If you configure a linux box so that it's not running network services, (except perhaps a copy of sshd for remote admin - with only specific machines allowed to connect) it's certainly not easy to break into. And I haven't ever had to worry about viruses. If they're a concern anything along the lines of tripwire can take care of them. More importantly, if a user has no write permissions on the software their using, they can't infect it.
Again, I think it depends on the person. You're right that breaking something, and seeing that it is fixable is a great way to get over fear. But I think there are people who it's difficult to get to use a computer at all because they're afraid of breaking it. If you tell them it's set up so they can't hurt it, they're more willing to try it.
Also, how are mac's safer from viruses then linux boxes?
I keep considering writing an e theme for people who have no computer experience. The goal would be to make it extremely obvious how to do everything. Often the buttons used on borders are obvious to us, but for the same reason that hyroglifics (sp) were obvious to anchient egyptians - they were used to them. Ideally a button should make it extremely clear what clicking on it will do. Buttons on the desktop can be made as large as is necessry, and also very obvious.
1) You've never gotten a call from a relative who knows less than you do about some program, going "how do I fix blah" or "how do I make it do blah"? Being able to connect to the machine can make these questions a lot easier to answer. 2) In my experience people who are uncomfortable with computers feel better when they know the system is set up in such a way that they can't hurt anything. 3) If the machine is powerful enough to run KDE comfortably everything the user sees can be icon based. In several ways KDE appears to be easier for new users than windows. 4) If the machine is powerful enough to run enlightenment comfortably (and you're willing to expend a little effort), it is possilbe to make an extremely easy to use enlightenment theme. Rather than an X symbolizing quit program, you can have a button that says quit program - it's annoying, but no one would miss the meaning.
Some people have the personality to tinker away, and some don't. My grandfather bought an XT years ago because my father and uncle were into computers, and tinkered until he could use it. But he wasn't afraid to mess around with it. On the other hand I have a grandmother who I had trouble showing how to play solitare on the computer because she was afriad she'd break it. There's no way that she could make it through an installation - particularly not if warnings like "autoprobing may damage hardware" appeared.
If someone wants to explore, by all means let them. But if they want to try out some of this stuff, but are scared, give them a system where they don't have to worry.
I think so. There have been articles about similar things recently. The only thing is that it's harder to get it declared a standard if you patent it. There need to be 2 independent implimentations for it to become a standard, even in name. In practice you have to convince people to use it, and if you charge a lot, that's hard.
The bayesian networks thing sounds like the situation is similar to that of RSA. RSA is based on a theorem in number theory that had existed for years before RSA was patented. But the RSA patent simply covers applying that theorm to cryptography on the computer. It was patentable because it was an application that no one had tried before. MS may be claiming the same thing with bayesian networks - that they were the first to apply them to computers.
It only talks about the benefits to internet.com, and how this gives them leverage in the open source community. Some mention of why this is good for linux today's readers would have been nice. As well as an explanation of what they think they can do for the OS community. This article sounded like "Look we just bought a big Linux site. Now we can exploit Linux users. Yay us."
It looks like their trying to provide a reasonably secure configuration for home users as a default. Other distributions I've seen have been pretty bad about this. The first time I installed Redhat I think I could have taken the place of any server in existence. Distinctions like "ftp client" and "ftp daemon" are confusing the first time around.
that their losing more customers this way them either decision would have cost them. I'm under the impression that they really need the G4 to succeed, but between the bug itself, and their lousy handling of it, that seems unlikely.
I'll agree that that is another cause, however if the price of software dropped by 50% I don't think the amount of piracy would remain constant. There are a fair nuimber of people who will but software if the price is reasonable. Right now for personal use the price of a lot of software is ridiculous. I'm not arguing that it isn't stealing. I'm argueing that MS is trying to stop it in the wrong way.
It's caused by high prices. This "event" failed because MS was assuming that people with pirated copies of MS products had bought them, and thought they were probably legal. Unless I'm badly mistaken most people with pirated copies of software got them from their friends. And the reason people do this is that software prices are incredibly high. I was told that Borland made their fortune when they released their development tools at a fraction of the going rate for software. People considered their prices fair and actually bought them. Paying $500 for an office suite isn't a big deal for a big company, but it is for an individual. When you're trying to sell software there are a couple of key questions: 1) How many people want this program? 2) How much are they all willing to pay? If software costs substantially more than people are willing to pay, and people want the software badly enough, there will be piracy. How much this piracy is hurting software companies is debatable. Often if they didn't have a pirated copy they wouldn't use the software at all. In that sense it costs them nothing.
