You have brought up a bit of a quandry here. When all is said and done and SCO is officially dead, and the lawyers have to find other work, will this be a feather in their cap with SCO on their resume or will they be seen as the bottom feeding scum that they are. You may not like the job they are doing but they are certainly doing it rather well.
Look. If you were the one hiring the lawyer, who would you rather have? A nice guy who plays nice and fair all the time, or a lawyer who knows every little legal loophole and how to exploit it?
Or, to put it in terms that most Slashdotters can understand, who would you rather have on your software security project? A nice programmer who might know a thing or two about breaking security, but has never done it, or a smart hacker who knows his way around every local and remote root exploit known to man and has done so?
As the saying goes, when you have to dig through dirt, you get yourself a worm.
While that's fairly bad, particularly from the point of view of debugging problems, it occurs that I've never really dealt with modules on server systems, only on desktop. On the servers, it's handled by the Linux distro, and we then don't touch anything, whereas on my own desktop installing new hardware with freaky module requirements isn't at all odd.
Kernel modules don't necessarily deal with hardware. We have some that, for example, deal with security software. We have others installed on various speciality boxes like routers and latency injectors. The fact that we are a lab environment means that we frequently need to change things around, and loading and unloading modules is actually a fairly routine task, even on 'servers'.
> Most thought you could find out that information by reading/etc/modules.conf! (No, I'm not kidding.) I dunno, if I didn't know about lsmod,/etc/modules.conf would seem a good guess...
A good guess, but not at all accurate.
Purely out of curiousity, what other sort of questions are you asking?
Mostly laying out real scenarios that we actually encounter and seeing if the thought process is there. For instance, if you had a Solaris or Linux application that was spitting a 'Abort. Errno 2.' or some such and there was no indication on what was happening in the log files for application (or in/var/adm/messages), what would you do?
1) You should obviously know that errno 2 is a ENOENT (file not found) on UNIX. If you don't, you should at least know that you could look in/usr/include/sys/errno.h to find that out.
2) You should obviously know how that you could use truss/strace to find out what file or directory is "missing."
Since in our lab we debug and test and custom internal-built and customized COTS applications for performance and scalability, and sometimes don't even have convenient access to the developers, we need a hard core administrator. Especially since we don't have time to train him or her -- we often get the application a couple of months before release.
We can't find good workers. I've interviewed repeatedly and found the new talent is terrible -- it seems that has technology becomes more "known," the amount of GOOD talent is dropping. I've interviewed some people from top colleges that just don't know their way around a business at all, and I have no desire to train them in exchange for a high 5 figure salary.
Yep. I've been interviewing people for a Unix systems administrator position for our group where I work, and I can tell you that while the vast majority of candidates can probably poke their way around a Unix box from a users' standpoint, actually understanding the toolset from an administration and debugging standpoint seems to be a rare skillset. I can't tell you how man people I asked about various simple day-to-day Unix administration operations who couldn't tell you how they were done because they followed some "procedures" that someone else wrote up or "it's been too long", or "we used [buzzword product X, Y or Z]" or other such nonsense. For instance, many so-called "Linux experts" did not know that the command to list the kernel modules loaded into a running kernel is 'lsmod'. Most thought you could find out that information by reading/etc/modules.conf! (No, I'm not kidding.)
Do you run with archive logs turned on or off? I run with them turned on (not my idea), and it's when Oracle tries to create an archive log that it crashes. If I turn off archive logging, the problem doesn't exist, and, you're right, it won't crash -- it'll just sit there in a somewhat disabled state, not being able to create records, etc., until some space is added or freed up.
As if heuristic analysis were the key to stopping all the malware on the Internet. If it were, everyone would install AVG on their Windows PCs and all the malware would just go away.
However, reality bites us in the arse and then we realize that heuristic analysis only goes so far.
The key to having freedom from malware is to have operating systems which do not make it easy for malware to thrive. OSes should not default to having users logging in with administrative priveleges. Applications should not be able to be installed with a view lines of embedded scripting code in a Web page or an e-mail. E-mail software shouldn't allow attachments to be executed immediately upon clicking -- users should have to save the attachment and then launch it from the shell. This requires a little more thought process than "if I click here, I will see naked pictures of [insert celebrity]!"
Of course, the OS where all of these bad security models exist is Windows and the e-mail application in question is Outlook. We need to stop looking at the problem of malware as a user education problem and start seeing it for what it really is -- a broken OS issue caused by the greed and stupidity of the largest software company in the world.
