Err perhaps someone could explain to me what is better about openZaurus
For one thing, it can have reliable TCP/IP. It may depend on the specific networking hardware (USB to Linux, or WinXP, or a CF ethernet or 802.11b card) you're trying to use, but the Sharp-supplied ROMs often simply give up on transmitting after 100kB or so (and then won't re-initialize until after a reboot).
Other kinds of hardware compatibilty and stability were also improved in OpenZaurus (although the recent release-candiates introduced several random bugs, hopefully the 3.2 release has fixed them)
The major design change with OpenZaurus is that the main filesystem is stored in Flash memory, rather than just RAM like on the original ROMs (or on a Palm). This means that if your system crashes (or loses power), files like your addressbook and network card settings are preserved.
(The normal Sharp ROMs only write to flash during the special "reflashing" process to upgrade the ROM, but OpenZaurus can modify it at any time. This could possibly create a risk of hardware failure, as flash memory has limited reusability)
a better jpg viewer that might allow me to see large pictures direct off my digital camera in full screen mode
The viewer can scale and rotate. However, it's decompression algorithm might not be efficient enough to unpack a full 1600x1200 JPG without exhausting your system RAM. (There are commerical viewers which can definately cope)
The problem I have with my zaurus is that is does not work with my wireless card.
OpenZaurus is known to work pretty well with this. Of course, the Sharp ROMs often handled it decently too (although they split the network configuration across several applets, making you go through more steps to set it up)
Check out the slick coloring- a cool translucent look for soothing, grey-on-grey symbols with gently blurred edges. They'll be especially good in bright conditions, when the LCD screens of a PDA will reflect back all the ambient light and remove all contrast.
Additionally, Twirlip might have implied owning patents- I can't recall exactly. There was also another post which gave a good clue, but I can't find it either. (And slashdot doesn't even have an approximation of a good search engine!)
Ron has been known to use the nickname "Twirlip of the Mists"
Other people could've read that book...
But it is very understandable that such a celebrity might want to anonymize himself on a forum like Slashdot, so as to focus responses on what he says, not who he is.
Not damning, but interesting evidence nonetheless.
Since Rivest is all about information security, it would be amusing if a "traffic analysis" attack could strip him of an assumed identity. Twirlip has a high posting frequency- an automatically collected log of his userpage would reveal his daily schedule. Rivest lives in Massachusetts (a known time-zone), and often teaches an undergraduate class, so part of his schedule is public knowledge.
So, a quick test would be to list times Rivest was known to be lecturing in the past few years, and then look for any post from Twirlip inside that envelope.
(Of course, an individual could confound this style of analysis by pre-scheduling a forum post at a later time...)
Re:Unproven = untrustworthy = low SAT scores!
on
Synthetic Vision
·
· Score: 1
That is a false analogy (Unproven = untrustworthy).
The Apple iPod is a good example.
A procurement officer for active military troops should have completely different standards of trustworthiness than civilians needing an MP3 player.
An unexpected failure of untested hardware can have drastically different costs depending on what you were trying to do when it conked out...
Additionally, the "innovations" of the IPod were in the areas of reliability and portability- the actual task of "playing MP3s from a hard disk" already had a decade of practice behind it. So far, Augmented Reality systems haven't worked well yet, even in laboratory situations.
Why do you think Twirlip of the Mists is Ron Rivest?
Twirlip doesn't seems to post on crypto subjects (although it's hard to tell, because the 24 message buffer doesn't last him a day). Maybe he's trying to avoid the topics of his day job?
This is a troll, right? How could anyone mod-up such an incorrect statistical manipulation?
To get statistics based on the entire developer population, one only needs to obtain the percentage of developers who use Linux to any extent and multiply all statistics against it.
Hmm, 97% of Catholic monks oppose war in Iraq. And they make up 0.0001% of the US population. So can we multiply those numbers to see if the US public supports the war?
(Hint: the opinions of a sample selected by any means other than total randomness can never be considered representative of the whole.)
Who else would you ask to find out where people are moving to Linux from?
But that's not the question. The right question isn't "Where did current Linux developers come from?", but "Where did old Windows & Unix developers go?"
If you don't ask the latter question, you're not justified in reaching the conclusions the study claims.
You want to know if people are moving to Linux from Unix or from Windows.
That's not what "taking developer marketshare" means. "Share" means percentage of total, not an absolute number.
Suppose in 1993 Unix had 10% share and Microsoft 90%, and that today Linux has 10%. If you find, as the survey did, that 60% of Linux devs came from Windows, and 30% from Unix, that means Unix & Windows now have (approx) 7% & 84%, respectively. Meaning Unix has lost 30% of its developers, but Microsoft is down just 7%.
