I think you'd be surprised what you can do with a VIA C3. You could get even smaller/lower power than a mini. A G4 is easily the better processor in general, but a C3 may be good enough (I dunno about this guy's project's needs), and it wins easily in areas like encryption (as it has encryption acceleration). It's also way cheaper if you're doing a lot of customization, those boards can be had for a lot less than a mini.
"Fedora was (by design) too fast-changing to support. I now run all my servers and desktops on Gentoo and it's working great for me."...
Okay...
What?
You're going to have to explain this to me.
How is Gentoo any slower-moving or stabler than Fedora? In my regretably amble experience with Gentoo, change was frequent and catyclysmic. Time saved because portage rocks did not make up for the hundreds of hours repairing the shattered remains of my system every time an untested ebuild was released to the stable branch.
At least Fedora sits still for a few months between uphevals.
"But if that were not sufficient, then let's consider the reality of where the sales are: the desktop. And so far, I haven't seen a Linux distro that didn't suck on the desktop. From sluggishness to ugly font painteing to buggy desktop apps (because let's face it, the average Linux freak could care less about the desktop)...."
I agree Linux isn't ready for the desktop, but I disagree with your reasons.
-Kernel 2.6 is very responsive. Buggy, but responsive. -The only distro I've seen screw up fonts in the last 3 years or so is Gentoo. And Gentoo is good at screwing up. -3rd party GUI stuff goes from outstanding (Firefox) to crap (xcdroast). In no way is this distinct from Windows.
The problem is that you have to drop to the command line to do some things and you have to know things you shouldn't have to know.
They could potentially go into a different "meltdown" state.
Say someone builds a battery big enough to power a laptop, or maybe a PDA. If it's sufficiently well insulated (say in packing material), thermal equilibrium might be high enough to damage the unit or injure someone that touches it. The packaging of the nuclear material would be robust enough to survive no problem up to hundreds of degrees, but your PDA would still be broken.
We can only have perfectly written software when the task is trivial and stationary. Web browsers are hilariously far from trivial, and the requirements change over time.
I have two disks purchased a year apart and they both reported 100, so I figured it meant 0 reallocated sectors... I guess 100 doesn't mean 0, just "not many".
"What surprised me is that the manufacturers have their own bad sector table, so when you get the disk it's fairly likely that there are already bad areas which have been mapped out."
"It's not Apple's fault the KDE development community doesn't have a lot of use for the code that makes WebCore."
Yes it is. Apple has been releasing the code without any context. There's no way to know what bug a patch fixes, or what feature it implements. Without that kind of information, the code is more trouble than it's worth.
AFAIK Apple is satisfying its legal obligations under the LGPL license, but little more. That means they won't get sued, but it also means people are going to speak up when Apple claims to be collaborating with the community. Yes, they get along better with other projects, but they've burned KHTML and I think it's unreasonable to expect the developers to ignore that.
The sides and bottom of the aquarium would have very poor thermal conductivity. There is a lot of surface area as you say, but it takes a lot of glass to make up for one sq cm of aluminium. I was thinking of the top of the aquarium, and assuming a fan was circulating air through the gap at the top.
After reading the other posts I think the temperature would probably reach equilibrium and everything but the CPU and GPU would be opperating warmer than usual, but still at safe temperatures. But those parts are what limit the performance of the system if you're overclocking, so that might not be too bad.
Well the other two are perfectly serviceable, and if it meets your needs it would likely be a waste of time to duplicate the same firewall with different firewall software.
It's the firewall maintained by the OpenBSD project. The other BSDs now support it because it's more powerful than the IPFW and IPF firewalls that have been used historically on the BSDs. MacOS uses IPFW with a GUI. It's perfectly good for a desktop machine, but it's not hard to imagine someone wanting more on a server.
That's just an example, but there are other reasons one might pick OpenBSD over the alternatives. Same goes for Linux, MacOS X, just about every OS out there.
minis are popular for testing software destined for embedded systems running Linux on a PowerPC processor.
MacOS X and Linux are similar enough that porting software is pretty easy, but they're different enough that the software has to be ported. If software has been ported, you're not running quite the same code. The libraries and system calls also have subtle differences.
It doesn't matter that you're wasting an OS X license if the alternative costs more than a mini. It also doesn't matter if all the hardware doesn't work if the mini ends up sitting in the server room with people SSHing to it.
Because it has software MacOS X doesn't. Linux and MacOS X are not source code compatible, just close enough that porting isn't that hard most of the time.
Note that "most software" being ported doesn't cut it, because we might be talking about in-house stuff.
When you're testing software, it has to be in the environment that it will run on in production. If it's expected to run on Linux PPC, you test it on Linux PPC. To make it portable, you need to mess around with the preprocessor and the code that runs isn't the same on the two platforms, so a test on MacOS X is not good enough. Also the libraries aren't quite the same, and some of the system calls are different.
The mini is one of the cheapest PowerPC boxes you can get new, so there is actually a reason to use it for testing. PowerMacs are some of the cheapest 64-bit PowerPC boxes you can get.
It doesn't matter that you don't get support for the airport and graphics when the thing is sitting on a shelf in the server room.
