comercial interests need to be met. (And I think we all agree that McDonalds restaruant and McDonalds heating can be in the same town without problem, but you can't have a second McDonalds restaruant in the world unless it has been McDonalds for 100 years)
On the other side we have personal sites. I want to be McDonalds.per because my family name is McDonald (okay, my last name is not McDonalds, but I know some McDonalds and it makes the point best) which when combined with the rest of my family allows email to bluGill@McDonalds.per and Jounior@McDonalds.per to get through easially. Since.per specificly dissallows anyone with a trademark from getting a domain we are free from the previous problems. Somehow however we need to solve the next one though: There are many families McDonald in the world, and most of them are not blood relatives. (you could up to Adam and Eve or the evoltion equevelent)
And a good choice was made too. This is interesting to me as a BSD fan, but it is still in the curiosity stage. If NetBSD would announce instead some break-through that allowed them to impliment SMP in 3 days that could run a single process 2x as fast on two proccessors as one, that would be front page worthy. (And most of us would wonder when the person who achived that is going to show the e!=mc^2, which should be about as easy a task)
I just wanted to post this so that it is known that not everyone thinks a bad decision was made.
The local home depot sells cat-5 cable, and female ends that snap into a special wall plate. They also have telephone, coax and speaker female ends for this plate. (look first to see what you can get before making plans) Got some of that, and then find a way to get up to each room without making a mess. This is the hardest part if your not doing new construction. The walls in our second story do not line up with the walls to the first, and I could only get to the basement. (Outside walls of course line up, but they are full of insulation so that isn't useful) I did manage to find where the toilet drain went thoguh a wall, which was great since normally you can't get through floors in walls without drilling. Look in your house, if you don't see what I mean your house is probably built different.
For new construction, run plenty of cable, the stuff is cheap. For older construction the hard part is running cable so I ended up skimping. (For ethernet you can run two jacks on a normal cable if you know how to wire it.)
Run caox and phone for sure when you do this, and I'd run it all to one central localation in a wireing closest.
BTW, don't make the mistake I did: make sure you have some way to tell which jack in the wireing relates to which elsewhere. I have labels in the closet, but I can't tell left from right. Not good.
Try again. IBM did not lose money on mainframes last year. It isn't a high growth market, but it still makes up a large portion of IBMs money stream. Maybe 83-93 IBM lost money in mainframes, but last I check they were shipping mainframes as fast as they could make them. (part of that was Y2K hit mainframes the hardest) Nothing can process the volumn of data a mainframe can. (Okay, a cray can do it, but your talking mainframe class price too)
Most dealers are onto the internet. That is they know that you know their price. Some (but by no means all) have realised that you know their fair profit and try to get things done quickly at a good price. After all, if you know their numbers they have to haggel different.
Two days ago I test drove a '95 S10. I liked it, so I drove it home, hit the internet and looked up the values. I then walked into the dealer with the knowledge of what they paid and what they expected to get it for. Since I knew my trade in better then them I was able to make them an offer that included a fair profit for them and a good price for me. I was out in an hour (and it only took that long because my trade-in was in my dads name yet, so I had wait for him to get that paperwork)
I would recomend using your local dealer first. That is do the research on the internet, and then walk into the lcoal dealer and make a fair offer. If they don't take it walk - if they don't call back in a day changing their mind (This is a common practice) check the internet. Although dealers are big buisness, they are local, and I like to support the local guy if possibal.
Supported Scanners Some USB scanners are now supprted, but expect USB support to be reworked in a year or two once people figgure out how USB works best. (ie right now USB is in a state of quick hacks and amazement that things work. Things are looking better, as work is going on, but I get the feeling that the programers are not sure how to do USB best/right
There are scsi scanners that are unsupported. There are one or two parrelle scanners that happen to work. There are usb scanners that will never work, and usb scanners where the manufacture hopes they will work soon.
I have a microtek V6USL which works just fine, though only were the very latests backend (which I don't belive is intigrated into any current versuion of SANE yet)
Slashdot Mirror
The deCSS source is out there. The current court arguments is it is primarly a pirate tool. Write your own program to read DVDs, with dvd as an integral part, and your tool is not a pirate tool it is a viewer. Courts are more likely to look at your program and say "WEll, reverse engineering is legal (THe PS emulator helps this case), and your program is obviously for viewing DVDs and doesn't provide a save function so it much be legal.
