The thing is, Python is not really any easier than C++. The only things it gives you are more run-time flexibility and a garbage collector (which is trivial to add to C++; although not as integrated).
Otherwise it's just as verbose and irksome as C++. However, C++ is way faster. I find Python to be rather annoying as far as scripting languages go.
For easy GUI development I like FLTK and its Fluid tool, Qt Designer, or just plain Visual Studio (C++, VB, or C#).
Put that sucker up in the air about 50ft and then tell me how far it goes. You'd be surprised at what a difference it makes. The difference is huge because inside your house everything is blocking and absorbing the radio energy.
Note that you need to keep the connection from the transmitter and the antenna extremely short because cable loses are very high at the frequencies used for 802.11. What this means is that you should put your router or wireless adapter up there at 50+ ft with the antenna connected directly to it. Then run your ethernet or USB cable back down to your computer(s).
A good high gain omni-directional antenna can make a difference. Just be careful about having too much gain as there are maximum legal limits for power (which includes gain from antennas).
Or better yet, get your amateur radio license and come join us for some real fun. We can use extremely high-gain antennas (like slot omni antennas) and enormous power for transmitting (although most people stick to less that 10 watts or so). Our access points go tens of miles, especially those mounted at 300+ft.
Of course different areas have different rules, but just about everywhere there are restrictions on antenna gain and power for 802.11. Especially if you are not a licensed amateur radio operator.
These things are unlicensed part 15 devices which have strict restrictions on power output (which includes any gain from directional antennas) and can not interfere with licensed devices like amateur radio operators.
I'll leave it up to the reader to Google for what the limits are in your area since it varies. Just remember that you can't just slap on any super-high-gain antenna and remain legal.
I mean why would a person NOT use the fan provided by Intel in the retail package?
I'll tell you why. I don't overclock (I've been known to underclock though) and I like quiet systems. The stock fan/heatsink combo doesn't cut it. I tend to use a larger sink than necessary and run a larger fan at a low voltage. I enjoy silence. You don't know what you're missing out on.
With that said, if it's a retail CPU then I keep the old sink and fan so I could send it all back if needed. I rarely buy retail packages any more though.
Re:What's that Arch thing the guy is talking about
on
Fedora Core 2 Review
·
· Score: 3, Informative
Arch Linux. It's an i686, 2.6 kernel, devfs, KDE 3.2.2, GNOME 2.6 using binary distro (similar to Debian except even more lightweight and up-to-date).
I use it all the time. My primary machine is still Debian but all my other machines and servers are running Arch. It requires a bit more setup work than Debian.
I like it because it is extremely lightweight but has an excellent packaging system (pacman). The packaging system (and all those packages) are pretty much the only reason I've stuck with Debian all these years and Arch is the first to come along that comes close (Gentoo is OK, but compiling is a waste of time). Although it doesn't have anywhere near the number of packages as Debian, I can see it growing rapidly.
An example of the sane thinking behind Arch: There is no "/usr/doc" directory. I always use manpages or go online to find documentation. I've never understood why so many distros include all that documentation. I mean you rarely use it (mostly just for setup), why make it take up disk space? Everything is online nowadays and manpages are easy/handy.
Also, the install is fairly raw (which is a good thing). It just works and is simple. They need to fix some stuff with regards to swapfile setup (like if you don't want a swap partition) but otherwise it is fairly easy. You almost don't even need the installer (just the boot CD). Too many distros go off with their crazy complex and broken installers that end up leaving you frustrated (*cough* Debian *cough*).
I agree that O'Caml is pretty good. However, it has some serious flaws holding it back.
One is that it suffers from an absolutely horrible default syntax. Erlang and the original SML that O'Caml is derived from present a much more sane syntax. Yes, you can change the syntax but no one wants to add that confusion!
Another is that O'Caml suffers from academia syndrome. Too many absolutely useless and/or confusing features and too much optimization for the least used cases. A common problem with inexperienced developers, academics, engineers pretending to be programmers, and/or other ivory tower builders.
Cyclone (a "safe" extension to C) is another project that could have been really great but was run into the ground by academic foolishness.
Sorry, but it sounds to me like Jobs realized that continuing development of OS9 would me more effort than it was worth.
