Slashdot Mirror


User: f97tosc

f97tosc's activity in the archive.

Stories
0
Comments
621
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 621

  1. Game theory... on Artificial Intelligence in Poker · · Score: 1

    To create a poker AI you just have to figure out what the odds are of getting something after your first hand. Then based on odds the computer will decide which cards to get rid of and which to go for ... It is able to calculate the odds of winning and is therefore able to make the best choice possible.

    Except, often the best strategy is not to play in a determinsistic way, but to have a probability distribution of different hands. In Poker this is essential, because if you play deterministically (i.e., always play the same way for the same hand) you are too predictable - you never bluff when you have a poor hand and when you have a good hand everyone else will know this so you don't get any action.

    Tor

  2. Re:Duh! on Wal-Mart Cancels RFID Trial · · Score: 1

    waiting for the cost to come down before using the RFIDs in the stores."

    Let's see.. completely revamp and replace all your cash registers and portable readers and software to use a product that you now have to pay $$$ for each item you sell...

    or stick with barcodes, your equipment already supports it and to put a barcode on a product is free (I.E. your products ALREADY comes with barcodes on them.)
    As I have consulted for large (Fortune 500) retailers on how they should use their inventory systems I feel qualified to respond to this post.

    The bar code idea is pretty obvious. You tag things when they get in. Quantities are deducted by the cash register when they sell; and thus you know your current quantity and when to order new stuff.

    In theory, that is. In practice, the guys checking in the stuff sometimes miss a box. Sometimes things get stolen. Sometimes something breaks and is thrown away. Sometimes there is a 2-for-1 offer that is incorrectly programmed - the customer gets the right price but instead of one strawberry and one orange, two strawberry are deducted from the system.

    The problem with this is that you lose data integrity. This messes up the ordering system. Say, for example, that the computer is set to generate an order when there are only 4 units left - and 5 units were stolen or broken. Then the shelf will stay empty, no-one will buy the product, and no new products will be orderded from the supplier. This is of course an economic catastrophy, much worse than the loss of the original 5.

    To avoid these problems you have to perform inventory checks. Yepp, you send out guys with scanners to physically coun't the number of products on the shelves. In the stores I worked with they did this once a month or so.

    Thus I too believe, that it is only a matter of time until these new tracking systems take over. Some day it will make business sense, because the cost of tags and revamped systems will be less than the labor costs of inventory controls and lost sales from missing products.

    Tor

  3. On air drag on NASA Test Shows Foam Could Be Culprit · · Score: 1

    Hum, there seems to be a lot of confusion around air drag. Let's clarify a few things.

    A good approximation for air drag is:

    F = C*S*D*V^2 (give and take some normalizing constant to get the right unit)
    Where C is a shape factor of the object (e.g., aerodynamic baseball - small C, broken foam - big C).
    S is the surface area of the object
    D is the density of the gas or liquid, in this case the athmosphere at the altitude of the accident
    V is the speed of the object, relative to the air. For the falling foam, this is initially the speed of the shuttle.

    Of course, the acceleration A is given by
    A=F/M = C*S*D*V^2 / M

    Thus, when I wrote 'massive' I really meant dense. What is interesting is the relationship between the surface area and the mass; the ratio S/M. This is no doubt greater for foam than for a baseball. And in a additon to this the shape factor is greater as well.

    Tor

  4. Re:This is the exact mistake NASA did on NASA Test Shows Foam Could Be Culprit · · Score: 1

    The foam probably wasn't falling, just slowing down due to air resistance. Whilst the shuttle was accelerating upwards.

    The impact speed was affected by gravity, air drag, and the acceleration of the shuttle. Of these three the first and last were of neglible importance. "falling" is admitively an imprecise term, perhaps "coming loose and being deaccelerated by air drag" is better.

    What's important here is the relative speed between foam and shuttle,which in turn is determined by the absolute air speed of the shuttle (and thus the foam, when it breaks loose). Because of the (non-linear) properties of air drag, different shuttle speeds causes vastly different impact scenarios for falling foam. For example, if the shuttle goes 50 times faster than a car, then a dropped item (such as loose foam) will experience a force 2500 times greater. Such a force is sufficient to cause a relative speed of 850 km/ hour in the distance before the foam hits the shuttle. On the other hand, if the foam had fallen off right after blast-off, when the shuttle was still at only say 100 mph, then this would never have happened.