All I can tell you is that my ora_smon, pmon, etc. processes die. The tables themselves are fine, but after I clear up the space, I have to reissue a 'startup' from sqlplus.
There's plenty of misinformation going around. For instance
PostgreSQL doesn't behave as nicely as Oracle when the system fills up, Goulet said. In those instances, the system tends to crash quickly.
I'm, among other things, an Oracle administrator. When the filesystem that holds the databse files fills up on Oracle 9i 9.2.0.4 on both Solaris and Linux, I can tell you for sure that the Oracle instance will crash suddenly, with nothing more than a notation in the log that the disk was full trying to write to file such-and-such.
That's not any different from what they describe with PostgreSQL.
VMware, bochs, Xen, and QEMU can all run Windows (and other OSes) that have their filesystems stored inside an image file on the host OS. Why would any other OS virtualizer be different?
Bah. Big government is the new libertarianism. Having a big, stupid government that can't do anything right is far superior to having a sleek, functioning government that can oppress its people at a moment's notice.
Except when the big, stupid government is a front for the sleek, functioning oligarchy really runs the show from behind the scenes.
This is actually not too far different from how human vision works -- we sort of guess about the landscape in front of us based on knowledge of other landscapes and the current landscape around us that we can see close-up. We just don't have lasers.;)
CAD/Drafting. Meaning *2D* CAD. This is possible. I was talking about 3D solids-based CAD, which is what I deal with on a day-to-day basis. Solid modeling requires a heck of a lot more power than 2D-CAD.
Plus you're running Win2K, which has far lower memory requirements than XP.
(BTW, isn't there a Donovan song about Pamela Jones?)
I think you're thinking of Pamela Jo, not that I'm a big Donovan fan or anything.
But Microsoft isn't joing the ODA, they're joining the ISO group that's responsible for ODF's ISO certification.
Interestingly enough, despite Microsoft's protests to the contrary, as the person who is quoted in the article (Groklaw's PJ) has stated on her site, the specific committee on which Microsoft sits just happens to be the one that's responsible for tallying up all the votes and giving the final yay or nay.
Sorry, but Microsoft's statements of 'we're not gonna do anything, we're just here to watch,' sound just a bit increduluous if you ask me.
PC with 300 megahertz or higher processor clock speed recommended; 233 MHz minimum required (single or dual processor system);* Intel Pentium/Celeron family, or AMD K6/Athlon/Duron family, or compatible processor recommended
128 megabytes (MB) of RAM or higher recommended (64 MB minimum supported; may limit performance and some features)
1.5 gigabytes (GB) of available hard disk space
Microsoft's minimum specifications are completely ludricous. I'm sure, you COULD run Windows XP on a machine of those specs, but forget about running anything with extensive memory requirements (Photoshop or CAD applications or some desktop publishing packages). And it takes up to 2 minutes to start a web browser (even IE).
Dream on. Windows XP sucks even a 600 or 800 MHz with 256 MB of RAM.
Antonio, you want a burning bush? Well here it is. I am God. Yes, the One! No, really. Listen. I'm tired of trying to reach you through indirect means. (Jeez, how many students do I have to be??) Creationism is true. I created the world in 6 days. How? How do you think, peabrain? I spoke the word and it was created. It's like magic, but when you're omnipotent it's not such a big deal. No what I mean? No, of course you don't... No one really does.
Huh. I'd think an omnipotent and omniscient being such as yourself would have a better grasp of the English language.
But what do I know? I'm just a mere mortal and you are obviously God.
Cray certainly has a lot of supercomputing experience, but relying on compiler improvements to make or break an architecture doesn't have a good track record.
Seriously. Just ask Transmeta. Or Intel (think Itanic).
Seriously, when I was jounger I always dreamed of being a game developer, I used to wander in gamedev.net and make cool opengl demos but seeing how underpriced are game programmers and the difficulty to get into the industry (requiring 10 published titles to get into a company?) I am opting to find a real work =o)
OpenGL? Heck, when I was younger, we aspiring game programmers didn't have any OpenGL. No siree! 3D what??? We had ASCII graphics! And, if you were REALLY lucky, you had one those new-fangled graphics cards that could do 320x200! 16 WHOLE COLORS! w00t!
OpenGL! Puh! I'm gonna wack you kids with my walker!
