** Flame disclaimer - comments below are my opinion and personal experience ** Question to you or anyone. Why would it not be a good desktop system? Just ease of use, or lack of apps, or what?
A combination of all of the above. Getting Gnome/KDE working on FreeBSD isn't quite as no-brainer-ish as it is on say RH9 or Suse. Getting it working on OpenBSD (as with most apps) is another notch up the difficultly ladder. Anyone who knows what ~/.xinitrc means will have OBSD or FBSD up and running with their desktop of choice in an evening, but expect more 'care and feeding' initially. Aren't most apps that most folks use on a typical desktop available? Browser, email, chat client, media players, editors, etc?
Yes. The vast majority are, and with ports you have the option of a Gentoo-like optimized compile for your hardware.
Just wondering because I keep threatening myself to switch from Linux, for better firewalling and a tighter but smaller community.
This goes both ways, sometimes its more difficult to find a solution to your specific problem in the community. Also, some of the OBSD folk have been known to be a bit....a....abrasive(?)
I like that the apps get relooked at,audited before inclusion, I like that part a LOT, because I didn't know they did that. That makes sense to me. I'd rather have fewer apps, but better quality apps. I take it this concept is unique to openBSD?
Again, its a choice you get to make as far as the trade-off. The OBSD folks only do the heavy-duty auditing on the core system components (look at the web site to find out what these are). You can build an outward-facing box running these applications and know you're really getting some of the best of what's out there. OTOH, you can download source and compile this that and the other thing with X and everything else, to some degree compromising the "bulletproof OS" idea, but trading that for usability. It comes down to the trade-offs you're willing to make, and to me, it seems like most Linux variants target maximum compatibility with HW and SW, and maximum usability, potentially at the expense of performance and security and reliability. FreeBSD seems to focus on reliability, and secondarily performance and security, with a fair bit of effort still spent on usability. OBSD focuses maximum effort on security, and everything else is secondary - if something has to be sacrificied in the name of security, it is done without question, regardless if it is a minor change or a major usability or friendly feature.
Reading the description in the article for installing and a few tweaks doesn't seem that difficult at first glance. I am impressed with their claim of only one remote exploit in many years.
You can install a base system in 30 minutes using 500MB of disk if you follow the handbook on the first page of the web site. Its worth at least an evening to investigate.
Last question, how does it run on older hardware in a GUI desktop environemnt? Acceptable, fast, dog slow, what? Similar to linux from one of the big vendors?
X is X. No getting away from that.
What is a practical minimum set of hardware specs for a good GUI environemnt?
I've run a "usable" system on a 500 Celeron with 256M of RAM without tweaks. I wouldn't build brains on it, but it did what I needed.
Sorry for all the questions, but I truly am interested. The more I am on the net, the more security I want, and this latest month has seen just a slew of potentially bad news exploits. I don't want to fool with it, I think it makes more sense to start out with the best and most secure system and learn and build from that, rather than patch and patch and patch all the time and sit and surf with your fingers crossed.
At least build one OBSD as a router/firewall and put your more vulnerable machines behind it. I think a few others here have done similarly with good results.
Actually, I've heard this is a myth too - the guy a few posts up has the reason: The airlines don't let you use phones in flight becuase they were ASKED not to by the cell phone companies. If the planes on 9/11 were flying at their normal crusing speeds, cell phone calls would have been impossible, and those supposed calls to family members would not have taken place (they could have been AirPhone calls though, and the media, families, etc., just got the details wrong, which would be understandable under the circumstances).
Actually, it does, Itanium runs 32bit x86 code in emulation mode, and will do so slower than a middle-of-the-road Xeon will today. It won't be able to touch the performance of the Intel Hybrid running 32-bit code.
The thread from the top has been a battle between the "Itanium is real 64-bit and 32/64 sucks" crowd, and the "32/64 addresses the only mass market requirement outside the scope of todays 32 bit Intel architecture, large memory access, and with that, Corporate America will largely be spared the Itanium" group. From a technical perspective, the first group are correct in their assumed technical superiority, in this they take their rightful place alongside Alpha, OS/2, Microchannel, Betamax, etc.
This is the one area in which Itanium has no benefit over a hybrid. All of the other things that are "better" about "true" 64-bit chips are still in effect, but the memory addressability is the same, so your point is a restatement of the argument in favor of the hybrids NOT in favor of "true" 64-bit computing; Hybrids will solve the #1 problem experienced by large requirements on current platforms, without requiring major architectural changes to the environment. This will stave off widespread migration of "enterprise" customers to the likes of Power and Itanium for another 3-5 years.