Companies would do better to stop trying to "educate" customers about piracy, and figure out how to keep the cost of software in the range that customers can afford.
Note: I'm not argueing that piracy is right. I'm commenting on what I think the real causes are.
First, perhaps next time before flaming a company everybody could wait a couple of days to see how they respond to complaints, and consider their intentions.
Second, everyone keeps commenting that we need a big court battle to uphold the GPL, and every time some big company messes up everybody goes "This is it!" No matter how sure they are that they can win no big company is going to go about it like that. If I were say MS, and I wanted to get the GPL ruled worthless I would take some little tiny subsidiary I don't care about, and have them try to redistribute "Tom's Tetris" or something of the sort. The GPL specifically states that if you violate it you lose all rights to the software whatsoever, including useing it. No large company is going to risk losing the right to use Linux ever, just in case it gets big.
Re:Note: I am Insured. So is Nate.
on
Hemos is Homeless
·
· Score: 1
It wasn't marked down - it was posted by an AC and hasn't been moderated at all (at least not as of when I'm typing this)
In AFS users can create groups. Each group has an owner, who is the only person who can add people to that group. In many situations this is extremely convenient. Issues with conflicting names are dealt with by nameing groups like this owner:mygroup
Here's how I'd do this in UNIX (this may be a really bad idea, that's why I'm throwing it out for comment):
Create a dir called/etc/usergroups which is writable by group groupcreators (obviously too long a name, you get the point)
Then any member of groupcreators can make files which contain a group name and members in this directory. They then run a program (mkusrgrps perhaps?) that goes through this directory, goes through each file, watches for anything that looks suspicious in the entries (attempts to set the gid of this group for instance) and if everything looks ok adds them to/etc/group, with each each group named something like this "username-groupname from file". The trick would be making sure if this group already exists to just update it, not create a new one, and to assign unique gid's to each new group, but that should be easy - one plus the greatest existing gid.
Actually come to think of it the dir/etc/usergroups is pointless here - just only give permission to run mkusrgrps to members of group groupcreators, and have them specify the file containing the info on the cmdline.
I think if the file was parsed very carefully this would provide convenient user managable groups.
Right now we may not like all the traits we've inherited from our parents, but at least we recognize that they did have much choice in them. What about a kid who knows their parents picked their traits?
ok, sorry, I was misinformed. My point was that Debian seemss to be taking th place of Slackware as the distro of choice for experienced users. Newbies tend to use RedHat. That makes offering tech support for Debian seem pointless, although I guess they may be trying to encourage use of it by newbies.
Debian is trying to use only Free Software. This is, while a laudable goal, making it difficult for newbies. Correct me if I'm wrong, but don't they include Mozilla as the browser? That won't go over well in the mainstream. In my experience the people using Debian don't need tech support, which makes this seem strange.
That's very difficult for a company like Intel to do. Here they are making an investment - they hope to earn money through it. They can't just donate foo millions of dollars to "the linux kernel developers." Redhat can give back to the community by hiring linux developers since expanding linux clearly benefits them. If even Redhat tried to donate money to any of these groups they would risk a lawsuit from shareholders.
They say they want to add proprietary software into their distro. I don't know if I like the sound of that. I guess it really depends how they do it, but in theory they can create a distro that's dysfunctional without a couple of proprietary components of theirs, and then say "well, sure, you can download all the GPL'd stuff from our site - it just won't work."
It's more likely that they're refering to adding stuff like word perfect. In that case it's ok, redhat does the same, but redhat's commitment to GPL the stuff they write makes me a lot more comfortable.
Slashdot Mirror
If you configure a linux box so that it's not running network services, (except perhaps a copy of sshd for remote admin - with only specific machines allowed to connect) it's certainly not easy to break into. And I haven't ever had to worry about viruses. If they're a concern anything along the lines of tripwire can take care of them. More importantly, if a user has no write permissions on the software their using, they can't infect it.