Why do you need to measure ALL possible endpoints? You just need to target the specific hosts/ports that you perceive a performance issue on and have a few control machines that you aren't finding a problem with. Even then, you what services a particular host is running, so you just need to do you timings based on valid requests for that protocol (i.e., http, ftp, bittorrent, etc.
For that matter, you could do all other sorts of network analysis using tools built into your OS. Of course, Linux has more of these tools already installed by most distributions than Windows, but there are Windows versions of many of these tools.
traceroute and ping by themselves can be pretty englightening. They don't give you the complete picture, and for that you need to measure the actual throughput of your transfers. wget can be useful for that. Furthermore, if the ISPs/telcos were to purposely choke certain sites, you would see it terms of either bandwidth throttling, QoS or perhaps even latency injection.
Experienced network admins will have no trouble in determining if these things are in place using available tools. For the average user, the tools you list can be useful, but interpreting their results is best left to the professionals, IMHO.
(Disclosure -- I'm a well-seasoned network and systems administrator)
Yeah. You'll need the dual core machine to do HiDef, though, and that's not $599, that's $799, plus all the of the "hidden" costs another poster pointed out, such as the need to buy Roxio Toast to do DVD burning.
You're easily into the $1000+ range, when you could have just thrown together a MiniITX Linux box with Knopmyth for half that price with no additional effort.
2. Plug a USB2 or Firewire tuner and the Keyspan USB remote sensor into it.
3. Install EyeTV software & Keyspan remote software (both included with the hardware.)
4. Set up your universal remote (your TV and/or receiver remote might be a programmable one. Otherwise there are plenty out there to choose from for about twenty bucks) to control both the TV tuner and all your Mac media apps.
5. Watch as your new HiDef movies run incredibly slow and jumpy, with dialogue out of sync with the picture, taking you back to the early days of video on desktop PCs, because of the Mac Mini's pokey 1.5Ghz processor and paltry 667 Mhz FSB. Oh, and the DVD drive is VERY loud.
Slashdot Mirror
While that's fairly bad, particularly from the point of view of debugging problems, it occurs that I've never really dealt with modules on server systems, only on desktop. On the servers, it's handled by the Linux distro, and we then don't touch anything, whereas on my own desktop installing new hardware with freaky module requirements isn't at all odd.
/etc/modules.conf! (No, I'm not kidding.) /etc/modules.conf would seem a good guess...
/var/adm/messages), what would you do?
/usr/include/sys/errno.h to find that out.
Kernel modules don't necessarily deal with hardware. We have some that, for example, deal with security software. We have others installed on various speciality boxes like routers and latency injectors. The fact that we are a lab environment means that we frequently need to change things around, and loading and unloading modules is actually a fairly routine task, even on 'servers'.
> Most thought you could find out that information by reading
I dunno, if I didn't know about lsmod,
A good guess, but not at all accurate.
Purely out of curiousity, what other sort of questions are you asking?
Mostly laying out real scenarios that we actually encounter and seeing if the thought process is there. For instance, if you had a Solaris or Linux application that was spitting a 'Abort. Errno 2.' or some such and there was no indication on what was happening in the log files for application (or in
1) You should obviously know that errno 2 is a ENOENT (file not found) on UNIX. If you don't, you should at least know that you could look in
2) You should obviously know how that you could use truss/strace to find out what file or directory is "missing."
Since in our lab we debug and test and custom internal-built and customized COTS applications for performance and scalability, and sometimes don't even have convenient access to the developers, we need a hard core administrator. Especially since we don't have time to train him or her -- we often get the application a couple of months before release.
We can't find good workers. I've interviewed repeatedly and found the new talent is terrible -- it seems that has technology becomes more "known," the amount of GOOD talent is dropping. I've interviewed some people from top colleges that just don't know their way around a business at all, and I have no desire to train them in exchange for a high 5 figure salary.
/etc/modules.conf! (No, I'm not kidding.)
Yep. I've been interviewing people for a Unix systems administrator position for our group where I work, and I can tell you that while the vast majority of candidates can probably poke their way around a Unix box from a users' standpoint, actually understanding the toolset from an administration and debugging standpoint seems to be a rare skillset. I can't tell you how man people I asked about various simple day-to-day Unix administration operations who couldn't tell you how they were done because they followed some "procedures" that someone else wrote up or "it's been too long", or "we used [buzzword product X, Y or Z]" or other such nonsense. For instance, many so-called "Linux experts" did not know that the command to list the kernel modules loaded into a running kernel is 'lsmod'. Most thought you could find out that information by reading
Do you run with archive logs turned on or off? I run with them turned on (not my idea), and it's when Oracle tries to create an archive log that it crashes. If I turn off archive logging, the problem doesn't exist, and, you're right, it won't crash -- it'll just sit there in a somewhat disabled state, not being able to create records, etc., until some space is added or freed up.
