I have to admit I like Gnome a lot. But even if I were a KDE user I would have to give Gnome credit for their release schedule and process. They come out with a good set of reasonably stable features every 6 months, reliably enough that Ubuntu (and Fedora?) base their distro releases on the Gnome schedule. They've guaranteed core ABI compatibility through the 2.x series, which has been out for 5+ years by my count. They're conservative in what they add and take away. And every release has a nice set of release notes which tells me exactly what to look for in terms of new features.
Software development ain't easy, especially not in the decentralized volunteer world of free software, but the Gnome guys seem to have it down pretty well. Kudos to them.
How are any of these better than using TrueCrypt in traveller mode? The only thing I can think of is that TrueCrypt requires administrator rights to use. And I suppose they may be easier to use for people who don't know much about computers or encryption. But I trust TrueCrypt a hell of a lot more than anything which comes preinstalled on these things.
Well, 4'33" did have sound. The performance consisted of a piano player opening, and later closing, the cover on the keys of a piano, and included whatever ambient noise (fans, cars, people shifting in their seats) was present in the room at the time. A recording of this would be perfectly copyrightable, whereas 4'33" of absolute silence wouldn't be.
Function+arrows produces home, end, pageup, and pagedown. Function+backspace produces delete. Apple shortcuts don't really rely on those keys (they use mostly option/control + arrows instead) but it's definitely annoying if you're used to those keys.
He said 'right', not 'perfect'. Doing DRM 'right' would be a scheme which allows consumers the rights they expect (watching / listening on all of their own devices, able to share with family and friends, no chance of DRM 'expiring' due to format changes or abandonware) while making it reasonably difficult to copy and re-distribute on a large scale.
As long as consumers get the rights they want at a fair price, and as long as it's nontrivial to get an illegal copy, most consumers will opt for the purchased DRM version. iTunes DRM does a pretty good job (though it could be better), its biggest drawback is that it's Apple-only--a standardized DRM format along the lines of iTunes would be a pretty good system overall. People have long since been able to crack iTunes DRM and yet they still buy stuff from iTMS in massive quantities.
Saying DRM is impossible because it can always be cracked is like saying government is impossible because people will break the law. Of course they will. You just have to structure people's incentives so that the majority of people are happy not doing so.
I think it ought to be illegal to change a convicted person's sentence after the fact, e.g. adding 10 more years of jail time or adding restrictions to what they can do when they're released. All terms of the sentence must be determined and specified at the time of sentencing. If NJ wants to do this, then fine, but it can only apply to sex offenders who are convicted after the bill is passed. Or, the state should be required to have a sentencing hearing for every person it wishes to change the sentence for, and allow the person to challenge it before a judge.
What's the difference between applying a blanket sentence to former child molesters and applying a blanket sentence to people of a particular race or political party? Any argument that "one is more right than the other" just boils down to "one is more threatening than the other", and if you allow that argument, then you're basically in a police state.
I'm too lazy to find the quote, but Bruce Schneier said that the NSA's comments on DES effectively started the field of modern academic cryptography, and that many researchers 'made their bones' analyzing why the changes to DES were important. Thus at the time, the NSA was over a decade ahead, but it's very unlikely they're that far ahead now.
Red Hat Enterprise Linux virtualization includes the ability to perform live migration, allowing customers to seamlessly move running applications from one server to another...
How good is the live migration support? Has anyone used it?
You can persuade the program, but you can't persuade the user. The program just needs to ask the user "did this test take 2 seconds or 20 seconds?" The user will know the difference between a 2s test and a 20s test. Of course, if you ask them to distinguish between 2 seconds and 4 seconds, they might have a harder time. And tests like that obviously don't help at all if you don't have a flesh-and-blood user to ask.
I'm no anti-virus expert--but does anyone need anti-virus software anymore? Gone are the days when viruses are spread by floppy. (Mostly) gone are the days when email clients were so brain-dead that they would automatically execute attachments. But most importantly, gone are the days when the main type of infection is viruses (which spread via some sort of user action). These days, worms (which require no user action) are the dominant threat. And anti-virus software, which relies on signatures, is nearly useless against worms which (by their automatic nature) spread far too quickly for even automatic signature updates to catch. Furthermore, worms generally cause most of their havoc just by spreading (and clogging the network), and by infecting PCs to use as bot-farms.
Perhaps I'm just isolated from the sort of users who are so stupid as to get viruses on their PCs...but are there any left? And does anti-virus software help these people?
Yes, but you can compare the local CPU clock with external clocks. If the CPU claims that the timing test you execute took 2 seconds, but 20 seconds have elapsed according to an external clock, then you know something is amiss.
The external clock doesn't even have to be accessed directly. The testing app could run a test and ask the user if it seemed to take 2 seconds or 20 seconds. I don't think a CPU can skew a human's perception of time...
