All the basic algorithms are very similar. In fact, the latest edition of Knuth's The Art of Computer Programming (the computer science book) still has information on optimizing tape operations! The algorithms used for operations like random number generation, searching and sorting are the same ones used in mainframes and minicomputers; granted, they have evolved to take advantage of more processing power (random number generation in particular) but the basic concepts remain the same.
There have been very few "complete upheavals" in computer science, just as there have been very few "complete upheavals" in EE in the last 50 years. Computers are faster now, but the algorithms they execute are the same.
I'm sorry, but why would they have to notify the copyright owners? There are thousands of radio stations that do this same exact thing.
Radio stations pay a lot of money for broadcast rights. While small labels may give out some songs free to college radio stations and the like, radio station owners can count on paying tens or hundreds of thousands of dollars for broadcast rights to a hot new single from a major label.
What's even funnier is that they chose the title to describe their feelings toward record company execs (who turned down their first title, "Metal Up Your Ass"). They changed their tune pretty fast when it was their money and not their music on the line; that's the very definition of selling out.
Note: I love early Metallica, but can't stand anything post-Black Album. About the time they started selling out.
First, VB may be good for rapid app development (I can crank out a DB frontend in a couple of hours in VB!), but it's nothing for cracking. Doing buffer overflows in VB is practically impossible; the language isn't designed with direct access to memory in mind. C is, and it's a much better language to crack systems in. While it's probably possible to do dangerous things in VB, given enough time, you're just reinventing the wheel. I could do much worse with access to lcc-win32 or gcc than I could with VB.
Regarding the "HR drones:" if an 31337 crax0r graduate of this program were to put "computer security training" on their resume, they would be telling the truth and not make the "evil hacker" association in the HR worker's mind. Putting the right spin on the truth is almost as important as the facts behind it in this case.
Nice troll, but the "Visual Basic development tools" reference put you over the top into unbelivability.
And to commence feeding: your comment on hacking experience being bad is totally groundless: I wouldn't trust an architect who couldn't tell me the points in a building vulnerable to bombing, and I wouldn't trust a sysadmin who didn't have at least a basic knowledge of hacking techniques.
Your argument is bullshit. Pi is the ratio of circumfrence to diameter of a circle, nothing more. It may show up in other places; this is coincidence. I was merely pointing out how the ratio of diameter to circumfrence could be =! ~3.14159.
This is the definition of pi. Pi may be a useful tool in other equations, but that doesn't change its definition.
It seems to me that tablet PCs could be made much, much lighter. While battery weight is something of an insoluble problem, getting rid of the hard drive would be a good start. You could probably get 256 MB of flash in there for the same price as a decent size laptop HDD, and the weight would go down and battery life up. The point of the tablet is to take quick notes and annotate existing documents; it would be able to do both with, say, 192 MB for user data and 64 MB for the OS. The notes can be converted to ASCII with a handwriting-recognition system, or just compressed (they're graphics; they'll compress well). Give it a CompactFlash slot, and users can add more storage as they see fit.
The point is that manufacturers should look at the tablets as a PC-PDA hybrid, not just a PC in a weird form factor. The design influences how people are going to use the system; in this case, they don't want to use it for movie editing or gaming, they want it to jot down notes. As such, give them a device to take notes, not a device that has all the weight, power consumption, and UI as a PC. They want a specialized, cheap device. Give it to them.
Re:Why the emphasis on a polished desktop?
on
Ximian's Back
·
· Score: 1
The problem with X is precisely that it is client-server. While being able to run anything across the network is great, it doesn't work too well with accelerated graphics, like are common today. The DRI is just a hack on top of X to get it to work with 3D acceleration; all of X runs directly opposite to the concept of intelligent terminals. According to the X philosophy, all apps should be run on a blazing-fast server, which then displays the apps on a dumb client. Unfortunately, today's machines don't work well with that idea. The client-server interface generates all sorts of unneeded cruft for 90% of systems; those that need X can still use it (or, if it's just to allow users to work remotely, something like VNC fills the gap nicely).
Also, what in hell are you talking about VMs for? Perhaps you meant "window manager" instead of "virtual machine?" Assuming you did mean WM instead of VM, there's no reason that the good features of X couldn't be adopted into a new windowing system. The reason for migrating would be to drop the workarounds and bugs that result from having to keep backwards compatability with the original versions of xterm, xclock and xlogo, and the bad assumptions that get in the way of things like font AA.
