So you re-image the computer. That takes all of what, 10 minutes. Just put in the student contract or whatever that if the student fubars the computer, then anything on said computer will be erased. Quite simple, really; the student loses all their mp3s, pr0n, whatever, you get a hassle-free life, everyone wins.
That's a well thought-out decision, however you're kidding yourself if you think that "momentum has nothing to do with it". Momentum has everything to do with it: that is precisely why win2k beats Linux for your roll-out. Linux being too expensive is just another way of saying "we don't have the skill set in-house to implement this", which is a perfectly understandable and valid business decision.
Pronounceable... interesting concept that. One would think that if pronounceability was a major concern then one would choose a slashdot ident with that property...
I just want kate and kdevelop available on windows. That way I can have the same dev environment at home and at work.
BTW, what closed-source media playback software do you use? I use mplayer (under linux, admittedly), and I find it much better than anything I have found under windows (without looking too hard).
Perhaps a decent measure would be to average the processing out over a week or so; eg each seti@home unit is, say, 1e9 floating point operations, calculate how many units are processed in a week, divide by seconds in a week, there's your number. This method allows for the redundancy of the @home method, ie, each unit will be computed a number of times, if only the completed units are counted it gives a measure of true (sustained) performance.
Yes, yes, yes, Open BSD is nice and all, but does it run on Alpha with turbochannel bus? Does it have downloadable install ISOs? Does it have pkgsrc? No, no and no, so it is not a direct replacement for NetBSD for me.
Of course, I should probably be running openBSD on my sparc web server, but netBSD had downloadable, bootable ISOs available, and openBSD didn't...
Great script; thanks for the rsync line.
BTW, if you wanted to use a for loop, here's one way:
for i in $(seq 6 -1 1) ; do if [ -d daily.${i} ] ; then mv -f daily.${i} daily.$(( $i + 1 )) fi done
Note that you don't really need to delete the daily.7 dir first; the mv -f should take care of
that. Also, I think the $((... )) operator might only apply to bash, so you may have to change the !/bin/sh to !/bin/bash at the top of the script.
Another suggestion is to use a version control system to stor your home directory. It's a really good way to keep old versions of stuff, and the bonus is that you only need to backup your repository. I'd suggest subversion, it works for me.
Re:Do they or do they not have the source legally?
on
Zeta Goes Gold
·
· Score: 1
I am replying twice because I have another point to make and didn't want it to get lost in the first post
You wrote:
Linux distros? How many of them make major changes to the kernel? (Honest question, I have no idea, but I'm guessing not many.) How many of them are released by a single person? Again, if you have a company to work on it, awesome. But a single developer would be overwhelmed.
Corollary: The only people who get to work on large, non-trivial programs are members of large teams (because a "single developer would be overwhelmed"). Therefore, in a closed-source-only world, the only non-trivial programs that get written are by large companies, meaning the only ones that get written are the ones that scratch a particular market itch.
Re:Do they or do they not have the source legally?
on
Zeta Goes Gold
·
· Score: 1
The only custom builds of FF I know of are Netscape 8, K-Meleon and some nightly builds that get compiled with optimizations on.
The only custom builds I know of IE are, well, umm, none.
And that is exactly the point. Users of IE are forced to upgrade to the MS step; they can't get security fixes without MS say-so, and they can't get enhancements without MS saying so, like, maybe, full CSS2 compliance.
If you are from the USA, you have this wonderful thing called "freedom of speech". It allows you to say pretty much anything in a public forum without lawful consequences (civil law notwithstanding). Now, how many people actually take the opportunity to exercise their freedom of speech rights in a meaningful way? Not many, I'd bet, however the right to speak is an important one. Similarly, the right to modify software, even if the rights are not directly exercised by the vast majority of users, is an important one.
Re:Is it YOUR own source code?
on
Zeta Goes Gold
·
· Score: 1
I found the example xorg.conf file supplied by nvidia to be more than enough to get twinview working. I had triple monitors at one stage, too, by virtue of a PCI geforce mx200 or somesuch, but I suspect that is not fast enough for what the OP asked for.