    Tor

  5. Re:This is the exact mistake NASA did on NASA Test Shows Foam Could Be Culprit · · Score: 1

    The point isn't the large speed of the shuttle, but the acceleration. After the foam detached, the shuttle accelerated up into it. By the time the shuttle reached the position of the foam, the shuttle had accelerated *a lot*, therefore the velocities were very different.

    Sorry, this is wrong. It is true that an accelerating shuttle causes a speed difference between falling foam, but the effect is very small. Think about it. The shuttle was travelling at what, say 1000m/s. If the foam fell 10 m, this would happen in 0.01 s. The shuttle could not possible accelerate by 850 km per hour ~= 300 m/ s during that period. That would imply an acceleration of 30,000 m/s^2, or 3000 g.

    No, what happened with the falling foam was that it hit air drag. And as I pointed out in my original post, this is proportional to the square of the speed. Thus if the shulle was travelling at say 50 times the speed of a car, then a released item will deaccelerate 2500 times faster than the baseball in the original example (give some for the low density of the foam, take some for the thinner atmosphere).

    Tor

  6. This is the exact mistake NASA did on NASA Test Shows Foam Could Be Culprit · · Score: 4, Informative

    They figured that falling foam could not be so fast; it isn't in everyday situations.

    But common sense only applies to common situtations. In exotic situtions you have to use math and computers. Your basic intuition simply does not work.

    And the difference here is that the shuttle was going extremely fast. I don't know the exact speed, but much faster than 850 km/ hour.

    The math of this is that air drag is proportional to the square of the speed. On top of this the foam is much lighter than the baseball. So if the shuttle was travelling in say 4000 km / hour (~Mach 4); what will be the speed of the foam by the time it hits the shuttle?

    You have to do math and simulations for this one. NASA did, after the disaster, and you should not throw out the results (that the foam had slowed down to say 3150 km per hour) because of your everyday experience with speeds below 100 km/ hour.

    Your post is illustrative of how easy these mistakes are to make. In rocket science, you have to think about and calculate everything; because your intution does not work.

    Tor

  7. Re:Minor curiosity... on NASA Test Shows Foam Could Be Culprit · · Score: 1

    There's a lot of talk of repairs but I haven't heard any predictions for scenarios where repair was impossible.

    They are looking into materials that could be used to patch up holes. It is unlikely but possible that they will find something sufficiently strong and workable.

    Perhaps NASA should start looking at new designs with potentially fatal flaws. Have they not been using this design for something like 15-20 years now?

    I agree, but maybe they should wait until they have a plan with somehting really revolutionary on the table. Perhaps in ten years they could build a scramjet thing with much greater capacity. Building something new with similar but improved designs is not very cost-efficient; and while security could potentially be improved they may also introduce new bugs that take another disaster to identify.

    Tor

  8. Re:Why Techs Are Dweebs From Another Planet on Public Confused by Tech Lingo · · Score: 1

    Technological jargon, like all other jargon, has two purposes:

    1) To allow for more efficient communication on a particular topic
    2) To make the group that understands the jargon feel special and keep them apart from others

    We must acknowledge both of these in any serious discussion. Significantly, I think there is some jargon that has a needless amount of 2) in it. Why say 810.11 (or whatever, I always forget the right number) when all you mean is wireless (of course, the more precise term can be useful when comparing wireless protocols but not really when you are only pointing out that you have a wireless card). Simlarly, phrases like DVR are likely to confuse people and it really isn't that essential to save one second from saying the full words.

    On the other hand, there are many instances when 1) actually is important. There is no simple subsitute for Megapixel. You can say kinda good and really crisp but if you are going to invest in a digital camera it is a very good idea to learn about the most basic measurements of resolution.

    Tor

  9. Re:hmm... on Distributed Computing Economics · · Score: 1

    interesting thought, but what is the difference between this and the age old concept of the cost/benefit relationship...? im not trolling, it seems that it is jsut that concept with a tech twist.. br>
    Sure, it is just cost/benefit but that tech twist is what has been missing for years. Important IT decisions have been left to the preferences of IT people instead of conducting a cost/ benefit analysis for the company as a whole. I think this tech twist is very imporant, and we will absolutely see more of it in the future.