I bought my iPhone on 8/24, which is 12 days before the annoucement on 9/5. It's great that I can get a $100 store credit, but do I qualify for the full $200 rebate (the standard Apple offer)? Is the window for that rebate 10 days or 14 days?
Excuse me, but asking legions of VB developers about VB 6 is like asking soccer moms about the fuel-injection system on their new minivan. VB.net is a sane language, VB 6 and before are steaming piles of codeshit. Anyone who sticks to VB 6 by choice has no clue what a language should be.
I work on an MMO which is also being published by SOE Platform Publishing. They are not involved at all in game design or implementation--they will be doing publishing only, i.e. box sales, marketing. In my opinion they're a very good choice for publishing MMOs.
If the game is good or bad, it's the developer's fault, not the publisher's. Don't lump games that are published by SOE in with games that are developed by SOE.
Jobs is probably not happy about his thunder being stolen right before for the June 11th keynote
I strongly doubt he didn't know about it. This is Jonathan Schwartz, not a OS X rumors blogger. At any rate, ZFS in OS X is Sun's thunder; Time Machine is Apple's thunder, and that's already announced. How many OS X users (other than slashdot readers) will care in the slightest about the underlying filesystem? What they care about are the features, like Time Machine, that it enables.
Not for drivers, it's not. The Linux userspace API is extremely stable, and the ABI somewhat less so. But anything inside the kernel can change drastically. That's why Linux hackers want all drivers in the kernel tree, so they can find anything which breaks due to an API change, and fix the problems.
Huh? Do you prepare your statements every time you execute them? That would certainly be slower, but if you prepare the statement once and execute it many times, the local overhead becomes minimal and the execution time in the database improves because the DB doesn't have to re-parse and re-plan the query.
The MMO industry is shaping up to be much like the movie industry. There's a ton of money to be made, and everyone knows it, and everyone wants a piece. But making a blockbuster, or even breaking-even, is HARD. Really hard. And expensive. And so the only way to be profitable is to make a lot of them, some good and some bad, and hope you come out ahead.
Worse, at least the movie business is rather mature. There are lots of people who know what they're doing, more or less. The MMO business is in its infancy. It's as if movies had been invented in 1970, then Jaws comes out in 1976, and you have a dozen production companies striving to reproduce that one huge success.
In this day and age, just getting an MMO out the door is basically a success.
It's a hard road that doesn't pay very well and has very poor job security. It's mindless, grueling work.
But I know several game designers who got their start in QA and used it to prove that they knew how to make good games. If you really do think you would excel as a game designer, if you're a bright, motivated person who can distinguish themselves from the other testers, and you can't find a way into the industry another way, it's not a bad way to go. If all this applies, and you live in the bay area, I know at least one company that would be interested in hiring you.
It's popular because it's designed to be embedded. Other scripting languages are designed around being run as the top-most context of executiong, e.g. being run from the command line or when a user requests a web page. Lua is very easy to embed in your own app, so that for example you can write a script which can access the variables and functions of your running application. This is extremely useful for games in particular.
You can easily hook up a Lua method to access a variable in your containing application. More importantly, any C function that you bind to a Lua method can return a special code which indicates that the function is "yielding". When that happens, the Lua interpreter pauses the script and returns to your application (the point at which you invoked the interpreter). Then later, you can "resume" the Lua script by providing the actual return value to the Lua method, which then picks up as if nothing had happened.
Python can do something similar with coroutines, but again, it's not designed (as much) around being embedded so they're not as easy to use in that way.
You're certainly right that shared memory is more of a pain in the ass than threads, but two of your points are wrong.
1) you can memory-map files instead of using SysV shmget.
3) you can put pthread mutexes and condition variables in shared memory instead of using SysV semaphores.
Your point #2 is correct, but that's the same thing the grandparent was saying--you get memory protection, meaning that you have to explicitly share data. The biggest pain is that you can't put pointers into shared memory since it will be mapped to different locations in different processes, so you have to use offsets and calculate pointer values per-process.
Creating processes on Linux is exactly the same overhead as threads (they both use the same syscall, clone()) and both are comparable in performance to creating threads on Windows and Solaris.
Shared memory isn't perfect for every problem (go figure, what is?), and are definitely more difficult to use, but they do offer better protection and robustness.
I'm a lead server developer at a MMOG company in the Bay Area. What do we want? We want talented programmers. If you can hack it, then education, experience, resume, all that crap is immaterial. The most important stuff you have to teach yourself. Learn what's out there, play with it, use it, fix it, rewrite it. If you know it, the job is yours. And by job, I mean that literally, since we're actively hiring. I'm not very hard to find.
