While there aren't any guarantees, Linux tends to oops rather than panic when a driver screws up; the effects of a kernel oops are usually more localized.
Also, Christopher Tolkien vigorously decried the movies with incredible vehemence -- so much that he even was willing to effectively disown his own son over the issue.
If Tolkien's heirs still had the movie rights, the movies would never have been made.
Also, as far as porting Linux applications to Windows goes, in my experience as a project leader of Inkscape, it's been worth it.
Not only has the Windows version attracted a lot of users and developers, it's also helped migrate a number of them to Linux.
The best way to migrate most people to Linux is not to throw them in the deep end of the pool and hope they learn to swim. A gradual transition works better.
Once all the apps they use on a regular basis are ones which are available on Linux, most people are pretty happy to switch.
That might sound a little naive, but from experience that's the way it works.
It does, however, require that the free applications be better than their proprietary equivalents (for the user's specific needs, at least). We can do that, right?
He is unusual, sadly. It's apparent that you haven't learned this area yet either.
Big-O is a measure of complexity that tells you the worst-case efficiency of a piece of code -- and you don't need a slide rule or a calculator, becuause the answer isn't a number, it's a formula you can determine by looking at the structure of your program.
Knowing how different "types" of formulas compare to one another, it helps you choose and optimize your algorithms.
Being aware of complexity issues ("...is this algorithm O(1)? O(log n)? O(n^2)?") really is essential to writing clean, efficient code -- particularly code that scales well. This applies to almost all non-trivial programs.
(it's also important to know what Big O does not mean -- e.g. when is bubble sort consistently among the fastest sort algorithms? it is if you know your list is very short or mostly sorted beforehand!)
Even if you don't get a formal CS education, it behooves you to learn this stuff. Hit places like Wikipedia and the Portland Pattern Repository Wiki when you see terms like "Big O" mentioned. Trust me, it'll make you a drastically better programmer.
(The Wikipedia article is here: Big O Notation, though it does assume a certain level of math background. Google might be good for finding more "layman's" explanations to get you started.)
While Mono is doing an implementation of Microsoft's extensions to the ECMA spec, they're also doing their own set in parallel.
If all they did was cloning, of course the best they could ever hope for was barely keeping up.
This means that if Microsoft torpedoes the.NET clone stuff, Mono still has a viable system built on top of the ECMA standard (Gtk#, etc...) that they've been encouraging people to target all along.
Note whose APIs Ximian is writing their apps to... they aren't Microsoft's...
I used to think Miguel was naive. Now I think he's a really shrewd bastard... They got Microsoft's support and then pulled an "embrace and extend" on MICROSOFT.
Well, the thing is, although they are implementing Microsoft's extensions, in a parallel effort, Mono's building their own set of standard libraries on top of the ECMA standard (which is what they're actually using for apps). This poster puts it pretty clearly.
There isn't any competition for the base ECMA stuff; it's shared by Microsoft's libraries/extensions and Mono's. The competition is between the two sets of standard libraries themselves.
However, I'm betting you've never seen programmers try to use inheritance to express relationships of a "has a" (or "other") nature, because they haven't thought it through clearly (the "if all you have is a hammer..." syndrome).
So, it's helpful to train programmers to think about the "is a"/"has a"/etc relationship up front, rather than just kind of fuzzily using inheritance because they know the classes are related (and then maybe deciding that there is an "is a" relationship ex post facto).
Actually there are in part security motivations for that (e.g. look at the reasons Java's String class is final).
It's better to be a little too agressive with final than not enough (they can safely back off later, but you can't make things final that weren't before without causing massive breakage).
Also, given C# has good language support for non-inheritance aggregation/delegation, all those final classes aren't as painful as they would be in Java (provided you aren't trying to write Java code in C#).
Most historic ls(1) implementations do munging of nonprintable and high-ascii characters (c.f. the -b and -q options) -- they would have had to disable that to make it UTF-8 friendly, and that might break shell scripts if they did it by default.
I don't have my laptop in front of me right now, but perhaps they added an ls option to turn off the munging.
except instead of haveing the small team of astronauts having to deal with this, you will have a large room filled with the ever so bright people from NASA
...and really, really lousy ping times. The latency makes improvisation nearly impossible (or at best very costly).