I think you'd be surprised what you can do with a VIA C3. You could get even smaller/lower power than a mini. A G4 is easily the better processor in general, but a C3 may be good enough (I dunno about this guy's project's needs), and it wins easily in areas like encryption (as it has encryption acceleration). It's also way cheaper if you're doing a lot of customization, those boards can be had for a lot less than a mini.
What's wrong with man? It's not like there's a viable replacement.
Well, there's info. But the best thing I can say about info is that it's almost always possible to avoid it.
OpenBSD's kernel is biglock. Works fine on systems with a limited number of CPUs, but I don't imagine it would do very well on a 64-way system.
Who doesn't have OpenSSH?
What, no mention of the RMS talk on the 18th? :)
"Fedora was (by design) too fast-changing to support. I now run all my servers and desktops on Gentoo and it's working great for me." ...
Okay...
What?
You're going to have to explain this to me.
How is Gentoo any slower-moving or stabler than Fedora? In my regretably amble experience with Gentoo, change was frequent and catyclysmic. Time saved because portage rocks did not make up for the hundreds of hours repairing the shattered remains of my system every time an untested ebuild was released to the stable branch.
At least Fedora sits still for a few months between uphevals.
So what you're saying is that the RFID chip they're planning to use is stronger than the one they want to put in my passport/RealID? ...
Yeah, I can believe that.
Doesn't SCO get any love?
"But if that were not sufficient, then let's consider the reality of where the sales are: the desktop. And so far, I haven't seen a Linux distro that didn't suck on the desktop. From sluggishness to ugly font painteing to buggy desktop apps (because let's face it, the average Linux freak could care less about the desktop)...."
I agree Linux isn't ready for the desktop, but I disagree with your reasons.
-Kernel 2.6 is very responsive. Buggy, but responsive.
-The only distro I've seen screw up fonts in the last 3 years or so is Gentoo. And Gentoo is good at screwing up.
-3rd party GUI stuff goes from outstanding (Firefox) to crap (xcdroast). In no way is this distinct from Windows.
The problem is that you have to drop to the command line to do some things and you have to know things you shouldn't have to know.
They could potentially go into a different "meltdown" state.
Say someone builds a battery big enough to power a laptop, or maybe a PDA. If it's sufficiently well insulated (say in packing material), thermal equilibrium might be high enough to damage the unit or injure someone that touches it. The packaging of the nuclear material would be robust enough to survive no problem up to hundreds of degrees, but your PDA would still be broken.
They've been bickering since before Open Source software had that name.
We can only have perfectly written software when the task is trivial and stationary. Web browsers are hilariously far from trivial, and the requirements change over time.
Okay.
I have two disks purchased a year apart and they both reported 100, so I figured it meant 0 reallocated sectors... I guess 100 doesn't mean 0, just "not many".
It works on SATA on my OpenBSD system...
"What surprised me is that the manufacturers have their own bad sector table, so when you get the disk it's fairly likely that there are already bad areas which have been mapped out."
Can't you get the count with SMART?
"It's not Apple's fault the KDE development community doesn't have a lot of use for the code that makes WebCore."
Yes it is. Apple has been releasing the code without any context. There's no way to know what bug a patch fixes, or what feature it implements. Without that kind of information, the code is more trouble than it's worth.
AFAIK Apple is satisfying its legal obligations under the LGPL license, but little more. That means they won't get sued, but it also means people are going to speak up when Apple claims to be collaborating with the community. Yes, they get along better with other projects, but they've burned KHTML and I think it's unreasonable to expect the developers to ignore that.
The sides and bottom of the aquarium would have very poor thermal conductivity. There is a lot of surface area as you say, but it takes a lot of glass to make up for one sq cm of aluminium. I was thinking of the top of the aquarium, and assuming a fan was circulating air through the gap at the top.
After reading the other posts I think the temperature would probably reach equilibrium and everything but the CPU and GPU would be opperating warmer than usual, but still at safe temperatures. But those parts are what limit the performance of the system if you're overclocking, so that might not be too bad.
It's just an example. There are plenty of others.
I'd say the surface area of the various heatsinks would be greater than the surface area of the oil exposed to air in that tank.
Well the other two are perfectly serviceable, and if it meets your needs it would likely be a waste of time to duplicate the same firewall with different firewall software.
"PF?"
It's the firewall maintained by the OpenBSD project. The other BSDs now support it because it's more powerful than the IPFW and IPF firewalls that have been used historically on the BSDs. MacOS uses IPFW with a GUI. It's perfectly good for a desktop machine, but it's not hard to imagine someone wanting more on a server.
That's just an example, but there are other reasons one might pick OpenBSD over the alternatives. Same goes for Linux, MacOS X, just about every OS out there.
Because Linux/PowerPC is really nice for embedded stuff?
minis are popular for testing software destined for embedded systems running Linux on a PowerPC processor.
MacOS X and Linux are similar enough that porting software is pretty easy, but they're different enough that the software has to be ported. If software has been ported, you're not running quite the same code. The libraries and system calls also have subtle differences.
It doesn't matter that you're wasting an OS X license if the alternative costs more than a mini. It also doesn't matter if all the hardware doesn't work if the mini ends up sitting in the server room with people SSHing to it.
"Gotta disagree with you there. But I guess you don't use java at all, which is a must for me."
Java support on OpenBSD is crap. PF support on MacOS X is crap. Depends on your needs.