I think the CAPS are meaningful. Parent is talking about the X11 "DISPLAY" environment variable. If you have more monitors than you can connect to one video card, you can either have hardware acceleration or all physical screens on the same DISPLAY. Or at least that was the case the last time I looked into it.
Debian stable is always horrifically out of date. If your machine is a server, you run Debian stable because it won't change out from underneath you and break something. You might have to install a couple packages from backports or build a few things from source to get the relevant versions of whatever it is you intend to use the server for. That's okay though, because it's rare that you have to manually update more than a few packages for any particular use case, and once you've done it the first time, you don't need to do it again until the next release of Debian. The stack, however, is consistent enough that you can make "aptitude update && aptitude upgrade -y" a cron job.
For systems that are used interactively, however, it is best to run Testing or Sid, where reasonably current software is only an aptitude install away. You can try out 2 or 3 different hipster interpreted languages and recent releases of 12 different libraries while using the latest VIM and a modern desktop environment and audio player. (You'll still have to install the web browser yourself though. Alas.) When it comes time to deploy, you only need to go through the make-it-work-on-stable's-ancient-stack dance with the 4 or 5 packages your product actually depends on.
So MWh are a unit of energy. MW are a unit of power. MW per hour (MW/h) are a uinit of accelerating power consumption. MWh per hour (MWh/h) are the same thing as MW (the hours cancel).
Its nice of you to rant about how someone else is wrong, but next time, calm down and actually get it right yourself. In your huff and puff, you turned volume into acceleration, probably in a typo. But it left me a pedantic place to respond;)
It seems to have fantastic specs in most departments, but why so little RAM? The most annoying issue on modern smartphones is the UI stalls that occur when you try to switch back and forth between more than 2 apps, or open more than 4 tabs in the web browser. 4 GiB would solve most of those problems.
You really don't want 64 bit Chrom(e/ium). All it does is waste more memory. I've seen 64 bit Chromium use over 1 GiB for only 20 tabs. At the moment, any possible performance benefit of the extra registers is offset by the greater cache footprint of a 64 bit build. The future is the x32 ABI. AMD64 only makes sense for things that mmap large files and work with huge amounts of data.
Do you get hardware acceleration on all of them? Do you have to do something weird like run multiple X servers? I was under the impression that bad things begin happening when you try to connect more monitors than you can drive with a single video card.
And I know a guy that made an RS-232 serial terminal out of a speak-and-spell. The ability to write code for other machines does not make a personal computer.
Which is entirely sufficient for what 95% of people do with their computers. The only thing that you have to be wary of is screen resolution. Cheaper Windows Windows machines and the Macbook Air tend to stick you with the godawful 1366x768.
That gives 15.076 bits per symbol as opposed to 5.17 for single-case alphanumeric (my usual choice for memorability and efficient entry). That means passwords of the form of correcthorsebatterystaple are in between an 11 character and a 12 character alphanumeric password in strength, assuming you use a good RNG for generating passwords and your attacker has full knowledge of your dictionary. To get a random english password of equivalent strength to a 20 character alphanumeric (my standard for encrypted disks), you need to use 7 words. That might be reasonable if you are a relatively fast typist with a relatively poor memory.
If you wanted your random english password to be as strong as the AES key you're deriving from it, you'd need 9 words.
Mixed case and special characters are a dumb idea anyway. If you use those, you are using the shift key. Adding the variable of whether you pressed the shift key or not doubles the number of possibilities for each symbol. That is, 2x possible symbols where x is the number of keys. Optionally pressing a second (lowercase) key gives x+x^2 possible symbols.
With even a slightly good password, the server response latency is enough of a brute force block. Anything more than that is to protect the server from the coincident DOS. The people cracking 10-12 character passwords are doing it locally on highly parallel hardware.
That gives 15.076 bits per symbol as opposed to 5.17 for single-case alphanumeric (my usual choice for memorability and efficient entry). That means passwords of the form of correcthorsebatterystaple are in between an 11 character and a 12 character alphanumeric password in strength, assuming you use a good RNG for generating passwords and your attacker has full knowledge of your dictionary. To get a random english password of equivalent strength to a 20 character alphanumeric (my standard for encrypted disks), you need to use 7 words. That might be reasonable if you are a relatively fast typist with a relatively poor memory.
If you wanted your random english password to be as strong as the AES key you're deriving from it, you'd need 9 words.
With the handicap of being an interpreter for a layout-oriented scripting language written in an interpreted scripting language, I suppose one might choose to call the performance amazing.
If the security risk of leaving your machine powered on is a serious concern, you should be using laptop with systemd and a fast SSD and keeping it in a tamper-evident safe. You might also consider potting an ultrabook in thermally-conductive epoxy.
Windows users still install programs by downloading executables from the internet and running them as root. It doesn't matter what we do to our windows and doors when one wall of our house is missing.
Only for the 13" model, which by all reports has the Intel HD4000 struggling to keep up
What reports? The integrated GPU in my Core 2 Duo handles my 3520x1200 frame buffer reasonably well on Linux. I find it hard to believe that 5 generations later Intel hasn't improved at all.
Slashdot Mirror
I think the CAPS are meaningful. Parent is talking about the X11 "DISPLAY" environment variable. If you have more monitors than you can connect to one video card, you can either have hardware acceleration or all physical screens on the same DISPLAY. Or at least that was the case the last time I looked into it.
Debian stable is always horrifically out of date. If your machine is a server, you run Debian stable because it won't change out from underneath you and break something. You might have to install a couple packages from backports or build a few things from source to get the relevant versions of whatever it is you intend to use the server for. That's okay though, because it's rare that you have to manually update more than a few packages for any particular use case, and once you've done it the first time, you don't need to do it again until the next release of Debian. The stack, however, is consistent enough that you can make "aptitude update && aptitude upgrade -y" a cron job.