Good luck, but you have a better leg to stand on.
Been done for Years. In the late '70s a Company called network Systems released the first router, for the purpose letting mainframe devices sit in anouther location. CNT is their big complitor, which does a similear thing.
Of course the orginial routers didn't do what you want (and didn't do IP until much latter) but it was the direct ancestor of what companies use now what they want disk or tape drives located in a different state or country.
Interesing story, in 1985 this company sent a VP to california to examing and perhaps buy a small company that was just getting started then. AFter doing due dilligence they decided this company was going nowhere. The small company that wasn't going anywhere was Cisco. (Accually most of those close to Netowrk Systems today conclude that the only people who would win from that deal going through is 3com, but that is a different story)
I agree with you (I'm not a big investor, but I refuse to ignore P/E) However you seem to have messed things up a little. P/E is Price/Earnings. The P does not stand for profit, it stands for price. AFter paying all the emploiess, depts, and power, water, and other bills, Earnings is what you have left over. Price is what your paying for your share of the earnings. Divide Price by Earnings, and you have a simple you historicly useful measuire of the company.
I'll agree that PE isn't the whole story, but if this figgure gets high (The historical average is about 15) that means someone is paying extra for the profit, if it gets low, that means your getting a bargin. The critics point out (correctly) that a company that is growing fast will have a high PE, because looking to the future you can say expect the earnings to grow to make the PE more reasonable, while a company with a low PE is generally in trouble, and you may only own a share of those earnings for a few years before bankruptcy. Of course this is simplified, but you get the idea. If your going to invest money do some more research.
(ap) Sherwin-Winniams has announced a new paint that when electrified resists gravity.
So said cheif researcher, who prefered to go by bluGill. He went on to remark how similear this appeared at to quantum mechancis at first.
Going back to the drawing board they decided to try something too obvious to start with: instead of a single static charge they applied a continious charge. Since the paint was always being activated it could not shut off without warning.
This paint would have been released years ago, but the first large test failed. They intended to levitate a sphere with humans in 100 feet off the gound (not exactly but they figgured they could get close), but the switching mechanism jamed when the stresses changed between grav and anti-grav areas.
Fortunatly with anti-grave so cheap we had put an entire hydrophonic garden onboard. Those onboard claim that they went as far as Mars, but our estimates of the speed they obtained suggest they only rose to the moon.
This paint is sure to bring out a new ear of space travel. Since mass is not an issue and the paint is cheep a backyard builder can build an inter-planity explorer in their backyard. The possibilities are endless.
Was I the only one who thought of Danny Dunn when I saw this story?
Sigh, replying to myself... some things that I just recalled
Things don't go away in the embedded world. One of our key products was designed in 1984. Several years ago the CPU supplier told us they were taling final orders. So we ordered a 15 year supply of CPUs. We had no choice, a 10 mhz cpu looks horridaly slow today, back then it was the best we could do, and the hardware depends on some timeing of when the CPU gets the bus to let other custom parts on the bus. We can't change CPUs (without spending $$$ on a product that is nearly obsete, there are no new cusotmers, but the old customers need more or replacements from time to time)
Anouther poster mentioned the Memory and CPU is generally constrained. We find this less true today then before. In fact we have found that we have to put more memory in than we need because that is all that is avaiable. Then we complain to the supplier that it uses to much power...
My compile/test/fix cycle is up to nearly an hour now. That is After fixing a big and recompiling (which doesn't take any longer) I now have to download software, burn proms, and reboot. Each reboot runs 20 minutes of hardware diagnostics (We are looking at getting that down, but we don't want to do a less then complete coverage) before I can simulate the situation I need. I'm doing some unusal endcases that take time to setup after that.
Still, I like embedded systems. It is a refreshing change form the microsoft crashes often, or linux it can crash though it normally doesn't world. We can crash, but only if we have something redundant to take over.
Sigh, I knew I should have put qualifies on this.
I'm aware of (but don't use) programs that would help the situation. My point was that the human contact is lacking at times. Those proposing a technical solution are sticking a finger in the dike and ignoreing the water pouring over the top.
When you can't be there technical solutions are better then noting, but the lack of human interaction of not being there is a price that you will have to pay no matter what.
