80 columns was created as a formatting standard because of punchcards.
Actually, I'm pretty sure the origin of the 80 column standard is that that's how many characters you get on A4 paper portrait with 2cm margins and 12pt Courier. (And, I believe, "Letter" paper with 1 inch margins is the same, although perhaps some US hacker will correct me on that one...)
One of my beefs with Java is that it seems impossible to write comfortable code in less than 120 columns.
I see no reason why you should have a problem with that. I write Java in my job, and rarely exceed my editor window's default width, which is 94 columns. Are you using 8 column indents, by any chance?
I have used a lot of editors over the years. And I've never used one that doesn't have a short key sequence for cycling through open buffers. It might not be alt-tab, but there is always something. My favourites were Borland's windows IDEs, which used to assign a numerical shortcut keys to the first 10 documents open at any time.
As a VB coder [...] If I went much over 60 chars per line
Last time I used VB, the IDE whined at you if you pressed enter before you had a complete statement (as it tried to compile it and got a syntax error), so formatting to any width other than "a complete statement on each line" was impossible. I take it this has been fixed in more recent versions?
Right... that just puts copyright owning artists even more at the mercy of publishing corporations.
Just because RIAA recording companies require their artists to produce 'work for hire' and take the copyright from them, not every publishing corporation treats its content generators so badly. Chances are that the copyright of the last book you read is owned by its author.
One of the interesting things about the book publishing industry is it has a concept similar, but much fairer than what you're talking about: if an author licenses his book to a publisher, but the publisher doesn't offer copies of it to the public for some period of time (usually about 3 years, I think) then the license is automatically terminated, meaning that the author is free to sell a new license (usually to a discount paperback publisher). This is a good system which works well.
So, I think what needs to happen is that the 'work for hire' system needs to be substantially modified. Some kind of protection needs to be put into place so that independent content producers don't have their copyrights screwed out of them by abusive monopolistic publishers. Then the artists would be free to license their own work on reasonable terms, and to revoke those licenses if the publisher isn't representing them.
See, for instance, on a Windows 2000 machine, Control Panel / Administrative Tools / Local Security Policies / Account Policies / Account Lockout Policies.
RAID-0 is not appropriate for general desktop use. You use it in applications which are disk-bound. The classic examples of this are databases and video editing.
Hate to point this out, but video editing is rapidly becoming 'general desktop use'. Admittedly, most desktop video editors use lossy compression formats with low bitrates that wouldn't benefit from RAID0, but I suspect the average bitrate is going to increase with storage capacity. People always want higher quality.
Even under perceived heavy I/O loads, the reality is often that the hard disk is under-used - I occasionally compress videos from miniDV to DVD, and my CPU would need a four or five fold increase in speed to even begin to put pressure on the single 7200 RPM hard disk.
Have you tried capturing video from a TV tuner card using a lossless codec (e.g. huffyuv) at high resolution (e.g. 720x576x25fps, PAL DVD resolution)? I'm not _certain_ that this would require a RAID0 array, but I know from experience that a 5400RPM hard disk can't keep up with it, and my benchmarks suggest that a single 7200 couldn't. But I know that two 5400s can.
Huh? Most people can remember 10 random characters. It takes a while to learn them (probably 20 minutes or so), but there are very few people who can't do it. The admin password for my company's web server, for instance, is a 10 character random password (case-sensitive alphanumeric), and none of our employees who require access have had trouble learning it.
The space shuttle doesn't need that much for its automated stuff.
Well, yeah. But the space shuttle runs on hardware designed in the 1970s. Building and maintaining the software they need to run on this costs $100 million per annum, according to this site.
I suspect if they were less constrained in their hardware, it would reduce that cost, at least a little. But that would require a reevaluation of the hardware from an engineering perspective, and would probably cost even more. NASA spends a lot of money on safety, and very tight, critically analysed software (probably audited at an assembly language level) is the only way they can achieve that. But it means a lot of extra work for them.
Why would it need a kernel hack? The fact that they managed to use XP embedded suggests that their hardware environment was pretty standard. You can't easily hack the XP kernel -- Microsoft only release the source code under very strict controls, and the NDA attached would probably be something that a university wouldn't want to get involved with.
Also, there's nothing stopping them from releasing their C# code. I suspect they're just as likely to do so in this case as they would if they'd written it under a Linux system.
