For general-purpose note taking, paper is great. Unless your brain works differently than mine does, you will find it too distracting and slow to take notes on a PDA, or even a laptop. And I agree with the people who said that transcribing your notes from scrawls on paper into computer notes will help you really fix the material in your head.
I think you should have a PDA. You can get a Visor Deluxe on eBay for $30 now, and that will work just fine. I really like my Palm Tungsten T, but it costs ten times as much, so don't get one unless you are really not worried about it getting stolen. You can listen to tunes with a Palm Tungsten T if you want, which is a major plus. (I listen to my Ogg files using AeroPlayer.)
For those times when you might want to take notes from a book in the library, a keyboard for your PDA will be great. I'd rather carry my Palm and a portable Palm keyboard, than carry a laptop around. And the keys are quiet on my PDA keyboard, so I don't make clicky-clicky noises to annoy other people.
As for a computer, I agree that you need to scope out the dorm rooms before you commit to a particular computer. If only a laptop will fit in your study space, do get a security cable to secure it. Don't plan on carrying the laptop around campus; it's heavy and annoying, and you are unlikely to actually do it for long.
You will want a nice printer. I suggest an HP DeskJet printer for all-around goodness. You can make very clean black-and-white pages for English papers, you can put color graphs in for science papers, you can put photos in for art history papers... with special paper you can even print photos. You probably don't want the tiny compact folding printer, because that holds very little paper. Just get a desk model that can hold a couple hundred sheets of paper.
Murphy's Law says you will run out of ink, or toner, or whatever your printer uses to print, in the middle of the night the day before a paper is due. With a deskjet, you can get a spare black ink cartridge for $35 or so. Cheap insurance. If your color runs out you can always print in greyscale, so if money is tight just get the black one; otherwise one black and one color cartridge would be good to get.
Used to be, there was this cool place in Seattle called Entros. For a while there was one in San Francisco, too. Now they are gone, with no trace save a "demo" web site.
It was a restaurant/bar that had about five entertainment areas around it. The entertainment areas were always cool and different.
The most famous was "Interface", where one person wears a blindfold and a camera, and the other person sits in front of a screen watching what the camera sees; using two-way radio, the second guy tells the first guy where to go and what to do. "Go left, step forward, reach down, no, left, no, LEFT, feel for the ball, YOU TOUCHED IT! GO BACK!" Within a set period of time you had to accomplish certain tasks. If you got them all done, you were allowed to enter the victory lounge. I never got to see that lounge...
Anyway, they had a sort of game show where you had to compete to see who was the calmest. They would hook up the players to biofeedback, and then they would do various things to try to shake the players' calm. For the winner, it played a recorded voice saying something like "YOU ARE THE BUDDHA".
The delays in the next one weren't nearly as bad as the delays for the one mentioned in the article. Some distros had the patch out before SCO, but SCO wasn't the last. It really wasn't worth commenting on the second sendmail patch.
This months issue of a US linux magazine (Probably "Linux Magazine", but I'd need to go home to check) has a pretty favorable review of SCO Linux in it.
That's the June 2003 issue of Linux Journal, page 78. And I didn't think it was "pretty favorable". It was as neutral as possible. The part about the delays in the sendmail security patch was not at all favorable.
The May issue of a magazine usually comes out in April. It probably goes to the printers 6 weeks before being released, so that would put the magazine being created in each March, before the lawsuit.
The final draft was submitted to the magazine about a week after SCO announced its lawsuit, but most of the writing was before that.
The solid rocket boosters were an afterthought, hacked onto the shuttle when it turned out not to have enough thrust from the main engines. Any mishap with the solid rocket boosters is not survivable: if one fails, the unbalanced thrust from the other one dooms the shuttle. You cannot shut them off, you cannot jettison them until after they burn out on their own, you can't do anything with them but light them and hope they work right.
Do you really think this is a good design?
The shuttle is designed to come screaming down from orbit at huge speeds, then glide in to a really long landing on a really long landing strip. A thin layer of thermal tiles keeps it from burning up. There is no second layer of tiles behind the first layer as a backup. There is no way to repair a damaged tile in orbit. There is no way to abort a landing, fly around, and come back for a second attempt. Stresses on the landing gear are high, and if two tires blow you will damage the shuttle beyond repair.
Do you really think this is a good design?
What I would like to see is a simple design. It should be a single stage if possible, two stage if necessary. It will have one set of engines: the built-in rocket engines. It will use those same engines to take off, and to land. It will have enough engines that it can have multiple engine failures and still land safely. Landing under rocket power, it will not fly so fast it is in danger of burning up, it won't require a special extremely long runway, and it can abort a landing for any reason and come back around for a second landing attempt.