I'd like to agree. If that Palm can go wireless he is set, just put a PCMCIA ethernet card in one (linux) laptop, and let the laptop route to wireless when you need to connect to someone else's machine/network. Just get two cables, one crossover, one normal and your set.
Many buisnesses nowadays have waveLan or similear installed (I think waveLan is biggest, but I'm not sure) which means with DHCP you can connect directly to their network as needed.
Again, the palm might be a problem, but AFAIK you can't connect ethernet to that either, so you should be able to go wireless. If you don't go wireless you will kick yoruself the first time one of your ethernet cables breaks just before a presentation and you can't do what you want. (Although I suppose some cusomters may have so much wireless congestion or RF noise that this isn't perfect either)
Where I work we are considering Linux for our systems. We attempting to get in on the embedded linux consortum (Not sure of the name) becuase there is only one other non-vender company in on that. We want some influence if we are to use it.
Embedded systems can be anything from a PC monitorying some hardware (I think IBM's mainframe printer uses OS/2 for that) Then there are embedded systems where the comptuer makes the decisions, like an ATM. And CNC (Computer control of metal working equipment) is different.
The product my company is working on is somewhat like a router (I can't give it away), and we have special hardware to get the packet speed up, so the OS only needs to monitor tempatures and change configurations. This is different from CNC where the comptuer has to change which windings on the moter power is going through at the right speed, which depends on how fast the work is spinning, amoung other things. Then there is medical equipment which needs to monitor the health of someone, and adjust drug doses, and call a doctor/nurse if things get out of hand.
Now look at requirements. Marketing assures me that in our case the customer wants relability before all else. Our MTBCF [if you don't know what this means you don't belong in this industry] is 500,000 hour. Some of our componants (harddrives in particular) are not rated for that long, even before you get into statistical analysis. Since our machine breaking could cause every computer in the company to go down we have to meet those numbers. The CNC machine is expected to be reliable, but if it breaks unexpectedly it only costs one person's days work, which isn't nearly as big a deal. The medical device is different. It cannot break in operation, whatsoever. Any problem could cause someone to die, and that is not acceptable. (I know that doctors are not perfect, and for that matter the machine isn't either, but that doesn't change the fact that no breakage is acceptable). That means the device must be capaiable of detecting all faults imeadiatly and sounding an alarm, and preventing any possibility of overdose. (That is if a medicina valves breaks open there better be some other way to shut off the flow) Fortunatly in most cases drug flow can be cut off for a few minutes (meaning that you can sound an alarm if a problem occures and the staff has time to replace the machine)
I could go on, but you quikcly realise that embedded systems have differing goals. I would say that if you want to get into them, write some error recovery. Get involved in detecting memory errors or something, and continuing on. A sig 11 in linux normally means bad memory, (often realted to overclocking) can you find a way around doing a core dump? Of course there is also the functionality. If you want to be involved in the medical mahcine, then I expect you to have some training in medacine, The CNC programer should spend some time in a metal shop. Of course when your linking computers you only need comptuer expirence, but this is a very small portion of the embedded systems market.
This guy has a doctorate, so I dont want to call him an idiot, but nothing in this interview gave me the idea that poplar trees were really that much better then say oak trees, wheet, or algee [in a pond]. He did in fact say they needed data. I looks to me though like he likes trees, so he is calling trees the solution. Now maybe they are, but I don't see any evidence.
Even if he manages to clean up polution, how is it better to have polution in the cells of trees that your burn and release into the air? Of course organics don't have this problem (much), but other chemicals just get absorbed into the plant and then what? A question I immeadiatly have that isn't answered.
Like I said, the guy has a doctorate in this, he must have done some studing, but I don't see anything that makes me belive he did something.
I work at on Office in MN, and can in theory tellecommute anytime. (THere are dial in servers and the like to get around the firewall. As a Unix orianted enviroment we have it easier) There is one guy who lives in Arizona (For those who don't know US geography, he is about 1700 KM away) He gets his work done, I get my work done. However when we have to work togather it is difficult. I cannot show him what is on my screen, he cannot see what is on mine. (not well anyway) I cannot point to a line of code and ask what that does. He cannot take control of my screen for a moment to write a line of code that he understands better then me. I need to look over his shoulder when he does because I understand the pig picture of that function better then him, but he knows the function I want to call better...