At that rate, Linux will kill all other Unices before Microsoft is really harmed. (That's the natural result of the survey- but, it doesn't have enough datapoints to really predict much of anything)
There are more factors- for instance, there is no "conservation of developers law"- old Unix devs can simply be fired and not respond to any surveys, but they should still count as "lost market". I wrote another comment on the situation.
We could discuss the Millenium Challenge all night
I have some first-hand experience. The Pentagon briefly employed me to code MC02 software. I did my job (which was a small fixup task), observed some of the overall planning, and went back home. I followed the remainder closely, though, and interviewed some full-time participants.
wasn't it really to test out tactics and various scenarios?
MC02 had 3 separate tasks:
Experiment- Supposedly, this was the overall goal: determine "What-If?" outcomes for missions conducted with new equipment, new techniques, or against new threats. Primarily done virtually- with combat units that are 100% computer simulated (maybe AI controlled, or maybe driven by a human pilot in a flight trainer)
Train- Allow warfighters to rehearse accomplishing their missions. Conducted by real troops on training grounds who marched, drove, or flew with GPS/INS telemitry inserting their positions (and gunshots) into the computer-simulation world.
Demonstrate- Occasionally, a new piece of hardware had an event scheduled for it. VIPs (local Congressmen) had appointments to watch the new gizmo do it's job at a preplanned time.
Plainly, those goals conflict with each other. Any scientist will confirm that an "experiment" loses all validity if it's forced to be either training or demonstration at the same time. In fact, to be strictly scientific, an experiment should only have 1 variable changed at a time.
The goals of training and demonstration interfered with experimentation not only by diverting dollars and focus from it, but worse, by pre-determining the outcomes of engagements. If troops are deployed to train and conduct fake missions, then CGFs ("Computer Generated Forces"- like NPC enemies) simply couldn't kill 100 of them and leave them lying on the dirt for 15 days of a 21 day exercise- those guys came to march and shoot, and that's what they'll do! And the CGF likewise couldn't flee the attackers entirely- they had to hold their positions so they can be defeated by trainees pretending to fire artillery shells.
Likewise, there were a few occasions where the deployment of an impressive new vehicle was scheduled so that the Representatives whose constituents would build it could come watch. So once again, the scenario had to give it enemies to chase at an exact time. And of course, no enemy was allowed to kill it earlier.
So, the requirements of training and demonstration really destroyed any opportunity to learn from computer simulation. The progress and outcome of the battles were determined ahead of time by a few "experts" using their best judgement and imaginations. Well, if these expert analysts were smart enough to predict what would happen, why did they need all that computer modeling? In the end, MC02 was designed to lend objective legitimacy ("the computer plays no favorites") to decisions about future-force composition that had already been made. It was a high-tech, expensive rubber-stamp.
(There were other things wrong with the experimentation too- inconsistencies on what guns can kill what tanks, how wide is the Persian Gulf, and how far you can see from the bridge of a CVN, for instance- but they were minor factors in comparison, and they might've been fixed if scheduling constraints had allowed time to pause the simulation and resolve it. But no- with live training participants, time of day must go on...)
What was the most postive outcome from MC02? Looking at it from a software developer's perspective, the prime task for MC02 was to simply integrate combat simulation software from all 5 services. That's 50+ independent programs that have to be made to run in the same WAN game. (Note that they considered SOCOM, not Coast Guard, to be the 5th service). Maybe in the future, MC02's legacy will be the ability to run similar exercises at less cost. That might really help us learn something new.
Lip service. Trigun talks a good line (better than most US shows with the same themes), but doesn't show it in practice.
At least (and this is a very big redeeming quality), the author has the decency to point out that Vash's ideals are stupid. This is primarily accomplished through other characters speaking rationally, and Vash being unable to articulate a response.
(I wonder if this theme will survive into the US version? It was subtle to begin with, and American cartoon audiences have a poor reputation for detecting nuance)
(see Wolfwood)... (see Legato.)
If you have to pick out 2-3 small incidents, that by definition is not what the series is about.
By and large, the show is about interrupting tense confrontations with deadly force, and yet not killing anyone.
(Now, one could say the whole show was just buildup or contrast to one or 2 key episodes. But in that case, it could've done with far less than 26 eps)
I could select 3 or 4 episodes of Sailor Moon to demonstrate it's about self-sacrifice and unrequited love- but that won't disguise the true nature of the other 197 episodes.