Why do you think Ebay prohibits the sale of firearms?
Because their lawyers haven't figured out a way to assure they couldn't be held liable in the ensuing lawsuits. The day they do they will displace WalMart as the world's #1 distributor...
I work almost exclusively with small businesses, 10 or 20 of which have many horrid access databases which have grown to the point of being in the state you'd expect in scenario #2, however, none of them have the money to do it, so they simply deal with whatever problems result. Its still better than option #1 and only a few have more than 100 employees so #3 isn't an option either.
The other 80 or so clients I've seen who use access have absolutely no problems. It does what it does quite well, and for small, relatively simple requirements (read: vast majority of small-business needs) it works fine. I'd love to be able to convince them to use Postgres or MySql, but it ain't gonna happen any time soon - there are enough things that are actually broken to suck up those resources...
I would venture to guess, the % of Americans with non-retirement, soluble $$'s in the stock market, that they can manage and access freely, is a very limited number. So, no, the majority of US citzens do not benefit from the stock market.
The fact the an investment is not liquid does not detract from its benefit. You are exactly wrong.
Even after your child care and mortgage, you have more money doing the same job than one of the folks in these so called "englightened" countries. The problem with everyone pointing to Germany or the Netherlands is they don't realize they're paying for it one way or another. 35 weeks of leave doesn't come free, its only "free at the time".
Socialized medicine in Canada results in a per-capita national debt second in the wolrd only to Italy, and cannot be sustained indefinitely without major changes or even MORE taxes, in a nation where anyone making more than about US$75,000 pays a total of 55% in taxes.
If you were to negotiate an annual %15 pay cut with your employer in exchange for 6 weeks vacation, they might entertain the thought. Any company in the US who would introduce 35 weeks paid leave for maternity would have to lower the salary of every employee in the company to pay for it, have no raises for like 5 years, or explain to shareholders (current employees 401K and retirees in large part) why their profits are cut 50% now and forever more.
The US didn't sign the Kyoto protocol because it could have been *devastating* to the economy. If you actually read any of it, the larger (US, Canada, EU, etc) nations were responsible for much more than the smaller nations, and the more wealthy nations would have to pay for what amounts to "clean up services" and "factory augmentation" for countries that couldn't afford it themselves when the years certain pollution restriction limits come due. Bottom Line: Mexico and the like need simply do nothing until the last year of a new requirement, then show some evidence that they can't afford to pay, then we'd be forced to retrofit their factories and clean up their lakes and streams. Screw that. God Bless George Bush for not signing this piece of crap (I don't like GWB and rarely say such things, perhaps one decent advisor snuck through). I want to help the environment, but I don't feel compelled to pay $1,000 a year in additional taxes in 2009 because they've got some factories puking sludge that need to be cleaned up.
In global agreements of any kind, America generally gets screwed, we should enter into very few of them.
Well, I think its fair to say that to be a "significant" expense, it would appear somewhere in our accounting system. I run a small business, and significant expenses include labor, auto maintenance, parts, etc. I don't have an account in Quicken for "counterfeit loss", yet my books balance and comparing one year to two years to five years back, things on the expense side don't look too much different. So I can say with a degree of certainty, in my case, counterfeit currency is NOT a significant expense for this small business.
Yeah, right:) You could have bought three of those servers for the price of a Mac. Don't get me wrong, I like Macs, but no Mac has ever been anywhere near as affordable as a comparably performing PC. You spend the extra money for the cool factor and specific software, not for performance.
Well, in my country there is a fifth amendment to our constitution that prevents the government from compelling us to incriminate ourselves in criminal court. This is a direct violation of that amendment.
The biggest argument against this, though, is that there is no third-party validation. The local cops have to calibrate their radar guns, on a regular basis, with some known standard. The black box on the other hand, is configured however GM wants it to be configured, and is held to no external standard. In fact, they could alter the data to protect themselves in certain error conditions and no one could do anything about it because they're accountable to no one. Other factors such as damage to the device, and more likely, damage to sensors, could alter the readings recorded by the device, and the poor driver would be convicted with almost no recourse - the car company probably couldn't even be forced to provide the source to the box code since its a trade secret.
