I'd say that C++ is remarkably compact, you can know everything there is about it easily.
I've been writing C++ code for the past 15 years, and I beg to differ. I still occasionally run across obscure features of the language that I'd missed before.
In fact, over my 20-year programming career, it's the least compact language that I've ever become proficient in (which includes Java, JavaScript, Prolog, Scheme, C, dBaseIII+, Perl, Ruby, posix sh, Scala, TCL, Robotic, VB, assorted "classic" BASIC flavors, Haskell, and probably some others I've forgotten).
Libraries can offset the simplicity of a language somewhat (e.g. Java), but even when you consider C++'s libraries, it's much harder to find or use C++ libraries in a uniform way than in most other languages. Don't even get me started on the ABI issues.
One needn't have faith that Silverlight will continue to be cross-platform to have more faith in Silverlight's cross-platform availability than Flash's.
Dark matter! Maybe dark energy! Even maybe we have to revise theories in astrophysics because we were wrong on something... sigh, why do scientists think they are right now when their forbears were wrong?
Because their theories better fit the data. When they find a place where their theories and those of their predecessors don't work (this planet may be such a case), they work on formulating more general theories based on what they already know. And when they do this, they don't start from scratch each time, but build instead on previous discovery.
That's what science does. It progresses. It works. Would you rather we abandon the scientific method and just make up random stuff without testing it against reality? Even dark matter and dark energy aren't arbitrary: they're provisional descriptions of stuff we're actually seeing happen.
I'm getting really sick of this "oh, we can't really ever know anything because no theory is perfect, so let's just give up on this science thing" attitude.
Actually, I'll take that back about Shadow of the Colossus; looking over a game script I guess it wasn't too bad, at least if the intent was to use the traditional plural/singular T-V distinction in English to convey differences in formality in the original Japanese subtitles (in which case Dormin's use of "we" would be the "royal we"). They did still misuse -st endings a lot.
Yeah, those were pretty cool, and some were really entertaining. They definitely are trying to do storytelling through the logs and lore scans in both MP1 and MP2, as well as dropping lots of storytelling clues in the design of the environments you explore.
Sadly (particularly in the Chozo and Luminoth lore) the logbook scans suffer from the same problem as many video games: when it comes to writing in-game text, the writers sound like pretentious English majors (Tycho of Penny Arcade is the most extreme example of this, though I think he's being deliberate about it). Even the dialogue in Shadow of the Colossus, a game with great overall storytelling, suffers a lot that way.
As far as I can tell, the current setup gives SoundExchange the legal right to collect all compulsory license fees, except where you have an explicit license in place which supersedes the default compulsory license.
Linus' argument seems very confused. A number of things strike me, but I'll just pick out two:
1. I don't think the GPLv3 authors would dispute that morality and law don't always coincide (RMS advocates civil disobedience in certain situations, for example), so raising the examples of Ghandi or conscientious objectors in WWII Germany seems to be a red herring on Linus' part.
2. That killing (for example) is sometimes right and sometimes wrong isn't a demonstration that morality is unimportant, since the moment we say it can be right or wrong we admit that the justifiability of killing is itself determined by moral principles.
Reading the article, my main impression was that I'd like to see Linus' quotes in context.
Without seeing the context, in general I would say the core disagreement between RMS and Linus (setting aside the frivolous GNU/Linux naming thing) lies in their respective notions about morality: RMS believes it to be essentially objective, whereas Linus considers it a subjective concern. This seems to be another manifestation of that disagreement.
As far as I can figure, the only legally sound thing (as users) is simply not to trade the media in the DRMed form. In practice they'll probably do so anyway.
If you wanted to put legal pressure on the manufacturer, I imagine you'd need to use some more indirect means than suing them for copyright violation, again because the user isn't being forced to use the manufacturer's devices.
The one way around this as an independent artist is to release your music under a general license that explicitly allows broadcasting for free or at some rate (Creative Commons would work); explicit licenses take priority over the default compulsory license.
That doesn't mean SoundExchange won't necessarily harass you if you (as an internet radio station) play such music though, just that theoretically you don't legally owe them anything.
There is no single standard behavior for Lisp garbage collection, nor even really is "Lisp" a single language. It's a wild forest of various implementations and dialects, some of them confederated under the Common Lisp specification.
Alternately, if they find an unapproved Open Source application on a lot of people's machines, they might ban it and forcibly remove it from people's PCs if it's found. That happened about a year ago with Firefox where I work; fortunately they don't know to look for Seamonkey yet so I can still use that for web development instead (though I still miss out on some stuff like Firebug).