Why the emphasis on a polished desktop?
on
Ximian's Back
·
· Score: 1, Interesting
It seems to me that instead of creating wonderful eyecandy (which this certainly is), the Linux community should focus on creating standards and moving away from antiquated systems like X. I would much rather have a windowing system that didn't have 20-odd years of cruft, but instead had native support for things like antialiasing and an X compatability layer.
Also, where are the easy configuration tools? Redhat is especially guilty of thinking that the user either wants everything configured either automagically or manually, but doesn't want a simple way to manage most devices in their system. Some basic standards on the locations of config files are desperately needed (just having/etc to dump everything into is not OK, as some packages put things in their own directory, some put them in the root, and some are crazy and keep them somewhere else entirely). Linuxconfig used to be good for setting up most components of a system, but it was only good for Redhat, who dumped it around 7.0, and not replaced. Debian has no central config system; this is slightly more acceptable as it is an "expert's" system, but I would still like to see some system that would shape itself to my configuration and run the configuration tools from the.deb package, if nothing else.
There seems to be a lot of great work going on in the kernel level, and a lot of great work in the desktop level, but we're still stuck on 20 year old tools in the middle. How would you like to use CDE instead of Ximian or KDE3? Yet we submit willingly to using outdated technology in the middle, for essentialy no reason at all. If Linux is ever to become a desktop powerhouse, it must be equally refined, with a good kernel, a good set of system config tools, and a good desktop. Just having two out of three is not good enough.
Lawyers don't deserve all the blame, but they do deserve a significant portion. Many corporations (especially large ones) will employ a legal team that makes recommendations to the upper management regarding legal matters like lawsuits. Thus, a lawyer may recommend a frivolous lawsuit to the higher-ups in a company, who don't understand the situation and don't care very much. They just sign what their legal team puts in front of them and let the lawyers run free. After all, if the law allows them to file these lawsuits, then it must be okay, right?
The job of these lawyers is to find and exploit legal loopholes to benefit the corporation they work for. A lot of them are scumballs. That's not to say that all lawyers are bad; that's an attitude I don't understand. There are many lawyers committed to things like bringing criminals to justice or fighting for the poor. They just aren't the ones driving brand-new Porsches. Still, many lawyers are willing to take a pay cut to work in the DA's office, because they simply can't stand the "do-it-for-money" attitude of many big law firms and corporate legal departments. If people were willing to look at the entire legal picture, they would find that it's like many others: a few high-profile assholes, and many people who toil away for good causes in relative obscurity.
By the way, I'm not a lawyer, but I do know several. I do not associate with assholes. End of story.
Let's imagine a sphere, and a two-dimensional being who lives on that sphere. If that two-dimensional being were to draw a circle (by taking a point, then marking all points a certain distance from that point) the ratio of radius to circumfrence is not necessarialy 2pi.
If it's a small circle on a large sphere, then the ratio is very near to 2pi, but if it's a large circle on a small sphere, then the ratio goes up. For example, imagine that our 2D friend draws a circle from the "north pole" of his sphere to the "equator." The ratio is higher than 1:2pi, because measuring along the surface of the sphere, we get a measurement longer than what it would take if the sphere were flattened. Or we could take the extreme case: the radius of our circle is the distance from the "north pole" to the "south pole" of the sphere. In this case, everything converges on the "south pole" and forms a circle with radius 0, and an undefined ratio. I hope this makes sense; I can't remember the equations for non-Euclidian space and can't be bothered to look them up. I decided to go with the intuitive approach, which works much better with illustrations. Oh well.
Why not? Pi is only ~3.14159 in flat space. In curved geometry (like on the surface of a sphere) the ratio of a circle's diameter to its circumfrence can be not equal to pi in "flat space" or even variable, depending on the size or location of the circle.
I suggest reading Feynman's Six Easy Pieces and Six Not-So-Easy Pieces (or, even better, the complete Lectures on Physics); they can be tough going, but you will come out with a solid understanding of the basics of relativity. Note that they were written for freshmen and sophomores at Caltech, so math up to basic calculus is required.
When you finish the book, you will understand how space could be curved, even 3-dimensionally. One piece of advice: don't try to map every equation to your perception of the "real world," or you'll go crazy halfway through Not-So-Easy Pieces. Just go through the math and think about it later.