Probably in the wrong place, but anyway, the other day I was doing a list of what I dislike about windows. Here it is:
No decent command-line shell container out-of-the-box (cf. xterm/rxvt/aterm/eterm/konsole etc)
Not true multi-user (one login at once -> cf. VNC+Linux, Xwindow system etc)
No sloppy focus
Taskbar doesn't span multiple monitors (At least on my setup)
taskbar is inflexible as a container for other apps (cf. KDE/gnome taskbar, fluxbox slit, wmaker)
Only one window manager available
No virtual desktops (Out-of-the-box)
No OOTB window roll-up
Only one focus model
That's all available out of the box on Linux. My lappy runs fluxbox cause it's only a PIII-600, while my desktop is a full-blown KDE job. The KDE environment is much more productive than my work WinXP box. Who says Linux isn't ready for the desktop?
I'm not quite so silly to believe that bootlegging in high-powered automobiles was routine behaviour...
Well, another datapoint to this behaviour is of course Dukes of Hazzard, so I'd venture to suggest that it was a little more common than you believe. After all, what could be more real that that show?
Surprisingly, the book (called "Schindler's Ark"), was written by a reasonably famous Australian author, Thomas Kenneally. The story goes that he was in the supermarket/hairdresser/video store/whatever in Melbourne, and this old jewish guy comes up to him and goes: "you're that author, right? Well, have I got a story for you". And that's how an Australian author with no ostensible connection to the holocaust wrote the (true) story.
If you actually go to the link in the parent post, it's Vanuatu. You can do your own research on whether he's telling the truth about that particular country.
Does it really matter that much if the free-as-in-beer tools from Altera and/or Xilinx can synthesise the design? We could have a philosophical discussion for a week about the relative merits of free-as-in-speech vs free-as-in-beer, but at the end of the day that's all that's available for synthesis, so you have to use it...
Yes, yes, yes, all that may be true, however the best thing about gentoo is that you install it a grand total of once. Then, like BSD, you just track current with "emerge -u world", and occasionally change a symlink to point to the latest tree (it even tells you when to do this).
No more installing the latest version when it comes out, taking the machine down, blah blah blah. It all can happen in the background; you don't even have to stop working while the system updates. Or, you can do it overnight.
Of course, a fast machine and distcc both help, but even without those it's not too much of a hassle once the main system is set up.
Speaking of tough call - didn't know whether to mod you up or reply, so decided on replying.
Agree completely about the embedded stuff, there's embedded, and then there's *really* embedded, like what you're talking about.
From what I can gather, though, the embedded development of the future will all happen on microprocessors embedded into FPGAs, which will also serve as multi-peripherals. So, the embedded developer of the future had better know VHDL or verilog, because he'll be customising a processor for his problem domain, then writing the software for his custom processor (+ peripherals). All on a chip that can run at many MHz while consuming a metric dogfart of power. It's exciting times ahead!
For examples of what I'm talking about, see the FPGAs from (eg) Altera (cyclone series) and Xilinx (Spartan 3/E); these are 100,000-plus-gate devices which consume very little power, and come in under around $20 ea in small quantities.
Would it be feasible to just send out all the GOP sets to the parallel processors, then if the GOP being processed relies on the next/previous GOP, re-preocess the GOP required on that processor? It would mean some duplication of effort, but computer time is cheap, and if only a small percentage of sections had to be re-processed, it'd be a net gain.
Slashdot Mirror
So you re-image the computer. That takes all of what, 10 minutes. Just put in the student contract or whatever that if the student fubars the computer, then anything on said computer will be erased. Quite simple, really; the student loses all their mp3s, pr0n, whatever, you get a hassle-free life, everyone wins.
That's a well thought-out decision, however you're kidding yourself if you think that "momentum has nothing to do with it". Momentum has everything to do with it: that is precisely why win2k beats Linux for your roll-out. Linux being too expensive is just another way of saying "we don't have the skill set in-house to implement this", which is a perfectly understandable and valid business decision.
Pronounceable... interesting concept that. One would think that if pronounceability was a major concern then one would choose a slashdot ident with that property...
BTW, what closed-source media playback software do you use? I use mplayer (under linux, admittedly), and I find it much better than anything I have found under windows (without looking too hard).
Understand Lightening? I can't even spell it!
Perhaps a decent measure would be to average the processing out over a week or so; eg each seti@home unit is, say, 1e9 floating point operations, calculate how many units are processed in a week, divide by seconds in a week, there's your number. This method allows for the redundancy of the @home method, ie, each unit will be computed a number of times, if only the completed units are counted it gives a measure of true (sustained) performance.
Cool computer. Tried to reach the web link in your .sig, but it timed out...
Of course, I should probably be running openBSD on my sparc web server, but netBSD had downloadable, bootable ISOs available, and openBSD didn't...