As long as we keep flying the shuttle, I'd like to see liquid-fuelled rocket boosters. At least if you have a problem, you can shut the fuel pumps off and get both boosters to instantly shut down.
[landing under rocket power] takes insane amount of fuel.
Not so. By the time you are landing, the rocket is very light because the fuel tanks are nearly empty. It takes a lot of fuel to fight gravity and get into space, and when you land most of that fuel has been used up.
The shuttle cannot and will not ever land under rocket power. But a future spacecraft, perhaps an SSTO one, can be designed to do it. And while I am not a rocket scientist, or even an insightful amateur, other people who have done the math say it's the right way to go.
The great part about this is that it doesn't mess up the layout as often as setting the minimum size font does. Also, it doesn't affect printing at all (I don't remember whether setting the minimum size affects printing or not, but this sure doesn't).
f you middle click on a link, open link in a new tab. If you middle click off of a link, activate the autoscroll. Simple as that.
And, in fact, Galeon has already implemented this in exactly the way you describe.
Plus, as a bonus, if you select a URL and then middle-click on the "New Tab" button on the toolbar, it will open a new tab with that URL loaded! I really love that.
That whole concept of actually using ambient "air" rather the having to carry liquid oxygen to get your vehicel off the ground makes a fair amount of sence to me.
The problem is that liquid oxygen is both cheap and pretty compact. And, once you burned it, it's gone. Your scramjet design needs scramjet engines, and it will carry those engines to orbit and back. They only help during the initial takeoff phase, but the weight penalty is there 100% of the time. The same applies to the wings.
You want your spaceship to get up out of the atmosphere as quick as possible, so it won't be able to use the scramjet engines all that long anyway.
You can build a rocket that will take off, fly in space, and land... all using the same (rocket) engines. This will be a simpler design than some complicated thing that starts out with jet engines, switches over to scramjet engines when speeds are high enough, then switches over to rocket engines when there is no ambient air around (i.e. in space). Simpler is good.
There is some merit to the idea of a two-stage design, where the first stage can be air-breathing and can carry the second stage up to thinner air. The proponents of single-stage-to-orbit (SSTO) think you don't need the air-breathing lower stage, and that the operational cost will be lower if you have the SSTO. One stage is less than two stages, so it should take less maintenance, less manpower, and so on. But no one has seriously tried to develop a complete SSTO design, other than on paper; if it turns out we can't quite get a useful payload on an SSTO, we could always have an extra stage to get it on its way.
Well, except that Space is much more dangerous than flying around in the air. It's magnitudes more difficult *and* dangerous!
It's not as hard as you think.
A rocket is a fundamentally simple device. You have fuel tanks, some sort of pump, and a reaction chamber. There is no reason to think that all space vehicles will be tremendously unsafe like the shuttle.
You can always increase risk with stupid design, and I'm afraid that's what the shuttle did. Columbia broke up because of compromised thermal tiles; there is no way to fix tiles while in orbit, and the design of the shuttle is to come screaming down from orbit at huge speeds and finally land like a giant brick with wings. (And shuttle has no ability to fly past the landing point, swing round, and land on a second pass; it has to land in one pass with no mistakes allowed. Brilliant design.)
It isn't all that hard to land under rocket power, using rockets to slow the descent to reasonable speeds, and then you don't run the risk of losing the spacecraft and all people on board because a few thermal tiles got damaged. And you can have enough extra rockets that if some malfunction, you can land with the others. And you can easily back off and come round again if something happens when you are trying to land.
The key is to make redundancy and survivability the primary design goal, and to build prototypes and fly them to refine the design before you build your fleet. Shuttle was designed on paper, and there were no X-vehicle prototypes anywhere along the way. Stupid bits in the design stayed in.
Official NASA documents estimated that you should be able to fly the shuttle 10,000 times before you lost one ("five nines reliability"). The reality is much closer to 100 times (two nines). This is very poor. If airplanes would kill you one time in a hundred, I sure wouldn't want to fly on an airplane... and there is nothing inherent to space operations which justifies the poor record of the shuttle.
We need to replace it with something safer, and that is possible.
It's very simple. What we need are reusable ships with a modest cargo capacity, plus maybe a few "big dumb boosters" for launching big things.
It's also very clear that NASA is not capable, as an organization, of doing this. NASA has some smart people working there, but any really large project will safely bury the smart ones under red tape where they can't do anything. If you want to convert money into piles of paper, have NASA attempt to make a follow-up to the shuttle.
