Stuff I like about OpenBSD: -The man pages. -The ports tree (yes, Gentoo has something similar) -Massive effort has gone towards making all the configuration move to one place and to make it of consistent style. -Shell builtins get their own man page. -Secure stuff is easier to do. For example, adduser by default gives every user their own group. -Only the required set of tools is included by default, rather than everything being the default. This is a matter of opinion, but to me it's easier to install something if I need it. If I decide I need something, the ports tree makes it very easy to install. -Man pages give much more detail, are always kept up to date, and describe all dependencies of everything including their use of environment variables.
For high level stuff, it just takes less time to figure out. I need books and time to get this stuff done with Linux, but I have not yet had to consult anything but man pages on OpenBSD. I bought "Absoloute OpenBSD" before I installed OpenBSD. It's excelent but useless, the man pages give too much detail about all the little tricks for a book to be necessary.
As far as MacOS X goes... too much stuff requires libraries or drivers. The libraries aren't always easy to get, and the drivers often don't exist. For high end stuff, this makes MacOS a bit toothless.
IMO, Linux has a loooong way to go. MacOS X wins (on ease of use) for trivial to low end power stuff, and OpenBSD wins (on ease of use) easily on everything above.
You haven't lived until you've dealt with OpenBSD man pages.
"I suspect much of the 95% of PCs you speak of are safely walled up in institutions, schools and corporations private networks, which are generally out of scope for a worm like blaster to target. "
Tell that to the good people behind the firewall in NYU's dorms. Someone brought an infected computer in, and now they're getting about 2 probes/min.
The only thing to do is to make the updates as infrequent as possible through extensive code audits before software is released, and as painless as possible when they are required. Apple has it half right, they have infrequent painful updates (more painful than Windows!), and from what I've seen that's enough to get most people to do it.
A month ago, people were making fun of me when I told them the sky was falling.
I have a friend that's been behind a firewall all summer. I told her to install updates and a software firewall before she moved back to her dorm, but it was too much effort. When she connected her computer in her dorm today, she had a worm within minutes, behind the uni's firewall. Apparently they had posted messages around the building asking people to make sure they don't have worms and viruses before the connected, but people don't care. It makes me angry sometimes.
I'll tell you one thing though. This friend of mine didn't argue when I told her to install a firewall this time. And now that she sees how much probing goes on, she might listen to me next time.
There are a lot of canditates, and some we don't even know about.
I think the most likely is SCO's motion for a preliminary injunction being rejected. Even idealistic shareholders know they can't win a war of attrition with IBM.
That same mechanism that's keeping their stock price up is very vulnerable to a massive cascade, ready to be triggered by some bad news. They have to keep making these press releases to keep it from doing that even now, and they won't be able to stop it as soon as something big happens.
That'll help in IBM's countersuit. I actually read the entire complaint (needed to procrastinate with a project) and they explicitly accused SCO of FUD to scare people into giving them money and to increase their stock price. Every time SCO does this, they put a nail in their coffin.
I don't know what the burden of proof is for a claim like that, and I don't know if IBM can meet it without something additional. Either way, I believe a part of IBM's strategy is to goad SCO into making statements like this. Each one will make SCO look worse at trial, and some of them might expose them to further lawsuits.
IBM is making sure that even if SCO can win the suit they'll be facing such massive retaliation from so many parties that three billion won't help. I'm not convinced IBM wants to buy SCO, but it will be cheap before they're done.
That just occured to me too! My response at this point would be to send SCO a letter saying "I have a Linux box. If you sue me, you'll have to reveal what the code is that you're talking about.".
There are only three problems with that plan.
a) They're going to go after the big companies, not me.
b) They're going to do it in the best legal environment they can get for it. And that environment is not anywhere in Canada.
c) I switched to OpenBSD before this happened, so I'm not in violation even if they're telling the truth. Damn.
You know, I thought their stock price was getting a little low.
They're making IBM's case for them.
Re:The organization has an obvious slant
on
Joining the ACLU?
·
· Score: 1
Those opinions are often found in the same people that subscribe to weird notions like "by killing people, we make their friends more likely to try to kill us". I'd be happy if such a person was the next president of the US, but I am not holding my breath.
