I agree that everything else you've said there is true. I simply disagree with your beliefs on how large the bases can be. A bitstream 125MB long would require two 524288000-bit numbers, or base 2^524288000. (Which is a hell of a number.)
Here, let's do this thing. Start with a binary number: 10110101 10011010 10010100 01101111 11010100. You'll notice this is 40 bits - far, far, far less than 10 kilobytes. Convert it into a base 32,640 number. Right now. Do it. You can even try converting it into two if you want. It's 40 bits, so that should be easy.
I'll give you a hint, before you start: You can't. You could convert it into two base 65536 digits, plus a base 256 digit. Or five base 256 digits. Or two base 32768 digits and a base 1024 digit. But you're confusing "bits" with "base". A 16-bit "digit" is the same as a base 65536 digit. A 32-bit "digit" is the same as a base 4294967296 digit. A 40-bit number - what we have here - would require a base 1099511627776 digit. In general, N bits takes a 2^N base digit to store, if you want to cram it into a single digit.
1000 bits cannot be stored in a base-1000 number. Not even close. 1000 bits wouldn't quite fit in a base-10^301 digit, in fact.
You overestimate the quality of average reporting and, again, fail to actually understand the mathematics involved. Please, follow the simple thought experiment I outlined in the last paragraph, then spend some time thinking about how that applies to your example.
An 8.5" by 11" sheet, at 300dpi with 24-bit color, can store at most 25,245,000 bytes of data. This assumes zero redundancy, the ability to detect 24-bit color, and no alignment needed. This is the maximum. "Geometrically encoded data" doesn't gain you anything on top of that - a raw scan of that sheet with those parameters will be that many bytes in size, and there's simply no way to synthesize more data magically out of the scan.
You do not seem to understand how mathematics works.
Your first step, for example, is hideously flawed. Take step 2. Each matrix can hold an image painted in one of 256 colors, on a background painted in one of 255 remaining colors. Okay, fine. However, this doesn't give us 85000*256*255 bits of data. In fact, this gives us 85000*log2(256)*log2(255) bits of data, or approximately 1.36 million bits.
Again, using an L shape, and rotating it, does let us have 4 distinct positions. However, that doesn't quadruple our data capacity. That quadruples the *number of possible states* the image is in . . . giving us a whopping two bits of extra storage. Adding another L gives us another two bits. Adding a center dot, or not, gives us another two bits. So now we've got approximately 1.360006 million bits. Congratulations.
If you disagree with this, here's a proof. Take a single byte. Obviously, we can store one byte of data in it. Now add a single bit. This can be in one of two positions, thus *doubling our storage*! Do this 32 times. Now, with a single byte, and 32 bits, we can store 4 gigabytes of data! Oddly, though, this doesn't work.
But he's got a financial reward for it - he's been selling a recording for fifty years.
That's not a bad reward at all. Considering how valuable shelf space is in music stores, if a CD isn't selling much it's not going to be selling at all. Fifty years of sales should make quite a tidy amount of cash.
I have very limited respect for authority - I assume they're in authority for a good reason, but if that's disproven they will very very rapidly lose that respect.
I agree that if they're willing to take full responsibility for their actions then that's an extremely good reason for respect. But as you mention, most people - in authority or not - simply aren't, and I'm not going to respect someone who wants the power with none of the responsibility.
Why would we boycott them? It's not like they're being forced to sit there and do anti-spam work - they're choosing to because (presumably) the pay or working conditions are better.
That's what the anti-sweatshop people fail to understand. It's not like high-priced lawyer jobs await these people if only they weren't being forced to make shoes for Nike. Working in this sweatshop is literally the best choice they have, often by quite a lot, and you want to . . . take it away from them?
As someone in the game industry, I care absolutely zero for what degree you have. Seriously. It makes no difference to me if you have a MS in game development or a PhD in agriculture. I simply don't care. If you wanted me to hire you, you'd have to have some proof of your skills - a game you worked on, a significant amount of code you'd done (or art, if you were an artist). Something that can prove you actually know what you're doing, and not simply that you have a piece of paper.