Maybe, but if you remember, Jobs left Apple and did the NeXT thing. I think he already knew the old MacOS was a dead-end.
NeXTOS was a Mach kernel/BSD system just like OS X. Essentially all he did was release a newer version of NeXTOS on Apple hardware with a more MacOS-like interface.
Yeah right, until MacOS becomes popular enough for someone to exploit some known hole and install some sinister daemons, keyboard loggers, or whatever.
And don't think the root/user separation in OSX is gonna save you. All the virus/trojan would have to do is wait until you ran something requiring privileges (like the software update) and either grab the root password or piggyback inside the timeout period.
Of course, this ignore something else: If you could give me a pill and give me a toned body right now, the odds are much greater that I'd engage in much more exercise then I do now, even if it weren't strictly necessary. The hump is what stops me; I've tried several times to start an exercise program, but I've got so far to go before it's really fun and not boring that I never make it over that hump. I mean, I feel all bad about it and stuff, but that doesn't help much.
I doubt you would really feel any different about exercising. Oh, you might at first, but then you'd revert. It's all mental and if you don't have the will now, you won't when you are in shape.
I'm in damn fine shape and I sometimes struggle to keep up the activity. It does not get any easier no matter what shape you are in. Again, it's all in your mind. A change of attitude is the only thing that will make a difference and that is one of the toughest things anyone can do.
I didn't read the whole paper but the first attack was dealing with the meta-data.
What I can't believe is that they would still leave so much stuff unencrypted. A very poor design decision.
I mean, how freaking hard is it to put a flag at the start of the archive saying it is encrypted and then just raw encrypt the rest of the data. That design seems obvious and would be as secure as AES can be (eg. just create a normal zip, encrypt it, add flag/tag at start).
65 dBa is quiet. Those crappy old desktop computers put out more sound than that. Most high quality cars have an interior road noise level of around 65 dBa at 60 MPH.
Did they mean 165 dBa or something? (now that that would be loud as hell)
If I could get drivers for my WinTV PVR250 and infrared remote control on FreeBSD-5.x, I'd be far happier.
Yeah, my MythTV machine is one of the machines running ReiserFS. Seems to work OK but I'm worried that the database may get corrupted some day (like when the beta IVTV driver locks up the system or something). I've been thinking of converting it to ext3 but that machine has so much storage that I can't copy everything off.
In my own testing I found FFS to be comparable to EXT2/3 which is probably the appropriate choice for a BSD-like Linux system.
Although I've never tried them I've heard of PVR250/350 drivers for FreeBSD. CXM or something like that?
What? I did the same thing. I bought the Windows version and downloaded the Linux stuff. Much easier that way and I can play on both platforms if I wish (even though I only ever use Linux).
First copy your Windows install of Q3 over to your Linux machine (or copy the stuff straight off your Q3 CD). Then download and install the latest Q3 point release for Linux. Install it in the same directory.
All you need are the Linux executable and the.pk3 files off the CDROM (which work on any system).
Just make sure you install the point release last because it updates/overwrites some of the pk3 files.
I never noticed it being as bad with ext2/3, even in writeback mode. I have lost data with them but not as often as Reiser (with which my heart sinks every time the system locks).
I didn't know that Reiser now had ordered mode. I just turned it on and I'll keep it that way! Thanks.
Seriously, all the time we see benchmarks like this that are just done on the same machine with the same setup. Who knows if there is some unseen problem or bottleneck (in this particular case the CPU is weak).
We need a large sample base. Different types of chipsets, CPU's, hard-drives, etc. Then we can better see the big picture or at least see how the filesystems might perform on a system similar to your own.
So I'm calling for a "filesystem benchmark" page were people can post their results from a standard set of benchmarks. Something where they can include their system specs/setup and everything.
Add me to the list of people who have lost data with Reiser.
Now, I still use Reiser. I currently have 4 or 5 systems running on pure ReiserFS. I use it for both small and large files(ystems).
The problem with Reiser is that there is a high probability of data corruption if your system dies (lock-up or power failure, etc.). Generally, any files that are open when the system goes down are subject to corruption.
