take MySQL for example First, MySQL does NOT attempt to take over my machine. While they do have precompiled binaries, they have no knowledge of or control of exactly which binaries are running on my machine. Bluntly, in the general case, they *cannot* update my machines. Second, MySQL is not the ONLY thing running on my machine. Securing MySQL only while leaving everything else open seems like a stupid waste of effort. MySQL is responsible for what they do. I am responsible for what I do. A compilation, such as RedHat does have some responsibility for keeping tabs on bug fixes, including security issues. If I install the RedHat RPM, I would expect RedHat not MySQL to be a timely source of information and fixes. If I install MySQL the old-fashioned way neither is in a position to automatically update it on my machines. In the latter case, I would expect to get information that there is a problem from RedHat and the actual source-level patches or diffs from MySQL. No silver bullets, but pretty straight-forward and effective.
They rely for 100% on the users coming to look for them on their own initiative You're missing the point. It is totally reasonable for them to rely on the users to supply the patches. "You have the source. Fix it. Send us the patches and if we like them (for whatever our reasons or lack thereof are) we might make them available for future use." Not even OpenBSD is that blunt, but that is the ultimate contract.
Always Blame Microsoft. Pretty effective actually as an initial response. At least it serves to make the scapegoat other than the user/victim.
Keep all your patches up-to-date? Seems like there is some very cruel mischief that can be done if you can get a bunch of lemmings to blindly install anything that looks like a patch. Not even Microsoft is actually dumb enough to religiously install all their patches.
In the end, if you're not happy with what you hear on the radio, change the channel! Or just turn it off! If they are not serving the public interest, you don't turn off the radio, you turn off the transmitter. They do not have any fundamental right to the radio frequencies they are transmitting on.
All bugs do the same damage. NOT. All fires do the same damage. NOT.
In both cases, the point is to minimize the extent of the damage. Actually OSS and full disclosure is as close to a solution as you will find.
If you know what's going on, you'd be amazed at what you can shut down without installing any bugfixes;-)
Look at Melissa and the Unix Honor Virus. The critical difference is that the recipient can see the Unix Honor Virus, while Melissa has lots of useful cloaking devices.
It was an "internet worm", originated by some haker in the internet for the internet. Bah! They fixed the "internet" in a few hours at most. It's the intranets and data centers that are still trying to put Humpty Dumpty back together again.
I can believe they were dancing around the news. Too much money from the software for the agile business. Well I guess we find out how agile those businesses really are;-)
Somehow I'm not all that concerned with Red Hat "dropping" support.
From current RedHat.com
All Users of Red Hat 4.0 Linux Please Read Important Red Hat, Inc. is not releasing security upgrades specific for this version of Red Hat Linux anymore. Users of Red Hat Linux 4.0 should upgrade to either Red Hat Linux 4.2 or Red Hat Linux 5.1, and apply the errata updates from ftp.redhat.com/pub/redhat/updates/.
Microsoft provide one of the easiest patching systems in the world No, he's right. It's the one degree of separation thing. Patching's easy. He didn't say anything about getting it to work afterwards.
So forget about finding a 'perfect' database with everything in it.. I'm not worried about a perfect database. I'm worried about an imperfect database that some idiot *will* insist on treating as if it were perfect. What happens with the cars with serial numbers which "don't exist"? I would assume that you handle it somehow, reasonably. The problem with automated systems is that while they can reason sylogistically, they are quite incapable of being reasonable. Correct logic from faulty premises leads to incorrect conclusions.
I couldn't find anything on SourceForge for tracking military pension payments, maintaining a 50-year history of agricultural surveys in the southeast region of the country, etc. These are apps that will be built in house, on top of whatever platform you're using Methinks you're missing something. Open Source doesn't mean you can find it on SourceForge. It doesn't even mean you can get the binary (at any price). It means that them as gets the binary can easily get the source (if they don't already have it) The examples you gave are exactly the kind of things for which you want/need Open Source. Come hell or high water, no matter what any third parties do, you want to ensure the survivability of these programs so you can keep on running them. I do not want to fork Linux. I will go through a lot of extra trouble so as to not fork Linux. But ultimately, if that is the only way I can do what I need to do, I can and I will.
If in the process they end up throwing away software that works that must be rewritten, that's a disservice to the taxpayers. I agree completely. But look at the reasons why the software would need to be rewritten. The most likely cause is gratuitous and incompatible changes in the platforms they are coded for. With Closed Source, you're very easily stuck between a rock and a hard place, with *no* options for a way out.