    Tor

  10. Re:What you think we dont know its illegal? on Freenet Creator Debates RIAA · · Score: 1

    So what, sharing is still right.

    I think that is for the author to decide.

    What boggles my mind in the case of music sharing is that the authors (artists) are still in bed with an expensive and copmletely obsolete middle-man (RIAA) instead of providing cheap high-quality downloads - where they themselves get all the money.

    Tor

  11. Quite so, on Estonia: Where the Internet is a Human Right · · Score: 1

    and you have a right to running water, and a right to a 40-hour work week, and a right to Internet access, and a right to a refrigerator, and a right to 99-cent cheeseburgers with your Super Club card", governments cheapen

    I like this point, and I would like to add a bit to it. The difference with "the right to free speach" and "the right to Internet" is in what obligations these give to other people.

    In the former case, there is no obligation to other people, but in the latter, all of a sudden someone has the obligation to provide Internet. So, if you are smart/ hardworking/ lucky it is your obligation to provide the Internet for me who is less so.

    Of course, this can be handled through an organized government, but the point still holds. By adding a laundry lists of nice things to have and calling the "rights", you are very much infringing on more basic rights, because you are forcing some people to provide certain things for others.

    Tor

  12. Re:Not on a Mac it 'aint on Motherboard Audio Comes Of Age · · Score: 3, Insightful

    Try the following:
    1) play mp3 through decent stereo straight from (Quicksilver) Mac.
    2) Burn same Mp3 to CD and play through same stereo.


    Now try the above but have a friend randomly switch sound sources while you look the other way. No trolling, but can you still sense the difference?

    There is a _lot_ of psychology in sound systems; oftentimes even the true audiophiles fail blind tests between pieces of equipment of which they have very different opinions.

    Tor

  13. Re:Hopefully it'll be trend. on Microsoft Considers $10 Billion Dividend · · Score: 1

    It'd be nice to see dividends once again be the main way most people expect to make money with stocks

    Well a company can do 2 things with profits.

    1) Give it to the shareholders

    2) Reinvest it in the company, so that they can do more of 1) in the future

    The decision rule is (or should be) that if you think that reinvesting it gives a greater payoff than average interest/ or other stock, then you do 2, otherwise 1.

    Doing a lot of 1) tells the market that MS does not see a lot of growth. They can keep making money the way things are, but they don't see themselves getting much bigger.

    THere is nothing wrong or unstable about reinvesting money in a company, it is a sign of growth and larger dividends in the future.

    But I agree with you that it is crazy how many people think they can outcompete the market by specutalive trading. But that has nothing to do with current vs future dividends, it is just whether you think you have a better predictions of future earnings than everybody else.

    Tor

  14. Re:Privacy implications are nill on Twist on DNA Privacy · · Score: 3, Insightful

    "This man is definately the murderer because his DNA matches", would you find him guilty? I think most people would.

    OJ walked.

    The problem is that we are dealing with statistics and they are often wrong.

    Whereas of course, an emotional witness has 100% accuracy? I think DNA as evidence is great, not because it is completely impossible to get it wrong, but because it is so superior to the alternatives.

    If it now turns out that incomplete DNA can give a false postive then this should be carefully studied. Next time they will no that the DNA shows that "it is either him or a close relative". THen you can start talking to close relatives and see if they have an alibi.

    Tor

  15. Re:Well which lasts longer? on Ink More Expensive Than Champagne · · Score: 1

    I mean, you drink some champaigne, take a piss an hour later, and it's gone.

    On the other hand, I picked up my fiance while drunk. That was two years ago; she is still around.

    Tor

  16. hum on Microsoft Patenting IM Translation? · · Score: 4, Insightful

    you are certainly right that it is the claims that matter. However, one thing should be noted about the claim. A claim that is phrased this way means that ALL the components listed have to be included for the patent to apply.

    A method for translating instant messages exchanged between two or more devices over a network by one or more users that communicate in different languages, the method comprising: establishing a user profile indicating at least one user language and one or more translation preferences of the one or more users; receiving a message as input composed by at least one of the users according to the user language; translating the message from the user language to at least one different language corresponding to the one or more translation preferences; and transmitting the message in translated form to at least one of the two or more devices.