OTOH, many of the guys around here will send a message once or twice a month "Working from home this afternoon". We find that once you know how a piece of code works, how it interacts with everything else, it is much more productive to go home where there is nobody around (Note, see the children comment someone else made) to get a lot of code written and tested quickly. Turn that around though, if you don't understand howyour part of the program works, or how it interacts with everything else, you are better off at the office where you can wonder the hallways: "Hey Dave, soon as your out of the restroom there, mike and I have a question for you." Meaning that if you need someone your better off face to face, maybe with a white board to sketch things on.
Telecommuting can work. Sometimes it works better, other times it is worse. My boss encourages us to tellecommute once in a while because he knows that once in a while you are more productive. (Besides, I'm in Minnesota, once in a while the snow gets bad enough that you cannot get to work, much less safely get to work. Tellecommuting doesn't cost him near as much.)
I find that the majority of my time in X is spent with something onscreen not running on the local machine. Even if we call the programs running on the same machine as my window manager this is true.
Maybe your desktop machine is fast enough, but not mine. I used to have an ultra sparc on my desktop. I hated it, genuine noisy fan, poor keyboard (You think Sun would know the controll key belongs next to the A. xmodmap is for more difficult tasks than that) And when I kicked the power cord all my compiles stoped. Now the server is in a back room, I have a NCD (A dumb terminal that can run X) on my desk.
As a devolper I find that my programs have to run on a machine in the lab. Pop an X window from the lab machine to my desktop and its like working in the lab. THen I can go home and pop an X window from the lab to my home FreeBSD machine. (I don't normally work from home, but if I can check my compile over night and fix the one typo after supper I save time at work the next day)
I used to have a sun3 in one room, and my main machine in anouther. In theory I could run programs on the sun3, but since the main machine is 400 mhz, and the sun is 16 (20? doesn't matter as m68k and x86 of different generations don't compare well by looking at mhz, but you can get the idea) who wants to? X still has life. Xfree86 is very fast on a local machine, and works just fine on a remote one.
Lemma: If csh doesn't have it, then it is evil.
Proff: This old foggy started when csh was the best avaiable and can't be bothered to switch
csh doesn't have tab completion.
Tab completion is evil.
QED
excercise 1: Use lemma to prove tcsh is evil.
exercise 2: prove all sh dervitives are evil.
Exercise 3 (for advanced students) Prove that sh is evil but not as evil as any other shell
Seriously though, you can't assume that your favorite shell is the one I use. For that matter your favorite shell may not be avaiable on the computers I use, which include version of UNIX from when Unix was direct decendant of Bell Labs code, and not standard that could be applied to any OS.
Note that the converse is not true, as csh is very evil.
Besides the above, the unix way is not 8.3 filenames, but smaller ones. Unix would have only two letter filenames allowed, but the limit to uniqe two letter filenames is smaller then a typical /ur/lc/bn, and the early test to make sure the new longer filename filesystem would work changes that path to /usr/local/bin.
P.S. I have this nice swam^h^h^h^h parcel of land in Florida if anyone is interested.
Bingo! I know many musicians who have gone from $1000 computers to $20,000 dedicated machines with less functionality. You see when you are recording with a group you get 70 tries. ON the 70th try the magic is suddenly there and the song moves right. This is not an explainable thing since the other 69 tries are note perfect, but the 70th somehow has a level of energy that didn't exists. The last thing you want is to get the magic (whatever it is) and then find out the recording device didn't capture it.