This is one of the worst ideas ever. At least word is out now, soon there will be a pissed off engineer who will figure out how to bypass it....
Yeah, but the difference is you *know* you're going to a horror movie. War of The Worlds did not inform the audience that it was for entertainment, it was deliberately produced to sound like a news broadcast, which is why so many people freaked. Imagine if you were watching World News Tonight and the wall behind Peter Jennings melted to reveal a UFO landing and a dozen aliens with ray-guns vaporizing stuff. Jennings stands up and says "Hey, what's..." and he gets blasted by one of them before 10 million viewers. This would probably cause some hysteria, even in the most sarcastic and cynical circles since everyone knows ABC would NEVER stage something like that. People would be freaking for sure.
If you are using Exchange, it even allows you to publicly expose your request list to the web.
Which will be nice after the first day or so, when 10 or 20 virus have infected your machine. People working for other companies can submit their own requests to your system to ask you to stop attacking them.
If you hire consultants, they'll go for maxing out their billable hours and their expense sheet. In-house guys are generally on salary and don't have much in the way of conflicts of interest (umm... IF, that is, you don't hire any fanatics).
If you manage consultants correctly you can account for their time and this is not an issue. The salaried guys OTOH are impossible to get to do *anything*, their biggest decisions of the day are whether to complain about the coffee, the company, or the weather, especially at large companies, where it costs $100k to fire someone who's been there for more than five years.
He has a significant other. He told us, he was playing snake on it.
Personally, I want to get this vest, and have 10 of my friends get one too. We'll all go down to the bar at JFK airport, never fly anywhere, just sit in the bar, acting really drunk, and talking about how much we love Allah. Then, we'll run Morpheus on them, throw them un unsuspecting passersby, and call the RIAA, what glorious fun.
The report said that reliance on a single software vendor exposes one to undue security risks. Do you have any basis whatsoever for disagreeig with that conclusion?
Yes. Reliance on a single vendor allows one to maximize the syngergy between products (if it exists, maybe MS is a bad example), minimize training costs (again, assuming product continuity), and in many cases, increase security across the board in cases where the products are designed to work together. Also, ROI for support contracts is an inverse proportion to the number of vendors in an environment as the amount of time on the phone spent convincing them "its not the other guy's shit" is reduced:)
Slashdot Mirror
** Flame disclaimer - comments below are my opinion and personal experience **
Question to you or anyone. Why would it not be a good desktop system? Just ease of use, or lack of apps, or what?
A combination of all of the above. Getting Gnome/KDE working on FreeBSD isn't quite as no-brainer-ish as it is on say RH9 or Suse. Getting it working on OpenBSD (as with most apps) is another notch up the difficultly ladder. Anyone who knows what ~/.xinitrc means will have OBSD or FBSD up and running with their desktop of choice in an evening, but expect more 'care and feeding' initially.
Aren't most apps that most folks use on a typical desktop available? Browser, email, chat client, media players, editors, etc?
Yes. The vast majority are, and with ports you have the option of a Gentoo-like optimized compile for your hardware.
Just wondering because I keep threatening myself to switch from Linux, for better firewalling and a tighter but smaller community.
This goes both ways, sometimes its more difficult to find a solution to your specific problem in the community. Also, some of the OBSD folk have been known to be a bit....a....abrasive(?)
I like that the apps get relooked at,audited before inclusion, I like that part a LOT, because I didn't know they did that. That makes sense to me. I'd rather have fewer apps, but better quality apps. I take it this concept is unique to openBSD?
Again, its a choice you get to make as far as the trade-off. The OBSD folks only do the heavy-duty auditing on the core system components (look at the web site to find out what these are). You can build an outward-facing box running these applications and know you're really getting some of the best of what's out there. OTOH, you can download source and compile this that and the other thing with X and everything else, to some degree compromising the "bulletproof OS" idea, but trading that for usability. It comes down to the trade-offs you're willing to make, and to me, it seems like most Linux variants target maximum compatibility with HW and SW, and maximum usability, potentially at the expense of performance and security and reliability. FreeBSD seems to focus on reliability, and secondarily performance and security, with a fair bit of effort still spent on usability. OBSD focuses maximum effort on security, and everything else is secondary - if something has to be sacrificied in the name of security, it is done without question, regardless if it is a minor change or a major usability or friendly feature.