It's often hard to tell that anything went wrong because everything will appear normal but your files will have bogus data in them. I've lost VMware machines (they still worked but all sorts of things were screwed up in Windows), all my GNOME settings (multiple times; I guess GNOME holds lots of files open or updates them all the time; I've since switched to KDE), source files, graphics files, etc...
I'm thinking of switching to something else. XFS chews too much CPU, JFS has been extremely unstable every time I've tried it (which I admit has been a while), ext3 is slow. Meh, that's why I've just stuck with Reiser. I like how fast it can delete huge directory trees and it generally performs very well except for the data loss thing.
My Dell C840 does 1600x1200 on its 15" screen... This thing is 17" and only 1440x900?! Sounds too much like a Mac.
The real stinker for this machine is that you can't get a nVidia card. Sorry, but I've been burned too many times by ATI. I only use nVidia and just that feature alone could determine if I buy a particular machine. nVidia "just works" whether I'm running Windows, Linux, whatever, and it performs well.
For that DR model (the one with the desktop CPU) it would be nice if you could get it with no processor and no RAM. Then I can buy my own (cheaper) and jam it in.
Re:SIlence is a pipe dream for me
on
A Silent PC Solution?
·
· Score: 4, Informative
hmmm... Yeah, you've got a lot of noise makers there.
Replace the power supply with a Zalman unit (check Newegg for "noise-free" versions). The Zalman isn't completely silent but it's a very high quality supply (heavy though!) and it adjusts the fan speed automatically. There may be more quiet units out there, I don't know. That Enermax is loud though.
You've got to get rid of your video card's fan. It will be loud when everything else is quieted down. I would seriously consider a fanless video card. This can make a huge difference even when you think it's fairly quiet.
Those hard-drives you're using are loud as hell (and you've got 2 of them!). Go with a single "quiet" drive. I use Seagate but you'll have to look around for what you need (maybe the quiet Maxtor). Last I checked, Western Digital drives are the loudest out there.
2 80mm case fans?? You might try taking one or both out and see how your system does. Improve air-flow through your case and let the power supply do most of the work. This is harder to do and takes lots of experimentation. It is a black art of sorts. You could also try a much larger fan running at low RPM (larger as in at least 160mm+).
Hmmm, there really isn't much to a driver. It's basically just a coil of wire attached to a suspended structure (the cone) that sits inside a permanent magnet. The energy is fed to the coil which makes it move inside the magnet which in turn moves the cone structure to create air pressure waves (sound).
Simple science-type experiments are super easy to do. No more complex than an electric motor experiment.
Although I haven't read it, this probably has everything you need.
It's certainly quite possible to design a C# vector class that's both more memory and processor efficient in most cases than C++. Here's how:
You mean you've implemented this in C#? Or are you saying what you could do?
Sorry but there's no way to do it in the current C# implementation. You can't even get access to the memory allocators. There is no C# data structure that can tightly pack data in memory while still being dynamic. The best you can do is an array of struct's which is about as efficient as C/C++ but is absolutely not dynamic without causing a ton of overhead (eg. there is no realloc in C#).
If you can do it in C# then show me some code.
C++ doubles the amount of space allocated for a vector (or queue, or list, or stack, or dequeue, or binary heap, etc) whenever a resize exceeds the amount already allocated, unless you know enough to tell it to do otherwise.
You're talking about the default vector allocator. STL is extremely dynamic, you do not have to use to the default allocator if you don't want you. You can be extremely flexible here for whatever you need and design anything you want.
If you've ever done much benchmarking of the C++ STL, you know that it's usually faster than otherwise identical code written with arrays, which shouldn't be possible, since the array access code can be done fairly easily in assembly without virtual function table lookups and such, but nonetheless is quite real.
Actually, I have, and you're wrong. So me an example and I'll show you why your array code is slower than the C++ version.
Virtual function lookups?! Do you even know how the STL is designed or how C++ works? Now I'm not sure why I bothered with the rest of what I wrote above.
Slashdot Mirror
You don't need to wait. You can currently "export" (print) to Postscript/EPS. Convert that to PDF, and you're done.
Going the other way is what I'd really like to see. That is, import ps and PDF into Inkscape.
The thing is, Python is not really any easier than C++. The only things it gives you are more run-time flexibility and a garbage collector (which is trivial to add to C++; although not as integrated).