Remember that videogames don't force you to do anything. Only someone who is deeply disturbed lacks the ability to differentiate between fantasy (like a computer or console game) and reality. This reminds me of the '80s worries over D&D "cults" that would supposedly do crazy stuff to people because it was in D&D. Not gonna happen, unless someone's fairly far out there already.
One other thing: I hear a lot about videogames training kids to be killers. Again, not gonna happen. While some videogame skills might transfer over to the real world, most don't. Nobody who plays Quake for 8 hours a day picks up any marksmanship skills at all, any more than playing Tetris prepares you for a job in mail-order packaging. Besides, anyone playing games obsessively will lack the physical fitness necessary in a combat environment. Videogames are for the most part designed to be unrealistic to a degree; apart from hardcore sim-heads, those kinds of games are seen as boring and don't sell. While small amounts of realism make a game fun (think Counterstrike), large amounts simply consist of players doing boring, repetitive things (just like in the real world). Games don't train anyone to be a killer.
In my experience, PC manufacturers won't do anything about 3rd party hardware; they tend to not do anything for tech support until it's gone. The only issue would be with things that failed so badly as to hurt other components in the computer; a rare case to say the least (although I have seen it happen). As for QC, it tends to be independent of number of products made. QC efficiency is probably similar: you pull a certain percentage of products off the line and test them, no matter what the number of products shipping is. Better QC comes from testing more products, but that cuts into your bottom line.
That's a percentage graph. For every 100 computers shipped, Apple had fewer of them come back than anyone else; thus, their QC beats Dell's.
Also, Apple's tech support tends to be freakishly fast. I had a friend get a Powerbook G4's mobo replaced in 3 buisiness days, including shipping. She thought that she mislabeled the package and Airborne was shipping it back to her; she called Apple and they asked if the problem was fixed. Believe it or not, it was. If it wasn't for the price, I would buy from Apple every time, and that iPod is giving me a serious case of geek-lust.
Worse yet, SCO's own CEO, who estimates that intellectual property accounts for about 1/3 of his sales, doesn't understand the difference between a patent and a trade secret. A patent is afforded legal protection for a limited time in exchange for disclosing all the information necessary to reproduce the patented object, be it an algorithm, gene, or device, to the public. Also, it's the responsibility of your competitors to check the Patent Office and make sure they don't violate your patent. If they do violate your patent, then you can sue and recover damages even if you didn't notify them about the patent violation beforehand. It's their responsibility to check out patents.
On the other hand, trade secrets don't have any legal protection, but you don't have to disclose anything to the public. The only way to protect a trade secret is through contracts, made with everyone that the secret code comes into contact with. If you make these contracts, then you can sue for damages if the secret is leaked (assuming you were smart enough to put that provisio in your contract).
SCO seems to want to have it both ways; they want the legal protection of a patent, and the ability to sue people associated with Linux for patent violations, but they also want to not disclose anything about their technology. In essence, they seem to want to sue Linus for something he had no control over, and no way to know that he was doing anything wrong (if there was any wrong committed; it could be that, as a poster on the LKML noted, the similarities are like having "her bosom heaved in anticipation" in two different cheesy romances; the similarity is coincidental and rather likely, given that the two OSes implement similar concepts).
Note that although they can't have patent protection on their software, SCO may still be able to sue IBM for contract violations. If IBM did indeed leak SCO's code into Linux or elsewhere, they could be liable for damages (although almost certainly not $1 billion; no sane judge or jury would award that much to a company who hasn't sold that much software over the past few years, even when it was dealing in Linux itself and thus getting the benefit of IBM's supposed evil machinations)
As an aside, does anyone else find it perversely enjoyable to refresh the Yahoo ticker constantly and realize that for every cent SCOX loses, the bastards who started this whole thing are probably out tens of thousands of dollars?
Just saw on CBS Marketwatch that SCO's CEO, Darl McBride, is threatening to sue Linus for patent infringement unless more companies license SCO products. To put it bluntly, what in the sweet merciful FUCK is going on?
I doubt there would be grounds to sue Linus even if there was clear-cut evidence of patent infringement in the Linux kernel. Linus didn't put the supposedly offending code in there, but he's still liable? SCO seems to be really grasping at straws here; they know that the bluff has been called. I just hope they don't manage to do any more FUD-based harm to Linux before they die a merciful death.
I'm not a lawyer, so nobody use this as legal advice. If you want legal advice, go pay a lawyer.