The "game degree" path may push you through making an actual game. Or it might not. I really don't know, and I honestly don't care. Pick your classes based on what you'll learn from them, not what your diploma will say.
This assumes you want to get a job at one of the smaller more personal companies, not a code-monkey job at a behemoth company.
My ISP has a web-based configuration utility that allows me to set a server-side firewall to one of several default values. One of their options blocks several commonly-exploitable ports on Windows. I don't use those ports for anything, and I have my own firewall so those ports shouldn't reach my Windows boxes in any way whatsoever, but I set it to block them anyway. (This was the default setting, actually.)
Something similar would work fine. Block port 25 to SMTP by default and have a web config utility to change it. If you really wanted, you could set it up to email the user if they tried accessing port 25 when it was blocked ("You might be trying to get past this firewall. Or, you might have a virus. Here's how you can find out, and here's how you can disable it if you need . . . ")
My own tastes run to his middle works. His early pulpy stuff is often too juvenile, and the sexual liberation that he examined in Stranger became rambling and unfocused in everything after that. (Though his finale, To Sail Beyond The Sunset, struck me as a remarkable throwback and a fitting capstone to his works.) Try Time Enough for Love and Starship Troopers; at the very least as light sci-fi you should be able to read them pretty quickly.
This is pretty much what I was going to say, although I suspect my tastes lean more towards the early stuff - even in Stranger I felt he was spending too much time with sex-for-sex's-sake. The guy's a dirty old man, basically, in much the same way as Piers Anthony is. His early work was more distinctive and interesting than Piers's, and he came in on the forefront of scifi instead of the trailing edge. (And he didn't get trapped into writing an endless series of Xanth books.)
But his late stuff . . . man, it's worth reading for the hilarious constant-sex factor, but I can't say I'd recommend it past that. I read a full bibliography of his once, and with a few extraordinarily notable exceptions I honestly don't think it was all that good.:(
I changed my motherboard and Windows told me I had to reauthorize. I tried to use the "authorize across network" option but it said it didn't have a network connection. Since it wouldn't let me boot into my user account I couldn't find the issue.
I would have just called, but I was fiddling with hardware for troubleshooting purposes and I had no interest in calling again every hour. So I found a crack and installed it (luckily it would let me boot into safe mode). Worked great. (And Internet worked instantly. I have no idea why it wouldn't let me authorize the first time.)
The same crack is now installed on all my computers. Blow me, MS.
I have no expectation of my software quitting in 6 months and my data being held ransom.
Why do you think that? There are several software packages out there moving into a subscription form. There are many that stop providing bugfixes once your version is "old enough". ("Yeah, of course you can keep using our old software! It won't work on any OS released recently, but that's not our problem. Want the new version? It'll cost you $lots.") And there are some - especially web services - that simply change their terms of use or featureset occasionally.
If your data is locked into those, sorry, sucks to be you.
Now, perhaps this software package didn't have that issue, but it's nowhere near an uncommon event, and very, very far away from being a strawman.
This is a pretty lousy justification:) Imagine if I offered to give you a free software package that saved all its files in a proprietary locked and encrypted format. In six months the software package will stop reading or writing and you'll have to pay $10,000 to get back into your files.
But it's free! I'll install it for you right now.
(I'm not saying this is the situation with you, just that your reasoning sucks;) )
I had to spend an hour and a half bypassing their "copy protection" software just to play it. I have Daemon Tools installed. It didn't like this. Now, if I'd pirated it I'd be saying "well played, EA". But I didn't. I paid for it.
As much as I want to play BF2142, I'm not doing that again.
I don't see why "when". Nuclear power is getting increasingly cheap and while the resources are technically limited, they're not running out anytime this millenia. Plenty of time for fusion to become practical.
Power has gotten cheaper since day one, why would it suddenly start getting more expensive? Short periods of time, yes. Long periods of time, no.
This would be more true if solar panels lasted longer, but sadly they slowly degenerate over time. I believe replacement is suggested after 25 years, at which point they have - assuming present energy prices - rarely paid for themselves. They might pay for themselves if electrical power gets more expensive.
All you need to do is trick someone into installing it as root/admin.