Otherwise it's just as verbose and irksome as C++. However, C++ is way faster. I find Python to be rather annoying as far as scripting languages go.
For easy GUI development I like FLTK and its Fluid tool, Qt Designer, or just plain Visual Studio (C++, VB, or C#).
Seriously, just build your arcade machines.
The dimensions are out there. No need to go to auctions and buy expensive and/or crappy old hardware. Build your own from scratch, use MAME and voilà.
You know, then you can actually play the arcade machines.
That's what I did anyway. I built my own air-hockey and pool table also.
Put that sucker up in the air about 50ft and then tell me how far it goes. You'd be surprised at what a difference it makes. The difference is huge because inside your house everything is blocking and absorbing the radio energy.
Note that you need to keep the connection from the transmitter and the antenna extremely short because cable loses are very high at the frequencies used for 802.11. What this means is that you should put your router or wireless adapter up there at 50+ ft with the antenna connected directly to it. Then run your ethernet or USB cable back down to your computer(s).
A good high gain omni-directional antenna can make a difference. Just be careful about having too much gain as there are maximum legal limits for power (which includes gain from antennas).
Or better yet, get your amateur radio license and come join us for some real fun. We can use extremely high-gain antennas (like slot omni antennas) and enormous power for transmitting (although most people stick to less that 10 watts or so). Our access points go tens of miles, especially those mounted at 300+ft.
This brings up the question:
Can someone not in Italy be charged under Italian this law?
Of course different areas have different rules, but just about everywhere there are restrictions on antenna gain and power for 802.11. Especially if you are not a licensed amateur radio operator.
These things are unlicensed part 15 devices which have strict restrictions on power output (which includes any gain from directional antennas) and can not interfere with licensed devices like amateur radio operators.
I'll leave it up to the reader to Google for what the limits are in your area since it varies. Just remember that you can't just slap on any super-high-gain antenna and remain legal.
I mean why would a person NOT use the fan provided by Intel in the retail package?
I'll tell you why. I don't overclock (I've been known to underclock though) and I like quiet systems. The stock fan/heatsink combo doesn't cut it. I tend to use a larger sink than necessary and run a larger fan at a low voltage. I enjoy silence. You don't know what you're missing out on.
With that said, if it's a retail CPU then I keep the old sink and fan so I could send it all back if needed. I rarely buy retail packages any more though.
Arch Linux. It's an i686, 2.6 kernel, devfs, KDE 3.2.2, GNOME 2.6 using binary distro (similar to Debian except even more lightweight and up-to-date).
I use it all the time. My primary machine is still Debian but all my other machines and servers are running Arch. It requires a bit more setup work than Debian.
I like it because it is extremely lightweight but has an excellent packaging system (pacman). The packaging system (and all those packages) are pretty much the only reason I've stuck with Debian all these years and Arch is the first to come along that comes close (Gentoo is OK, but compiling is a waste of time). Although it doesn't have anywhere near the number of packages as Debian, I can see it growing rapidly.
An example of the sane thinking behind Arch: There is no "/usr/doc" directory. I always use manpages or go online to find documentation. I've never understood why so many distros include all that documentation. I mean you rarely use it (mostly just for setup), why make it take up disk space? Everything is online nowadays and manpages are easy/handy.
Also, the install is fairly raw (which is a good thing). It just works and is simple. They need to fix some stuff with regards to swapfile setup (like if you don't want a swap partition) but otherwise it is fairly easy. You almost don't even need the installer (just the boot CD). Too many distros go off with their crazy complex and broken installers that end up leaving you frustrated (*cough* Debian *cough*).
I agree that O'Caml is pretty good. However, it has some serious flaws holding it back.
One is that it suffers from an absolutely horrible default syntax. Erlang and the original SML that O'Caml is derived from present a much more sane syntax. Yes, you can change the syntax but no one wants to add that confusion!
Another is that O'Caml suffers from academia syndrome. Too many absolutely useless and/or confusing features and too much optimization for the least used cases. A common problem with inexperienced developers, academics, engineers pretending to be programmers, and/or other ivory tower builders.
Cyclone (a "safe" extension to C) is another project that could have been really great but was run into the ground by academic foolishness.