I'd hardly think so. The entire storyline is contained in just 2 episodes. (the flashback episode, one half of the conclusion, and any random 15 minutes from anyplace else in the series).
"I'm bad! Before you can fight me, you must get through my 10+ exotically named and costumed warriors! Only then can the arch-rivals do battle!"
Compare to something like Escaflowne or even 1 single episode of Cowboy Bebop, and Trigun has virtually no story. Why, even Evangelion had more of a storyline.
(And Trigun didn't even finish the story- we don't know if V.S and M.K get married!)
The wandering hero carries a silver revolver and a strange secret, but none can fathom his dark past and guess the implications...
Until it is dramatically revealed that all this time, he was hiding- another gun! And it's larger than the previous one!
What further suprise could await us in the exciting conclusion? It couldn't possibly be another firearm- is there any way an even larger gun could possibly exist, without threatening the entire planet with it's destructive blast?
I don't like it either- there's a whole series of anime where the hero starts out as vastly more powerful than any opponent, and stays that way.
There's no uncertainty, challenge, or opportunity for growth. It's just wish-fullfillment role-playing for young viewers who can't stand up to challenges on their own, and want to emphasize with someone who's always victorious. Anime like Trigun and Ruronin Kenshin are like this. So are US shows like, oh, A-Team and Knight Rider. And any movie starring Jean-Claude VanDamme.
Some people believe that television should have a moral or educational message- and from that standpoint, Trigun is especially bad. True, the hero constantly speechifies about not killing enemies- but he leaves situations in his wake where somebody's going to be shot, lynched, or simply starve. (At least the other characters occasionally call him on it.)
And from a safety perspective, it's doubly bad. The hero constantly uses gunfire as a non-lethal weapon. A few years ago, US censors even took toy, rubber guns out of imported anime... regardless of whether they were right or wrong, the content of Trigun should offend those interests vastly more strongly.
Oh, and it is going to be dubbed... bleh... not worth spending your time on anyway.
Right on. Adequate dubbing is possible, but it hasn't happened for a TV anime yet.
Cowboy Bebop is the one exception (among what I belive CN shows). Several of it's (better) episodes weren't even televised in Japan, because they were considered too violent.
The most important part of the quote is that the individual acts in full view of the public and government, and thus serves as a messenger of truth and disclosure.
In either the Bezos case, or traditional civil disobedience, by stringently obeying the letter of the law, you can call attentions to shortcomings in that law.
On a related note, the majority of music-swapping copyright infringers are not practicing civil disobedience, because their actions are hidden. But Dr. Jack Kevorkian was, as he publicized his actions.
but my morals tell me that even in a time of war, assasination is wrong
Not just your morals. Ronald Regan said it was wrong too, as did most every US President. The current rule in effect is Executive Order 12333, sections 2.11 and 2.12.
Of course, while most US Presidents pay lip-service to the principle of no assasinations, hardly a term expired without the Commander-in-Chief ordering something that looks exactly like one. (Regan tried to kill Quadafi, Clinton tried to kill Bin Laden... they all do it)
(Bush attempted to rescind this order in specific regards to alQuaeda)
You can get entertaining articles on the subject by searching for 12333.
What are the implications?
Actually, the implications could be positive. Traditional "no assasination" rules are a way of making warfare more paletable to heads of state. It gives them some assurance that just their soldiers' lives are risked, not their own. The US claims to have such a rule as a way to protect it's own president (whose exposure to assault from a single operative is much higher than Saddam Hussein's, or other enemy leaders)
While even the native americans had some conflicts, they were nothing like have been in the past couple hundred years
Population-wise, the majority of "native americans" weren't the peaceful woodland tribes of (what is today) Canada and the US. They were citzens or subjects of the Aztec empire in mesoamerica, and their wars were of a similar (or worse) ferocity than those in 19th century Europe.
In a way, just looking at population density is a better way of predicting belligerence than cultural traditions. (It is more accurate to use "density" to mean "people per unit of life-sustaining resource", rather than "people per square meter")
From that perspective, it is natural that "Indians" of North America had little incentive for lethal conflict, since their land was capable of supporting many more people. But put a large number of people in a barren desert (either 1600s Mexico, or 1900s Middle East), and we should expect strife as water resources (and the oil which can buy more food/water) are contested.
middle-east based culture (those that spawend from Mesopatamia)
Very nearly all civilized cultures spawned from Mesopotamia. (And GW Bush claims that his favorite philosopher was from the Middle East...)
If most people claim to be good Christians/Jews/Muslims/Buddhists, why is there still violence?