Of course musicans have also discovered that for all the promise of digital, the orginal masters (running at 15 ips on 1/2 inch tape) can be recorded and edited in analog without anymore (noticable) hiss then then digital and analog editing has advantages. (mind you digital has advantages too, unless your are an engineer with expirence I wouldn't trust your judgement on which is better)
Security is all nice, but you have to first consider how it will be used
So you get a ticket to the next linux convention out of town. (Nice of the boss to declair it buisness related) you live your comptuer on with all this monitoring equipment. Local theif who has been casing your home for a while notices that you didn't come home last night, your a geek, and there is a linux conf... late the next night breaks in and takes your comptuers. You come back to find a broken window, but the computer that you stored the images from the camera is gone. (And the theif has probably raped your computer with an illegal copy of windows on the machine)
Now lets go the other way. Your arrive home with your security system on. Three buddies with their comptuers in hand are behind you for a network party. In your excitement you forget to turn the alarm off. Just as you get the comptuers booting 120db blasts in your ears. Opps, forget to turn the alarm off. After a few times you quit turning on the alarm, or the neighbors get used to it going off at all hours when you forget to turn it off. (Remember break-ins are rare)
What I'm getting at is two points: The alarm system must be useful in either preventing crime, or capturing the criminal after words. If the alarm is not on or the neighbors are traned to ignore it the theif is home free. If the theif in the later case steals the VCR/comptuer recorder with his picture on you have nothing to give the police and you are free.
Seriously consider paying a big company to monitor your alarm (which they will insist on installing). It costs money, but if your really going to deal with the hasstle you may as well have them call the police. (Far as I know it is illegal for an alarm system to call the police, but the alarm system can call the monitoring station, and then they call the police) And the big advantage is by connecting the fire alarms to the alarm system you can accually get the fire dept out when that neighbor kid figgured he can't break in so he burns the house down.
Lets assume (for discussion) that Solaris is more accurate then linux. That does not make solaris accerate enough.
I did poorly in my numerical anaysis class, but I do remember a couple things: rounding errors (which floating point almost always has) increase greatly. Often you can get a result from a calculation that after analysis has zero significant digets. That is you could get your result from a random number generator and have just as much confidence in the answer. This of course depends on the calculations involved.
Anouther place the professor pointed out that adding more data to your real world sample increases the error out of all bounds. (This was related to finding a polynomial for some line, and basicly if the line is right, there may be an equation, but not a polynomial. In that case they can prove there does not exist a function that can tell you how far off you are when calculating a data point)
In summery, differences between what solaris calculates and what linux comes up with should not be your first concern. First you should understand all the errors involved, and most of them are fundamental to a computer with a finite number of bits.
(In addition to what the AC said about the quality of old glass) When they made those old cathedrals they put the widest edge down.
So, yes glass flows (I think anouther person estimated it at 200 microns for some large quanity of time), but the quality of glass back then was not that great therefore any measurments made to old glass are inconclusive.
See Wine Hq. The license on Wine just (is in the process of?) changed from GPL to BSD to fit the goals of Wine better.
First of all, all old GPL code is still GPLed. They simply contacted all the authors they could find (and in fact gave them some time decide) Results were about 95% in favor, 2% opposed, and the rest unfindable. The 2% of coders were deemed insignificant in that it wouldn't be too difficult to re-code their contributions. The other will probably end up re-coded eventially if the authoers cannot be found.
There are pros and cons to all licenses, please do not get into an argument of should Wine have changed. They did, and unless you contribute code you have no voice in the change. The topic is can it be done, and obviously they have proven it can. If you want to do the same things read their archives, there is no point in repeating mistakes they made.
Great points, but missing something: traffic.
I live in Minnesota, and drive 20 miles to work. (Lets assume that works provides a recharging station so I have a full charge when I leave) Take a cold winter morning, -25, with a coule traffic accidents. It now takes not 30 minutes to get to work, but 75. Since the tempature is so cold I expect the heater to be running the entire time. The electric car does not have the range I need already, just running the heater on full.
I can write with either hand (not well, but I can) as an expiriment I signed all my credit cards with my right hand, and then all the sales slips with my left hand. Onle ONE clerk has noticed that my signature didn't match the one on the card and demanded further identification. (Once noting my drivers license had the same name, and signature and my picture on it he accepted the sale)
Somehow the above tidbit fits into this topic, and it forms some arguement in here. I don't know what though.
Really there are two problems, and nobody admits they have to be addressed seperatly.
First there is commercial tradmarks and the area the trademark covers. For instance:
McDonalds.com - the fast food restaruant we all know
McDonalds.com - the orginal restaruant that Ray Kroc bought the idea from (Are they still around, in any case the rights of that deal should handle this)
McDonalds.com - a local electrical company in some small city run by some brothers McDonald.
McDonalds.com Same as the last one, but diffierenty city. Maybe these guys are plumbers, whaterver.