How much you want to bet the people responsible are core wars champions?:)
If not for the fact that it was so costly and dangerous, I'd say it was pretty cool. Well... it's cool anyway. The rules are not as well defined as core wars, and they can change as firewalls are thrown up and new worms and counter-worms emerge in the middle of the contest.
People are going to have to start preparing worm templates and efficient spreading algorithms in preparation for the next vulnerability discovery, so they can launch the initial attack first.
I know. But that's one of the most used optional services. Let's not let the reality distortion feild change our reaction. This is a major fuckup on Apple's part, and the fact that they've handled it well only partially mitigates the damage.
MacOS beats Windows handily in a lot of areas, but there are some things where it's not all there yet. Since not everything many businesses need is available for MacOS, they'd need to support two OSes, which would be much worse.
As an OS X user, I can say that I need my experience to keep the system running smoothly. A novice would still need help, I think. It's only a little better than Windows in that respect.
Now, if these semi-anual superworms don't stop, then I think it'll be more cost effective in the long run, but it's going to take a long time for that to get noticed.
Speaking of which... Is this a Software Update window that's just popped up? It is. Wow. A remote root exploit in OS X. I think I'll install that. Excuse me, this will require a reboot.
I'm pretty happy with it. I'm just getting into OpenBSD, and this book makes it pretty easy to get started on complicated things. It's not for people with no experience, but neither is OpenBSD.
That would provide very high quality random numbers. The analog world creeps in all over the place. The analog world affects the digital equipment, and people use the Internet. That's a pretty good source, I think.
Of course, an attacker can tap your network and find out what the ping times are as well. They're very hard to predict, but the attacker wouldn't have to predict them.
I don't know why IBM hasn't filed for a temporary injunction against the license termination.
They might just ignore the license termination and treat it as though it's not credible. IBM might just be egging SCO on to do increasingly stupid things that will ultimately hurt them in the lawsuit and expose them to liability.
Before this is over, IBM's various countersuits are going to award IBM more in damages than SCO is worth. Even if SCO wins the lawsuit, they're fucked. IBM will pound them into rubble until the only way to settle up will SCO stock... all of it.
In general, though, I agree with your sentiments. One only has to look at GNOME and KDE to see how much the Unix landscape has been infiltrated by people that just don't get it.
I don't really agree... For any UNIX system to perform well as a desktop system, it needs a powerful desktop environment. It doesn't have to be big, but that helps.
For a lot of people, probably most of them, a really powerful desktop environment must be able to do what they want transparently. MacOS and yes, Windows, can do this. KDE, Gnome, AfterStep, these can do it, but not as well as MacOS or Windows. Other wm's like fvwm fail. If you like them, fine, but they don't work for a lot of people.
KDE et all don't infect the rest of the system with complexity, and they're plenty extensible. As a program they're big, but visually it's a lot of small wigits that do one thing well.
It's no worse than having commands built into the shell, or having one big kernel that does everything. Or, God forbid, busybox. Refusing to use big and powerful even when it's the best way and the complexity doesn't spread doesn't make sense.
I think it would be very difficult to stop, say, a thousand planes all launched at once. Obviously there would be issues with launching that many at once, but there would also be issues stopping enough of them to matter.
Hm... they'd have to be automated, but microcontrollers are cheap. They could be programmed to attempt a time on target style arrival, so that only a tiny fraction could be investigated. They could also be set up to drop the payload at a predetermined spot and continue on until they ran out of gas, that would make it difficult to know where people were picking up the drugs from.
I imagine the best way to deal with GPS would be to use it until it was jammed and then fly by dead reckoning until it became available again. A really big GPS blackout would be required to stop the planes if their drop sites were well inside the country.
My objection is that it would tell the spammer that your e-mail address is live, and it would screw up stuff like mailing lists that have URLs to click to confirm you want to be on the list.
I have no problem with loading the page in and of itself. They've sent you the URL, they're soliciting a visit.
Slashdot Mirror
Stuff I like about OpenBSD:
-The man pages.
-The ports tree (yes, Gentoo has something similar)
-Massive effort has gone towards making all the configuration move to one place and to make it of consistent style.
