Some people think because they've got the best known asymptotic run time that they are done. Others reach that point and then tweak the implementation and compiler options to go faster yet. I had some nice code that ran decent, but I spent months (a free time project) to get it faster. The theoretical run time has not changed, but in practice it is now 2x to 3x faster than when I started.
What did I do?
optimized data structures (not algorithms)
moved a critical function from one class to another - still puzzles me why 25%
considered branch penalties
one batch of auto-generated C code - refined several times
rewrote some functions as plain C
rewrote one complex expression in asm
It's a lot of work, but I spent at least as much time considering algorithms prior to the optimization effort. Both strategies are important for making fast code.
Besides, in these benchmarks everything was the same except the compiler version...
If you look at the original Hebrew, the word translated "day"...
Is there a place that collects these common mistranslations? I recall Wikipedia addressing the ten commandments and claiming that "Thou shall not kill" would be more accurately translated as "Thou shall not murder". This is also more consistent, since killing is not considerd bad in certain cases - like punishment. I've often wondered if there are a lot of mistranslated statements in there that could make the whole book more sensible if they were corrected.
If the earth was created in 6 time periods of more than a day, they should not use the word "day". Use "phase" or "period" or "era". I've also wondered if the people who lived to 900 years actually lived for 900 lunar cycles. When one reads about amazing stuff, he should question the sources or at least the translation in the case of this book.
"It will not be an easy sell, even though OpenOffice should more than satisfy all curricular needs"
Unfortunately the "curriculum" usually includes MS office applications. Yes, the skills are transferable, but try to explain that in a PTA meeting where people think their kids need to learn programs that are used in the "real" world. I'd still try for OOo, but expect a lot of resistance for reasons you can't begin to imagine.
From TFA "Did this lead the committee to wonder for a moment whether Europe should weaken its rights? No. Their response was that this showed we had to make the European rights much stronger."
IP laws are not about rights. They take freedom (to copy, to innovate, to use an algorithm, etc) from the public and grant exclusive permission to a private entity or individual. It's clear that if you want widespread innovation the solution is to loosen the restraints on the public - i.e. strengthen their rights.
I'm not totally against "IP". I think 14 years was an OK term for copyright, and that patents shouldn't apply to software or trivial ideas.
Do not be ashamed. Any effort that gets you laid (especially an automated one) is to be commended. I merely thought of this idea, but you implemented it. Then again, I already had a wife when it occured to me. Just another example of "if you can think of it, someone has probably already done it".
" I remember a while ago I heard that the largest data transport method was the US mail, and by a wide margin."
I get more physical junk mail than spam, so I would say this is true. Cheap bulk mail is subsidised by more expensive important stuff. What's a stamp cost these days? 35 cents? What's bulk? 5 Cents? I've heard all the arguments that this is appropriate, but I just don't buy them. The cost of the postman coming to the box is far higher than any other portion of it, and he wouldn't have to stop 5 out of 6 days if not for the junk mail.
" Nerds all over the world has been doing this for years in varous MMOG's attempting to get in touch with girls."
I always thought a "real" nerd would make a bot to pick up the girls for him and then just read the logs from the ones that actually email him afterward. This would seem to be a big time saver.
"We like open source because many OS programs are good or even very good, not because they are open source. Or don't we?"
I like Free Software (GPL) because of the license. As a consequence of this license, many programs are good or very good. I actually prefer Free Software to other open source. This attitude is rather common, but so is yours. In the end, most of this stuff exists because of the licensing model. One should respect that. Should we call it the "best" feature? Probably not. GPL or just OSS does not imply quality automatically.
Dude, if you want to follow a potential big disaster just go over to google news and search for Marburg. The current outbreak is on the verge of exploding in Angola and it's running with a 100% fatality rate this time. Read about the health care system there, the customs, the possible mis-reporting of the number of cases, the blog someone is doing... It may come under control soon. Let's hope so, because this asteroid thing sounds like a much better way to die.
I had the same idea about 2 years ago. I checked last night and it's written down in one of my notebooks. Just goes to show that if you think up an idea, chances are someone else has thought of it, or will shortly.