Again, I think it depends on the person. You're right that breaking something, and seeing that it is fixable is a great way to get over fear. But I think there are people who it's difficult to get to use a computer at all because they're afraid of breaking it. If you tell them it's set up so they can't hurt it, they're more willing to try it.
Also, how are mac's safer from viruses then linux boxes?
I keep considering writing an e theme for people who have no computer experience. The goal would be to make it extremely obvious how to do everything. Often the buttons used on borders are obvious to us, but for the same reason that hyroglifics (sp) were obvious to anchient egyptians - they were used to them. Ideally a button should make it extremely clear what clicking on it will do.
Buttons on the desktop can be made as large as is necessry, and also very obvious.
1) You've never gotten a call from a relative who knows less than you do about some program, going "how do I fix blah" or "how do I make it do blah"? Being able to connect to the machine can make these questions a lot easier to answer.
2) In my experience people who are uncomfortable with computers feel better when they know the system is set up in such a way that they can't hurt anything.
3) If the machine is powerful enough to run KDE comfortably everything the user sees can be icon based. In several ways KDE appears to be easier for new users than windows.
4) If the machine is powerful enough to run enlightenment comfortably (and you're willing to expend a little effort), it is possilbe to make an extremely easy to use enlightenment theme. Rather than an X symbolizing quit program, you can have a button that says quit program - it's annoying, but no one would miss the meaning.
Some people have the personality to tinker away, and some don't. My grandfather bought an XT years ago because my father and uncle were into computers, and tinkered until he could use it. But he wasn't afraid to mess around with it. On the other hand I have a grandmother who I had trouble showing how to play solitare on the computer because she was afriad she'd break it. There's no way that she could make it through an installation - particularly not if warnings like "autoprobing may damage hardware" appeared.
If someone wants to explore, by all means let them. But if they want to try out some of this stuff, but are scared, give them a system where they don't have to worry.
I think so. There have been articles about similar things recently. The only thing is that it's harder to get it declared a standard if you patent it. There need to be 2 independent implimentations for it to become a standard, even in name. In practice you have to convince people to use it, and if you charge a lot, that's hard.
The bayesian networks thing sounds like the situation is similar to that of RSA. RSA is based on a theorem in number theory that had existed for years before RSA was patented. But the RSA patent simply covers applying that theorm to cryptography on the computer. It was patentable because it was an application that no one had tried before. MS may be claiming the same thing with bayesian networks - that they were the first to apply them to computers.
It only talks about the benefits to internet.com, and how this gives them leverage in the open source community. Some mention of why this is good for linux today's readers would have been nice. As well as an explanation of what they think they can do for the OS community. This article sounded like "Look we just bought a big Linux site. Now we can exploit Linux users. Yay us."
It looks like their trying to provide a reasonably secure configuration for home users as a default. Other distributions I've seen have been pretty bad about this. The first time I installed Redhat I think I could have taken the place of any server in existence. Distinctions like "ftp client" and "ftp daemon" are confusing the first time around.
that their losing more customers this way them either decision would have cost them.
I'm under the impression that they really need the G4 to succeed, but between the bug itself, and their lousy handling of it, that seems unlikely.
> Q: I read that Gates is worth $85 billion and that he's starting
> to give some of it away. Do you think he'd give me a million or two?
How about if I ask his new foundation for a million or two if I'll stop using Linux? (I can always switch to a BSD)
I'll agree that that is another cause, however if the price of software dropped by 50% I don't think the amount of piracy would remain constant. There are a fair nuimber of people who will but software if the price is reasonable. Right now for personal use the price of a lot of software is ridiculous.
I'm not arguing that it isn't stealing. I'm argueing that MS is trying to stop it in the wrong way.
Somehow I not frightened at the possiblity that somebody using an unlicensed copy of windows will crash their computer into me.
It's caused by high prices.
This "event" failed because MS was assuming that people with pirated copies of MS products had bought them, and thought they were probably legal. Unless I'm badly mistaken most people with pirated copies of software got them from their friends.
And the reason people do this is that software prices are incredibly high. I was told that Borland made their fortune when they released their development tools at a fraction of the going rate for software. People considered their prices fair and actually bought them. Paying $500 for an office suite isn't a big deal for a big company, but it is for an individual.