If they started using that scanning tool here, I'd probably resign; I rely on Open Source tools pretty heavily to do my job. I could probably manage without, but I'd be pretty miserable.
Of course, on the other side of the coin, there's also the issue of a pedophile working for GeekSquad acquiring pictures and personal information on your children.
All they would need is a more robust run-time type information system.
No, not at all. The current C/C++ specifications permit compilers to transform code in ways that can interfere with a garbage collector. Fortunately compilers do not do that as often as they could, but it seems like something important that should be addressed.
It doesn't need to be part of the core, but they do need to further define certain areas of the language to permit garbage collection to be implemented in an entirely reliable fashion. That's the main part Hans cares about.
Note, we had to build up a lot of our own infrastructure to enable a gradual transition (see all the src/gc-* files). At minimum, you a way to pin gc-managed objects which are referenced from the non-gc heap.
Granted, that may not be an issue for some projects -- many can probably get away with using the boehm collector in malloc-replacement mode, where you just twiddle the build flags and you're done. But we have an awful lot of RGBA buffers allocated by libraries we don't necessarily control, which would look like giant arrays of bogus pointers if they were blindly allocated as non-atomic buffers, so we had to leave the default malloc in place, and begin using the boehm allocator explicitly here and there.
I'd be inclined to agree, actually, though that has do with massive structs like SPStyle (ugh) and poorly written manual memory management (livarot) rather than the boehm collector.
Generally the garbage-collection-enabled code has done fairly well, and where we switched to using the bohem collector it often allowed us to share data structures, where previously extra copies were made to avoid memory management issues.
The commenter you quoted understates his point -- not only are collectively-authored myths and Shakespeare's adaptations a human norm, but even what we would properly call "fan fiction" today has a longer history than many people suppose. I first realized this when someone pointed me to this book about the widespread nature of what could only be called 18th-century fanfic.
Slashdot Mirror
I've been writing C++ code for the past 15 years, and I beg to differ. I still occasionally run across obscure features of the language that I'd missed before.
In fact, over my 20-year programming career, it's the least compact language that I've ever become proficient in (which includes Java, JavaScript, Prolog, Scheme, C, dBaseIII+, Perl, Ruby, posix sh, Scala, TCL, Robotic, VB, assorted "classic" BASIC flavors, Haskell, and probably some others I've forgotten).
Libraries can offset the simplicity of a language somewhat (e.g. Java), but even when you consider C++'s libraries, it's much harder to find or use C++ libraries in a uniform way than in most other languages. Don't even get me started on the ABI issues.
You must be using a 32-bit browser binary then. At least it works for you. I've had considerably less luck on my PPC laptop...
One needn't have faith that Silverlight will continue to be cross-platform to have more faith in Silverlight's cross-platform availability than Flash's.
Because their theories better fit the data. When they find a place where their theories and those of their predecessors don't work (this planet may be such a case), they work on formulating more general theories based on what they already know. And when they do this, they don't start from scratch each time, but build instead on previous discovery.
That's what science does. It progresses. It works. Would you rather we abandon the scientific method and just make up random stuff without testing it against reality? Even dark matter and dark energy aren't arbitrary: they're provisional descriptions of stuff we're actually seeing happen.
I'm getting really sick of this "oh, we can't really ever know anything because no theory is perfect, so let's just give up on this science thing" attitude.
You're thinking penal scanner. The OP meant penile which is, uh, something else...
How does one deal with latency when recording over the net like that, I wonder?
Actually, I'll take that back about Shadow of the Colossus; looking over a game script I guess it wasn't too bad, at least if the intent was to use the traditional plural/singular T-V distinction in English to convey differences in formality in the original Japanese subtitles (in which case Dormin's use of "we" would be the "royal we"). They did still misuse -st endings a lot.
Yeah, those were pretty cool, and some were really entertaining. They definitely are trying to do storytelling through the logs and lore scans in both MP1 and MP2, as well as dropping lots of storytelling clues in the design of the environments you explore.
Sadly (particularly in the Chozo and Luminoth lore) the logbook scans suffer from the same problem as many video games: when it comes to writing in-game text, the writers sound like pretentious English majors (Tycho of Penny Arcade is the most extreme example of this, though I think he's being deliberate about it). Even the dialogue in Shadow of the Colossus, a game with great overall storytelling, suffers a lot that way.