The US government should make iron-clad promises to buy launches. Station re-supply launches for the International Space Station would be a great place to start. If John Carmack's company, or any other company, can get a vehicle going that can run supplies to orbit, the government should hire them to do it. In other words, pay for results but for nothing else, and don't have any part of the government (especially NASA) trying to help design the ships.
Soldiers have always been trained to be desensitized to violence. They have a job to do, and they need to be doing it, not standing around going "Whoa! Violence! Eeeek!"
Perhaps the military uses video games now as part of the training, but I suspect they mostly still rely on the old-fashioned training techniques.
By the way, soldiers are not just killing machines who go around shooting anything that moves. There are rules, harsh ones, that govern what soldiers do. Not to say that no American soldiers have ever committed atrocities, but that's very much the exception rather than the rule.
once you do, APL is a much clearer way to write some hairy multidimensional array operations.
Maybe so, but the set of problems for which APL is the perfect solution is small enough that few are willing to master it.
Python, with serious math libraries (written in C) grafted onto it, can do hairy array operations too. And once you master Python, you can use it for more than just matrix math; you can do pretty much anything. I read an article about how a science lab was switching away from FORTRAN to Python for their matrix math; there was a quote something like "after trying Python, no one wants to go back to FORTRAN." Alas, I didn't keep the URL and I can't find the article now.
If you like APL, that's fine with me. I still like FORTH, even though some people call it a Write-Only Language.
Spoken like someone who hasn't coded much assembly...
In fact, I haven't coded much assembly, although I have done some. But you are missing the point here, dude. Would you use Applesoft BASIC (or assembly) to teach an Introduction to Programming class to high-school kids?
No. You want a language that lets you focus on the basic concepts of programming, without having to learn tons of annoying syntax. Pascal is better than BASIC for this. Java is better than Pascal. Python is probably best.
It could have been worse: about 20 years ago, my high school had an intro to programming course using Applesoft BASIC on Apple ][ computers. Students had to learn a whole lot about how broken Applesoft BASIC is. Just one example, in both C and Applesoft BASIC:
if (flag)
DoSomething(3); else
DoSomeOtherThing(4);
// next statement
120 IF FLAG = 0 GOTO 160 130 GLOBAL1 = 3 140 GOSUB 600 150 GOTO 180 160 GLOBAL2 = 4 170 GOSUB 700 180 REM next statement
600 REM DoSomething() -- argument is GLOBAL1
700 REM DoSomeOtherThing() -- argument is GLOBAL2
No block structuring: you can only have a single statement after an IF. For anything else, you have to test for the opposite of what you really want to test for, and GOTO the line number of the ELSE case. Then you fall through to the IF case if you didn't take the GOTO. At the end of the IF case, don't forget the GOTO to skip over the ELSE case to the next statement! If you want a subroutine that will have some sort of argument, you need to use a global variable. (You could actually use the same global for both functions, since there is no danger of multiple threads of execution or anything like that.)
Modern languages are so much nicer than the evil old BASIC languages.
Better would be languages which are self-documenting...
There is no language that will force perfect code. There is always room for a poor programmer to produce hard-to-understand code. Functions that do two unrelated things, confusing control flow, bad variable names, broken code that was repeatedly patched instead of being cleaned up... the possibilities are endless.
Nonetheless, some languages have been designed with self-documenting code in mind; sometimes it even works.
If you look at languages like COBOL, they have long descriptive keyword names designed to make the code easy to read. But you get tired of looking at those long keywords.
I haven't used ADA, but I understand that it is somewhat designed for self-documenting code, and that as a result you are hemmed in on all sides by language rules. (ADA fans please comment here.)
The best language I have seen for this is Python. As a rule there is exactly one way to do things, so you don't trip over obscure hackish tricks that you have to puzzle out. The language doesn't force self-documenting or comments, but it does force indentation; everyone indents their Python pretty much the same (compare with the mess that is C indentation). The language is high-level enough, with lots of libraries, so you don't need to write 10 lines of code just to do one simple thing.
Python was designed by a guy who is both a computer geek and a math geek. The math geek in him led to a very tidy language design, and I like it very much. I think schools ought to be using Python to teach introductory programming classes.
I stand corrected. I don't know where I read that the revision A would not support OS X, but clearly it does. The official upgrade requirements page clearly shows the "Bondi Blue" iMac as a supported platform.
I note that the page is for version 10.2 of the OS. Has the revision A been supported all along, and my info was completely wrong, or was it not supported in 10.0?
Did that ad actually convince people to buy Apple computers, or was it just a "we're so cool" thing that was a waste of advertising budget? I suspect the latter.
We have seen some UI innovation, and we will see more.