As if heuristic analysis were the key to stopping all the malware on the Internet. If it were, everyone would install AVG on their Windows PCs and all the malware would just go away.
However, reality bites us in the arse and then we realize that heuristic analysis only goes so far.
The key to having freedom from malware is to have operating systems which do not make it easy for malware to thrive. OSes should not default to having users logging in with administrative priveleges. Applications should not be able to be installed with a view lines of embedded scripting code in a Web page or an e-mail. E-mail software shouldn't allow attachments to be executed immediately upon clicking -- users should have to save the attachment and then launch it from the shell. This requires a little more thought process than "if I click here, I will see naked pictures of [insert celebrity]!"
Of course, the OS where all of these bad security models exist is Windows and the e-mail application in question is Outlook. We need to stop looking at the problem of malware as a user education problem and start seeing it for what it really is -- a broken OS issue caused by the greed and stupidity of the largest software company in the world.
All I can tell you is that my ora_smon, pmon, etc. processes die. The tables themselves are fine, but after I clear up the space, I have to reissue a 'startup' from sqlplus.
There's plenty of misinformation going around. For instance
PostgreSQL doesn't behave as nicely as Oracle when the system fills up, Goulet said. In those instances, the system tends to crash quickly.
I'm, among other things, an Oracle administrator. When the filesystem that holds the databse files fills up on Oracle 9i 9.2.0.4 on both Solaris and Linux, I can tell you for sure that the Oracle instance will crash suddenly, with nothing more than a notation in the log that the disk was full trying to write to file such-and-such.
That's not any different from what they describe with PostgreSQL.
VMware, bochs, Xen, and QEMU can all run Windows (and other OSes) that have their filesystems stored inside an image file on the host OS. Why would any other OS virtualizer be different?
Bah. Big government is the new libertarianism. Having a big, stupid government that can't do anything right is far superior to having a sleek, functioning government that can oppress its people at a moment's notice.
Except when the big, stupid government is a front for the sleek, functioning oligarchy really runs the show from behind the scenes.
You know, kinda like here in the U.S.
The human eye and visual cortex are an amazingly complex and complicated system. Interestingly enough, I remember about reading about the a while back in December. According to the wired article, IIRC, Stanley used a combination of lasers (short range) and cameras (long range). It then took what it learned about the short range view of the landscape and it applied that knowledge against what it was seeing on the long range from the cameras.
;)
This is actually not too far different from how human vision works -- we sort of guess about the landscape in front of us based on knowledge of other landscapes and the current landscape around us that we can see close-up. We just don't have lasers.
CAD/Drafting. Meaning *2D* CAD. This is possible. I was talking about 3D solids-based CAD, which is what I deal with on a day-to-day basis. Solid modeling requires a heck of a lot more power than 2D-CAD.
Plus you're running Win2K, which has far lower memory requirements than XP.
(BTW, isn't there a Donovan song about Pamela Jones?)
I think you're thinking of Pamela Jo, not that I'm a big Donovan fan or anything.
But Microsoft isn't joing the ODA, they're joining the ISO group that's responsible for ODF's ISO certification.
Interestingly enough, despite Microsoft's protests to the contrary, as the person who is quoted in the article (Groklaw's PJ) has stated on her site, the specific committee on which Microsoft sits just happens to be the one that's responsible for tallying up all the votes and giving the final yay or nay.
Sorry, but Microsoft's statements of 'we're not gonna do anything, we're just here to watch,' sound just a bit increduluous if you ask me.
PC with 300 megahertz or higher processor clock speed recommended; 233 MHz minimum required (single or dual processor system);* Intel Pentium/Celeron family, or AMD K6/Athlon/Duron family, or compatible processor recommended 128 megabytes (MB) of RAM or higher recommended (64 MB minimum supported; may limit performance and some features) 1.5 gigabytes (GB) of available hard disk space Microsoft's minimum specifications are completely ludricous. I'm sure, you COULD run Windows XP on a machine of those specs, but forget about running anything with extensive memory requirements (Photoshop or CAD applications or some desktop publishing packages). And it takes up to 2 minutes to start a web browser (even IE). Dream on. Windows XP sucks even a 600 or 800 MHz with 256 MB of RAM.