As far as I can tell, the current setup gives SoundExchange the legal right to collect all compulsory license fees, except where you have an explicit license in place which supersedes the default compulsory license.
Wow, okay. Looks like I hit the nail on the head.
Linus' argument seems very confused. A number of things strike me, but I'll just pick out two:
1. I don't think the GPLv3 authors would dispute that morality and law don't always coincide (RMS advocates civil disobedience in certain situations, for example), so raising the examples of Ghandi or conscientious objectors in WWII Germany seems to be a red herring on Linus' part.
2. That killing (for example) is sometimes right and sometimes wrong isn't a demonstration that morality is unimportant, since the moment we say it can be right or wrong we admit that the justifiability of killing is itself determined by moral principles.
Reading the article, my main impression was that I'd like to see Linus' quotes in context.
Without seeing the context, in general I would say the core disagreement between RMS and Linus (setting aside the frivolous GNU/Linux naming thing) lies in their respective notions about morality: RMS believes it to be essentially objective, whereas Linus considers it a subjective concern. This seems to be another manifestation of that disagreement.
As far as I can figure, the only legally sound thing (as users) is simply not to trade the media in the DRMed form. In practice they'll probably do so anyway.
If you wanted to put legal pressure on the manufacturer, I imagine you'd need to use some more indirect means than suing them for copyright violation, again because the user isn't being forced to use the manufacturer's devices.
So, if I've got some Public Domain or CC-licensed songs, they're probably going to fall into the "may have come from file sharing sites" bin.
"Those are some nice Creative Commons media files you've got there. It'd be a shame if something happened to them..."
The one way around this as an independent artist is to release your music under a general license that explicitly allows broadcasting for free or at some rate (Creative Commons would work); explicit licenses take priority over the default compulsory license.
That doesn't mean SoundExchange won't necessarily harass you if you (as an internet radio station) play such music though, just that theoretically you don't legally owe them anything.
Thank you for pointing this out. Sadly, part of the problem is that (contemporary) Americans generally don't understand the concept of natural rights.
There is no single standard behavior for Lisp garbage collection, nor even really is "Lisp" a single language. It's a wild forest of various implementations and dialects, some of them confederated under the Common Lisp specification.
I'm certainly thinking about it.
Alternately, if they find an unapproved Open Source application on a lot of people's machines, they might ban it and forcibly remove it from people's PCs if it's found. That happened about a year ago with Firefox where I work; fortunately they don't know to look for Seamonkey yet so I can still use that for web development instead (though I still miss out on some stuff like Firebug).
If they started using that scanning tool here, I'd probably resign; I rely on Open Source tools pretty heavily to do my job. I could probably manage without, but I'd be pretty miserable.
Of course, on the other side of the coin, there's also the issue of a pedophile working for GeekSquad acquiring pictures and personal information on your children.
Heh. I wrote a Ruby Quiz problem on just that a while back.
:(
Sadly, nobody submitted a solution involving time travel.
No, not at all. The current C/C++ specifications permit compilers to transform code in ways that can interfere with a garbage collector. Fortunately compilers do not do that as often as they could, but it seems like something important that should be addressed.
See Hans' paper Simple Garbage-Collector Safety for details.
It doesn't need to be part of the core, but they do need to further define certain areas of the language to permit garbage collection to be implemented in an entirely reliable fashion. That's the main part Hans cares about.
Note, we had to build up a lot of our own infrastructure to enable a gradual transition (see all the src/gc-* files). At minimum, you a way to pin gc-managed objects which are referenced from the non-gc heap.
Granted, that may not be an issue for some projects -- many can probably get away with using the boehm collector in malloc-replacement mode, where you just twiddle the build flags and you're done. But we have an awful lot of RGBA buffers allocated by libraries we don't necessarily control, which would look like giant arrays of bogus pointers if they were blindly allocated as non-atomic buffers, so we had to leave the default malloc in place, and begin using the boehm allocator explicitly here and there.
I'd be inclined to agree, actually, though that has do with massive structs like SPStyle (ugh) and poorly written manual memory management (livarot) rather than the boehm collector.
Generally the garbage-collection-enabled code has done fairly well, and where we switched to using the bohem collector it often allowed us to share data structures, where previously extra copies were made to avoid memory management issues.
The commenter you quoted understates his point -- not only are collectively-authored myths and Shakespeare's adaptations a human norm, but even what we would properly call "fan fiction" today has a longer history than many people suppose. I first realized this when someone pointed me to this book about the widespread nature of what could only be called 18th-century fanfic.