So I need some advice on what goes into a network. It won't be that large right now, but it has to be scalable for up to a couple of hundred machines.
1) You had better find some damn fine PCs to replace those Suns, because a couple hundred PCs can make your life miserable due to lots of random breakage.
...you can get a cheaper Linux x86 machine that is four times faster.
2) This is not true (unless you found Pentiums with SPECfp of over 3000!). If you buy the right-sized computers for your task, the hardware costs won't be a dominating part of your budget. Human costs and non-OS commercial licensing will be, regardless of your platform choice.
Whenever people say that Linux is absolutely outright cheaper then commercial UNIX, then I'm pretty convinced they haven't figured out all the costs involved. Also, I'm not convinced they understand just how simple maintaining a Solaris box can be, for example, due to sunsolve.sun.com, ample documentation, optional support out the wazoo, etc.
Before you go blazing these new trails, just stop and think for a minute. Put aside the zealotry and really think hard about what is and is not cost effective. Regardless of your choice, you really need to be convinced it is the right one.
I know this might sound dumb, but I'm always amazed that car engines after 100,000 miles or so have gone through over a quarter of a billion revolutions. That's about a billion controlled explosions and about 4000 gallons of gasoline (give or take a few).
As near as I can make out from their site, SCO is claiming that Linux is derived from Minix (which is sort of true) which was derived from BSD (again, a stretch) which came from AT&T Unix (true for once) which SCO now owns.
Yeah, and like, me and Kevin Bacon are like brothers, man!
I really like Netscape 7 for Solaris, then I tried it on Windows.
Netscape 7 on Windows is a blatant spy-ware rigged ad-infested Real-hindered pile of poo. I spent a good half-hour going through all of the options for Netscape 7 + the RealPlayer plugin + the WinAmp plugin to turn off all of their "value added" spyware garbage.
That said, Netscape 7 for Solaris has been quite nice. It's amazing how the packaging changes the product.
10K SCSI drives tend to be pretty loud and run quite hot.
Not any more. The latest 10K SCSI drives purr like a sleeping kitten (aww...). I just bought a Ultra160 Seagate Cheetah, and I can only hear it by putting my ear up to the computer's case. The fans are way louder.
Your example of software to control/program computer-controlled milling machines. If someone abstracted the market enough so that it made sense to write an interface to the abstract API, instead of worrying about interacting with CADCAM producers to ensure that their/your drivers were compatible with their software/your hardware...
I really wasn't talking about low-level stuff like the good old DOS printer driver days. I was talking about domain expertise--the stuff that lies between the programmers' ears. Manufacturing (from paper/CAD to finished part) is no trivial task, and finding a person with both solid engineering/manufacturing expertise and programming expertise would be quite an accomplishment.
This kind of software isn't likely to become part of the next Red Hat distribution nor a download from SourceForge. The market is specialized, the learning curve is high, and the financial barriers to entry are very high.
When you compare all enterprise commercial apps against the most mature and most turnkey opensource ones, you'll find a lot of projects with good intentions but little functionality compared to commercial offerings.
Open Source software is the final destination of products that are destined to become commodities. Operating systems, word processing, personal finance, and some games, for example.
The software products that will likey never become open source serves domains so specialized, complex, or competitive that only businesses can drive them. In other words, no one would want to put up with such software in their spare time. A good example of this would be high-end computer-aided manufacturing and process planning. The problem domain and the hardware, such as multi-axis milling machines, are so expensive and complex that the cost and risk associated with proprietary software isn't that big of a deal. Also, there are so few people who can write such software well, that they deserve to recieve a salary for their work.
It's all these other "me too" products, such as Microsoft Office, Microsoft Windows, etc., that belong in the public domain:).
Recently, I walked by the arcade at the mall, and it was filled with rather insecure-looking teenagers trying to socialize. It certainly wasn't like the stereotypes we see in Tron, but more like something I remember seeing in Mad Magazine.
This sad bunch would probably cling onto anything that looks well presented and potentially trendy, so Microsoft's "get-em while they're young" approach just might enslave a few more into the MS Horde. Unfortunately, Microsoft is very successful at capturing those first seeking direction in their technological lives...but more like a cult leader than a genuine visionary.
They are not the cheap alternative to others, but rather they are the expensive option that people are now looking for alternatives to.
How things do change in ten years! I believe the cycle goes: Mainframes -> UNIX -> Windows -> Linux -> Linux on Mainframes -> resurgence of UNIX (maybe Plan 9's appearance?) -> who knows (Windows is out of the question this time around)
I'm glad that I don't use a browser that permits (by it's default settings) malicious html code to wipe my entire hard disk drive, implant viruses and or trojans and permits the hijacking of my data.