The orthodox teachings of Judiaism and Buddhism are hardly pacficist- they realistically grant violence will be part of a normal society. Truely orthodox Christianity is non-violent, but it's hardly ever practiced. (One modern example are the Quakers)
Interestingly, the modern, high-membership religion that appears to signficantly restrain it's members from violence is one you didn't mention: Hinduism.
It is quite painful to come across an error message like this. It means that your program is correct, if "program" is defined as "C++ code". But in the broader sense- that the "program" is source code, build scripts, makefiles, and anything else needed to create a working executable- it is not correct. But the error comes not from C++ directly, but from the inadequate support for C++ in your development environment.
The underlying problem here is that those 2 g++ commands are actually running different programs. The first one is running a C++ compiler, as anyone would expect. But the 2nd call (the one that creates the executable file) is falling through to a linker (ld).
When C++ was introduced, C compilers were replaced with C++ compilers. But C linkers were basically kept around with few modifications. This was done to allow compatibilty with existing ("legacy") libraries, and other languages (which had all been usinc C linkage as a lowest-common-denominator for multiple-language development). However, the linker's ignorance of C++ template syntax means that not all linkers can cope with things that are completely valid according to the C++ language.
There are workarounds, of course- and the widespread acceptance of these workarounds as a normal part of building programs is part of the reason we're still stuck with old linkers.
Here's a page of GCC Documentation on the problem's cause, and some ways to address it. It gives this line of introduction:
"C++ templates are the first language feature to require more intelligence from the environment than one usually finds on a UNIX system."
Unless we could someday get significantly smarter linker programs with C++ awareness, and a corresponding minor improvement to compiler software so it can feed those linkers a little extra data, problems like this one won't go away.
If you'd like to kill some people, then you're free to go to Washington and ask for permission. It may sound like I'm making a joke, but depending on your proposal, they could be quite willing to support you. Maybe you'll even get a gun and ammunition provided! The easiest way to start is by submitting these forms.
It is by no means a "kludge". The intended use of a tool is never a kludge, even if some people find it personally unintuitive. (The only kludgey thing in that article was the extra magic needed to convince MS VC++ to deal with templates sanely)
The goal of adding templates to C++ was to provide the same power macros provided to C, but as part of the language, rather than a preprocessor to it. That way, they can be typesafe, produce better error messages, and be more flexible (since templates live inside of namespaces, unlike macros which exist at the parser level).
C programmers often performed fast math evaluations with macros, so replacing them with templates is simply a natural step.
However, looking at templates from an entirely different perspective- it is possible to reasonably classify most all uses of templates as "kludges".
Why? Because they are often used to force an evaluation to happen at compile-time, rather than runtime. But, if compilers made full use of the static program information that is present during compilation, they could determine that many conditional checks are often tautologies, and elminate them from the output code.
The reason C++ compilers can't usually make these throughout checks is because of backwards compatibility- to support interoperation with other programming languages, individual C++ files are compiled independently and then linked together by a stupider program. The linker has little knowledge of the static program elements, and less ability to modify the output code. So many potentially useful optimizations are impossible- unless the programmer uses something like templates to force them to occur.
So, from that perspective, it can be argued that nearly every template is a kludge- a workaround for the legacy compiler/linker split. (Many uses of macros were a similar kludge- a way to force an expression to be computed inline, rather than spending a function dispatch on it)
a decision of *when* do you want to handle the "type" issue.
Not just *when*, but also *who*. Or actually, *what*.
With a mechanism like templates, type checking is accomplished by the compiler, not the programmer. That means no extra statements in the source to bulk up the code, no extra conditional checks at runtime to slow it down, and most importantly, no opportunity for a human programmer to forget some of the checks.
That is the fundamental difference between a C++ STL vector, and a Java Vector which happens to only contain Integers.
An additional problem with the Java Vector (and something that'll be fixed when templates are introduced in Java 1.5) is that it can't contain things which don't inherit from Object (like int and float).
(Note that if a compiler is very smart, it is theoretically possible for both methods to be equally efficient at runtime. But that degree of optimization is rare even today- and depending on the class linking style, it may be impossible for the compiler to get all the data needed to optimize like that.)
Slashdot Mirror
Err perhaps someone could explain to me what is better about openZaurus
For one thing, it can have reliable TCP/IP. It may depend on the specific networking hardware (USB to Linux, or WinXP, or a CF ethernet or 802.11b card) you're trying to use, but the Sharp-supplied ROMs often simply give up on transmitting after 100kB or so (and then won't re-initialize until after a reboot).