Slashdot Mirror
I have to admit I like Gnome a lot. But even if I were a KDE user I would have to give Gnome credit for their release schedule and process. They come out with a good set of reasonably stable features every 6 months, reliably enough that Ubuntu (and Fedora?) base their distro releases on the Gnome schedule. They've guaranteed core ABI compatibility through the 2.x series, which has been out for 5+ years by my count. They're conservative in what they add and take away. And every release has a nice set of release notes which tells me exactly what to look for in terms of new features.
Software development ain't easy, especially not in the decentralized volunteer world of free software, but the Gnome guys seem to have it down pretty well. Kudos to them.
How are any of these better than using TrueCrypt in traveller mode? The only thing I can think of is that TrueCrypt requires administrator rights to use. And I suppose they may be easier to use for people who don't know much about computers or encryption. But I trust TrueCrypt a hell of a lot more than anything which comes preinstalled on these things.
Well, 4'33" did have sound. The performance consisted of a piano player opening, and later closing, the cover on the keys of a piano, and included whatever ambient noise (fans, cars, people shifting in their seats) was present in the room at the time. A recording of this would be perfectly copyrightable, whereas 4'33" of absolute silence wouldn't be.
Function+arrows produces home, end, pageup, and pagedown. Function+backspace produces delete. Apple shortcuts don't really rely on those keys (they use mostly option/control + arrows instead) but it's definitely annoying if you're used to those keys.
He said 'right', not 'perfect'. Doing DRM 'right' would be a scheme which allows consumers the rights they expect (watching / listening on all of their own devices, able to share with family and friends, no chance of DRM 'expiring' due to format changes or abandonware) while making it reasonably difficult to copy and re-distribute on a large scale.
As long as consumers get the rights they want at a fair price, and as long as it's nontrivial to get an illegal copy, most consumers will opt for the purchased DRM version. iTunes DRM does a pretty good job (though it could be better), its biggest drawback is that it's Apple-only--a standardized DRM format along the lines of iTunes would be a pretty good system overall. People have long since been able to crack iTunes DRM and yet they still buy stuff from iTMS in massive quantities.
Saying DRM is impossible because it can always be cracked is like saying government is impossible because people will break the law. Of course they will. You just have to structure people's incentives so that the majority of people are happy not doing so.
I think it ought to be illegal to change a convicted person's sentence after the fact, e.g. adding 10 more years of jail time or adding restrictions to what they can do when they're released. All terms of the sentence must be determined and specified at the time of sentencing. If NJ wants to do this, then fine, but it can only apply to sex offenders who are convicted after the bill is passed. Or, the state should be required to have a sentencing hearing for every person it wishes to change the sentence for, and allow the person to challenge it before a judge.
What's the difference between applying a blanket sentence to former child molesters and applying a blanket sentence to people of a particular race or political party? Any argument that "one is more right than the other" just boils down to "one is more threatening than the other", and if you allow that argument, then you're basically in a police state.
I'm too lazy to find the quote, but Bruce Schneier said that the NSA's comments on DES effectively started the field of modern academic cryptography, and that many researchers 'made their bones' analyzing why the changes to DES were important. Thus at the time, the NSA was over a decade ahead, but it's very unlikely they're that far ahead now.
Red Hat Enterprise Linux virtualization includes the ability to perform live migration, allowing customers to seamlessly move running applications from one server to another...
How good is the live migration support? Has anyone used it?
Is this based on Xen or something else?
Perhaps it's Rama?
You can persuade the program, but you can't persuade the user. The program just needs to ask the user "did this test take 2 seconds or 20 seconds?" The user will know the difference between a 2s test and a 20s test. Of course, if you ask them to distinguish between 2 seconds and 4 seconds, they might have a harder time. And tests like that obviously don't help at all if you don't have a flesh-and-blood user to ask.
I'm no anti-virus expert--but does anyone need anti-virus software anymore? Gone are the days when viruses are spread by floppy. (Mostly) gone are the days when email clients were so brain-dead that they would automatically execute attachments. But most importantly, gone are the days when the main type of infection is viruses (which spread via some sort of user action). These days, worms (which require no user action) are the dominant threat. And anti-virus software, which relies on signatures, is nearly useless against worms which (by their automatic nature) spread far too quickly for even automatic signature updates to catch. Furthermore, worms generally cause most of their havoc just by spreading (and clogging the network), and by infecting PCs to use as bot-farms.
Perhaps I'm just isolated from the sort of users who are so stupid as to get viruses on their PCs...but are there any left? And does anti-virus software help these people?
Yes, but you can compare the local CPU clock with external clocks. If the CPU claims that the timing test you execute took 2 seconds, but 20 seconds have elapsed according to an external clock, then you know something is amiss.
The external clock doesn't even have to be accessed directly. The testing app could run a test and ask the user if it seemed to take 2 seconds or 20 seconds. I don't think a CPU can skew a human's perception of time...