The answer would have been to use something that only a few people are big fans of. QNX springs to mind, and would have been a more appropriate system than Linux, too.
I believe its primary design function is to serve mail/groupware data to users who don't have necessarily have a login on the server -- this is certainly the most common usage scenario for groupware. In this case, it will probably require all e-mail to be stored in a central database. It's the only way it can realistically do ACL type stuff anyway.
Besides, you'll be accessing your mail via IMAP. Why do you care where its stored?
Anyone else find it interesting that Gamma, Helm, Johnson and Vlissides (Design Patterns) is cited more frequently than any individual volume of Knuth's Art of Computer Programming, yet the latter is generally considered a more important book?
There are dozens of RAM defragmentation software for PC; forget them all. Mem Doubler has a unique function which is able to determine when your RAM needs to be defragmented! Just tick a checkbox, and Mem Doubler will adapt to your computing style.
Excuse me? RAM defragmentation? They've got to be joking. There are two possible kinds of fragmentation, external and internal. External is irrelevant with a paged memory manager (i.e. on any modern CPU). Internal can only be solved with cooperation of the applications.
Our interface fully respects Microsoft's guidelines and is very intuitive.
That's why in the screenshot of one of their dialog boxes they have buttons ordered 'Quit', 'OK', where the MS guidelines are clearly that they should be 'OK', 'Close'.
If MS guidelines are so important, where's their conformance testing results? What, no Windows logo?
Because triple buffering is a buzzword from modern graphics processing: you have one buffer being displayed, one buffer for the hardware renderer to be working on and one where the CPU is preparing data for the hardware renderer. They've just taken this and applied it to an area where it doesn't make any sense, probably because the software does nothing other than a few flashy graphics to convince users that it really has made a difference.
Login to a Windows XP machine and it can be up to 30 seconds before the harddrive quiets down enough to get any work done. Yeah, the Windows GUI feels pretty zippy once it is loaded, but I can't tell you how much even the occassional hourglass outside of the working app pisses me off. And what is with that hourglass/arrow combination pointer? Is the computer too busy or not? Make up your mind!
Log in to a KDE/Linux machine (or at least mine) and it can be 20 seconds before the desktop is displayed. At least with Windows XP you can actually do useful work while the background stuff is loading.
Why would an hourglass icon from an application you're not currently using disturb you? That seems a little strange to me. If you're not using it, why do you even pay attention to its existance?
Here's a point you seem to have missed: the hourglass cursor is controlled by the application, not the operating system. The application tells windows to use it when it is busy & therefore can't respond to user requests. Windows also uses it if an application doesn't have a predefined cursor and isn't responding to events.
Oh, and the hourglass/arrow is a feedback indicator. It's telling you that the computer has responded to your request, but you can now go on to do something else if you want. The lack of a similar system in Linux desktops is really annoying, I find. KDE has an almost acceptable approach where a box is displayed on the panel to indicate an application is starting, but it isn't nearly as good (because the hourglass arrow cursor is an OS level feature of windows, so can achieve things KDE/Gnome/whatever couldn't achieve without a kernel patch).
I think Microsoft has made the GUI zippy at the expense of proper multitasking.
Windows multitasks as well as Linux for me. I don't understand why you have a problem with it.
Does gentoo do prelinking? If so, that's probably what did it for you -- prelinking has been the only real speed advantage windows has had over linux for quite a while. I've not set up a prelinked distribution for myself yet, but looking at the benchmarks of where my current system is spending its time and how much benefit prelinking should get, it should take my KDE performance from slower than windows to faster.
There's a reason this isn't being reported... what do you expect? The game is the event that, more than any other, gamers have been waiting for for _years_. And a very high proportion of pirates are gamers. Of course they're all going to be trading copies of Doom 3. I'd be shocked if they weren't.
128 Mb works fine on my WinXP machine, with little or no swapping. If you need more than that, I'd blame the applications you're running, not the operating system.
Slashdot Mirror
80 columns was created as a formatting standard because of punchcards.
Actually, I'm pretty sure the origin of the 80 column standard is that that's how many characters you get on A4 paper portrait with 2cm margins and 12pt Courier. (And, I believe, "Letter" paper with 1 inch margins is the same, although perhaps some US hacker will correct me on that one...)