Other kinds of hardware compatibilty and stability were also improved in OpenZaurus (although the recent release-candiates introduced several random bugs, hopefully the 3.2 release has fixed them)
The major design change with OpenZaurus is that the main filesystem is stored in Flash memory, rather than just RAM like on the original ROMs (or on a Palm). This means that if your system crashes (or loses power), files like your addressbook and network card settings are preserved.
(The normal Sharp ROMs only write to flash during the special "reflashing" process to upgrade the ROM, but OpenZaurus can modify it at any time. This could possibly create a risk of hardware failure, as flash memory has limited reusability)
a better jpg viewer that might allow me to see large pictures direct off my digital camera in full screen mode
The viewer can scale and rotate. However, it's decompression algorithm might not be efficient enough to unpack a full 1600x1200 JPG without exhausting your system RAM. (There are commerical viewers which can definately cope)
The problem I have with my zaurus is that is does not work with my wireless card.
OpenZaurus is known to work pretty well with this. Of course, the Sharp ROMs often handled it decently too (although they split the network configuration across several applets, making you go through more steps to set it up)
Hurray! A slick GUI, in the style of Microsoft Windows XP(tm).
Just look at the new icons compared to the old ones.
Check out the slick coloring- a cool translucent look for soothing, grey-on-grey symbols with gently blurred edges. They'll be especially good in bright conditions, when the LCD screens of a PDA will reflect back all the ambient light and remove all contrast.
Additionally, Twirlip might have implied owning patents- I can't recall exactly. There was also another post which gave a good clue, but I can't find it either. (And slashdot doesn't even have an approximation of a good search engine!)
Ron has been known to use the nickname "Twirlip of the Mists"
Other people could've read that book...
But it is very understandable that such a celebrity might want to anonymize himself on a forum like Slashdot, so as to focus responses on what he says, not who he is.
Not damning, but interesting evidence nonetheless.
Since Rivest is all about information security, it would be amusing if a "traffic analysis" attack could strip him of an assumed identity. Twirlip has a high posting frequency- an automatically collected log of his userpage would reveal his daily schedule. Rivest lives in Massachusetts (a known time-zone), and often teaches an undergraduate class, so part of his schedule is public knowledge.
So, a quick test would be to list times Rivest was known to be lecturing in the past few years, and then look for any post from Twirlip inside that envelope.
(Of course, an individual could confound this style of analysis by pre-scheduling a forum post at a later time...)
That is a false analogy (Unproven = untrustworthy).
The Apple iPod is a good example.
A procurement officer for active military troops should have completely different standards of trustworthiness than civilians needing an MP3 player.
An unexpected failure of untested hardware can have drastically different costs depending on what you were trying to do when it conked out...
Additionally, the "innovations" of the IPod were in the areas of reliability and portability- the actual task of "playing MP3s from a hard disk" already had a decade of practice behind it. So far, Augmented Reality systems haven't worked well yet, even in laboratory situations.
Why do you think Twirlip of the Mists is Ron Rivest?
Twirlip doesn't seems to post on crypto subjects (although it's hard to tell, because the 24 message buffer doesn't last him a day). Maybe he's trying to avoid the topics of his day job?
That observation would be specific, valid, and useless.
The survey is reaching a conclusion that is general, invalid, and useful (except for being wrong).
This is a troll, right? How could anyone mod-up such an incorrect statistical manipulation?
To get statistics based on the entire developer population, one only needs to obtain the percentage of developers who use Linux to any extent and multiply all statistics against it.
Hmm, 97% of Catholic monks oppose war in Iraq. And they make up 0.0001% of the US population. So can we multiply those numbers to see if the US public supports the war?
(Hint: the opinions of a sample selected by any means other than total randomness can never be considered representative of the whole.)
Who else would you ask to find out where people are moving to Linux from?
But that's not the question. The right question isn't "Where did current Linux developers come from?", but "Where did old Windows & Unix developers go?"
If you don't ask the latter question, you're not justified in reaching the conclusions the study claims.
You want to know if people are moving to Linux from Unix or from Windows.
That's not what "taking developer marketshare" means. "Share" means percentage of total, not an absolute number.
Suppose in 1993 Unix had 10% share and Microsoft 90%, and that today Linux has 10%. If you find, as the survey did, that 60% of Linux devs came from Windows, and 30% from Unix, that means Unix & Windows now have (approx) 7% & 84%, respectively. Meaning Unix has lost 30% of its developers, but Microsoft is down just 7%.