Reading the description in the article for installing and a few tweaks doesn't seem that difficult at first glance. I am impressed with their claim of only one remote exploit in many years.
You can install a base system in 30 minutes using 500MB of disk if you follow the handbook on the first page of the web site. Its worth at least an evening to investigate.
Last question, how does it run on older hardware in a GUI desktop environemnt? Acceptable, fast, dog slow, what? Similar to linux from one of the big vendors?
X is X. No getting away from that.
What is a practical minimum set of hardware specs for a good GUI environemnt?
I've run a "usable" system on a 500 Celeron with 256M of RAM without tweaks. I wouldn't build brains on it, but it did what I needed.
Sorry for all the questions, but I truly am interested. The more I am on the net, the more security I want, and this latest month has seen just a slew of potentially bad news exploits. I don't want to fool with it, I think it makes more sense to start out with the best and most secure system and learn and build from that, rather than patch and patch and patch all the time and sit and surf with your fingers crossed.
At least build one OBSD as a router/firewall and put your more vulnerable machines behind it. I think a few others here have done similarly with good results.
Actually, I've heard this is a myth too - the guy a few posts up has the reason: The airlines don't let you use phones in flight becuase they were ASKED not to by the cell phone companies. If the planes on 9/11 were flying at their normal crusing speeds, cell phone calls would have been impossible, and those supposed calls to family members would not have taken place (they could have been AirPhone calls though, and the media, families, etc., just got the details wrong, which would be understandable under the circumstances).
Actually, it does, Itanium runs 32bit x86 code in emulation mode, and will do so slower than a middle-of-the-road Xeon will today. It won't be able to touch the performance of the Intel Hybrid running 32-bit code.
The thread from the top has been a battle between the "Itanium is real 64-bit and 32/64 sucks" crowd, and the "32/64 addresses the only mass market requirement outside the scope of todays 32 bit Intel architecture, large memory access, and with that, Corporate America will largely be spared the Itanium" group. From a technical perspective, the first group are correct in their assumed technical superiority, in this they take their rightful place alongside Alpha, OS/2, Microchannel, Betamax, etc.
This is the one area in which Itanium has no benefit over a hybrid. All of the other things that are "better" about "true" 64-bit chips are still in effect, but the memory addressability is the same, so your point is a restatement of the argument in favor of the hybrids NOT in favor of "true" 64-bit computing; Hybrids will solve the #1 problem experienced by large requirements on current platforms, without requiring major architectural changes to the environment. This will stave off widespread migration of "enterprise" customers to the likes of Power and Itanium for another 3-5 years.
Because their lawyers haven't figured out a way to assure they couldn't be held liable in the ensuing lawsuits. The day they do they will displace WalMart as the world's #1 distributor...
So, in other words......yes.
I work almost exclusively with small businesses, 10 or 20 of which have many horrid access databases which have grown to the point of being in the state you'd expect in scenario #2, however, none of them have the money to do it, so they simply deal with whatever problems result. Its still better than option #1 and only a few have more than 100 employees so #3 isn't an option either. The other 80 or so clients I've seen who use access have absolutely no problems. It does what it does quite well, and for small, relatively simple requirements (read: vast majority of small-business needs) it works fine. I'd love to be able to convince them to use Postgres or MySql, but it ain't gonna happen any time soon - there are enough things that are actually broken to suck up those resources...
Great elitist attitude! Yes, we're ashamed of our dates! As we were ashamed we didn't have pretty red uniforms in that war....
The fact the an investment is not liquid does not detract from its benefit. You are exactly wrong.
Even after your child care and mortgage, you have more money doing the same job than one of the folks in these so called "englightened" countries. The problem with everyone pointing to Germany or the Netherlands is they don't realize they're paying for it one way or another. 35 weeks of leave doesn't come free, its only "free at the time".
Socialized medicine in Canada results in a per-capita national debt second in the wolrd only to Italy, and cannot be sustained indefinitely without major changes or even MORE taxes, in a nation where anyone making more than about US$75,000 pays a total of 55% in taxes.
If you were to negotiate an annual %15 pay cut with your employer in exchange for 6 weeks vacation, they might entertain the thought. Any company in the US who would introduce 35 weeks paid leave for maternity would have to lower the salary of every employee in the company to pay for it, have no raises for like 5 years, or explain to shareholders (current employees 401K and retirees in large part) why their profits are cut 50% now and forever more.
Be careful what you wish for.