Before you can raise the bar, you have to reach the bar where it is. The Mac has been shipping, with a UI, since 1984. Windows has been around in one form or another about as long. KDE and GNOME have had much less time to get their acts together.
Both KDE and GNOME now are decent and usable. From there, they can try new things and perhaps advance introduce some UI innovations. In the past, most of the energy has been going into just getting things working; now, more energy will go into experiments and new ideas.
P.S. UI innovations tend to be subtle things. If you do radical, weird stuff, people don't like it.
Xerox came out with the GUI. Apple and Microsoft both started working on a GUI. Apple asked Microsoft to support MacOS (by releasing apps such as Word and Microplan for the Mac). Microsoft agreed, but required Apple to sign an agreement that Apple not sue MS over Windows. Apple signed the agreement. Later, Apple sued MS over Windows.
In the suit, Apple claimed that Windows infringed on a nebulous concept called "look and feel". The judge threw out all "look and feel" claims, and would only consider specific claims. Then the judge went down the list, throwing out any claim that was covered by the agreement. About a dozen claims were left. Then the judge went down the list, and threw out any claim that Apple didn't own (common GUI things that anyone could do without Apple's permission). Exactly zero claims were now left, and the judge dismissed Apple's suit.
Short version: Apple agreed not to sue, sued anyway, and the judge ruled that the agreement was perfectly valid and threw out the suit. Seems fair to me.
And, by the way, we would not now have either GNOME or KDE had Apple won that suit. Apple winning that suit would mean Apple owns GUI desktops, and that would be Very Bad Indeed. Forget free desktops for Linux. Even forget non-free desktops for Linux; Apple would insist you buy an Apple computer running an Apple OS if you wanted a GUI desktop. (And any Apple fans who want to claim otherwise: how many licenses has Apple sold to the TrueType patents?)
Hastings's Law: Cheaper and adequate wins against more expensive and better.
SCSI is better than IDE, but IDE is adequate. Beta had better quality than VHS, but VHS was adequate (and had much longer recording time on a tape). Apple was able to charge rapacious margins for a Mac in the days when x86 PCs were much harder to use; once Windows became adequate, customers started buying Windows versus the Mac.
A BMW 7 Series is a better car than a Honda Accord. But an Accord is adequate. Which sells more?
If you invested in Apple 15 years ago, they still honor your investment. I can't say that the same is true of MS
Are you talking about Apple Computer? The same Apple Computer that used to make the Newton, and the eMate, and just shotgunned those products and abandoned their users? (Remember how some companies tried to buy rights to the Newton OS so they could support the Newton community, and Apple refused to sell?)
Is this the same Apple Computer that promised that the original iMac would run OS X, and then quietly walked away from that promise?
Is this the same Apple Computer that allowed third party companies to make Mac clones, and then suddenly unilaterally shut down the lincensing deals? (That's more of a "screwing the clone vendors" thing than a "screwing their own customers" thing, but I'm still not happy about it.)
Apple makes some cool stuff, and the premium for it isn't as high as it used to be. (I remember the days when Apple was scoring 40% margins on their hardware! Gouging their own customers for the short-term profit.) For many people, a Mac is a reasonable thing to buy -- after all, if what you do is surf the net, you don't care that your computer is only about a gigaHertz instead of 3 gigaHertz. But to hold up Apple as some shining example of taking care of their loyal customers is a bit much.
I don't think Microsoft has ever just walked away from a group of customers. Even when Microsoft walked away from OS/2, it was because IBM was there to support the OS/2 customers. I have my issues with Microsoft these days, but on this particular issue I think they are the better company.
Spy Hunter for the PS2 or X-Box is a cool driving game. Not only do you drive, but you have machine guns, oil slicks, missiles, and other toys, and you use them to blow up dozens of bad guys.
When your car goes into the water, it morphs into a boat. If your car takes too much damage, it jettisons you in a motorcycle escape vehicle. If the motorcycle goes into the water, it morphs into a jet ski!
I feel this is a bit of a sleeper game -- hardly ever mentioned on websites I visit, yet it is really fun. I love locking missiles on the bad guys and blowing them up. I also really like the music. This is one of my favorite games; I finished the 13 levels long ago, but I don't get tired of going back and playing them again.
Slashdot Mirror
For general-purpose note taking, paper is great. Unless your brain works differently than mine does, you will find it too distracting and slow to take notes on a PDA, or even a laptop. And I agree with the people who said that transcribing your notes from scrawls on paper into computer notes will help you really fix the material in your head.