Why do you need to do that? Last time I checked, it was possible to send email from a normal user account.
I do agree that Linux makes root access much harder than Windows does, but it turns out that root access really isn't needed for most interesting (read: profitable) exploits.
The only reason the name has to be changed to "Iceweasel" is because the Debian team wanted to make changes to the package. If it was closed-source, that wouldn't be possible in the first place.
So it looks to me like open-source only gives more abilities in this case, not less.
(Yes, I realize that the reason they wanted package changes was because it conflicted with their license. That's rather tangential to the discussion.)
Well duh, of course it's caused by a programmer. Doesn't mean there's a good reason for it, it just means that someone was lazy or not particularly thoughtful.
Good to know that that's what's actually occuring, though. Even if there isn't an obvious way to fix it. Sigh.
I don't think Java's even installed on this computer:D but Javascript is.
I think it's the image looping, honestly, because a lot of pages are on a forum with avatars, and some people have animated avatars. But I kind of like the animations - many of them are clever and amusing. So I want to leave those on. But there's no reason they should be eating CPU when the window is minimized or when I'm looking at a different tab.
In fairness, I do have literally a hundred tabs open. But I still don't understand the constant CPU usage. (60%, 480MB RAM. The RAM usage goes up and down moderately unpredictably, the CPU sometimes does but it's rarer.)
What about its constant excessive CPU usage? Right now Firefox is using about 60% of my CPU, and it's minimized. (Before you ask, I also have Flashblock, so there shouldn't be any Flash apps running.)
Not necessarily. My code is littered with TODOs and FIXMEs - things that should technically be changed, either for speed, cleanness, or correctness's sake. However, the first two aren't likely to break a program, and the latter I always put traps around so if the "incorrect" cases are tickled the program shuts down cleanly.
Often, it turns out that the "incorrect" cases simply never get touched.
Programmers, it turns out, have a limited amount of time available, and have to prioritize and decide what fixes are most important. Making perfect 100% bulletproof beautiful code is rarely worth the time unless you're getting paid by the hour and your employer has nothing better for you to do.
Slashdot Mirror
Why will it be the 181st glyph?
I agree that everything else you've said there is true. I simply disagree with your beliefs on how large the bases can be. A bitstream 125MB long would require two 524288000-bit numbers, or base 2^524288000. (Which is a hell of a number.)
Your math is just as flawed now as it was before.
Here, let's do this thing. Start with a binary number: 10110101 10011010 10010100 01101111 11010100. You'll notice this is 40 bits - far, far, far less than 10 kilobytes. Convert it into a base 32,640 number. Right now. Do it. You can even try converting it into two if you want. It's 40 bits, so that should be easy.
I'll give you a hint, before you start: You can't. You could convert it into two base 65536 digits, plus a base 256 digit. Or five base 256 digits. Or two base 32768 digits and a base 1024 digit. But you're confusing "bits" with "base". A 16-bit "digit" is the same as a base 65536 digit. A 32-bit "digit" is the same as a base 4294967296 digit. A 40-bit number - what we have here - would require a base 1099511627776 digit. In general, N bits takes a 2^N base digit to store, if you want to cram it into a single digit.
1000 bits cannot be stored in a base-1000 number. Not even close. 1000 bits wouldn't quite fit in a base-10^301 digit, in fact.
You overestimate the quality of average reporting and, again, fail to actually understand the mathematics involved. Please, follow the simple thought experiment I outlined in the last paragraph, then spend some time thinking about how that applies to your example.
An 8.5" by 11" sheet, at 300dpi with 24-bit color, can store at most 25,245,000 bytes of data. This assumes zero redundancy, the ability to detect 24-bit color, and no alignment needed. This is the maximum. "Geometrically encoded data" doesn't gain you anything on top of that - a raw scan of that sheet with those parameters will be that many bytes in size, and there's simply no way to synthesize more data magically out of the scan.
The article is bunk. That's all there is to it.
You do not seem to understand how mathematics works.
Your first step, for example, is hideously flawed. Take step 2. Each matrix can hold an image painted in one of 256 colors, on a background painted in one of 255 remaining colors. Okay, fine. However, this doesn't give us 85000*256*255 bits of data. In fact, this gives us 85000*log2(256)*log2(255) bits of data, or approximately 1.36 million bits.