McDonalds.com Again a small local buisness. Maybe a hairstylists that takes appointments on the web
You get the idea, a major recignised worldwide name, but it cannot do anything about the small reginal companies that are not compittion. Somehow the needs of all these comercial interests need to be met. (And I think we all agree that McDonalds restaruant and McDonalds heating can be in the same town without problem, but you can't have a second McDonalds restaruant in the world unless it has been McDonalds for 100 years)
On the other side we have personal sites. I want to be McDonalds.per because my family name is McDonald (okay, my last name is not McDonalds, but I know some McDonalds and it makes the point best) which when combined with the rest of my family allows email to bluGill@McDonalds.per and Jounior@McDonalds.per to get through easially. Since .per specificly dissallows anyone with a trademark from getting a domain we are free from the previous problems. Somehow however we need to solve the next one though: There are many families McDonald in the world, and most of them are not blood relatives. (you could up to Adam and Eve or the evoltion equevelent)
And a good choice was made too. This is interesting to me as a BSD fan, but it is still in the curiosity stage. If NetBSD would announce instead some break-through that allowed them to impliment SMP in 3 days that could run a single process 2x as fast on two proccessors as one, that would be front page worthy. (And most of us would wonder when the person who achived that is going to show the e!=mc^2, which should be about as easy a task)
I just wanted to post this so that it is known that not everyone thinks a bad decision was made.
I did this myself, not hard.
The local home depot sells cat-5 cable, and female ends that snap into a special wall plate. They also have telephone, coax and speaker female ends for this plate. (look first to see what you can get before making plans) Got some of that, and then find a way to get up to each room without making a mess. This is the hardest part if your not doing new construction. The walls in our second story do not line up with the walls to the first, and I could only get to the basement. (Outside walls of course line up, but they are full of insulation so that isn't useful) I did manage to find where the toilet drain went thoguh a wall, which was great since normally you can't get through floors in walls without drilling. Look in your house, if you don't see what I mean your house is probably built different.
For new construction, run plenty of cable, the stuff is cheap. For older construction the hard part is running cable so I ended up skimping. (For ethernet you can run two jacks on a normal cable if you know how to wire it.)
Run caox and phone for sure when you do this, and I'd run it all to one central localation in a wireing closest.
BTW, don't make the mistake I did: make sure you have some way to tell which jack in the wireing relates to which elsewhere. I have labels in the closet, but I can't tell left from right. Not good.
Try again. IBM did not lose money on mainframes last year. It isn't a high growth market, but it still makes up a large portion of IBMs money stream. Maybe 83-93 IBM lost money in mainframes, but last I check they were shipping mainframes as fast as they could make them. (part of that was Y2K hit mainframes the hardest) Nothing can process the volumn of data a mainframe can. (Okay, a cray can do it, but your talking mainframe class price too)
I'll second the Edmunds advice. I also used Kelly's blue book when I bought my car
Most dealers are onto the internet. That is they know that you know their price. Some (but by no means all) have realised that you know their fair profit and try to get things done quickly at a good price. After all, if you know their numbers they have to haggel different.
Two days ago I test drove a '95 S10. I liked it, so I drove it home, hit the internet and looked up the values. I then walked into the dealer with the knowledge of what they paid and what they expected to get it for. Since I knew my trade in better then them I was able to make them an offer that included a fair profit for them and a good price for me. I was out in an hour (and it only took that long because my trade-in was in my dads name yet, so I had wait for him to get that paperwork)
I would recomend using your local dealer first. That is do the research on the internet, and then walk into the lcoal dealer and make a fair offer. If they don't take it walk - if they don't call back in a day changing their mind (This is a common practice) check the internet. Although dealers are big buisness, they are local, and I like to support the local guy if possibal.
Supported Scanners Some USB scanners are now supprted, but expect USB support to be reworked in a year or two once people figgure out how USB works best. (ie right now USB is in a state of quick hacks and amazement that things work. Things are looking better, as work is going on, but I get the feeling that the programers are not sure how to do USB best/right
There are scsi scanners that are unsupported. There are one or two parrelle scanners that happen to work. There are usb scanners that will never work, and usb scanners where the manufacture hopes they will work soon.
I have a microtek V6USL which works just fine, though only were the very latests backend (which I don't belive is intigrated into any current versuion of SANE yet)