I think you should have a PDA. You can get a Visor Deluxe on eBay for $30 now, and that will work just fine. I really like my Palm Tungsten T, but it costs ten times as much, so don't get one unless you are really not worried about it getting stolen. You can listen to tunes with a Palm Tungsten T if you want, which is a major plus. (I listen to my Ogg files using AeroPlayer.)
For those times when you might want to take notes from a book in the library, a keyboard for your PDA will be great. I'd rather carry my Palm and a portable Palm keyboard, than carry a laptop around. And the keys are quiet on my PDA keyboard, so I don't make clicky-clicky noises to annoy other people.
As for a computer, I agree that you need to scope out the dorm rooms before you commit to a particular computer. If only a laptop will fit in your study space, do get a security cable to secure it. Don't plan on carrying the laptop around campus; it's heavy and annoying, and you are unlikely to actually do it for long.
You will want a nice printer. I suggest an HP DeskJet printer for all-around goodness. You can make very clean black-and-white pages for English papers, you can put color graphs in for science papers, you can put photos in for art history papers... with special paper you can even print photos. You probably don't want the tiny compact folding printer, because that holds very little paper. Just get a desk model that can hold a couple hundred sheets of paper.
Murphy's Law says you will run out of ink, or toner, or whatever your printer uses to print, in the middle of the night the day before a paper is due. With a deskjet, you can get a spare black ink cartridge for $35 or so. Cheap insurance. If your color runs out you can always print in greyscale, so if money is tight just get the black one; otherwise one black and one color cartridge would be good to get.
steveha
Used to be, there was this cool place in Seattle called Entros. For a while there was one in San Francisco, too. Now they are gone, with no trace save a "demo" web site.
It was a restaurant/bar that had about five entertainment areas around it. The entertainment areas were always cool and different.
The most famous was "Interface", where one person wears a blindfold and a camera, and the other person sits in front of a screen watching what the camera sees; using two-way radio, the second guy tells the first guy where to go and what to do. "Go left, step forward, reach down, no, left, no, LEFT, feel for the ball, YOU TOUCHED IT! GO BACK!" Within a set period of time you had to accomplish certain tasks. If you got them all done, you were allowed to enter the victory lounge. I never got to see that lounge...
Anyway, they had a sort of game show where you had to compete to see who was the calmest. They would hook up the players to biofeedback, and then they would do various things to try to shake the players' calm. For the winner, it played a recorded voice saying something like "YOU ARE THE BUDDHA".
I miss Entros.
steveha
The delays in the next one weren't nearly as bad as the delays for the one mentioned in the article. Some distros had the patch out before SCO, but SCO wasn't the last. It really wasn't worth commenting on the second sendmail patch.
steveha
This months issue of a US linux magazine (Probably "Linux Magazine", but I'd need to go home to check) has a pretty favorable review of SCO Linux in it.
That's the June 2003 issue of Linux Journal, page 78. And I didn't think it was "pretty favorable". It was as neutral as possible. The part about the delays in the sendmail security patch was not at all favorable.
The May issue of a magazine usually comes out in April. It probably goes to the printers 6 weeks before being released, so that would put the magazine being created in each March, before the lawsuit.
The final draft was submitted to the magazine about a week after SCO announced its lawsuit, but most of the writing was before that.
I know this stuff because I wrote that article.
steveha
No.
The solid rocket boosters were an afterthought, hacked onto the shuttle when it turned out not to have enough thrust from the main engines. Any mishap with the solid rocket boosters is not survivable: if one fails, the unbalanced thrust from the other one dooms the shuttle. You cannot shut them off, you cannot jettison them until after they burn out on their own, you can't do anything with them but light them and hope they work right.
Do you really think this is a good design?
The shuttle is designed to come screaming down from orbit at huge speeds, then glide in to a really long landing on a really long landing strip. A thin layer of thermal tiles keeps it from burning up. There is no second layer of tiles behind the first layer as a backup. There is no way to repair a damaged tile in orbit. There is no way to abort a landing, fly around, and come back for a second attempt. Stresses on the landing gear are high, and if two tires blow you will damage the shuttle beyond repair.
Do you really think this is a good design?
What I would like to see is a simple design. It should be a single stage if possible, two stage if necessary. It will have one set of engines: the built-in rocket engines. It will use those same engines to take off, and to land. It will have enough engines that it can have multiple engine failures and still land safely. Landing under rocket power, it will not fly so fast it is in danger of burning up, it won't require a special extremely long runway, and it can abort a landing for any reason and come back around for a second landing attempt.
As long as we keep flying the shuttle, I'd like to see liquid-fuelled rocket boosters. At least if you have a problem, you can shut the fuel pumps off and get both boosters to instantly shut down.
steveha
[landing under rocket power] takes insane amount of fuel.