Should have gone for the patent back then;-) Actually, my problem isn't a lack of ideas, it's not having experience with starting a startup...
Let them compete with others. Once the government grants special privlege to one company, they become the local monopoly and prices go up. Yes, I'd rather have no network than a government granted monopoly acting as a revenue source. As you can tell, I've already got a network, so your question is not really valid. The first sign of trouble is when a government gets money by providing some service rather than having to figure out how to fund it. They become addicted to the income and then approve of anti-competetive actions taken by the "operator of the service" because it increases revenue without increasing taxes.
"Standard business practice. Get rid of the costly customers, or charge them more."
The ISP can already do as you suggest without signing an agreement with **AA. The question is what benefit do they get from signing with the **AA? I think the ISPs presently benefit by charging more to high bandwidth downloaders. Cutting them off would be a net loss unless **AA have something to offer to the ISP.
If government is going to proved a service (roads, parks, network, etc...) to the people, I think it should be available to everyone and paid for by taxes. Charging individuals for use is like going into business. Governments should not be in business. In a case like this, I can see the anti-competetive arguments from broadband providers making a lot of sense.
Provide it free and open with our tax dollars or not at all.
"if you want the most efficient solution then take a look at your hourly wage and make the hard decision."
That presumes you have the option of working that time to get extra money. Many of us are capped at 40 hours per week. My effective rate for evenings and weekends is $0.00. When I do contract work it goes back up, but that hourly rate argument falls appart unless you're talking about taking off paid work time to do whatever activity you're talking about.
If you've got more insight that makes the argument valid I'm listening - I have used it myself but felt I was just a scamming myself to get me to buy myself something expensive.
I like that the distribution originally picked one desktop (gnome) rather than burden the install media with duplicate packages for both. It's nice that they also now support the other (KDE) with a different CD. Me? I'm a gnome fan and don't want all that extra stuff to download, but it's nice that they support the KDE folks the same way now.
You're talking about (what's left of) downtown. The burbs are far better all around. The roads are getting lots of improvements just in time for the super bowl. I went to lunch today without my coat - it's almost 70 here already. Besides, Michigan will be the technology hot-spot for the robot building generation - just look at the results of the regional FIRST competitions. If you want electro-mechanical stuff, Detroit (or the surrounding area) is the place to design and build it.
"About the only good thing Detroit has for it is decent radio."
What if you only clicked the "I agree" button as part of your reverse engineering efforts, not because you really agree? Only the text of the EULA gives the button meaning - and only if you actually agree with it. Say some software archeologist 200 years from now manages to get one of those ancient PCs up and running and pops in an old CD to see what he's got. He'd be clicking "I agree" purely for research purposes....
I'm running FC3 and I wonder... After installing software from the net, or updating packages, how would you restore a machine to that same state if the HD failed? What if the packages were no longer available? I really like the idea of having everything on physical media. It'd be nice if everything you install or update over the net also gets added to a local "repository" that you can backup to a CD and "reinstall" at will to the same point. Automatically rebuilding from source is the ultimate version of this. Some day, I may want to reload an old disto with a bunch of upgrades onto what will then be "old" hardware. Depending on the net even in 2006 for FC3 seems a stretch. Other distos may vary, but the problem is the same.
Everything I've read says the star velocity should decrease with radius but observations show that it is nearly constant outside the galactic core. This "expected" decrease is because they model a galaxy as a bunch of stars orbiting a core mass and ignore the gravitation between stars - an incomplete model. My model (which is also incomplete) was a uniform disk of stars (at lattice points - also incorrect) and star velocity would actually increase all the way to the edge - in fact the rate of increase goes up near the edge. Obviously my model is wrong too, but I can certainly see that giving the disk some thickness and perhaps changing density (or thickness) with radius may result in a flat rotation curve. I dunno, but a flat uniform disk of stars exhibits the opposite problem with the curve.
Why should we suggest dark matter when varying the distribution of mass in our model can give the expected result? Does someone have a really good measurement of the density distribution in a galaxy available?