Just because it compiles doesn't mean you should do it. Programs should return an integer so that the calling program can know whether it failed or not (0 for success, anything else is failure and may indicate a specific problem). If everyone adhered to your "standard," writing shellscripts would become a living hell. Where were you when they were passing out standards, anyway?
What do you suggest as a replacement for floppies? CD-RWs are cheap, but take forever to format and require special software (unless you're using Win XP, which is a major headache to roll out). Iomega Zip and Jaz are decent, except that the standard is controlled by one company and media is ridiculously expensive ($10 for 100 megs? You've got to be kidding, I can get a cakebox of CD-Rs for $25!). Floppies are still a "lowest common denominator"; they can be found on almost all systems, they're cheap, and have enough room for a few Word documents or similar files.
Although you argue that students shouldn't be expected to have floppy drives, you don't produce any alternative. Floppies remain the cheapest and easiest method for transporting small amounts of data.
Trade secrets are still valuable as long as you keep them secret; if an NDAed employee were to run off with your source code and give it to another company, who implemented your algorithms, you would have legal recourse under trade secret law.
This is SCO's case; they argue that Linux uses some of their secret methods. What ESR's project is trying to do is establish that SCO's "secrets" were already widely available through lax security on SCO's part. If SCO failed to protect their trade secret, then they have no recourse; if they did protect it and it illegally got into Linux (like if an NDAed employee put SCO code into Linux) then SCO has a legal recourse.
Of course, I am not a lawyer and this is not legal advice. If you need legal advice, contact a lawyer in your jurisdiction.
Color lasers run $2K-4K US for a fairly basic model. They can give you a very high-quality print (lasers are still better than inkjets), which just happens to be waterproof. Although the initial bite is quite steep, I would imagine being able to print your own money would defray the costs somewhat.
Although it wouldn't pass a rigorous inspection, spending the printed money at a grocery store or similar location would probably be really easy (nobody would check it for the special security features, and previous posters have mentioned brands of paper that have that "money feel" to them).
As someone studying for a private pilot's license, I can understand where the divers are coming from. Basically, if you screw up once, your life is over. This might not be the case every single time, but one mistake at the wrong time could easily end your life and put whatever bits of you rescuers could find in a little pouch 10 feet underground.
Thus, hobbyists and amateurs use methods very similar to those of the "pros"; both need to ensure the utmost level of safety. Most of the time that caution isn't warranted and things are fine without double-checking everything, but one time in a hundred or a thousand, you are very glad you did. Basically, the extensive checklists become second nature and take only a few minutes to perform; the small amount of time "wasted" is no trade-off at all considering what might otherwise happen.
I take it you've never had to slog through thousands and thousands of lines of awful spaghetti code. COBOL is like any other programming language in that it's possible to write easy-to-read code or write awful code; it's all up to the programmer. COBOL's use of global variables and GOTOs, however, makes it very easy to write bad and unreadable code. As for object-orientation, it's possible to pick up the basic concept in a few hours, and to follow well-written C++ or Java code (assuming you already know how to program in the first place). The difficult part of OO is figuring out how to implement it in the first place; well-written OO code is quite readable.
Slashdot Mirror
There have been very few "complete upheavals" in computer science, just as there have been very few "complete upheavals" in EE in the last 50 years. Computers are faster now, but the algorithms they execute are the same.
Note: I love early Metallica, but can't stand anything post-Black Album. About the time they started selling out.
Regarding the "HR drones:" if an 31337 crax0r graduate of this program were to put "computer security training" on their resume, they would be telling the truth and not make the "evil hacker" association in the HR worker's mind. Putting the right spin on the truth is almost as important as the facts behind it in this case.
And to commence feeding: your comment on hacking experience being bad is totally groundless: I wouldn't trust an architect who couldn't tell me the points in a building vulnerable to bombing, and I wouldn't trust a sysadmin who didn't have at least a basic knowledge of hacking techniques.
This is the definition of pi. Pi may be a useful tool in other equations, but that doesn't change its definition.
The point is that manufacturers should look at the tablets as a PC-PDA hybrid, not just a PC in a weird form factor. The design influences how people are going to use the system; in this case, they don't want to use it for movie editing or gaming, they want it to jot down notes. As such, give them a device to take notes, not a device that has all the weight, power consumption, and UI as a PC. They want a specialized, cheap device. Give it to them.