Again, using an L shape, and rotating it, does let us have 4 distinct positions. However, that doesn't quadruple our data capacity. That quadruples the *number of possible states* the image is in . . . giving us a whopping two bits of extra storage. Adding another L gives us another two bits. Adding a center dot, or not, gives us another two bits. So now we've got approximately 1.360006 million bits. Congratulations.
If you disagree with this, here's a proof. Take a single byte. Obviously, we can store one byte of data in it. Now add a single bit. This can be in one of two positions, thus *doubling our storage*! Do this 32 times. Now, with a single byte, and 32 bits, we can store 4 gigabytes of data! Oddly, though, this doesn't work.
But he's got a financial reward for it - he's been selling a recording for fifty years.
That's not a bad reward at all. Considering how valuable shelf space is in music stores, if a CD isn't selling much it's not going to be selling at all. Fifty years of sales should make quite a tidy amount of cash.
I have very limited respect for authority - I assume they're in authority for a good reason, but if that's disproven they will very very rapidly lose that respect.
I agree that if they're willing to take full responsibility for their actions then that's an extremely good reason for respect. But as you mention, most people - in authority or not - simply aren't, and I'm not going to respect someone who wants the power with none of the responsibility.
Why would we boycott them? It's not like they're being forced to sit there and do anti-spam work - they're choosing to because (presumably) the pay or working conditions are better.
That's what the anti-sweatshop people fail to understand. It's not like high-priced lawyer jobs await these people if only they weren't being forced to make shoes for Nike. Working in this sweatshop is literally the best choice they have, often by quite a lot, and you want to . . . take it away from them?
I remain confused.
Not yet, unfortunately, and not for quite a while unless you're an artist. ;)
As someone in the game industry, I care absolutely zero for what degree you have. Seriously. It makes no difference to me if you have a MS in game development or a PhD in agriculture. I simply don't care. If you wanted me to hire you, you'd have to have some proof of your skills - a game you worked on, a significant amount of code you'd done (or art, if you were an artist). Something that can prove you actually know what you're doing, and not simply that you have a piece of paper.
The "game degree" path may push you through making an actual game. Or it might not. I really don't know, and I honestly don't care. Pick your classes based on what you'll learn from them, not what your diploma will say.
This assumes you want to get a job at one of the smaller more personal companies, not a code-monkey job at a behemoth company.
My ISP has a web-based configuration utility that allows me to set a server-side firewall to one of several default values. One of their options blocks several commonly-exploitable ports on Windows. I don't use those ports for anything, and I have my own firewall so those ports shouldn't reach my Windows boxes in any way whatsoever, but I set it to block them anyway. (This was the default setting, actually.)
Something similar would work fine. Block port 25 to SMTP by default and have a web config utility to change it. If you really wanted, you could set it up to email the user if they tried accessing port 25 when it was blocked ("You might be trying to get past this firewall. Or, you might have a virus. Here's how you can find out, and here's how you can disable it if you need . . . ")
This is pretty much what I was going to say, although I suspect my tastes lean more towards the early stuff - even in Stranger I felt he was spending too much time with sex-for-sex's-sake. The guy's a dirty old man, basically, in much the same way as Piers Anthony is. His early work was more distinctive and interesting than Piers's, and he came in on the forefront of scifi instead of the trailing edge. (And he didn't get trapped into writing an endless series of Xanth books.)
But his late stuff . . . man, it's worth reading for the hilarious constant-sex factor, but I can't say I'd recommend it past that. I read a full bibliography of his once, and with a few extraordinarily notable exceptions I honestly don't think it was all that good.
I changed my motherboard and Windows told me I had to reauthorize. I tried to use the "authorize across network" option but it said it didn't have a network connection. Since it wouldn't let me boot into my user account I couldn't find the issue.
I would have just called, but I was fiddling with hardware for troubleshooting purposes and I had no interest in calling again every hour. So I found a crack and installed it (luckily it would let me boot into safe mode). Worked great. (And Internet worked instantly. I have no idea why it wouldn't let me authorize the first time.)