When someone says the curve should go down in a particular way (without dark matter), ask why they assume that.
Your editor has adjustable tabs. Some have automatic tools to reformat the code to a given indentation style. I set my editor to substitute spaces for tabs whenever possible - it's the only way I can use different editors and keep my sanity.
do you use Python? I think the opinions on spacing are less important when braces are not involved, although I still prefer 4.
After thinking about my post, I can now say it clearly:
Most of the style debate in C is over where to put the {braces}. Python simply eliminates the braces, so there is nothing to debate.
If you want to leave a blank line where your braces used to go that's fine. If you want to indent a different number of characters that's fine too. Python doesn't care how FAR you indent, just that you do. Given that, braces are wasted characters.
I haven't seen anyone who does not indent C code (ioccc aside). So just imagine your existing style with all the braces left out. That's the Python way.
After calling bullshit on a recent/. thread related to dark matter, I did the math myself. There is no reason to "invent" dark matter to account for the flat galactic rotation curves. A recent/. story also says physicists have decided there is no need for "dark energy" either. These guys need to stop making shit up. Anyone who mentions dark energy or matter automatically gets put on my "quack" list. Not to mention people who submit them to slashdot.
Slashdot Mirror
What did I do?
optimized data structures (not algorithms)
moved a critical function from one class to another - still puzzles me why 25%
considered branch penalties
one batch of auto-generated C code - refined several times
rewrote some functions as plain C
rewrote one complex expression in asm
It's a lot of work, but I spent at least as much time considering algorithms prior to the optimization effort. Both strategies are important for making fast code.
Besides, in these benchmarks everything was the same except the compiler version...
Is there a place that collects these common mistranslations? I recall Wikipedia addressing the ten commandments and claiming that "Thou shall not kill" would be more accurately translated as "Thou shall not murder". This is also more consistent, since killing is not considerd bad in certain cases - like punishment. I've often wondered if there are a lot of mistranslated statements in there that could make the whole book more sensible if they were corrected.
If the earth was created in 6 time periods of more than a day, they should not use the word "day". Use "phase" or "period" or "era". I've also wondered if the people who lived to 900 years actually lived for 900 lunar cycles. When one reads about amazing stuff, he should question the sources or at least the translation in the case of this book.
Should they transfer 1024 Terabytes and store it all in one chunk, they will have a PetaFile.
Unfortunately the "curriculum" usually includes MS office applications. Yes, the skills are transferable, but try to explain that in a PTA meeting where people think their kids need to learn programs that are used in the "real" world. I'd still try for OOo, but expect a lot of resistance for reasons you can't begin to imagine.
IP laws are not about rights. They take freedom (to copy, to innovate, to use an algorithm, etc) from the public and grant exclusive permission to a private entity or individual. It's clear that if you want widespread innovation the solution is to loosen the restraints on the public - i.e. strengthen their rights.
I'm not totally against "IP". I think 14 years was an OK term for copyright, and that patents shouldn't apply to software or trivial ideas.
BTW is she hot?
I get more physical junk mail than spam, so I would say this is true. Cheap bulk mail is subsidised by more expensive important stuff. What's a stamp cost these days? 35 cents? What's bulk? 5 Cents? I've heard all the arguments that this is appropriate, but I just don't buy them. The cost of the postman coming to the box is far higher than any other portion of it, and he wouldn't have to stop 5 out of 6 days if not for the junk mail.
I always thought a "real" nerd would make a bot to pick up the girls for him and then just read the logs from the ones that actually email him afterward. This would seem to be a big time saver.
I like Free Software (GPL) because of the license. As a consequence of this license, many programs are good or very good. I actually prefer Free Software to other open source. This attitude is rather common, but so is yours. In the end, most of this stuff exists because of the licensing model. One should respect that. Should we call it the "best" feature? Probably not. GPL or just OSS does not imply quality automatically.