0 is true, everything else is false. It makes sense to do it that way; if a function completes successfully it doesn't really matter how it does it, but it can fail in any number of ways and it's important to know which one.
If a function completes successfully, it generally returns a VALUE. If a subprogram completes successfully it uses the unique value 0 to indicate success and some non-zero value to indicate WHICH FAILURE. If a variable is FALSE it is convenient to use 0 as FALSE and any non-zero to indicate TRUE.
0 AND anything is zero (with maybe some quibbles about anything actually having a value) true AND anything is not necessarily true.
Three licenses per machine? No wonder Microsoft sales are up.
Re:FUD and MISINFORMATION ABOUND!! see link
on
Maine School & Linux
·
· Score: 1
when this guy who "knew nothing about linux" was able to do some pretty advanced things. Like using "dd" to duplicate disks. Hmmmm, maybe he can read, as in Read The Fine Manual. Once you have the idea that it ought to be possible, what's so hard about dd if=/dev/hda of=/dev/hdd
Maybe distro docs are more advanced than I give them credit for Advanced??? Dunno about Minix, but it's very well documented in something as old and obsolete as SCO XENIX.
To make sense of it, you have to realize than most Germans understand German rather than Latin and most Brits understand English rather than Latin. Of course the true literati would read it in Greek and Hebrew.
I think you can get 5 9's scheduled out of a Yugo (or a Chevette). Think of a little boy in his pedal car dreaming of Le Mans. The execs have been had and like any victim of a scam want to delay the realization as long as possible.
This guy is very smart at how to play the game the wrong way and come out a winner. The one thing this guy can do that matters is to crack through the idea that the only possible desktop is Microsoft Windows. Windows --> Lindows --> Linux --> BSD --> Solaris --> MVS seems like a natural progression. (I'm sure I stepped on a few toes with that one) I don't even care if Lindows works, or how well. Once the breach is made, you're gonna see a lot of Linux desktops, several of which will actually be pretty good. Once it's known just what it takes to make a good one, you'll be hard pressed to find anything that isn't good. (Anybody else remember the first electronic calculators that came out. Some early ones were pretty strange, including one that represented negative numbers by 10's-complement (-1 was all 9's).
Hehe. 5 9's reliability. Sounds good. I can claim 5 9's (particularly if I don't know what I'm talking about;) (and if there's no penalty for failure;) 5 9's means that you've got the problems of 4 9's covered and can handle without interruption. For the remaining freaks that *do* happen, you have contingency plans which will include rousting out a small gaggle of executives. The blunt reality is that even if you do have the necessary calibre of equipment, you will almost always not go the 5 9's route. The thing is, 5 9's means 5 9's at 5 9's standards, not 5 9's at 1 or 2 9's standards. Wishful thinking is not a substitute for competence. Won't keep 'em from trying though:-(
Most programers need to be retaught how to program anything reliable. From an old fart, I'd go further and say that programmers should be taught that at least some things maybe should be reliable. I'd lay the blame squarely on the operating systems, with Microsoft Windows taking the lions share of blame. In practice, an application system will almost never be more reliable or secure than the operating system on which it is based. Further, an operating system telegraphs in subtle (ok, sometimes not so subtle) ways an attitude toward reliability, etc. This is a persistent relentless pressure. If the operating system is a Mickey Mouse wannabe, the surprise is that things aren't even worse than they are. (so give it time;)
I think history has shown that certain individuals HAVE had that much power. I'm not so sure. The individual/leader/dictator/whatever *does* sit on top of a pyramid which does have that much power. How far he can deviate from normalcy is a bit dubious. His orders will be "interpreted" by underlings. His meetings and agendas will be scheduled by underlings. His actions will be analyzed by underlings. You and I can pretty much go where we like, when and as we like. The president's choices will be severly constrained by the secret service who must first "arrange" things.
Slashdot Mirror
If you can't be anything else, at least be consistent...
Why?
take MySQL for example
First, MySQL does NOT attempt to take over my machine. While they do have precompiled binaries, they have no knowledge of or control of exactly which binaries are running on my machine. Bluntly, in the general case, they *cannot* update my machines.
Second, MySQL is not the ONLY thing running on my machine. Securing MySQL only while leaving everything else open seems like a stupid waste of effort.
MySQL is responsible for what they do. I am responsible for what I do.