One of my beefs with Java is that it seems impossible to write comfortable code
in less than 120 columns.
I see no reason why you should have a problem with that. I write Java in my job, and rarely exceed my editor window's default width, which is 94 columns. Are you using 8 column indents, by any chance?
I have used a lot of editors over the years. And I've never used one that doesn't have a short key sequence for cycling through open buffers. It might not be alt-tab, but there is always something. My favourites were Borland's windows IDEs, which used to assign a numerical shortcut keys to the first 10 documents open at any time.
As a VB coder [...] If I went much over 60 chars per line
Last time I used VB, the IDE whined at you if you pressed enter before you had a complete statement (as it tried to compile it and got a syntax error), so formatting to any width other than "a complete statement on each line" was impossible. I take it this has been fixed in more recent versions?
Right... that just puts copyright owning artists even more at the mercy of publishing corporations.
Just because RIAA recording companies require their artists to produce 'work for hire' and take the copyright from them, not every publishing corporation treats its content generators so badly. Chances are that the copyright of the last book you read is owned by its author.
One of the interesting things about the book publishing industry is it has a concept similar, but much fairer than what you're talking about: if an author licenses his book to a publisher, but the publisher doesn't offer copies of it to the public for some period of time (usually about 3 years, I think) then the license is automatically terminated, meaning that the author is free to sell a new license (usually to a discount paperback publisher). This is a good system which works well.
So, I think what needs to happen is that the 'work for hire' system needs to be substantially modified. Some kind of protection needs to be put into place so that independent content producers don't have their copyrights screwed out of them by abusive monopolistic publishers. Then the artists would be free to license their own work on reasonable terms, and to revoke those licenses if the publisher isn't representing them.
Thank you. And even better, include an option to play tracks from my MP3 collection.
Why do you describe this idea as 'new'?
See, for instance, on a Windows 2000 machine, Control Panel / Administrative Tools / Local Security Policies / Account Policies / Account Lockout Policies.
If its an example password, why doesn't it say in big letters "EXAMPLE - DO NOT USE THIS PASSWORD ON YOUR SYSTEM"? It should be clearer.
RAID-0 is not appropriate for general desktop use. You use it in applications which are disk-bound. The classic examples of this are databases and video editing.
Hate to point this out, but video editing is rapidly becoming 'general desktop use'. Admittedly, most desktop video editors use lossy compression formats with low bitrates that wouldn't benefit from RAID0, but I suspect the average bitrate is going to increase with storage capacity. People always want higher quality.
Even under perceived heavy I/O loads, the reality is often that the hard disk is under-used - I occasionally compress videos from miniDV to DVD, and my CPU would need a four or five fold increase in speed to even begin to put pressure on the single 7200 RPM hard disk.
Have you tried capturing video from a TV tuner card using a lossless codec (e.g. huffyuv) at high resolution (e.g. 720x576x25fps, PAL DVD resolution)? I'm not _certain_ that this would require a RAID0 array, but I know from experience that a 5400RPM hard disk can't keep up with it, and my benchmarks suggest that a single 7200 couldn't. But I know that two 5400s can.
Huh? Most people can remember 10 random characters. It takes a while to learn them (probably 20 minutes or so), but there are very few people who can't do it. The admin password for my company's web server, for instance, is a 10 character random password (case-sensitive alphanumeric), and none of our employees who require access have had trouble learning it.
Sorry, but at GBP 50 per token, plus about GBP 2500 for an authorisation server, SecurID is too expensive for many of us.
The space shuttle doesn't need that much for its automated stuff.
Well, yeah. But the space shuttle runs on hardware designed in the 1970s. Building and maintaining the software they need to run on this costs $100 million per annum, according to this site.
I suspect if they were less constrained in their hardware, it would reduce that cost, at least a little. But that would require a reevaluation of the hardware from an engineering perspective, and would probably cost even more. NASA spends a lot of money on safety, and very tight, critically analysed software (probably audited at an assembly language level) is the only way they can achieve that. But it means a lot of extra work for them.
Why would it need a kernel hack? The fact that they managed to use XP embedded suggests that their hardware environment was pretty standard. You can't easily hack the XP kernel -- Microsoft only release the source code under very strict controls, and the NDA attached would probably be something that a university wouldn't want to get involved with.