If the trojans are applied first, then virus implantation would be prevented...
My question with all of this is how is it going to make my life better by having Direct(whatever) running my automobile, or having a browser in my freezer at home?
1) It will make your life worse. For example, I would never go to the lowest bidder for something as important as an oil change! Who knows what kind of oil filter they use, or even if they put in the right amount of oil. Then the sleazebags try to sell new winshield wipers, breather elements, transmission fluid, etc. and leave fingerprints on the hood and fenders. Good old-fashioned word of mouth is still the only reliable way to find mechanics, contractors, etc.
2) Microsoft's habit of producing products of unbounded complexity has no business being in any of my household appliances, or my car (or my PC, for that matter).
I think the recent drive to automate and integrate everything will end when the public realizes that simplicity in most things is a virtue not a limitation, and computers make useful tools and that's about it.
No, not really. The only real difference (ignorning frilly graphics) is that Internet Exploder and Netscrape crash a whole lot more often than a 3270-type mainframe terminal:-)
It is orders of magnitude harder to guarantee the operation of a 50-million-line program than it is the operation of a tens-of-kilobyte terminal app.
Many people have yet to learn that complexity is the enemy.
I'm not sure why these people reach architect level positions.
Sympathy. They weren't good programmers, probably didn't get along in teams...but otherwise were nice people and we'd hate to see them go...okay, promotion!
Suddenly, the application is slower than molasses, going up hill on a cold day. No one knows how to navigate the new interface and productivity takes a major dive.
I think every big company has done this somewhere. I saw a beautiful one-page web form for entering labor hours and billing codes get transformed into a behemoth web app that literally takes ten clicks of the mouse where there was one. On top of that, it all runs on one Win NT server that gets trounced when thousands of employees "attempt" to enter their time at the end of the week. I've wasted hours just trying to enter the hours I've worked! Ugh.
It is so obvious that free code with it's transparency and freedom of modification solves all of the problems the author can dream up and that others suffered. Free software is modular, replacable and never dies.
I agree very much with what you said, but I have to troll about this one thing. This weekend I had a losing fight with GNOME and compiling a GNOME application on an non-GNU/Linux platform. The cob-web of dependencies is sickening, and even when I got the app to compile without errors, it crashed without a core dump (wonderful, time for gdb-land).
I love how far Free & OS software has come, but there are still interoperability issues that even autoconf falls far short on. Actually, I am wondering if autoconf is a liability (false sense of security, perhaps). Regardless, there are instances of GNU/Linux lock-in and even distribution lock-in, which aren't nearly as painful as Windows lock-in, but still puts a poker in the ass occassionally.
I assume this is some sort of visual studio bastardry you're refering to??
Yes. One of my co-workers told me a story about rebuilding a whole project from scratch when one of Visual Studio's pseudo-makefiles became corrupted. Without ample backups, Visual Studio is not a reliable development tool.
Another really bad thing about binary makefiles is that they are very inefficient to maintain in a version control system. Additionally, I consider them doubly bad, because I view makefiles as part of the source code of the system (configuration management and automation), which must be maintained in version control.
Sorting through gigabytes of Microsoft legacy code that was written under the pretense that nobody would ever see it. Now there's a scary thought. I'd hate to be the guy with THAT job...
The only reason China wants the source code is to torture people captured for interrogation. The only thing worse than the Windows source code would be bamboo under the nails (ouch!).
If the genuine Lexmark ink cartridges are that good, then they shouldn't have a problem convincing people to buy genuine ones.
What might scare Lexmark is that it does become like replacement auto parts. There are many cases where third-party auto parts are both superior in quality and cheaper. If they are trumped by third-party ink refills, they might have to step on their own feet and increase the prices of printers. I think this whole fiasco is the printer companies covering their asses after driving down the prices of the printers way too far.
Slashdot Mirror
...the one thing that I find downright irritating.
All these "kirritating" responses are making me kry!
So I need some advice on what goes into a network. It won't be that large right now, but it has to be scalable for up to a couple of hundred machines.
...you can get a cheaper Linux x86 machine that is four times faster.
1) You had better find some damn fine PCs to replace those Suns, because a couple hundred PCs can make your life miserable due to lots of random breakage.