Ah, excellent. The last time I did serious (GUI) Windows programming was in the flaky Win95 (OSR1) era, where GDI resources weren't always freed in my experience... (and obviously they never were on Win16). On early Win95 it may have been that there was still a lot of 16 bit stuff going on behind the scenes; I'm not sure.
I did think you were complaining that memory was being leaked after a crash, though...
If the memory a process allocated isn't unconditionally freed when the process exits (crash or no), that's an OS bug, not a bug in the application.
It's probably GDI resources and things, though, not memory per se. The Windows GDI design is such that there is no way a process' GDI resources to be freed if the process crashes.
Keep in mind also that Plato talked about Atlantis in The Republic, the same book where he discusses a magic ring that makes people invisible. But no one goes searching for the Ring of Gyges.
Well duh! That's because everyone knows Frodo destroyed it.
I don't intend to dismiss your anger at having your country messed with, because you should be angry.
But what safeguards against the abuse of power would a world government have? I don't think abuses of power have ever been limited to particular religions or economic system; why would it be different this time?
Look at what's currently happening in the EU with software patents (for example). No rabid Christians or Turbo-Capitalists involved (well, maybe Turbo-Capitalists).
What prevents a meta-government (whether the US [remember, it also started out as a union of sovereign states!], the EU, or a hypothetical world government) from behaving in equally high-handed ways in other areas of policy, or falling under the influence of unsavory politicians?
Slashdot Mirror
While there aren't any guarantees, Linux tends to oops rather than panic when a driver screws up; the effects of a kernel oops are usually more localized.
Regarding multiplexing -- doesn't WaitForMultipleObjects work?
Not that there aren't some really stupid inconsistencies, but I don't think this is one of them.
Also, Christopher Tolkien vigorously decried the movies with incredible vehemence -- so much that he even was willing to effectively disown his own son over the issue.
If Tolkien's heirs still had the movie rights, the movies would never have been made.
Bleah. Yes. One of my big hollywood pet peeves.
That was one of the things that killed the scene at the end of RotK for me.
I don't see much of a problem with that myself either.
Worse, however, is the pernicious phenomenon of the "book based on the film based on the book".
Screenplay novelisations of films based on books are invariably rank.
Also, as far as porting Linux applications to Windows goes, in my experience as a project leader of Inkscape, it's been worth it.
Not only has the Windows version attracted a lot of users and developers, it's also helped migrate a number of them to Linux.
The best way to migrate most people to Linux is not to throw them in the deep end of the pool and hope they learn to swim. A gradual transition works better.
Once all the apps they use on a regular basis are ones which are available on Linux, most people are pretty happy to switch.
That might sound a little naive, but from experience that's the way it works.
It does, however, require that the free applications be better than their proprietary equivalents (for the user's specific needs, at least). We can do that, right?
I guess you haven't noticed that "Mono" apps are being written to Gtk#, not System.Windows.Forms or modern equivalent.
Most Mono apps run on Windows, but only due to the windows port of Gtk#.
When confronted with computers, most people lose their capacity for rational thought and fall back to wishful thinking and superstition.
Try manning a helpdesk for a while if you don't believe me.
Me too. Looks like people are finally taking cues from places like infrastructures.org.
Ahh, ok. Maybe I misconstrued both your post and the OP slightly.
OOC, I was not aware algorithmic proofs had been debunked? Or did you mean that some standard accepted proofs were demonstrated to be wrong?
He is unusual, sadly. It's apparent that you haven't learned this area yet either.
Big-O is a measure of complexity that tells you the worst-case efficiency of a piece of code -- and you don't need a slide rule or a calculator, becuause the answer isn't a number, it's a formula you can determine by looking at the structure of your program.
Knowing how different "types" of formulas compare to one another, it helps you choose and optimize your algorithms.
Being aware of complexity issues ("...is this algorithm O(1)? O(log n)? O(n^2)?") really is essential to writing clean, efficient code -- particularly code that scales well. This applies to almost all non-trivial programs.
(it's also important to know what Big O does not mean -- e.g. when is bubble sort consistently among the fastest sort algorithms? it is if you know your list is very short or mostly sorted beforehand!)