Not so. By the time you are landing, the rocket is very light because the fuel tanks are nearly empty. It takes a lot of fuel to fight gravity and get into space, and when you land most of that fuel has been used up.
The shuttle cannot and will not ever land under rocket power. But a future spacecraft, perhaps an SSTO one, can be designed to do it. And while I am not a rocket scientist, or even an insightful amateur, other people who have done the math say it's the right way to go.
steveha
I don't use the "minimum font size" thing. Instead, I have a custom CSS file which sets the size of fonts.
s s under Linux):
Here is the secret formula. Put this in your userContent.css file (~/.mozilla/default/<random>/chrome/userContent.c
@media screen {
* {
font-size: 28px !important;
line-height: 30px !important;
}
}
The great part about this is that it doesn't mess up the layout as often as setting the minimum size font does. Also, it doesn't affect printing at all (I don't remember whether setting the minimum size affects printing or not, but this sure doesn't).
steveha
f you middle click on a link, open link in a new tab. If you middle click off of a link, activate the autoscroll. Simple as that.
And, in fact, Galeon has already implemented this in exactly the way you describe.
Plus, as a bonus, if you select a URL and then middle-click on the "New Tab" button on the toolbar, it will open a new tab with that URL loaded! I really love that.
steveha
That whole concept of actually using ambient "air" rather the having to carry liquid oxygen to get your vehicel off the ground makes a fair amount of sence to me.
The problem is that liquid oxygen is both cheap and pretty compact. And, once you burned it, it's gone. Your scramjet design needs scramjet engines, and it will carry those engines to orbit and back. They only help during the initial takeoff phase, but the weight penalty is there 100% of the time. The same applies to the wings.
You want your spaceship to get up out of the atmosphere as quick as possible, so it won't be able to use the scramjet engines all that long anyway.
You can build a rocket that will take off, fly in space, and land... all using the same (rocket) engines. This will be a simpler design than some complicated thing that starts out with jet engines, switches over to scramjet engines when speeds are high enough, then switches over to rocket engines when there is no ambient air around (i.e. in space). Simpler is good.
There is some merit to the idea of a two-stage design, where the first stage can be air-breathing and can carry the second stage up to thinner air. The proponents of single-stage-to-orbit (SSTO) think you don't need the air-breathing lower stage, and that the operational cost will be lower if you have the SSTO. One stage is less than two stages, so it should take less maintenance, less manpower, and so on. But no one has seriously tried to develop a complete SSTO design, other than on paper; if it turns out we can't quite get a useful payload on an SSTO, we could always have an extra stage to get it on its way.
steveha
Well, except that Space is much more dangerous than flying around in the air. It's magnitudes more difficult *and* dangerous!
It's not as hard as you think.
A rocket is a fundamentally simple device. You have fuel tanks, some sort of pump, and a reaction chamber. There is no reason to think that all space vehicles will be tremendously unsafe like the shuttle.
You can always increase risk with stupid design, and I'm afraid that's what the shuttle did. Columbia broke up because of compromised thermal tiles; there is no way to fix tiles while in orbit, and the design of the shuttle is to come screaming down from orbit at huge speeds and finally land like a giant brick with wings. (And shuttle has no ability to fly past the landing point, swing round, and land on a second pass; it has to land in one pass with no mistakes allowed. Brilliant design.)
It isn't all that hard to land under rocket power, using rockets to slow the descent to reasonable speeds, and then you don't run the risk of losing the spacecraft and all people on board because a few thermal tiles got damaged. And you can have enough extra rockets that if some malfunction, you can land with the others. And you can easily back off and come round again if something happens when you are trying to land.
The key is to make redundancy and survivability the primary design goal, and to build prototypes and fly them to refine the design before you build your fleet. Shuttle was designed on paper, and there were no X-vehicle prototypes anywhere along the way. Stupid bits in the design stayed in.
steveha
Only two out of five have failed.
Official NASA documents estimated that you should be able to fly the shuttle 10,000 times before you lost one ("five nines reliability"). The reality is much closer to 100 times (two nines). This is very poor. If airplanes would kill you one time in a hundred, I sure wouldn't want to fly on an airplane... and there is nothing inherent to space operations which justifies the poor record of the shuttle.
We need to replace it with something safer, and that is possible.
steveha
It's very simple. What we need are reusable ships with a modest cargo capacity, plus maybe a few "big dumb boosters" for launching big things.
It's also very clear that NASA is not capable, as an organization, of doing this. NASA has some smart people working there, but any really large project will safely bury the smart ones under red tape where they can't do anything. If you want to convert money into piles of paper, have NASA attempt to make a follow-up to the shuttle.