    Thus if you only change one of these parts you have successfully circumvented the patent. Long claims like these ones may seem powerful, but in fact the opposite is true. Generally it is the short ones that have the biggest coverage.

    One thing that comes to mind is that the message must be transmitted in translated form. If you transmit it first with a language tag and have the other user translate it then you are OK. And since that solution has now been discussed in a public forum it can never be patented.

    Tor

  17. Re:Rubbish on Leave Outer Space to the Millionaires · · Score: 1

    Given the infrastructure it takes for space exploration of any significant magnitude, how many individuals are going to pursue it just because they can? I would suspect not many.

    Not many, no surely they would be a fart in space compared to the myriads of government sponsored manned trips all over the solar system. ;-)

    Tor

  18. Re:Space should be left to corperations on Leave Outer Space to the Millionaires · · Score: 1

    it seems the profit motivation leads to things which are counter-productive..

    Time for some empiric observations. During the last 100-150 some countries have permitted and encouraged profit motivation, others have outlawed it and put the development in the hands of an intelligent elite. So where did we get the highest standard of living? The most efficent use of resources? The most scientific breakthroughs?

    ...

    I rest my case.

    Tor

  19. Re:Spaceflight? on DARPA Looking into Hypersonic Bombers · · Score: 1

    How the heck is a supersonic _combustion ramjet_ going to help you in space? Unless you're carrying a metric crapload of O2...

    Well not in space proper, but in getting there which is the main part of the high costs of spaceflight. Reusable scramjets burning air could in principle replace first stage rockets as we know them today.

    Tor

  20. Scramjets on DARPA Looking into Hypersonic Bombers · · Score: 1

    are really cool, they go quick as h3ll but need much less fuel than rockets.

    If they actually manage to develop scramjets there are a lot of more applications than bombers.

    Cheap space travel comes to mind.

    Tor

  21. Re:When is business speak going to die? on Java Database Best Practices · · Score: 2, Insightful

    "Best practices" is common sense for 300$ an hr brought to you by an IBM consultant.

    This is true, but many IT departments don't have people with common sense, so this can still be a worthwhile investment. Better, say, than getting new hardware or software based on the IT departments own recommendations.

    Tor

  22. Game theoretic analysis on eBay Provides No Privacy For Sellers · · Score: 2, Interesting

    1 Buyers prefer that the order can be traced -> higher demand where this is possible

    2 Some sellers may prefer that it can't

    3 Thus if there are several possible sales portals, sellers have to chose between higher demand and prices, or the feature that they cannot be traced

    4 Positive feedback: what kind of sellers prefer intracability over more money?

    I think this is here to stay.

    Tor

  23. If there were 2 Ebays, on eBay Provides No Privacy For Sellers · · Score: 2, Interesting

    one were police could get names and addresses of sellers and one not, where would you buy and sell?

    I think I would settle for the one where they could. It seems like some protection against fraud - and I don't really mind if they get a hold of my name and address, or that I sold some used computer book.

    Tor

  24. Translations on Bill Gates On Linux · · Score: 1

    USA TODAY: Nobody used OS/2.

    BG: Are you kidding? I mean, let's be serious. That was IBM, a company 15 times our size.

    BG Translation: We owned them. I cannot say this, less it appears as though we have been a monoply for a long time

    USA TODAY: There has been some criticism of the way in which you're been competing against Linux,...about undercutting Linux at any cost, per se.

    BG: Well I'm not sure what you mean by undercutting.

    BG Translation:I know exactly what you mean by undercutting. (How the h3ll did this come out?)

    USA TODAY: Is there a scenario by which you would at some point consider porting Microsoft applications into Linux?

    USA TODAY Translation: I am a moron.

    BG: There's no [need for] consideration of that at this point.

  25. Re:It's a JOKE people, sheesh on The Bug · · Score: 3, Funny

    I'm amazed at the responses I've been getting. My post was a JOKE

    Sorry, you can't joke about everything. Joking with nerds about their lack of girlfriends is a no-no. You really hit a sensitive spot.

    Tor