At that rate, Linux will kill all other Unices before Microsoft is really harmed. (That's the natural result of the survey- but, it doesn't have enough datapoints to really predict much of anything)
There are more factors- for instance, there is no "conservation of developers law"- old Unix devs can simply be fired and not respond to any surveys, but they should still count as "lost market". I wrote another comment on the situation.
I have some first-hand experience. The Pentagon briefly employed me to code MC02 software. I did my job (which was a small fixup task), observed some of the overall planning, and went back home. I followed the remainder closely, though, and interviewed some full-time participants.
wasn't it really to test out tactics and various scenarios?
MC02 had 3 separate tasks:
Plainly, those goals conflict with each other. Any scientist will confirm that an "experiment" loses all validity if it's forced to be either training or demonstration at the same time. In fact, to be strictly scientific, an experiment should only have 1 variable changed at a time.
The goals of training and demonstration interfered with experimentation not only by diverting dollars and focus from it, but worse, by pre-determining the outcomes of engagements. If troops are deployed to train and conduct fake missions, then CGFs ("Computer Generated Forces"- like NPC enemies) simply couldn't kill 100 of them and leave them lying on the dirt for 15 days of a 21 day exercise- those guys came to march and shoot, and that's what they'll do! And the CGF likewise couldn't flee the attackers entirely- they had to hold their positions so they can be defeated by trainees pretending to fire artillery shells.
Likewise, there were a few occasions where the deployment of an impressive new vehicle was scheduled so that the Representatives whose constituents would build it could come watch. So once again, the scenario had to give it enemies to chase at an exact time. And of course, no enemy was allowed to kill it earlier.
So, the requirements of training and demonstration really destroyed any opportunity to learn from computer simulation. The progress and outcome of the battles were determined ahead of time by a few "experts" using their best judgement and imaginations. Well, if these expert analysts were smart enough to predict what would happen, why did they need all that computer modeling? In the end, MC02 was designed to lend objective legitimacy ("the computer plays no favorites") to decisions about future-force composition that had already been made. It was a high-tech, expensive rubber-stamp.
(There were other things wrong with the experimentation too- inconsistencies on what guns can kill what tanks, how wide is the Persian Gulf, and how far you can see from the bridge of a CVN, for instance- but they were minor factors in comparison, and they might've been fixed if scheduling constraints had allowed time to pause the simulation and resolve it. But no- with live training participants, time of day must go on...)
What was the most postive outcome from MC02? Looking at it from a software developer's perspective, the prime task for MC02 was to simply integrate combat simulation software from all 5 services. That's 50+ independent programs that have to be made to run in the same WAN game. (Note that they considered SOCOM, not Coast Guard, to be the 5th service). Maybe in the future, MC02's legacy will be the ability to run similar exercises at less cost. That might really help us learn something new.
It's about a tormented super anime hero,
... (see Legato.)
Lip service. Trigun talks a good line (better than most US shows with the same themes), but doesn't show it in practice.
At least (and this is a very big redeeming quality), the author has the decency to point out that Vash's ideals are stupid. This is primarily accomplished through other characters speaking rationally, and Vash being unable to articulate a response.
(I wonder if this theme will survive into the US version? It was subtle to begin with, and American cartoon audiences have a poor reputation for detecting nuance)
(see Wolfwood)
If you have to pick out 2-3 small incidents, that by definition is not what the series is about.
By and large, the show is about interrupting tense confrontations with deadly force, and yet not killing anyone.
(Now, one could say the whole show was just buildup or contrast to one or 2 key episodes. But in that case, it could've done with far less than 26 eps)
I could select 3 or 4 episodes of Sailor Moon to demonstrate it's about self-sacrifice and unrequited love- but that won't disguise the true nature of the other 197 episodes.
Come on, please don't post like that without making a few suggestions.
I thought self-described otaku preferred something with continuity. And shorter series tend to be higher quality anyway.
(You should've just gone outside for some exercise when the 298347923873th rerun started! Or at least muted the sound and made your own dub...)
a good flowing storyline
I'd hardly think so. The entire storyline is contained in just 2 episodes. (the flashback episode, one half of the conclusion, and any random 15 minutes from anyplace else in the series).
"I'm bad! Before you can fight me, you must get through my 10+ exotically named and costumed warriors! Only then can the arch-rivals do battle!"
Compare to something like Escaflowne or even 1 single episode of Cowboy Bebop, and Trigun has virtually no story. Why, even Evangelion had more of a storyline.
(And Trigun didn't even finish the story- we don't know if V.S and M.K get married!)
Scroll to the end.