A compilation, such as RedHat does have some responsibility for keeping tabs on bug fixes, including security issues. If I install the RedHat RPM, I would expect RedHat not MySQL to be a timely source of information and fixes. If I install MySQL the old-fashioned way neither is in a position to automatically update it on my machines. In the latter case, I would expect to get information that there is a problem from RedHat and the actual source-level patches or diffs from MySQL. No silver bullets, but pretty straight-forward and effective.
They rely for 100% on the users coming to look for them on their own initiative
You're missing the point. It is totally reasonable for them to rely on the users to supply the patches. "You have the source. Fix it. Send us the patches and if we like them (for whatever our reasons or lack thereof are) we might make them available for future use." Not even OpenBSD is that blunt, but that is the ultimate contract.
Always Blame Microsoft. Pretty effective actually as an initial response. At least it serves to make the scapegoat other than the user/victim.
Keep all your patches up-to-date? Seems like there is some very cruel mischief that can be done if you can get a bunch of lemmings to blindly install anything that looks like a patch. Not even Microsoft is actually dumb enough to religiously install all their patches.
misguided attempts at fixing corruption by introducing even more _government_ regulations which are the very reason corruption even exists.
Sorry, without government regulations, corruption is *ALL* that exists.
In the end, if you're not happy with what you hear on the radio, change the channel! Or just turn it off!
If they are not serving the public interest, you don't turn off the radio, you turn off the transmitter. They do not have any fundamental right to the radio frequencies they are transmitting on.
Bugs happen.
;-)
Fires happen.
Bugs do damage.
Fires do damage.
All bugs do the same damage. NOT.
All fires do the same damage. NOT.
In both cases, the point is to minimize the extent of the damage.
Actually OSS and full disclosure is as close to a solution as you will find.
If you know what's going on, you'd be amazed at what you can shut down without installing any bugfixes
Look at Melissa and the Unix Honor Virus. The critical difference is that the recipient can see the Unix Honor Virus, while Melissa has lots of useful cloaking devices.
It was an "internet worm", originated by some haker in the internet for the internet.
;-)
Bah! They fixed the "internet" in a few hours at most. It's the intranets and data centers that are still trying to put Humpty Dumpty back together again.
I can believe they were dancing around the news. Too much money from the software for the agile business. Well I guess we find out how agile those businesses really are
Maybe, but consider what you would do if you had a 'sploit of Microsoft Windows for Workgroups. I know I'd never admit it.
Somehow I'm not all that concerned with Red Hat "dropping" support.
From current RedHat.com
All Users of Red Hat 4.0 Linux Please Read
Important Red Hat, Inc. is not releasing security upgrades specific for this version of Red Hat Linux anymore. Users of Red Hat Linux 4.0 should upgrade to either Red Hat Linux 4.2 or Red Hat Linux 5.1, and apply the errata updates from ftp.redhat.com/pub/redhat/updates/.
Older errata items for 4.0 are listed below.
MS had had this bug patched since June last year.
Yeah, they did a real good job didn't they?
Don't get so upset that MSFT continues to destroy *nix, mac, and linux.
So that's what this MSSQL worm is all about.
Microsoft provide one of the easiest patching systems in the world
No, he's right. It's the one degree of separation thing. Patching's easy. He didn't say anything about getting it to work afterwards.
So forget about finding a 'perfect' database with everything in it..
I'm not worried about a perfect database. I'm worried about an imperfect database that some idiot *will* insist on treating as if it were perfect.
What happens with the cars with serial numbers which "don't exist"? I would assume that you handle it somehow, reasonably. The problem with automated systems is that while they can reason sylogistically, they are quite incapable of being reasonable. Correct logic from faulty premises leads to incorrect conclusions.
I couldn't find anything on SourceForge for tracking military pension payments, maintaining a 50-year history of agricultural surveys in the southeast region of the country, etc. These are apps that will be built in house, on top of whatever platform you're using
Methinks you're missing something. Open Source doesn't mean you can find it on SourceForge. It doesn't even mean you can get the binary (at any price). It means that them as gets the binary can easily get the source (if they don't already have it)
The examples you gave are exactly the kind of things for which you want/need Open Source. Come hell or high water, no matter what any third parties do, you want to ensure the survivability of these programs so you can keep on running them. I do not want to fork Linux. I will go through a lot of extra trouble so as to not fork Linux. But ultimately, if that is the only way I can do what I need to do, I can and I will.
If in the process they end up throwing away software that works that must be rewritten, that's a disservice to the taxpayers.