-Shell builtins get their own man page.
-Secure stuff is easier to do. For example, adduser by default gives every user their own group.
-Only the required set of tools is included by default, rather than everything being the default. This is a matter of opinion, but to me it's easier to install something if I need it. If I decide I need something, the ports tree makes it very easy to install.
-Man pages give much more detail, are always kept up to date, and describe all dependencies of everything including their use of environment variables.
For high level stuff, it just takes less time to figure out. I need books and time to get this stuff done with Linux, but I have not yet had to consult anything but man pages on OpenBSD. I bought "Absoloute OpenBSD" before I installed OpenBSD. It's excelent but useless, the man pages give too much detail about all the little tricks for a book to be necessary.
As far as MacOS X goes... too much stuff requires libraries or drivers. The libraries aren't always easy to get, and the drivers often don't exist. For high end stuff, this makes MacOS a bit toothless.
IMO, Linux has a loooong way to go. MacOS X wins (on ease of use) for trivial to low end power stuff, and OpenBSD wins (on ease of use) easily on everything above.
You haven't lived until you've dealt with OpenBSD man pages.
"I suspect much of the 95% of PCs you speak of are safely walled up in institutions, schools and corporations private networks, which are generally out of scope for a worm like blaster to target. " Tell that to the good people behind the firewall in NYU's dorms. Someone brought an infected computer in, and now they're getting about 2 probes/min.
The only thing to do is to make the updates as infrequent as possible through extensive code audits before software is released, and as painless as possible when they are required. Apple has it half right, they have infrequent painful updates (more painful than Windows!), and from what I've seen that's enough to get most people to do it.
A month ago, people were making fun of me when I told them the sky was falling.
I have a friend that's been behind a firewall all summer. I told her to install updates and a software firewall before she moved back to her dorm, but it was too much effort. When she connected her computer in her dorm today, she had a worm within minutes, behind the uni's firewall. Apparently they had posted messages around the building asking people to make sure they don't have worms and viruses before the connected, but people don't care. It makes me angry sometimes.
I'll tell you one thing though. This friend of mine didn't argue when I told her to install a firewall this time. And now that she sees how much probing goes on, she might listen to me next time.
There are a lot of canditates, and some we don't even know about.
I think the most likely is SCO's motion for a preliminary injunction being rejected. Even idealistic shareholders know they can't win a war of attrition with IBM.
This isn't a cheap shot:
That same mechanism that's keeping their stock price up is very vulnerable to a massive cascade, ready to be triggered by some bad news. They have to keep making these press releases to keep it from doing that even now, and they won't be able to stop it as soon as something big happens.
That'll help in IBM's countersuit. I actually read the entire complaint (needed to procrastinate with a project) and they explicitly accused SCO of FUD to scare people into giving them money and to increase their stock price. Every time SCO does this, they put a nail in their coffin.
I don't know what the burden of proof is for a claim like that, and I don't know if IBM can meet it without something additional. Either way, I believe a part of IBM's strategy is to goad SCO into making statements like this. Each one will make SCO look worse at trial, and some of them might expose them to further lawsuits.
IBM is making sure that even if SCO can win the suit they'll be facing such massive retaliation from so many parties that three billion won't help. I'm not convinced IBM wants to buy SCO, but it will be cheap before they're done.
Because learning is fun and you can't learn everything you want to know by coding alone.
And he might want to teach.
That just occured to me too! My response at this point would be to send SCO a letter saying "I have a Linux box. If you sue me, you'll have to reveal what the code is that you're talking about.".
There are only three problems with that plan.
a) They're going to go after the big companies, not me.
b) They're going to do it in the best legal environment they can get for it. And that environment is not anywhere in Canada.
c) I switched to OpenBSD before this happened, so I'm not in violation even if they're telling the truth. Damn.
You know, I thought their stock price was getting a little low.
They're making IBM's case for them.
Those opinions are often found in the same people that subscribe to weird notions like "by killing people, we make their friends more likely to try to kill us". I'd be happy if such a person was the next president of the US, but I am not holding my breath.