The same crack is now installed on all my computers. Blow me, MS.
I have no expectation of my software quitting in 6 months and my data being held ransom.
Why do you think that? There are several software packages out there moving into a subscription form. There are many that stop providing bugfixes once your version is "old enough". ("Yeah, of course you can keep using our old software! It won't work on any OS released recently, but that's not our problem. Want the new version? It'll cost you $lots.") And there are some - especially web services - that simply change their terms of use or featureset occasionally.
If your data is locked into those, sorry, sucks to be you.
Now, perhaps this software package didn't have that issue, but it's nowhere near an uncommon event, and very, very far away from being a strawman.
In my case the SW was free so why not use it
:) Imagine if I offered to give you a free software package that saved all its files in a proprietary locked and encrypted format. In six months the software package will stop reading or writing and you'll have to pay $10,000 to get back into your files.
;) )
This is a pretty lousy justification
But it's free! I'll install it for you right now.
(I'm not saying this is the situation with you, just that your reasoning sucks
I bought Battlefield 2 when it came out.
I had to spend an hour and a half bypassing their "copy protection" software just to play it. I have Daemon Tools installed. It didn't like this. Now, if I'd pirated it I'd be saying "well played, EA". But I didn't. I paid for it.
As much as I want to play BF2142, I'm not doing that again.
I don't see why "when". Nuclear power is getting increasingly cheap and while the resources are technically limited, they're not running out anytime this millenia. Plenty of time for fusion to become practical.
Power has gotten cheaper since day one, why would it suddenly start getting more expensive? Short periods of time, yes. Long periods of time, no.
This would be more true if solar panels lasted longer, but sadly they slowly degenerate over time. I believe replacement is suggested after 25 years, at which point they have - assuming present energy prices - rarely paid for themselves. They might pay for themselves if electrical power gets more expensive.
Why do you need to do that? Last time I checked, it was possible to send email from a normal user account.
I do agree that Linux makes root access much harder than Windows does, but it turns out that root access really isn't needed for most interesting (read: profitable) exploits.
The only reason the name has to be changed to "Iceweasel" is because the Debian team wanted to make changes to the package. If it was closed-source, that wouldn't be possible in the first place.
So it looks to me like open-source only gives more abilities in this case, not less.
(Yes, I realize that the reason they wanted package changes was because it conflicted with their license. That's rather tangential to the discussion.)
Well duh, of course it's caused by a programmer. Doesn't mean there's a good reason for it, it just means that someone was lazy or not particularly thoughtful.
Good to know that that's what's actually occuring, though. Even if there isn't an obvious way to fix it. Sigh.
I don't think Java's even installed on this computer :D but Javascript is.
I think it's the image looping, honestly, because a lot of pages are on a forum with avatars, and some people have animated avatars. But I kind of like the animations - many of them are clever and amusing. So I want to leave those on. But there's no reason they should be eating CPU when the window is minimized or when I'm looking at a different tab.
In fairness, I do have literally a hundred tabs open. But I still don't understand the constant CPU usage. (60%, 480MB RAM. The RAM usage goes up and down moderately unpredictably, the CPU sometimes does but it's rarer.)
On my system, it's currently using 600MB of RAM (and 60% of the entire system CPU) while minimized.
:)
So I guess on average, it uses about 325MB of RAM?
What about its constant excessive CPU usage? Right now Firefox is using about 60% of my CPU, and it's minimized. (Before you ask, I also have Flashblock, so there shouldn't be any Flash apps running.)
Oh, and 570MB of RAM. Yeesh.
Not necessarily. My code is littered with TODOs and FIXMEs - things that should technically be changed, either for speed, cleanness, or correctness's sake. However, the first two aren't likely to break a program, and the latter I always put traps around so if the "incorrect" cases are tickled the program shuts down cleanly.
Often, it turns out that the "incorrect" cases simply never get touched.
Programmers, it turns out, have a limited amount of time available, and have to prioritize and decide what fixes are most important. Making perfect 100% bulletproof beautiful code is rarely worth the time unless you're getting paid by the hour and your employer has nothing better for you to do.