The US government should make iron-clad promises to buy launches. Station re-supply launches for the International Space Station would be a great place to start. If John Carmack's company, or any other company, can get a vehicle going that can run supplies to orbit, the government should hire them to do it. In other words, pay for results but for nothing else, and don't have any part of the government (especially NASA) trying to help design the ships.
steveha
Soldiers have always been trained to be desensitized to violence. They have a job to do, and they need to be doing it, not standing around going "Whoa! Violence! Eeeek!"
Perhaps the military uses video games now as part of the training, but I suspect they mostly still rely on the old-fashioned training techniques.
By the way, soldiers are not just killing machines who go around shooting anything that moves. There are rules, harsh ones, that govern what soldiers do. Not to say that no American soldiers have ever committed atrocities, but that's very much the exception rather than the rule.
steveha
once you do, APL is a much clearer way to write some hairy multidimensional array operations.
Maybe so, but the set of problems for which APL is the perfect solution is small enough that few are willing to master it.
Python, with serious math libraries (written in C) grafted onto it, can do hairy array operations too. And once you master Python, you can use it for more than just matrix math; you can do pretty much anything. I read an article about how a science lab was switching away from FORTRAN to Python for their matrix math; there was a quote something like "after trying Python, no one wants to go back to FORTRAN." Alas, I didn't keep the URL and I can't find the article now.
If you like APL, that's fine with me. I still like FORTH, even though some people call it a Write-Only Language.
steveha
Spoken like someone who hasn't coded much assembly...
In fact, I haven't coded much assembly, although I have done some. But you are missing the point here, dude. Would you use Applesoft BASIC (or assembly) to teach an Introduction to Programming class to high-school kids?
No. You want a language that lets you focus on the basic concepts of programming, without having to learn tons of annoying syntax. Pascal is better than BASIC for this. Java is better than Pascal. Python is probably best.
steveha
if (flag)
DoSomething(3);
else
DoSomeOtherThing(4);
120 IF FLAG = 0 GOTO 160
130 GLOBAL1 = 3
140 GOSUB 600
150 GOTO 180
160 GLOBAL2 = 4
170 GOSUB 700
180 REM next statement
600 REM DoSomething() -- argument is GLOBAL1
700 REM DoSomeOtherThing() -- argument is GLOBAL2
No block structuring: you can only have a single statement after an IF. For anything else, you have to test for the opposite of what you really want to test for, and GOTO the line number of the ELSE case. Then you fall through to the IF case if you didn't take the GOTO. At the end of the IF case, don't forget the GOTO to skip over the ELSE case to the next statement! If you want a subroutine that will have some sort of argument, you need to use a global variable. (You could actually use the same global for both functions, since there is no danger of multiple threads of execution or anything like that.)
Modern languages are so much nicer than the evil old BASIC languages.
steveha
In Ada, I'll write the code, look at it, realize that it says what it does, and not need to write comments.
An example of such code would be welcome here. As I said, I have never used Ada.
Ever seen APL? It's essentially encrypted at the source.
APL: the Write-Only Language. (People call Perl a WOL but APL is in a league all its own.)
Fortunately, Guido Van Rossum used his math geek powers for Good, rather than for Evil when he created Python.
steveha
Better would be languages which are self-documenting...
There is no language that will force perfect code. There is always room for a poor programmer to produce hard-to-understand code. Functions that do two unrelated things, confusing control flow, bad variable names, broken code that was repeatedly patched instead of being cleaned up... the possibilities are endless.
Nonetheless, some languages have been designed with self-documenting code in mind; sometimes it even works.
If you look at languages like COBOL, they have long descriptive keyword names designed to make the code easy to read. But you get tired of looking at those long keywords.
I haven't used ADA, but I understand that it is somewhat designed for self-documenting code, and that as a result you are hemmed in on all sides by language rules. (ADA fans please comment here.)
The best language I have seen for this is Python. As a rule there is exactly one way to do things, so you don't trip over obscure hackish tricks that you have to puzzle out. The language doesn't force self-documenting or comments, but it does force indentation; everyone indents their Python pretty much the same (compare with the mess that is C indentation). The language is high-level enough, with lots of libraries, so you don't need to write 10 lines of code just to do one simple thing.
Python was designed by a guy who is both a computer geek and a math geek. The math geek in him led to a very tidy language design, and I like it very much. I think schools ought to be using Python to teach introductory programming classes.
steveha
I stand corrected. I don't know where I read that the revision A would not support OS X, but clearly it does. The official upgrade requirements page clearly shows the "Bondi Blue" iMac as a supported platform.