When you're trying to sell software there are a couple of key questions:
1) How many people want this program?
2) How much are they all willing to pay?
If software costs substantially more than people are willing to pay, and people want the software badly enough, there will be piracy.
How much this piracy is hurting software companies is debatable. Often if they didn't have a pirated copy they wouldn't use the software at all. In that sense it costs them nothing.
Companies would do better to stop trying to "educate" customers about piracy, and figure out how to keep the cost of software in the range that customers can afford.
Note: I'm not argueing that piracy is right. I'm commenting on what I think the real causes are.
First, perhaps next time before flaming a company everybody could wait a couple of days to see how they respond to complaints, and consider their intentions.
Second, everyone keeps commenting that we need a big court battle to uphold the GPL, and every time some big company messes up everybody goes "This is it!" No matter how sure they are that they can win no big company is going to go about it like that. If I were say MS, and I wanted to get the GPL ruled worthless I would take some little tiny subsidiary I don't care about, and have them try to redistribute "Tom's Tetris" or something of the sort. The GPL specifically states that if you violate it you lose all rights to the software whatsoever, including useing it. No large company is going to risk losing the right to use Linux ever, just in case it gets big.
It wasn't marked down - it was posted by an AC and hasn't been moderated at all (at least not as of when I'm typing this)
I'm not sure exactly what he means, but . . .
/etc/usergroups which is writable by group groupcreators (obviously too long a name, you get the point)
/etc/group, with each each group named something like this "username-groupname from file". The trick would be making sure if this group already exists to just update it, not create a new one, and to assign unique gid's to each new group, but that should be easy - one plus the greatest existing gid.
/etc/usergroups is pointless here - just only give permission to run mkusrgrps to members of group groupcreators, and have them specify the file containing the info on the cmdline.
In AFS users can create groups. Each group has an owner, who is the only person who can add people to that group. In many situations this is extremely convenient. Issues with conflicting names are dealt with by nameing groups like this owner:mygroup
Here's how I'd do this in UNIX (this may be a really bad idea, that's why I'm throwing it out for comment):
Create a dir called
Then any member of groupcreators can make files which contain a group name and members in this directory. They then run a program (mkusrgrps perhaps?) that goes through this directory, goes through each file, watches for anything that looks suspicious in the entries (attempts to set the gid of this group for instance) and if everything looks ok adds them to
Actually come to think of it the dir
I think if the file was parsed very carefully this would provide convenient user managable groups.
Right now we may not like all the traits we've inherited from our parents, but at least we recognize that they did have much choice in them. What about a kid who knows their parents picked their traits?
ok, sorry, I was misinformed. My point was that Debian seemss to be taking th place of Slackware as the distro of choice for experienced users. Newbies tend to use RedHat. That makes offering tech support for Debian seem pointless, although I guess they may be trying to encourage use of it by newbies.
Debian is trying to use only Free Software. This is, while a laudable goal, making it difficult for newbies. Correct me if I'm wrong, but don't they include Mozilla as the browser? That won't go over well in the mainstream. In my experience the people using Debian don't need tech support, which makes this seem strange.
I'm concerned that they plan to be a parasite. From the sound of things they plan to write stuff, but not GPL it.
I didn't say there was anything wrong with it - it's just that Intel can't do it. They'd risk a lawsuit from their shareholders.
According to this article they plan to make proprietary additions.
That's very difficult for a company like Intel to do. Here they are making an investment - they hope to earn money through it. They can't just donate foo millions of dollars to "the linux kernel developers." Redhat can give back to the community by hiring linux developers since expanding linux clearly benefits them. If even Redhat tried to donate money to any of these groups they would risk a lawsuit from shareholders.
They say they want to add proprietary software into their distro. I don't know if I like the sound of that. I guess it really depends how they do it, but in theory they can create a distro that's dysfunctional without a couple of proprietary components of theirs, and then say "well, sure, you can download all the GPL'd stuff from our site - it just won't work."
It's more likely that they're refering to adding stuff like word perfect. In that case it's ok, redhat does the same, but redhat's commitment to GPL the stuff they write makes me a lot more comfortable.
Isn't Debian non-profit? You can't invest in a non-profit organization - only donate them money.