I think rockbox will overcome this, because the check for the hash is done in the default firmware, which rockbox replaces.
I bought my iPhone on 8/24, which is 12 days before the annoucement on 9/5. It's great that I can get a $100 store credit, but do I qualify for the full $200 rebate (the standard Apple offer)? Is the window for that rebate 10 days or 14 days?
Excuse me, but asking legions of VB developers about VB 6 is like asking soccer moms about the fuel-injection system on their new minivan. VB.net is a sane language, VB 6 and before are steaming piles of codeshit. Anyone who sticks to VB 6 by choice has no clue what a language should be.
I work on an MMO which is also being published by SOE Platform Publishing. They are not involved at all in game design or implementation--they will be doing publishing only, i.e. box sales, marketing. In my opinion they're a very good choice for publishing MMOs.
If the game is good or bad, it's the developer's fault, not the publisher's. Don't lump games that are published by SOE in with games that are developed by SOE.
Jobs is probably not happy about his thunder being stolen right before for the June 11th keynote
I strongly doubt he didn't know about it. This is Jonathan Schwartz, not a OS X rumors blogger. At any rate, ZFS in OS X is Sun's thunder; Time Machine is Apple's thunder, and that's already announced. How many OS X users (other than slashdot readers) will care in the slightest about the underlying filesystem? What they care about are the features, like Time Machine, that it enables.
Not for drivers, it's not. The Linux userspace API is extremely stable, and the ABI somewhat less so. But anything inside the kernel can change drastically. That's why Linux hackers want all drivers in the kernel tree, so they can find anything which breaks due to an API change, and fix the problems.
Huh? Do you prepare your statements every time you execute them? That would certainly be slower, but if you prepare the statement once and execute it many times, the local overhead becomes minimal and the execution time in the database improves because the DB doesn't have to re-parse and re-plan the query.
The MMO industry is shaping up to be much like the movie industry. There's a ton of money to be made, and everyone knows it, and everyone wants a piece. But making a blockbuster, or even breaking-even, is HARD. Really hard. And expensive. And so the only way to be profitable is to make a lot of them, some good and some bad, and hope you come out ahead.
Worse, at least the movie business is rather mature. There are lots of people who know what they're doing, more or less. The MMO business is in its infancy. It's as if movies had been invented in 1970, then Jaws comes out in 1976, and you have a dozen production companies striving to reproduce that one huge success.
In this day and age, just getting an MMO out the door is basically a success.
It's a hard road that doesn't pay very well and has very poor job security. It's mindless, grueling work.
But I know several game designers who got their start in QA and used it to prove that they knew how to make good games. If you really do think you would excel as a game designer, if you're a bright, motivated person who can distinguish themselves from the other testers, and you can't find a way into the industry another way, it's not a bad way to go. If all this applies, and you live in the bay area, I know at least one company that would be interested in hiring you.
It's popular because it's designed to be embedded. Other scripting languages are designed around being run as the top-most context of executiong, e.g. being run from the command line or when a user requests a web page. Lua is very easy to embed in your own app, so that for example you can write a script which can access the variables and functions of your running application. This is extremely useful for games in particular.
You can easily hook up a Lua method to access a variable in your containing application. More importantly, any C function that you bind to a Lua method can return a special code which indicates that the function is "yielding". When that happens, the Lua interpreter pauses the script and returns to your application (the point at which you invoked the interpreter). Then later, you can "resume" the Lua script by providing the actual return value to the Lua method, which then picks up as if nothing had happened.
Python can do something similar with coroutines, but again, it's not designed (as much) around being embedded so they're not as easy to use in that way.
You're certainly right that shared memory is more of a pain in the ass than threads, but two of your points are wrong.
1) you can memory-map files instead of using SysV shmget.
3) you can put pthread mutexes and condition variables in shared memory instead of using SysV semaphores.
Your point #2 is correct, but that's the same thing the grandparent was saying--you get memory protection, meaning that you have to explicitly share data. The biggest pain is that you can't put pointers into shared memory since it will be mapped to different locations in different processes, so you have to use offsets and calculate pointer values per-process.
Creating processes on Linux is exactly the same overhead as threads (they both use the same syscall, clone()) and both are comparable in performance to creating threads on Windows and Solaris.
Shared memory isn't perfect for every problem (go figure, what is?), and are definitely more difficult to use, but they do offer better protection and robustness.
I'm a lead server developer at a MMOG company in the Bay Area. What do we want? We want talented programmers. If you can hack it, then education, experience, resume, all that crap is immaterial. The most important stuff you have to teach yourself. Learn what's out there, play with it, use it, fix it, rewrite it. If you know it, the job is yours. And by job, I mean that literally, since we're actively hiring. I'm not very hard to find.
How did you get PG&E to do the energy check? My two-bedroom apartment with gas heat is around $100 per month.