Sorry, but it sounds to me like Jobs realized that continuing development of OS9 would me more effort than it was worth.
Maybe, but if you remember, Jobs left Apple and did the NeXT thing. I think he already knew the old MacOS was a dead-end.
NeXTOS was a Mach kernel/BSD system just like OS X. Essentially all he did was release a newer version of NeXTOS on Apple hardware with a more MacOS-like interface.
Thats why I'm such a FreeBSD/Mac advocate.
Yeah right, until MacOS becomes popular enough for someone to exploit some known hole and install some sinister daemons, keyboard loggers, or whatever.
And don't think the root/user separation in OSX is gonna save you. All the virus/trojan would have to do is wait until you ran something requiring privileges (like the software update) and either grab the root password or piggyback inside the timeout period.
Of course, this ignore something else: If you could give me a pill and give me a toned body right now, the odds are much greater that I'd engage in much more exercise then I do now, even if it weren't strictly necessary. The hump is what stops me; I've tried several times to start an exercise program, but I've got so far to go before it's really fun and not boring that I never make it over that hump. I mean, I feel all bad about it and stuff, but that doesn't help much.
I doubt you would really feel any different about exercising. Oh, you might at first, but then you'd revert. It's all mental and if you don't have the will now, you won't when you are in shape.
I'm in damn fine shape and I sometimes struggle to keep up the activity. It does not get any easier no matter what shape you are in. Again, it's all in your mind. A change of attitude is the only thing that will make a difference and that is one of the toughest things anyone can do.
I didn't read the whole paper but the first attack was dealing with the meta-data.
What I can't believe is that they would still leave so much stuff unencrypted. A very poor design decision.
I mean, how freaking hard is it to put a flag at the start of the archive saying it is encrypted and then just raw encrypt the rest of the data. That design seems obvious and would be as secure as AES can be (eg. just create a normal zip, encrypt it, add flag/tag at start).
Am I missing something or what?
65 dBa is quiet. Those crappy old desktop computers put out more sound than that. Most high quality cars have an interior road noise level of around 65 dBa at 60 MPH.
Did they mean 165 dBa or something? (now that that would be loud as hell)
Knowledge and Wisdom are not the same thing.
If I could get drivers for my WinTV PVR250 and infrared remote control on FreeBSD-5.x, I'd be far happier.
Yeah, my MythTV machine is one of the machines running ReiserFS. Seems to work OK but I'm worried that the database may get corrupted some day (like when the beta IVTV driver locks up the system or something). I've been thinking of converting it to ext3 but that machine has so much storage that I can't copy everything off.
In my own testing I found FFS to be comparable to EXT2/3 which is probably the appropriate choice for a BSD-like Linux system.
Although I've never tried them I've heard of PVR250/350 drivers for FreeBSD. CXM or something like that?
What? I did the same thing. I bought the Windows version and downloaded the Linux stuff. Much easier that way and I can play on both platforms if I wish (even though I only ever use Linux).
.pk3 files off the CDROM (which work on any system).
First copy your Windows install of Q3 over to your Linux machine (or copy the stuff straight off your Q3 CD). Then download and install the latest Q3 point release for Linux. Install it in the same directory.
All you need are the Linux executable and the
Just make sure you install the point release last because it updates/overwrites some of the pk3 files.
I never noticed it being as bad with ext2/3, even in writeback mode. I have lost data with them but not as often as Reiser (with which my heart sinks every time the system locks).
I didn't know that Reiser now had ordered mode. I just turned it on and I'll keep it that way! Thanks.
Seriously, all the time we see benchmarks like this that are just done on the same machine with the same setup. Who knows if there is some unseen problem or bottleneck (in this particular case the CPU is weak).
We need a large sample base. Different types of chipsets, CPU's, hard-drives, etc. Then we can better see the big picture or at least see how the filesystems might perform on a system similar to your own.
So I'm calling for a "filesystem benchmark" page were people can post their results from a standard set of benchmarks. Something where they can include their system specs/setup and everything.
Then maybe we'll get useful information.
Add me to the list of people who have lost data with Reiser.
Now, I still use Reiser. I currently have 4 or 5 systems running on pure ReiserFS. I use it for both small and large files(ystems).