How much you want to bet the people responsible are core wars champions? :)
If not for the fact that it was so costly and dangerous, I'd say it was pretty cool. Well... it's cool anyway. The rules are not as well defined as core wars, and they can change as firewalls are thrown up and new worms and counter-worms emerge in the middle of the contest.
People are going to have to start preparing worm templates and efficient spreading algorithms in preparation for the next vulnerability discovery, so they can launch the initial attack first.
I know. But that's one of the most used optional services. Let's not let the reality distortion feild change our reaction. This is a major fuckup on Apple's part, and the fact that they've handled it well only partially mitigates the damage.
What?
The default terminal application that comes with OS X allows an arbitrary number of terminal windows.
MacOS beats Windows handily in a lot of areas, but there are some things where it's not all there yet. Since not everything many businesses need is available for MacOS, they'd need to support two OSes, which would be much worse.
As an OS X user, I can say that I need my experience to keep the system running smoothly. A novice would still need help, I think. It's only a little better than Windows in that respect.
Now, if these semi-anual superworms don't stop, then I think it'll be more cost effective in the long run, but it's going to take a long time for that to get noticed.
Speaking of which... Is this a Software Update window that's just popped up? It is. Wow. A remote root exploit in OS X. I think I'll install that. Excuse me, this will require a reboot.
I'm pretty happy with it. I'm just getting into OpenBSD, and this book makes it pretty easy to get started on complicated things. It's not for people with no experience, but neither is OpenBSD.
That would provide very high quality random numbers. The analog world creeps in all over the place. The analog world affects the digital equipment, and people use the Internet. That's a pretty good source, I think.
Of course, an attacker can tap your network and find out what the ping times are as well. They're very hard to predict, but the attacker wouldn't have to predict them.
Mitochondria don't leave their host cells. They're foreign anyway, they're basically little bacteria, even the ones that come from humans.
I don't know why IBM hasn't filed for a temporary injunction against the license termination.
They might just ignore the license termination and treat it as though it's not credible. IBM might just be egging SCO on to do increasingly stupid things that will ultimately hurt them in the lawsuit and expose them to liability.
Before this is over, IBM's various countersuits are going to award IBM more in damages than SCO is worth. Even if SCO wins the lawsuit, they're fucked. IBM will pound them into rubble until the only way to settle up will SCO stock... all of it.
In general, though, I agree with your sentiments. One only has to look at GNOME and KDE to see how much the Unix landscape has been infiltrated by people that just don't get it.
I don't really agree... For any UNIX system to perform well as a desktop system, it needs a powerful desktop environment. It doesn't have to be big, but that helps.
For a lot of people, probably most of them, a really powerful desktop environment must be able to do what they want transparently. MacOS and yes, Windows, can do this. KDE, Gnome, AfterStep, these can do it, but not as well as MacOS or Windows. Other wm's like fvwm fail. If you like them, fine, but they don't work for a lot of people.
KDE et all don't infect the rest of the system with complexity, and they're plenty extensible. As a program they're big, but visually it's a lot of small wigits that do one thing well.
It's no worse than having commands built into the shell, or having one big kernel that does everything. Or, God forbid, busybox. Refusing to use big and powerful even when it's the best way and the complexity doesn't spread doesn't make sense.
I think it would be very difficult to stop, say, a thousand planes all launched at once. Obviously there would be issues with launching that many at once, but there would also be issues stopping enough of them to matter.
Hm... they'd have to be automated, but microcontrollers are cheap. They could be programmed to attempt a time on target style arrival, so that only a tiny fraction could be investigated. They could also be set up to drop the payload at a predetermined spot and continue on until they ran out of gas, that would make it difficult to know where people were picking up the drugs from.
I imagine the best way to deal with GPS would be to use it until it was jammed and then fly by dead reckoning until it became available again. A really big GPS blackout would be required to stop the planes if their drop sites were well inside the country.
I had about a fivefold increase in spam once when I accidently let an image load.
My objection is that it would tell the spammer that your e-mail address is live, and it would screw up stuff like mailing lists that have URLs to click to confirm you want to be on the list.
I have no problem with loading the page in and of itself. They've sent you the URL, they're soliciting a visit.
Well they could, but instead of load testing the thing people would just bitch about the duplicate.