I note that the page is for version 10.2 of the OS. Has the revision A been supported all along, and my info was completely wrong, or was it not supported in 10.0?
steveha
Did that ad actually convince people to buy Apple computers, or was it just a "we're so cool" thing that was a waste of advertising budget? I suspect the latter.
steveha
We have seen some UI innovation, and we will see more.
Before you can raise the bar, you have to reach the bar where it is. The Mac has been shipping, with a UI, since 1984. Windows has been around in one form or another about as long. KDE and GNOME have had much less time to get their acts together.
Both KDE and GNOME now are decent and usable. From there, they can try new things and perhaps advance introduce some UI innovations. In the past, most of the energy has been going into just getting things working; now, more energy will go into experiments and new ideas.
P.S. UI innovations tend to be subtle things. If you do radical, weird stuff, people don't like it.
steveha
Ah, yes, I remember this well.
Xerox came out with the GUI. Apple and Microsoft both started working on a GUI. Apple asked Microsoft to support MacOS (by releasing apps such as Word and Microplan for the Mac). Microsoft agreed, but required Apple to sign an agreement that Apple not sue MS over Windows. Apple signed the agreement. Later, Apple sued MS over Windows.
In the suit, Apple claimed that Windows infringed on a nebulous concept called "look and feel". The judge threw out all "look and feel" claims, and would only consider specific claims. Then the judge went down the list, throwing out any claim that was covered by the agreement. About a dozen claims were left. Then the judge went down the list, and threw out any claim that Apple didn't own (common GUI things that anyone could do without Apple's permission). Exactly zero claims were now left, and the judge dismissed Apple's suit.
Short version: Apple agreed not to sue, sued anyway, and the judge ruled that the agreement was perfectly valid and threw out the suit. Seems fair to me.
And, by the way, we would not now have either GNOME or KDE had Apple won that suit. Apple winning that suit would mean Apple owns GUI desktops, and that would be Very Bad Indeed. Forget free desktops for Linux. Even forget non-free desktops for Linux; Apple would insist you buy an Apple computer running an Apple OS if you wanted a GUI desktop. (And any Apple fans who want to claim otherwise: how many licenses has Apple sold to the TrueType patents?)
steveha
Hastings's Law: Cheaper and adequate wins against more expensive and better.
SCSI is better than IDE, but IDE is adequate. Beta had better quality than VHS, but VHS was adequate (and had much longer recording time on a tape). Apple was able to charge rapacious margins for a Mac in the days when x86 PCs were much harder to use; once Windows became adequate, customers started buying Windows versus the Mac.
A BMW 7 Series is a better car than a Honda Accord. But an Accord is adequate. Which sells more?
steveha
If you invested in Apple 15 years ago, they still honor your investment. I can't say that the same is true of MS
Are you talking about Apple Computer? The same Apple Computer that used to make the Newton, and the eMate, and just shotgunned those products and abandoned their users? (Remember how some companies tried to buy rights to the Newton OS so they could support the Newton community, and Apple refused to sell?)
Is this the same Apple Computer that promised that the original iMac would run OS X, and then quietly walked away from that promise?
Is this the same Apple Computer that allowed third party companies to make Mac clones, and then suddenly unilaterally shut down the lincensing deals? (That's more of a "screwing the clone vendors" thing than a "screwing their own customers" thing, but I'm still not happy about it.)
Apple makes some cool stuff, and the premium for it isn't as high as it used to be. (I remember the days when Apple was scoring 40% margins on their hardware! Gouging their own customers for the short-term profit.) For many people, a Mac is a reasonable thing to buy -- after all, if what you do is surf the net, you don't care that your computer is only about a gigaHertz instead of 3 gigaHertz. But to hold up Apple as some shining example of taking care of their loyal customers is a bit much.
I don't think Microsoft has ever just walked away from a group of customers. Even when Microsoft walked away from OS/2, it was because IBM was there to support the OS/2 customers. I have my issues with Microsoft these days, but on this particular issue I think they are the better company.
steveha
Spy Hunter for the PS2 or X-Box is a cool driving game. Not only do you drive, but you have machine guns, oil slicks, missiles, and other toys, and you use them to blow up dozens of bad guys.
When your car goes into the water, it morphs into a boat. If your car takes too much damage, it jettisons you in a motorcycle escape vehicle. If the motorcycle goes into the water, it morphs into a jet ski!
I feel this is a bit of a sleeper game -- hardly ever mentioned on websites I visit, yet it is really fun. I love locking missiles on the bad guys and blowing them up. I also really like the music. This is one of my favorite games; I finished the 13 levels long ago, but I don't get tired of going back and playing them again.
Recommended!
steveha