The problem with Reiser is that there is a high probability of data corruption if your system dies (lock-up or power failure, etc.). Generally, any files that are open when the system goes down are subject to corruption.
It's often hard to tell that anything went wrong because everything will appear normal but your files will have bogus data in them. I've lost VMware machines (they still worked but all sorts of things were screwed up in Windows), all my GNOME settings (multiple times; I guess GNOME holds lots of files open or updates them all the time; I've since switched to KDE), source files, graphics files, etc...
I'm thinking of switching to something else. XFS chews too much CPU, JFS has been extremely unstable every time I've tried it (which I admit has been a while), ext3 is slow. Meh, that's why I've just stuck with Reiser. I like how fast it can delete huge directory trees and it generally performs very well except for the data loss thing.
Yeah, I was going to mention the same thing.
My Dell C840 does 1600x1200 on its 15" screen... This thing is 17" and only 1440x900?! Sounds too much like a Mac.
The real stinker for this machine is that you can't get a nVidia card. Sorry, but I've been burned too many times by ATI. I only use nVidia and just that feature alone could determine if I buy a particular machine. nVidia "just works" whether I'm running Windows, Linux, whatever, and it performs well.
For that DR model (the one with the desktop CPU) it would be nice if you could get it with no processor and no RAM. Then I can buy my own (cheaper) and jam it in.
hmmm... Yeah, you've got a lot of noise makers there.
Replace the power supply with a Zalman unit (check Newegg for "noise-free" versions). The Zalman isn't completely silent but it's a very high quality supply (heavy though!) and it adjusts the fan speed automatically. There may be more quiet units out there, I don't know. That Enermax is loud though.
You've got to get rid of your video card's fan. It will be loud when everything else is quieted down. I would seriously consider a fanless video card. This can make a huge difference even when you think it's fairly quiet.
Those hard-drives you're using are loud as hell (and you've got 2 of them!). Go with a single "quiet" drive. I use Seagate but you'll have to look around for what you need (maybe the quiet Maxtor). Last I checked, Western Digital drives are the loudest out there.
2 80mm case fans?? You might try taking one or both out and see how your system does. Improve air-flow through your case and let the power supply do most of the work. This is harder to do and takes lots of experimentation. It is a black art of sorts. You could also try a much larger fan running at low RPM (larger as in at least 160mm+).
('bollards', for our British readers)
What? That's an odd name. I'd have called them "chazzwazzers".
Hmmm, there really isn't much to a driver. It's basically just a coil of wire attached to a suspended structure (the cone) that sits inside a permanent magnet. The energy is fed to the coil which makes it move inside the magnet which in turn moves the cone structure to create air pressure waves (sound).
Simple science-type experiments are super easy to do. No more complex than an electric motor experiment.
Although I haven't read it, this probably has everything you need.
It's certainly quite possible to design a C# vector class that's both more memory and processor efficient in most cases than C++. Here's how:
You mean you've implemented this in C#? Or are you saying what you could do?
Sorry but there's no way to do it in the current C# implementation. You can't even get access to the memory allocators. There is no C# data structure that can tightly pack data in memory while still being dynamic. The best you can do is an array of struct's which is about as efficient as C/C++ but is absolutely not dynamic without causing a ton of overhead (eg. there is no realloc in C#).
If you can do it in C# then show me some code.
C++ doubles the amount of space allocated for a vector (or queue, or list, or stack, or dequeue, or binary heap, etc) whenever a resize exceeds the amount already allocated, unless you know enough to tell it to do otherwise.
You're talking about the default vector allocator. STL is extremely dynamic, you do not have to use to the default allocator if you don't want you. You can be extremely flexible here for whatever you need and design anything you want.
If you've ever done much benchmarking of the C++ STL, you know that it's usually faster than otherwise identical code written with arrays, which shouldn't be possible, since the array access code can be done fairly easily in assembly without virtual function table lookups and such, but nonetheless is quite real.
Actually, I have, and you're wrong. So me an example and I'll show you why your array code is slower than the C++ version.
Virtual function lookups?! Do you even know how the STL is designed or how C++ works? Now I'm not sure why I bothered with the rest of what I wrote above.