2) This is not true (unless you found Pentiums with SPECfp of over 3000!). If you buy the right-sized computers for your task, the hardware costs won't be a dominating part of your budget. Human costs and non-OS commercial licensing will be, regardless of your platform choice.
Whenever people say that Linux is absolutely outright cheaper then commercial UNIX, then I'm pretty convinced they haven't figured out all the costs involved. Also, I'm not convinced they understand just how simple maintaining a Solaris box can be, for example, due to sunsolve.sun.com, ample documentation, optional support out the wazoo, etc.
Before you go blazing these new trails, just stop and think for a minute. Put aside the zealotry and really think hard about what is and is not cost effective. Regardless of your choice, you really need to be convinced it is the right one.
I know this might sound dumb, but I'm always amazed that car engines after 100,000 miles or so have gone through over a quarter of a billion revolutions. That's about a billion controlled explosions and about 4000 gallons of gasoline (give or take a few).
As near as I can make out from their site, SCO is claiming that Linux is derived from Minix (which is sort of true) which was derived from BSD (again, a stretch) which came from AT&T Unix (true for once) which SCO now owns.
Yeah, and like, me and Kevin Bacon are like brothers, man!
netscape 7
I really like Netscape 7 for Solaris, then I tried it on Windows.
Netscape 7 on Windows is a blatant spy-ware rigged ad-infested Real-hindered pile of poo. I spent a good half-hour going through all of the options for Netscape 7 + the RealPlayer plugin + the WinAmp plugin to turn off all of their "value added" spyware garbage.
That said, Netscape 7 for Solaris has been quite nice. It's amazing how the packaging changes the product.
10K SCSI drives tend to be pretty loud and run quite hot.
Not any more. The latest 10K SCSI drives purr like a sleeping kitten (aww...). I just bought a Ultra160 Seagate Cheetah, and I can only hear it by putting my ear up to the computer's case. The fans are way louder.
Your example of software to control/program computer-controlled milling machines. If someone abstracted the market enough so that it made sense to write an interface to the abstract API, instead of worrying about interacting with CADCAM producers to ensure that their/your drivers were compatible with their software/your hardware...
I really wasn't talking about low-level stuff like the good old DOS printer driver days. I was talking about domain expertise--the stuff that lies between the programmers' ears. Manufacturing (from paper/CAD to finished part) is no trivial task, and finding a person with both solid engineering/manufacturing expertise and programming expertise would be quite an accomplishment.
This kind of software isn't likely to become part of the next Red Hat distribution nor a download from SourceForge. The market is specialized, the learning curve is high, and the financial barriers to entry are very high.
When you compare all enterprise commercial apps against the most mature and most turnkey opensource ones, you'll find a lot of projects with good intentions but little functionality compared to commercial offerings.
:).
Open Source software is the final destination of products that are destined to become commodities. Operating systems, word processing, personal finance, and some games, for example.
The software products that will likey never become open source serves domains so specialized, complex, or competitive that only businesses can drive them. In other words, no one would want to put up with such software in their spare time. A good example of this would be high-end computer-aided manufacturing and process planning. The problem domain and the hardware, such as multi-axis milling machines, are so expensive and complex that the cost and risk associated with proprietary software isn't that big of a deal. Also, there are so few people who can write such software well, that they deserve to recieve a salary for their work.
It's all these other "me too" products, such as Microsoft Office, Microsoft Windows, etc., that belong in the public domain
"Besides, everyone knows that MS doesn't get it right until the 3rd or 4th try."
n
-
\ MS_try = still ain't got it right
/ i
-
i=1
Recently, I walked by the arcade at the mall, and it was filled with rather insecure-looking teenagers trying to socialize. It certainly wasn't like the stereotypes we see in Tron, but more like something I remember seeing in Mad Magazine.
This sad bunch would probably cling onto anything that looks well presented and potentially trendy, so Microsoft's "get-em while they're young" approach just might enslave a few more into the MS Horde. Unfortunately, Microsoft is very successful at capturing those first seeking direction in their technological lives...but more like a cult leader than a genuine visionary.
23GB should be enough for anybody!
Yeah. Now it'll take only 79 of these to back up that 1.8TB array we're all setting up this weekend.
They are not the cheap alternative to others, but rather they are the expensive option that people are now looking for alternatives to.
How things do change in ten years! I believe the cycle goes: Mainframes -> UNIX -> Windows -> Linux -> Linux on Mainframes -> resurgence of UNIX (maybe Plan 9's appearance?) -> who knows (Windows is out of the question this time around)
I'm glad that I don't use a browser that permits (by it's default settings) malicious html code to wipe my entire hard disk drive, implant viruses and or trojans and permits the hijacking of my data.