I agree completely. But look at the reasons why the software would need to be rewritten. The most likely cause is gratuitous and incompatible changes in the platforms they are coded for. With Closed Source, you're very easily stuck between a rock and a hard place, with *no* options for a way out.
0 is true, everything else is false. It makes sense to do it that way; if a function completes successfully it doesn't really matter how it does it, but it can fail in any number of ways and it's important to know which one.
If a function completes successfully, it generally returns a VALUE.
If a subprogram completes successfully it uses the unique value 0 to indicate success and some non-zero value to indicate WHICH FAILURE.
If a variable is FALSE it is convenient to use 0 as FALSE and any non-zero to indicate TRUE.
0 AND anything is zero (with maybe some quibbles about anything actually having a value)
true AND anything is not necessarily true.
Techtronix 4014 had a display resolution of 4096 x 3072. Was generally attached to a 16-bit minicomputer (back when 64k *was* enough).
Windows gives each user their own directory in which to save their files called "My Documents".
And keeps on giving.
Thanks to Microsoft Windows XP, the president of our company wound up with about 6 different directories for his email and documents.
Three licenses per machine?
No wonder Microsoft sales are up.
when this guy who "knew nothing about linux" was able to do some pretty advanced things. Like using "dd" to duplicate disks.
Hmmmm, maybe he can read, as in Read The Fine Manual.
Once you have the idea that it ought to be possible, what's so hard about
dd if=/dev/hda of=/dev/hdd
Maybe distro docs are more advanced than I give them credit for
Advanced??? Dunno about Minix, but it's very well documented in something as old and obsolete as SCO XENIX.
To make sense of it, you have to realize than most Germans understand German rather than Latin and most Brits understand English rather than Latin. Of course the true literati would read it in Greek and Hebrew.
I think you can get 5 9's scheduled out of a Yugo (or a Chevette).
Think of a little boy in his pedal car dreaming of Le Mans. The execs have been had and like any victim of a scam want to delay the realization as long as possible.
This guy is very smart at how to play the game the wrong way and come out a winner.
The one thing this guy can do that matters is to crack through the idea that the only possible desktop is Microsoft Windows. Windows --> Lindows --> Linux --> BSD --> Solaris --> MVS seems like a natural progression. (I'm sure I stepped on a few toes with that one)
I don't even care if Lindows works, or how well. Once the breach is made, you're gonna see a lot of Linux desktops, several of which will actually be pretty good. Once it's known just what it takes to make a good one, you'll be hard pressed to find anything that isn't good. (Anybody else remember the first electronic calculators that came out. Some early ones were pretty strange, including one that represented negative numbers by 10's-complement (-1 was all 9's).
Hehe. 5 9's reliability. :-(
Sounds good. I can claim 5 9's (particularly if I don't know what I'm talking about;) (and if there's no penalty for failure;)
5 9's means that you've got the problems of 4 9's covered and can handle without interruption. For the remaining freaks that *do* happen, you have contingency plans which will include rousting out a small gaggle of executives. The blunt reality is that even if you do have the necessary calibre of equipment, you will almost always not go the 5 9's route.
The thing is, 5 9's means 5 9's at 5 9's standards, not 5 9's at 1 or 2 9's standards. Wishful thinking is not a substitute for competence. Won't keep 'em from trying though
Always Blame Microsoft. ;)
( Almost as good as actually knowing what you're doing
Most programers need to be retaught how to program anything reliable.
From an old fart, I'd go further and say that programmers should be taught that at least some things maybe should be reliable. I'd lay the blame squarely on the operating systems, with Microsoft Windows taking the lions share of blame.
In practice, an application system will almost never be more reliable or secure than the operating system on which it is based. Further, an operating system telegraphs in subtle (ok, sometimes not so subtle) ways an attitude toward reliability, etc. This is a persistent relentless pressure. If the operating system is a Mickey Mouse wannabe, the surprise is that things aren't even worse than they are. (so give it time;)
I think history has shown that certain individuals HAVE had that much power.
I'm not so sure.
The individual/leader/dictator/whatever *does* sit on top of a pyramid which does have that much power. How far he can deviate from normalcy is a bit dubious. His orders will be "interpreted" by underlings. His meetings and agendas will be scheduled by underlings. His actions will be analyzed by underlings.
You and I can pretty much go where we like, when and as we like. The president's choices will be severly constrained by the secret service who must first "arrange" things.