For systems that are used interactively, however, it is best to run Testing or Sid, where reasonably current software is only an aptitude install away. You can try out 2 or 3 different hipster interpreted languages and recent releases of 12 different libraries while using the latest VIM and a modern desktop environment and audio player. (You'll still have to install the web browser yourself though. Alas.) When it comes time to deploy, you only need to go through the make-it-work-on-stable's-ancient-stack dance with the 4 or 5 packages your product actually depends on.
"Per" means "/".
So MWh are a unit of energy. MW are a unit of power. MW per hour (MW/h) are a uinit of accelerating power consumption. MWh per hour (MWh/h) are the same thing as MW (the hours cancel).
Its nice of you to rant about how someone else is wrong, but next time, calm down and actually get it right yourself. In your huff and puff, you turned volume into acceleration, probably in a typo. But it left me a pedantic place to respond ;)
Alright, pot.
it's the first time I've heard of Oxygen described as flashy.
The window shadow is a giant blue glow.
It seems to have fantastic specs in most departments, but why so little RAM? The most annoying issue on modern smartphones is the UI stalls that occur when you try to switch back and forth between more than 2 apps, or open more than 4 tabs in the web browser. 4 GiB would solve most of those problems.
Seeing as human vision is capable of higher resolution in lumiance than chroma, why are you calling the inefficient RGB layout "proper"?
You really don't want 64 bit Chrom(e/ium). All it does is waste more memory. I've seen 64 bit Chromium use over 1 GiB for only 20 tabs. At the moment, any possible performance benefit of the extra registers is offset by the greater cache footprint of a 64 bit build. The future is the x32 ABI. AMD64 only makes sense for things that mmap large files and work with huge amounts of data.
Do you get hardware acceleration on all of them? Do you have to do something weird like run multiple X servers? I was under the impression that bad things begin happening when you try to connect more monitors than you can drive with a single video card.
And I know a guy that made an RS-232 serial terminal out of a speak-and-spell. The ability to write code for other machines does not make a personal computer.
You paid $1500 for a laptop with a 1366x768 screen? You poor bastard.
Which is entirely sufficient for what 95% of people do with their computers. The only thing that you have to be wary of is screen resolution. Cheaper Windows Windows machines and the Macbook Air tend to stick you with the godawful 1366x768.
How much time do you spend messing around with bullshit caused by not having a real package manager?
I get 24592 3-7 letter words, which is about what you would want to use for a passphrase.
/usr/share/dict/american-english | wc -l
24592
$ grep -E '^[a-z]{3,7}$'
That gives 15.076 bits per symbol as opposed to 5.17 for single-case alphanumeric (my usual choice for memorability and efficient entry). That means passwords of the form of correcthorsebatterystaple are in between an 11 character and a 12 character alphanumeric password in strength, assuming you use a good RNG for generating passwords and your attacker has full knowledge of your dictionary. To get a random english password of equivalent strength to a 20 character alphanumeric (my standard for encrypted disks), you need to use 7 words. That might be reasonable if you are a relatively fast typist with a relatively poor memory.
If you wanted your random english password to be as strong as the AES key you're deriving from it, you'd need 9 words.
Mixed case and special characters are a dumb idea anyway. If you use those, you are using the shift key. Adding the variable of whether you pressed the shift key or not doubles the number of possibilities for each symbol. That is, 2x possible symbols where x is the number of keys. Optionally pressing a second (lowercase) key gives x+x^2 possible symbols.
With even a slightly good password, the server response latency is enough of a brute force block. Anything more than that is to protect the server from the coincident DOS. The people cracking 10-12 character passwords are doing it locally on highly parallel hardware.
I get 24592 3-7 letter words, which a reasonable set to use for a passphrase:
/usr/share/dict/american-english | wc -l
24592
$ grep -E '^[a-z]{3,7}$'
That gives 15.076 bits per symbol as opposed to 5.17 for single-case alphanumeric (my usual choice for memorability and efficient entry). That means passwords of the form of correcthorsebatterystaple are in between an 11 character and a 12 character alphanumeric password in strength, assuming you use a good RNG for generating passwords and your attacker has full knowledge of your dictionary. To get a random english password of equivalent strength to a 20 character alphanumeric (my standard for encrypted disks), you need to use 7 words. That might be reasonable if you are a relatively fast typist with a relatively poor memory.
If you wanted your random english password to be as strong as the AES key you're deriving from it, you'd need 9 words.
And the really good 3D glasses use circular polarization, so the picture doesn't change when you tilt your head.
speed of PDF.js are simply amazing.
With the handicap of being an interpreter for a layout-oriented scripting language written in an interpreted scripting language, I suppose one might choose to call the performance amazing.
Laptops are already under 25.5W, as long as you stay away from oversized screens and discrete GPUs.
On: 70W
Off: 0W
S3 Suspend: 2W
If the security risk of leaving your machine powered on is a serious concern, you should be using laptop with systemd and a fast SSD and keeping it in a tamper-evident safe. You might also consider potting an ultrabook in thermally-conductive epoxy.
Windows users still install programs by downloading executables from the internet and running them as root. It doesn't matter what we do to our windows and doors when one wall of our house is missing.
>20" Viewable
>0.24mm aperture grille
No.
Ah, good old glossy 1366x768.
I assure you, you don't have a 2048x1536 CRT. What you actually have is a 1600x1200 CRT, being overdriven so as to be so blurry it's useless.
Only for the 13" model, which by all reports has the Intel HD4000 struggling to keep up
What reports? The integrated GPU in my Core 2 Duo handles my 3520x1200 frame buffer reasonably well on Linux. I find it hard to believe that 5 generations later Intel hasn't improved at all.