"TriGun":
The wandering hero carries a silver revolver and a strange secret, but none can fathom his dark past and guess the implications...
Until it is dramatically revealed that all this time, he was hiding- another gun! And it's larger than the previous one!
What further suprise could await us in the exciting conclusion? It couldn't possibly be another firearm- is there any way an even larger gun could possibly exist, without threatening the entire planet with it's destructive blast?
You'll just have to tune in and find out!
I don't like it either- there's a whole series of anime where the hero starts out as vastly more powerful than any opponent, and stays that way.
There's no uncertainty, challenge, or opportunity for growth. It's just wish-fullfillment role-playing for young viewers who can't stand up to challenges on their own, and want to emphasize with someone who's always victorious. Anime like Trigun and Ruronin Kenshin are like this. So are US shows like, oh, A-Team and Knight Rider. And any movie starring Jean-Claude VanDamme.
Some people believe that television should have a moral or educational message- and from that standpoint, Trigun is especially bad. True, the hero constantly speechifies about not killing enemies- but he leaves situations in his wake where somebody's going to be shot, lynched, or simply starve. (At least the other characters occasionally call him on it.)
And from a safety perspective, it's doubly bad. The hero constantly uses gunfire as a non-lethal weapon. A few years ago, US censors even took toy, rubber guns out of imported anime... regardless of whether they were right or wrong, the content of Trigun should offend those interests vastly more strongly.
Oh, and it is going to be dubbed... bleh... not worth spending your time on anyway.
Right on. Adequate dubbing is possible, but it hasn't happened for a TV anime yet.
(for the unititiated: Vash doesn't believe in killing, and doesn't do so until the last few episodes).
Shame on you. That was a HUGE spoiler.
"Will he or won't he?" is supposed to keep viewers guessing until right to the end.
Cowboy Bebop is the one exception (among what I belive CN shows). Several of it's (better) episodes weren't even televised in Japan, because they were considered too violent.
And even that is aimed at young adults.
Exactly what I meant.
The most important part of the quote is that the individual acts in full view of the public and government, and thus serves as a messenger of truth and disclosure.
In either the Bezos case, or traditional civil disobedience, by stringently obeying the letter of the law, you can call attentions to shortcomings in that law.
On a related note, the majority of music-swapping copyright infringers are not practicing civil disobedience, because their actions are hidden. But Dr. Jack Kevorkian was, as he publicized his actions.
but my morals tell me that even in a time of war, assasination is wrong
Not just your morals. Ronald Regan said it was wrong too, as did most every US President. The current rule in effect is Executive Order 12333, sections 2.11 and 2.12.
Of course, while most US Presidents pay lip-service to the principle of no assasinations, hardly a term expired without the Commander-in-Chief ordering something that looks exactly like one. (Regan tried to kill Quadafi, Clinton tried to kill Bin Laden... they all do it)
(Bush attempted to rescind this order in specific regards to alQuaeda)
You can get entertaining articles on the subject by searching for 12333.
What are the implications?
Actually, the implications could be positive. Traditional "no assasination" rules are a way of making warfare more paletable to heads of state. It gives them some assurance that just their soldiers' lives are risked, not their own. The US claims to have such a rule as a way to protect it's own president (whose exposure to assault from a single operative is much higher than Saddam Hussein's, or other enemy leaders)
While even the native americans had some conflicts, they were nothing like have been in the past couple hundred years
Population-wise, the majority of "native americans" weren't the peaceful woodland tribes of (what is today) Canada and the US. They were citzens or subjects of the Aztec empire in mesoamerica, and their wars were of a similar (or worse) ferocity than those in 19th century Europe.
In a way, just looking at population density is a better way of predicting belligerence than cultural traditions. (It is more accurate to use "density" to mean "people per unit of life-sustaining resource", rather than "people per square meter")
From that perspective, it is natural that "Indians" of North America had little incentive for lethal conflict, since their land was capable of supporting many more people. But put a large number of people in a barren desert (either 1600s Mexico, or 1900s Middle East), and we should expect strife as water resources (and the oil which can buy more food/water) are contested.
middle-east based culture (those that spawend from Mesopatamia)
Very nearly all civilized cultures spawned from Mesopotamia. (And GW Bush claims that his favorite philosopher was from the Middle East...)
If most people claim to be good Christians/Jews/Muslims/Buddhists, why is there still violence?
The orthodox teachings of Judiaism and Buddhism are hardly pacficist- they realistically grant violence will be part of a normal society. Truely orthodox Christianity is non-violent, but it's hardly ever practiced. (One modern example are the Quakers)
Interestingly, the modern, high-membership religion that appears to signficantly restrain it's members from violence is one you didn't mention: Hinduism.