Also, what in hell are you talking about VMs for? Perhaps you meant "window manager" instead of "virtual machine?" Assuming you did mean WM instead of VM, there's no reason that the good features of X couldn't be adopted into a new windowing system. The reason for migrating would be to drop the workarounds and bugs that result from having to keep backwards compatability with the original versions of xterm, xclock and xlogo, and the bad assumptions that get in the way of things like font AA.
Also, where are the easy configuration tools? Redhat is especially guilty of thinking that the user either wants everything configured either automagically or manually, but doesn't want a simple way to manage most devices in their system. Some basic standards on the locations of config files are desperately needed (just having /etc to dump everything into is not OK, as some packages put things in their own directory, some put them in the root, and some are crazy and keep them somewhere else entirely). Linuxconfig used to be good for setting up most components of a system, but it was only good for Redhat, who dumped it around 7.0, and not replaced. Debian has no central config system; this is slightly more acceptable as it is an "expert's" system, but I would still like to see some system that would shape itself to my configuration and run the configuration tools from the .deb package, if nothing else.
There seems to be a lot of great work going on in the kernel level, and a lot of great work in the desktop level, but we're still stuck on 20 year old tools in the middle. How would you like to use CDE instead of Ximian or KDE3? Yet we submit willingly to using outdated technology in the middle, for essentialy no reason at all. If Linux is ever to become a desktop powerhouse, it must be equally refined, with a good kernel, a good set of system config tools, and a good desktop. Just having two out of three is not good enough.
The job of these lawyers is to find and exploit legal loopholes to benefit the corporation they work for. A lot of them are scumballs. That's not to say that all lawyers are bad; that's an attitude I don't understand. There are many lawyers committed to things like bringing criminals to justice or fighting for the poor. They just aren't the ones driving brand-new Porsches. Still, many lawyers are willing to take a pay cut to work in the DA's office, because they simply can't stand the "do-it-for-money" attitude of many big law firms and corporate legal departments. If people were willing to look at the entire legal picture, they would find that it's like many others: a few high-profile assholes, and many people who toil away for good causes in relative obscurity.
By the way, I'm not a lawyer, but I do know several. I do not associate with assholes. End of story.
If it's a small circle on a large sphere, then the ratio is very near to 2pi, but if it's a large circle on a small sphere, then the ratio goes up. For example, imagine that our 2D friend draws a circle from the "north pole" of his sphere to the "equator." The ratio is higher than 1:2pi, because measuring along the surface of the sphere, we get a measurement longer than what it would take if the sphere were flattened. Or we could take the extreme case: the radius of our circle is the distance from the "north pole" to the "south pole" of the sphere. In this case, everything converges on the "south pole" and forms a circle with radius 0, and an undefined ratio. I hope this makes sense; I can't remember the equations for non-Euclidian space and can't be bothered to look them up. I decided to go with the intuitive approach, which works much better with illustrations. Oh well.
I suggest reading Feynman's Six Easy Pieces and Six Not-So-Easy Pieces (or, even better, the complete Lectures on Physics); they can be tough going, but you will come out with a solid understanding of the basics of relativity. Note that they were written for freshmen and sophomores at Caltech, so math up to basic calculus is required.
When you finish the book, you will understand how space could be curved, even 3-dimensionally. One piece of advice: don't try to map every equation to your perception of the "real world," or you'll go crazy halfway through Not-So-Easy Pieces. Just go through the math and think about it later.
One other thing: I hear a lot about videogames training kids to be killers. Again, not gonna happen. While some videogame skills might transfer over to the real world, most don't. Nobody who plays Quake for 8 hours a day picks up any marksmanship skills at all, any more than playing Tetris prepares you for a job in mail-order packaging. Besides, anyone playing games obsessively will lack the physical fitness necessary in a combat environment. Videogames are for the most part designed to be unrealistic to a degree; apart from hardcore sim-heads, those kinds of games are seen as boring and don't sell. While small amounts of realism make a game fun (think Counterstrike), large amounts simply consist of players doing boring, repetitive things (just like in the real world). Games don't train anyone to be a killer.
In my experience, PC manufacturers won't do anything about 3rd party hardware; they tend to not do anything for tech support until it's gone. The only issue would be with things that failed so badly as to hurt other components in the computer; a rare case to say the least (although I have seen it happen). As for QC, it tends to be independent of number of products made. QC efficiency is probably similar: you pull a certain percentage of products off the line and test them, no matter what the number of products shipping is. Better QC comes from testing more products, but that cuts into your bottom line.