If the trojans are applied first, then virus implantation would be prevented...
My question with all of this is how is it going to make my life better by having Direct(whatever) running my automobile, or having a browser in my freezer at home?
1) It will make your life worse. For example, I would never go to the lowest bidder for something as important as an oil change! Who knows what kind of oil filter they use, or even if they put in the right amount of oil. Then the sleazebags try to sell new winshield wipers, breather elements, transmission fluid, etc. and leave fingerprints on the hood and fenders. Good old-fashioned word of mouth is still the only reliable way to find mechanics, contractors, etc.
2) Microsoft's habit of producing products of unbounded complexity has no business being in any of my household appliances, or my car (or my PC, for that matter).
I think the recent drive to automate and integrate everything will end when the public realizes that simplicity in most things is a virtue not a limitation, and computers make useful tools and that's about it.
The winning approach is clone PLUS provide added value.
Added value == lock-in.
For example, gmake vs. make, bash vs. sh, RPM vs. pkgadd or tar, GNOME vs. X-Windows+fvwm, etc.
GNOME, for example, can be argued as a step forward, but the others are usually just steps sideways (or backwards, yes RPM, I mean you).
No, not really. The only real difference (ignorning frilly graphics) is that Internet Exploder and Netscrape crash a whole lot more often than a 3270-type mainframe terminal :-)
It is orders of magnitude harder to guarantee the operation of a 50-million-line program than it is the operation of a tens-of-kilobyte terminal app.
Many people have yet to learn that complexity is the enemy.
I'm not sure why these people reach architect level positions.
Sympathy. They weren't good programmers, probably didn't get along in teams...but otherwise were nice people and we'd hate to see them go...okay, promotion!
Suddenly, the application is slower than molasses, going up hill on a cold day. No one knows how to navigate the new interface and productivity takes a major dive.
I think every big company has done this somewhere. I saw a beautiful one-page web form for entering labor hours and billing codes get transformed into a behemoth web app that literally takes ten clicks of the mouse where there was one. On top of that, it all runs on one Win NT server that gets trounced when thousands of employees "attempt" to enter their time at the end of the week. I've wasted hours just trying to enter the hours I've worked! Ugh.
It is so obvious that free code with it's transparency and freedom of modification solves all of the problems the author can dream up and that others suffered. Free software is modular, replacable and never dies.
I agree very much with what you said, but I have to troll about this one thing. This weekend I had a losing fight with GNOME and compiling a GNOME application on an non-GNU/Linux platform. The cob-web of dependencies is sickening, and even when I got the app to compile without errors, it crashed without a core dump (wonderful, time for gdb-land).
I love how far Free & OS software has come, but there are still interoperability issues that even autoconf falls far short on. Actually, I am wondering if autoconf is a liability (false sense of security, perhaps). Regardless, there are instances of GNU/Linux lock-in and even distribution lock-in, which aren't nearly as painful as Windows lock-in, but still puts a poker in the ass occassionally.
I assume this is some sort of visual studio bastardry you're refering to??
Yes. One of my co-workers told me a story about rebuilding a whole project from scratch when one of Visual Studio's pseudo-makefiles became corrupted. Without ample backups, Visual Studio is not a reliable development tool.
Another really bad thing about binary makefiles is that they are very inefficient to maintain in a version control system. Additionally, I consider them doubly bad, because I view makefiles as part of the source code of the system (configuration management and automation), which must be maintained in version control.
Sorting through gigabytes of Microsoft legacy code that was written under the pretense that nobody would ever see it. Now there's a scary thought. I'd hate to be the guy with THAT job...
The only reason China wants the source code is to torture people captured for interrogation. The only thing worse than the Windows source code would be bamboo under the nails (ouch!).
Typing is archaic.
So are proprietary binary "makefiles" that can ruin a whole project when they become corrupted.
If the genuine Lexmark ink cartridges are that good, then they shouldn't have a problem convincing people to buy genuine ones.
What might scare Lexmark is that it does become like replacement auto parts. There are many cases where third-party auto parts are both superior in quality and cheaper. If they are trumped by third-party ink refills, they might have to step on their own feet and increase the prices of printers. I think this whole fiasco is the printer companies covering their asses after driving down the prices of the printers way too far.
I collected a whole bunch of it, but then all I got was this armor...
What would you put to him, in that position?
Why are they so naive and lazy to think this sort of scheme would not land them in trouble?