Even if you don't get a formal CS education, it behooves you to learn this stuff. Hit places like Wikipedia and the Portland Pattern Repository Wiki when you see terms like "Big O" mentioned. Trust me, it'll make you a drastically better programmer.
(The Wikipedia article is here: Big O Notation, though it does assume a certain level of math background. Google might be good for finding more "layman's" explanations to get you started.)
While Mono is doing an implementation of Microsoft's extensions to the ECMA spec, they're also doing their own set in parallel.
.NET clone stuff, Mono still has a viable system built on top of the ECMA standard (Gtk#, etc...) that they've been encouraging people to target all along.
If all they did was cloning, of course the best they could ever hope for was barely keeping up.
This means that if Microsoft torpedoes the
Note whose APIs Ximian is writing their apps to... they aren't Microsoft's...
I used to think Miguel was naive. Now I think he's a really shrewd bastard... They got Microsoft's support and then pulled an "embrace and extend" on MICROSOFT.
Well, the thing is, although they are implementing Microsoft's extensions, in a parallel effort, Mono's building their own set of standard libraries on top of the ECMA standard (which is what they're actually using for apps). This poster puts it pretty clearly.
There isn't any competition for the base ECMA stuff; it's shared by Microsoft's libraries/extensions and Mono's. The competition is between the two sets of standard libraries themselves.
That is the way things ought to be.
However, I'm betting you've never seen programmers try to use inheritance to express relationships of a "has a" (or "other") nature, because they haven't thought it through clearly (the "if all you have is a hammer..." syndrome).
So, it's helpful to train programmers to think about the "is a"/"has a"/etc relationship up front, rather than just kind of fuzzily using inheritance because they know the classes are related (and then maybe deciding that there is an "is a" relationship ex post facto).
Actually there are in part security motivations for that (e.g. look at the reasons Java's String class is final).
It's better to be a little too agressive with final than not enough (they can safely back off later, but you can't make things final that weren't before without causing massive breakage).
Also, given C# has good language support for non-inheritance aggregation/delegation, all those final classes aren't as painful as they would be in Java (provided you aren't trying to write Java code in C#).
Celestia doesn't really do anything with orbital mechanics. You just click and zoop around. It's very pretty though.
That would be my guess.
Most historic ls(1) implementations do munging of nonprintable and high-ascii characters (c.f. the -b and -q options) -- they would have had to disable that to make it UTF-8 friendly, and that might break shell scripts if they did it by default.
I don't have my laptop in front of me right now, but perhaps they added an ls option to turn off the munging.
Actually, no. These are adult stem cells. Not banned in the US.
...and really, really lousy ping times. The latency makes improvisation nearly impossible (or at best very costly).
Ah, excellent. The last time I did serious (GUI) Windows programming was in the flaky Win95 (OSR1) era, where GDI resources weren't always freed in my experience ... (and obviously they never were on Win16). On early Win95 it may have been that there was still a lot of 16 bit stuff going on behind the scenes; I'm not sure.
...
I did think you were complaining that memory was being leaked after a crash, though
Ignore me then, I guess.
Have you considered using OGG as the container format?
No, OGG is the streaming container. Theora is a (heavily, at this point) modified version of VP3.
If the memory a process allocated isn't unconditionally freed when the process exits (crash or no), that's an OS bug, not a bug in the application.
It's probably GDI resources and things, though, not memory per se. The Windows GDI design is such that there is no way a process' GDI resources to be freed if the process crashes.
Which, again, is an OS (design) issue.
Well duh! That's because everyone knows Frodo destroyed it.
I don't intend to dismiss your anger at having your country messed with, because you should be angry.
But what safeguards against the abuse of power would a world government have? I don't think abuses of power have ever been limited to particular religions or economic system; why would it be different this time?
Look at what's currently happening in the EU with software patents (for example). No rabid Christians or Turbo-Capitalists involved (well, maybe Turbo-Capitalists).
What prevents a meta-government (whether the US [remember, it also started out as a union of sovereign states!], the EU, or a hypothetical world government) from behaving in equally high-handed ways in other areas of policy, or falling under the influence of unsavory politicians?