Great script; thanks for the rsync line. BTW, if you wanted to use a for loop, here's one way:
Note that you don't really need to delete the daily.7 dir first; the mv -f should take care of that. Also, I think the $(( ... )) operator might only apply to bash, so you may have to change the !/bin/sh to !/bin/bash at the top of the script.
Another suggestion is to use a version control system to stor your home directory. It's a really good way to keep old versions of stuff, and the bonus is that you only need to backup your repository. I'd suggest subversion, it works for me.
I am replying twice because I have another point to make and didn't want it to get lost in the first post
You wrote:
Linux distros? How many of them make major changes to the kernel? (Honest question, I have no idea, but I'm guessing not many.) How many of them are released by a single person? Again, if you have a company to work on it, awesome. But a single developer would be overwhelmed.
Corollary: The only people who get to work on large, non-trivial programs are members of large teams (because a "single developer would be overwhelmed"). Therefore, in a closed-source-only world, the only non-trivial programs that get written are by large companies, meaning the only ones that get written are the ones that scratch a particular market itch.
The only custom builds I know of IE are, well, umm, none.
And that is exactly the point. Users of IE are forced to upgrade to the MS step; they can't get security fixes without MS say-so, and they can't get enhancements without MS saying so, like, maybe, full CSS2 compliance.
If you are from the USA, you have this wonderful thing called "freedom of speech". It allows you to say pretty much anything in a public forum without lawful consequences (civil law notwithstanding). Now, how many people actually take the opportunity to exercise their freedom of speech rights in a meaningful way? Not many, I'd bet, however the right to speak is an important one. Similarly, the right to modify software, even if the rights are not directly exercised by the vast majority of users, is an important one.
decss?
I found the example xorg.conf file supplied by nvidia to be more than enough to get twinview working. I had triple monitors at one stage, too, by virtue of a PCI geforce mx200 or somesuch, but I suspect that is not fast enough for what the OP asked for.
I believe they did check the code, however the code to stop arbitrary image upload was encumbered by a patent, so they couldn't use it...
That's all available out of the box on Linux. My lappy runs fluxbox cause it's only a PIII-600, while my desktop is a full-blown KDE job. The KDE environment is much more productive than my work WinXP box. Who says Linux isn't ready for the desktop?
Well, another datapoint to this behaviour is of course Dukes of Hazzard, so I'd venture to suggest that it was a little more common than you believe. After all, what could be more real that that show?
Surprisingly, the book (called "Schindler's Ark"), was written by a reasonably famous Australian author, Thomas Kenneally. The story goes that he was in the supermarket/hairdresser/video store/whatever in Melbourne, and this old jewish guy comes up to him and goes: "you're that author, right? Well, have I got a story for you". And that's how an Australian author with no ostensible connection to the holocaust wrote the (true) story.
If you actually go to the link in the parent post, it's Vanuatu. You can do your own research on whether he's telling the truth about that particular country.
Does it really matter that much if the free-as-in-beer tools from Altera and/or Xilinx can synthesise the design? We could have a philosophical discussion for a week about the relative merits of free-as-in-speech vs free-as-in-beer, but at the end of the day that's all that's available for synthesis, so you have to use it...
I don't know the answer to your question, but I also wonder if their FPGA design will also be open-sourced?
No.
No more installing the latest version when it comes out, taking the machine down, blah blah blah. It all can happen in the background; you don't even have to stop working while the system updates. Or, you can do it overnight.
Of course, a fast machine and distcc both help, but even without those it's not too much of a hassle once the main system is set up.
Agree completely about the embedded stuff, there's embedded, and then there's *really* embedded, like what you're talking about.
From what I can gather, though, the embedded development of the future will all happen on microprocessors embedded into FPGAs, which will also serve as multi-peripherals. So, the embedded developer of the future had better know VHDL or verilog, because he'll be customising a processor for his problem domain, then writing the software for his custom processor (+ peripherals). All on a chip that can run at many MHz while consuming a metric dogfart of power. It's exciting times ahead!
For examples of what I'm talking about, see the FPGAs from (eg) Altera (cyclone series) and Xilinx (Spartan 3/E); these are 100,000-plus-gate devices which consume very little power, and come in under around $20 ea in small quantities.
Would it be feasible to just send out all the GOP sets to the parallel processors, then if the GOP being processed relies on the next/previous GOP, re-preocess the GOP required on that processor? It would mean some duplication of effort, but computer time is cheap, and if only a small percentage of sections had to be re-processed, it'd be a net gain.