Also, Apple's tech support tends to be freakishly fast. I had a friend get a Powerbook G4's mobo replaced in 3 buisiness days, including shipping. She thought that she mislabeled the package and Airborne was shipping it back to her; she called Apple and they asked if the problem was fixed. Believe it or not, it was. If it wasn't for the price, I would buy from Apple every time, and that iPod is giving me a serious case of geek-lust.
On the other hand, trade secrets don't have any legal protection, but you don't have to disclose anything to the public. The only way to protect a trade secret is through contracts, made with everyone that the secret code comes into contact with. If you make these contracts, then you can sue for damages if the secret is leaked (assuming you were smart enough to put that provisio in your contract).
SCO seems to want to have it both ways; they want the legal protection of a patent, and the ability to sue people associated with Linux for patent violations, but they also want to not disclose anything about their technology. In essence, they seem to want to sue Linus for something he had no control over, and no way to know that he was doing anything wrong (if there was any wrong committed; it could be that, as a poster on the LKML noted, the similarities are like having "her bosom heaved in anticipation" in two different cheesy romances; the similarity is coincidental and rather likely, given that the two OSes implement similar concepts).
Note that although they can't have patent protection on their software, SCO may still be able to sue IBM for contract violations. If IBM did indeed leak SCO's code into Linux or elsewhere, they could be liable for damages (although almost certainly not $1 billion; no sane judge or jury would award that much to a company who hasn't sold that much software over the past few years, even when it was dealing in Linux itself and thus getting the benefit of IBM's supposed evil machinations)
As an aside, does anyone else find it perversely enjoyable to refresh the Yahoo ticker constantly and realize that for every cent SCOX loses, the bastards who started this whole thing are probably out tens of thousands of dollars?
I doubt there would be grounds to sue Linus even if there was clear-cut evidence of patent infringement in the Linux kernel. Linus didn't put the supposedly offending code in there, but he's still liable? SCO seems to be really grasping at straws here; they know that the bluff has been called. I just hope they don't manage to do any more FUD-based harm to Linux before they die a merciful death.
I'm not a lawyer, so nobody use this as legal advice. If you want legal advice, go pay a lawyer.
FNORD.
Just because it compiles doesn't mean you should do it. Programs should return an integer so that the calling program can know whether it failed or not (0 for success, anything else is failure and may indicate a specific problem). If everyone adhered to your "standard," writing shellscripts would become a living hell. Where were you when they were passing out standards, anyway?
Although you argue that students shouldn't be expected to have floppy drives, you don't produce any alternative. Floppies remain the cheapest and easiest method for transporting small amounts of data.
This is SCO's case; they argue that Linux uses some of their secret methods. What ESR's project is trying to do is establish that SCO's "secrets" were already widely available through lax security on SCO's part. If SCO failed to protect their trade secret, then they have no recourse; if they did protect it and it illegally got into Linux (like if an NDAed employee put SCO code into Linux) then SCO has a legal recourse.
Of course, I am not a lawyer and this is not legal advice. If you need legal advice, contact a lawyer in your jurisdiction.
Although it wouldn't pass a rigorous inspection, spending the printed money at a grocery store or similar location would probably be really easy (nobody would check it for the special security features, and previous posters have mentioned brands of paper that have that "money feel" to them).
Thus, hobbyists and amateurs use methods very similar to those of the "pros"; both need to ensure the utmost level of safety. Most of the time that caution isn't warranted and things are fine without double-checking everything, but one time in a hundred or a thousand, you are very glad you did. Basically, the extensive checklists become second nature and take only a few minutes to perform; the small amount of time "wasted" is no trade-off at all considering what might otherwise happen.
You'll rue the day you crossed me!
I take it you've never had to slog through thousands and thousands of lines of awful spaghetti code. COBOL is like any other programming language in that it's possible to write easy-to-read code or write awful code; it's all up to the programmer. COBOL's use of global variables and GOTOs, however, makes it very easy to write bad and unreadable code. As for object-orientation, it's possible to pick up the basic concept in a few hours, and to follow well-written C++ or Java code (assuming you already know how to program in the first place). The difficult part of OO is figuring out how to implement it in the first place; well-written OO code is quite readable.