Dude, if you want to follow a potential big disaster just go over to google news and search for Marburg. The current outbreak is on the verge of exploding in Angola and it's running with a 100% fatality rate this time. Read about the health care system there, the customs, the possible mis-reporting of the number of cases, the blog someone is doing... It may come under control soon. Let's hope so, because this asteroid thing sounds like a much better way to die.
That should be the first rule in hacking stuff like this. Premature hacks (err optimization) is the root of all evil.
Should have gone for the patent back then ;-) Actually, my problem isn't a lack of ideas, it's not having experience with starting a startup...
Let them compete with others. Once the government grants special privlege to one company, they become the local monopoly and prices go up. Yes, I'd rather have no network than a government granted monopoly acting as a revenue source. As you can tell, I've already got a network, so your question is not really valid. The first sign of trouble is when a government gets money by providing some service rather than having to figure out how to fund it. They become addicted to the income and then approve of anti-competetive actions taken by the "operator of the service" because it increases revenue without increasing taxes.
The ISP can already do as you suggest without signing an agreement with **AA. The question is what benefit do they get from signing with the **AA? I think the ISPs presently benefit by charging more to high bandwidth downloaders. Cutting them off would be a net loss unless **AA have something to offer to the ISP.
Provide it free and open with our tax dollars or not at all.
That presumes you have the option of working that time to get extra money. Many of us are capped at 40 hours per week. My effective rate for evenings and weekends is $0.00. When I do contract work it goes back up, but that hourly rate argument falls appart unless you're talking about taking off paid work time to do whatever activity you're talking about.
If you've got more insight that makes the argument valid I'm listening - I have used it myself but felt I was just a scamming myself to get me to buy myself something expensive.
I like that the distribution originally picked one desktop (gnome) rather than burden the install media with duplicate packages for both. It's nice that they also now support the other (KDE) with a different CD. Me? I'm a gnome fan and don't want all that extra stuff to download, but it's nice that they support the KDE folks the same way now.
"About the only good thing Detroit has for it is decent radio."
And don't forget about the Windsor ballet...
What if you only clicked the "I agree" button as part of your reverse engineering efforts, not because you really agree? Only the text of the EULA gives the button meaning - and only if you actually agree with it. Say some software archeologist 200 years from now manages to get one of those ancient PCs up and running and pops in an old CD to see what he's got. He'd be clicking "I agree" purely for research purposes....
I'm running FC3 and I wonder... After installing software from the net, or updating packages, how would you restore a machine to that same state if the HD failed? What if the packages were no longer available? I really like the idea of having everything on physical media. It'd be nice if everything you install or update over the net also gets added to a local "repository" that you can backup to a CD and "reinstall" at will to the same point. Automatically rebuilding from source is the ultimate version of this. Some day, I may want to reload an old disto with a bunch of upgrades onto what will then be "old" hardware. Depending on the net even in 2006 for FC3 seems a stretch. Other distos may vary, but the problem is the same.
Nature is now on my quack list.
I know no one publishes original work to /. I'm not that dumb, though you couldn't tell from my posts.
"The only dark matter is between your ears" -- Me
Why should we suggest dark matter when varying the distribution of mass in our model can give the expected result? Does someone have a really good measurement of the density distribution in a galaxy available?
When someone says the curve should go down in a particular way (without dark matter), ask why they assume that.
do you use Python? I think the opinions on spacing are less important when braces are not involved, although I still prefer 4.
Most of the style debate in C is over where to put the {braces}. Python simply eliminates the braces, so there is nothing to debate.
If you want to leave a blank line where your braces used to go that's fine. If you want to indent a different number of characters that's fine too. Python doesn't care how FAR you indent, just that you do. Given that, braces are wasted characters.
I haven't seen anyone who does not indent C code (ioccc aside). So just imagine your existing style with all the braces left out. That's the Python way.
After calling bullshit on a recent /. thread related to dark matter, I did the math myself. There is no reason to "invent" dark matter to account for the flat galactic rotation curves. A recent /. story also says physicists have decided there is no need for "dark energy" either. These guys need to stop making shit up. Anyone who mentions dark energy or matter automatically gets put on my "quack" list. Not to mention people who submit them to slashdot.