Also, there's nothing stopping them from releasing their C# code. I suspect they're just as likely to do so in this case as they would if they'd written it under a Linux system.
The answer would have been to use something that only a few people are big fans of. QNX springs to mind, and would have been a more appropriate system than Linux, too.
I believe its primary design function is to serve mail/groupware data to users who don't have necessarily have a login on the server -- this is certainly the most common usage scenario for groupware. In this case, it will probably require all e-mail to be stored in a central database. It's the only way it can realistically do ACL type stuff anyway.
Besides, you'll be accessing your mail via IMAP. Why do you care where its stored?
Anyone else find it interesting that Gamma, Helm, Johnson and Vlissides (Design Patterns) is cited more frequently than any individual volume of Knuth's Art of Computer Programming, yet the latter is generally considered a more important book?
There are dozens of RAM defragmentation software for PC; forget them all. Mem Doubler has a unique function which is able to determine when your RAM needs to be defragmented! Just tick a checkbox, and Mem Doubler will adapt to your computing style.
Excuse me? RAM defragmentation? They've got to be joking. There are two possible kinds of fragmentation, external and internal. External is irrelevant with a paged memory manager (i.e. on any modern CPU). Internal can only be solved with cooperation of the applications.
Our interface fully respects Microsoft's guidelines and is very intuitive.
That's why in the screenshot of one of their dialog boxes they have buttons ordered 'Quit', 'OK', where the MS guidelines are clearly that they should be 'OK', 'Close'.
If MS guidelines are so important, where's their conformance testing results? What, no Windows logo?
Because triple buffering is a buzzword from modern graphics processing: you have one buffer being displayed, one buffer for the hardware renderer to be working on and one where the CPU is preparing data for the hardware renderer. They've just taken this and applied it to an area where it doesn't make any sense, probably because the software does nothing other than a few flashy graphics to convince users that it really has made a difference.
No, you're right, there is nothing. And the SSE registers, being 128 bits, would be more appropriate for this purpose anyway.
True. But executing a HLT to wait for the next interrupt is usually better.
Login to a Windows XP machine and it can be up to 30 seconds before the harddrive quiets down enough to get any work done. Yeah, the Windows GUI feels pretty zippy once it is loaded, but I can't tell you how much even the occassional hourglass outside of the working app pisses me off. And what is with that hourglass/arrow combination pointer? Is the computer too busy or not? Make up your mind!
Log in to a KDE/Linux machine (or at least mine) and it can be 20 seconds before the desktop is displayed. At least with Windows XP you can actually do useful work while the background stuff is loading.
Why would an hourglass icon from an application you're not currently using disturb you? That seems a little strange to me. If you're not using it, why do you even pay attention to its existance?
Here's a point you seem to have missed: the hourglass cursor is controlled by the application, not the operating system. The application tells windows to use it when it is busy & therefore can't respond to user requests. Windows also uses it if an application doesn't have a predefined cursor and isn't responding to events.
Oh, and the hourglass/arrow is a feedback indicator. It's telling you that the computer has responded to your request, but you can now go on to do something else if you want. The lack of a similar system in Linux desktops is really annoying, I find. KDE has an almost acceptable approach where a box is displayed on the panel to indicate an application is starting, but it isn't nearly as good (because the hourglass arrow cursor is an OS level feature of windows, so can achieve things KDE/Gnome/whatever couldn't achieve without a kernel patch).
I think Microsoft has made the GUI zippy at the expense of proper multitasking.
Windows multitasks as well as Linux for me. I don't understand why you have a problem with it.
Does gentoo do prelinking? If so, that's probably what did it for you -- prelinking has been the only real speed advantage windows has had over linux for quite a while. I've not set up a prelinked distribution for myself yet, but looking at the benchmarks of where my current system is spending its time and how much benefit prelinking should get, it should take my KDE performance from slower than windows to faster.
There's a reason this isn't being reported... what do you expect? The game is the event that, more than any other, gamers have been waiting for for _years_. And a very high proportion of pirates are gamers. Of course they're all going to be trading copies of Doom 3. I'd be shocked if they weren't.
128 Mb works fine on my WinXP machine, with little or no swapping. If you need more than that, I'd blame the applications you're running, not the operating system.