The entire computer industry has been stifled for years. We need competition, and we need it badly.
Yeah. Because Microsoft has no real competition at all in desktop operating systems.
Antonio, you want a burning bush? Well here it is. I am God. Yes, the One! No, really. Listen. I'm tired of trying to reach you through indirect means. (Jeez, how many students do I have to be??) Creationism is true. I created the world in 6 days. How? How do you think, peabrain? I spoke the word and it was created. It's like magic, but when you're omnipotent it's not such a big deal. No what I mean? No, of course you don't... No one really does.
Huh. I'd think an omnipotent and omniscient being such as yourself would have a better grasp of the English language.
But what do I know? I'm just a mere mortal and you are obviously God.
Cray certainly has a lot of supercomputing experience, but relying on compiler improvements to make or break an architecture doesn't have a good track record.
Seriously. Just ask Transmeta. Or Intel (think Itanic).
Another interesting bit of trivia. Apple Macintoshes have been designed using a Cray. What's even more ironic is that according to that same link, Seymour Cray used a Mac to design the next Cray.
Oh, my! Snort is open source? But Bill Gates and Steve Ballmer both said that open source software is insecure!
I had no idea! We're going to have stop using Snort immediately!
-- The FBI
Seriously, when I was jounger I always dreamed of being a game developer, I used to wander in gamedev.net and make cool opengl demos but seeing how underpriced are game programmers and the difficulty to get into the industry (requiring 10 published titles to get into a company?) I am opting to find a real work =o)
OpenGL? Heck, when I was younger, we aspiring game programmers didn't have any OpenGL. No siree! 3D what??? We had ASCII graphics! And, if you were REALLY lucky, you had one those new-fangled graphics cards that could do 320x200! 16 WHOLE COLORS! w00t!
OpenGL! Puh! I'm gonna wack you kids with my walker!
Why do you need to measure ALL possible endpoints? You just need to target the specific hosts/ports that you perceive a performance issue on and have a few control machines that you aren't finding a problem with. Even then, you what services a particular host is running, so you just need to do you timings based on valid requests for that protocol (i.e., http, ftp, bittorrent, etc.
And the name ain't Cumin, pal.
It also excels in one hand operation because of the clickable scroll wheel.
;-)
For most Slashdotters, this is, of course, the most important feature any Web-enabled mobile device.
For that matter, you could do all other sorts of network analysis using tools built into your OS. Of course, Linux has more of these tools already installed by most distributions than Windows, but there are Windows versions of many of these tools.
traceroute and ping by themselves can be pretty englightening. They don't give you the complete picture, and for that you need to measure the actual throughput of your transfers. wget can be useful for that. Furthermore, if the ISPs/telcos were to purposely choke certain sites, you would see it terms of either bandwidth throttling, QoS or perhaps even latency injection.
Experienced network admins will have no trouble in determining if these things are in place using available tools. For the average user, the tools you list can be useful, but interpreting their results is best left to the professionals, IMHO.
(Disclosure -- I'm a well-seasoned network and systems administrator)
Yeah. You'll need the dual core machine to do HiDef, though, and that's not $599, that's $799, plus all the of the "hidden" costs another poster pointed out, such as the need to buy Roxio Toast to do DVD burning.
You're easily into the $1000+ range, when you could have just thrown together a MiniITX Linux box with Knopmyth for half that price with no additional effort.
It's Web-native. Tellico is a KDE app, and Alexandria runs on Windows.
1. Buy a Mac Mini
2. Plug a USB2 or Firewire tuner and the Keyspan USB remote sensor into it.
3. Install EyeTV software & Keyspan remote software (both included with the hardware.)
4. Set up your universal remote (your TV and/or receiver remote might be a programmable one. Otherwise there are plenty out there to choose from for about twenty bucks) to control both the TV tuner and all your Mac media apps.
5. Watch as your new HiDef movies run incredibly slow and jumpy, with dialogue out of sync with the picture, taking you back to the early days of video on desktop PCs, because of the Mac Mini's pokey 1.5Ghz processor and paltry 667 Mhz FSB. Oh, and the DVD drive is VERY loud.
No, thanks.