It is quite painful to come across an error message like this. It means that your program is correct, if "program" is defined as "C++ code". But in the broader sense- that the "program" is source code, build scripts, makefiles, and anything else needed to create a working executable- it is not correct. But the error comes not from C++ directly, but from the inadequate support for C++ in your development environment.
The underlying problem here is that those 2 g++ commands are actually running different programs. The first one is running a C++ compiler, as anyone would expect. But the 2nd call (the one that creates the executable file) is falling through to a linker (ld).
When C++ was introduced, C compilers were replaced with C++ compilers. But C linkers were basically kept around with few modifications. This was done to allow compatibilty with existing ("legacy") libraries, and other languages (which had all been usinc C linkage as a lowest-common-denominator for multiple-language development). However, the linker's ignorance of C++ template syntax means that not all linkers can cope with things that are completely valid according to the C++ language.
There are workarounds, of course- and the widespread acceptance of these workarounds as a normal part of building programs is part of the reason we're still stuck with old linkers.
Here's a page of GCC Documentation on the problem's cause, and some ways to address it. It gives this line of introduction:
"C++ templates are the first language feature to require more intelligence from the environment than one usually finds on a UNIX system."
Unless we could someday get significantly smarter linker programs with C++ awareness, and a corresponding minor improvement to compiler software so it can feed those linkers a little extra data, problems like this one won't go away.
Agreed that profit is his primary motive.
However, Bezos has stated a belief that the US patent system is broken, and that by publicizing the problem, Amazon creates pressure to fix it.
(That letter was provoked by Tim O'Reilley.
If its Bezos's job to take advantage of the government due to poor enforcement for as long as he can get away with it,
Bezo has explicitly asked the government for permission to behave this way- and they agreed (that's what the patent application process is).
He's not being sneaky or hiding anything, and thus is "in reality expressing the highest respect for the law". (My apologies to MLK, for slightly twisting his quote. But it's still applicable)
If you'd like to kill some people, then you're free to go to Washington and ask for permission. It may sound like I'm making a joke, but depending on your proposal, they could be quite willing to support you. Maybe you'll even get a gun and ammunition provided! The easiest way to start is by submitting these forms.
It is by no means a "kludge". The intended use of a tool is never a kludge, even if some people find it personally unintuitive. (The only kludgey thing in that article was the extra magic needed to convince MS VC++ to deal with templates sanely)
The goal of adding templates to C++ was to provide the same power macros provided to C, but as part of the language, rather than a preprocessor to it. That way, they can be typesafe, produce better error messages, and be more flexible (since templates live inside of namespaces, unlike macros which exist at the parser level).
C programmers often performed fast math evaluations with macros, so replacing them with templates is simply a natural step.
However, looking at templates from an entirely different perspective- it is possible to reasonably classify most all uses of templates as "kludges".
Why? Because they are often used to force an evaluation to happen at compile-time, rather than runtime. But, if compilers made full use of the static program information that is present during compilation, they could determine that many conditional checks are often tautologies, and elminate them from the output code.
The reason C++ compilers can't usually make these throughout checks is because of backwards compatibility- to support interoperation with other programming languages, individual C++ files are compiled independently and then linked together by a stupider program. The linker has little knowledge of the static program elements, and less ability to modify the output code. So many potentially useful optimizations are impossible- unless the programmer uses something like templates to force them to occur.
So, from that perspective, it can be argued that nearly every template is a kludge- a workaround for the legacy compiler/linker split. (Many uses of macros were a similar kludge- a way to force an expression to be computed inline, rather than spending a function dispatch on it)
a decision of *when* do you want to handle the "type" issue.
Not just *when*, but also *who*. Or actually, *what*.
With a mechanism like templates, type checking is accomplished by the compiler, not the programmer. That means no extra statements in the source to bulk up the code, no extra conditional checks at runtime to slow it down, and most importantly, no opportunity for a human programmer to forget some of the checks.
That is the fundamental difference between a C++ STL vector, and a Java Vector which happens to only contain Integers.
An additional problem with the Java Vector (and something that'll be fixed when templates are introduced in Java 1.5) is that it can't contain things which don't inherit from Object (like int and float).
(Note that if a compiler is very smart, it is theoretically possible for both methods to be equally efficient at runtime. But that degree of optimization is rare even today- and depending on the class linking style, it may be impossible for the compiler to get all the data needed to optimize like that.)