The US didn't sign the Kyoto protocol because it could have been *devastating* to the economy. If you actually read any of it, the larger (US, Canada, EU, etc) nations were responsible for much more than the smaller nations, and the more wealthy nations would have to pay for what amounts to "clean up services" and "factory augmentation" for countries that couldn't afford it themselves when the years certain pollution restriction limits come due. Bottom Line: Mexico and the like need simply do nothing until the last year of a new requirement, then show some evidence that they can't afford to pay, then we'd be forced to retrofit their factories and clean up their lakes and streams. Screw that. God Bless George Bush for not signing this piece of crap (I don't like GWB and rarely say such things, perhaps one decent advisor snuck through). I want to help the environment, but I don't feel compelled to pay $1,000 a year in additional taxes in 2009 because they've got some factories puking sludge that need to be cleaned up.
In global agreements of any kind, America generally gets screwed, we should enter into very few of them.
Which changes the answer to the question in no way, still Bill Clinton
Well, I think its fair to say that to be a "significant" expense, it would appear somewhere in our accounting system. I run a small business, and significant expenses include labor, auto maintenance, parts, etc. I don't have an account in Quicken for "counterfeit loss", yet my books balance and comparing one year to two years to five years back, things on the expense side don't look too much different. So I can say with a degree of certainty, in my case, counterfeit currency is NOT a significant expense for this small business.
Actually, I read on Slashdot that VxWorks is dying.....
PF
Yeah, right :) You could have bought three of those servers for the price of a Mac. Don't get me wrong, I like Macs, but no Mac has ever been anywhere near as affordable as a comparably performing PC. You spend the extra money for the cool factor and specific software, not for performance.
Well, in my country there is a fifth amendment to our constitution that prevents the government from compelling us to incriminate ourselves in criminal court. This is a direct violation of that amendment. The biggest argument against this, though, is that there is no third-party validation. The local cops have to calibrate their radar guns, on a regular basis, with some known standard. The black box on the other hand, is configured however GM wants it to be configured, and is held to no external standard. In fact, they could alter the data to protect themselves in certain error conditions and no one could do anything about it because they're accountable to no one. Other factors such as damage to the device, and more likely, damage to sensors, could alter the readings recorded by the device, and the poor driver would be convicted with almost no recourse - the car company probably couldn't even be forced to provide the source to the box code since its a trade secret. This is one of the worst ideas ever. At least word is out now, soon there will be a pissed off engineer who will figure out how to bypass it....
Yeah, but the difference is you *know* you're going to a horror movie. War of The Worlds did not inform the audience that it was for entertainment, it was deliberately produced to sound like a news broadcast, which is why so many people freaked. Imagine if you were watching World News Tonight and the wall behind Peter Jennings melted to reveal a UFO landing and a dozen aliens with ray-guns vaporizing stuff. Jennings stands up and says "Hey, what's..." and he gets blasted by one of them before 10 million viewers. This would probably cause some hysteria, even in the most sarcastic and cynical circles since everyone knows ABC would NEVER stage something like that. People would be freaking for sure.
Its like Tracy Lords
Which will be nice after the first day or so, when 10 or 20 virus have infected your machine. People working for other companies can submit their own requests to your system to ask you to stop attacking them.
If you manage consultants correctly you can account for their time and this is not an issue. The salaried guys OTOH are impossible to get to do *anything*, their biggest decisions of the day are whether to complain about the coffee, the company, or the weather, especially at large companies, where it costs $100k to fire someone who's been there for more than five years.
DIRK: "but that's a whole nother thing"
FADE
I never suggested that security was increased by reliance on a single vendor. What I said was security was decreased by using more than one vendor.
He has a significant other. He told us, he was playing snake on it. Personally, I want to get this vest, and have 10 of my friends get one too. We'll all go down to the bar at JFK airport, never fly anywhere, just sit in the bar, acting really drunk, and talking about how much we love Allah. Then, we'll run Morpheus on them, throw them un unsuspecting passersby, and call the RIAA, what glorious fun.
Yes. Reliance on a single vendor allows one to maximize the syngergy between products (if it exists, maybe MS is a bad example), minimize training costs (again, assuming product continuity), and in many cases, increase security across the board in cases where the products are designed to work together. Also, ROI for support contracts is an inverse proportion to the number of vendors in an environment as the amount of time on the phone spent convincing them "its not the other guy's shit" is reduced