"One of the few popular languages with first-class functions"? Allow me to disagree but almost every dynamic language I know of has first-class functions.
Other languages, like C, C++, C#, also allow you to use functions as a data type.
I agree that, in some of them, their syntax does not make it easy to define functions as, for example, an argument, but you can define the function first, and pass it as an argument later if needed.
Closures are another thing altogether, but they are supported on many dynamic languages.
When I talk about an object oriented programming language I'm referring to a language that allows you to use the concepts of OOP in a *natural* and *homogeneous* way. I don't want to write a library and helper methods to write an OO program, I want to use the language.
It's OK if it doesn't has classes, and therefore inheritance does not have a place in Javascript, just stop trying to force it to be something that it was not meant to be (that is a general purpose language to write medium to large scale applications).
We already have a good number of established scripting languages that fill the niche. What does Javascript brings into this world that makes it interesting to work again on compilers, module platforms, optimizers, etc.?
The arguments posted in the article about what changed to consider moving javascript out of the browser, are in my opinion, pretty weak:
* We discovered AJAX: besides doesn't having anything to do with the argument, we might say that more than AJAX, browsers started to be a *little* more standard compliant, so designing complex HTML application became less painful. AJAX is really a so simple thing that I really don't believe is the responsible for our buzzed web 2.0 (besides... we always had iframes). Heavy support for CSS, fixing of layout issues, etc., that's what probably brought our web interfaces as we know them today. * Its included on every consumer computer: Yes, in the browser. You will not use the browser to run these javascript programs, since there are limitations, and for a reason, on what you can do from the browser. Probably, you will need to download a javascript runtime to execute this new javascript programs anyway. * Designers know how to program javascript: And that is why it's in the browser and not running on your server or free on your computer. Have you ever looked at the average javascript source code? People program as if they needed to save every byte on their source code, avoiding white space and having tons of a,b,c variables. Have they never heard of minimizers? And if size is such a problem, standarize an optimized intermediate representation instead.
What remains is that javascript is cool. That's probably right if you feel cool when you write ugly hacks to make things work.
And regarding problems with javascript to be used on large applications and not as glue code, I would say prototype programming is one of my main concerns. Weakly typed languages already have the disadvantage of lacking compile time type checking, and the difficulties to perform automated refactoring since you don't know to what a variable will refer. But with prototype languages you also add the difficulty to know what's the structure of an object.. In other dynamic languages you can also do it (ie. changing the structure of a class during runtime), but, being there and doing that, it's a probable road to "WTF is going on" (with exceptions, of course).
After reading a thread about why C++ wasn't used for developing git (basically, apparently it's used by bad programmers), I decided to take a look at git source code, first file I check and I find this:
http://gitorious.org/git/mainline/blobs/master/commit.c#line144
I suppose that one wasn't written by Linus, otherwise I would tend to say that even wiz programmers tend to misuse goto:P.
Agreed, every time I have to develop complex web based interfaces I feel like I have to hack my way into something that wasn't designed for what I need to do. Truly hate it. It seldom feels right and clean, and I do consider myself a clean programmer.
I don't think the issue is if you can do it with HTML or not, but wether you should. Its entretaining to push a technology to its limits and think you are a wizard, but its also a waste of time. I really hope XUL kicks off (its great, but its still hackish). It doesn't make sense to keep trying to fix HTML by adding new out of the hat components that try to make a file cabinet work as a database. They are just another hack.
well, not exactly our experience here. We have been trying to develop a system using XUL and JavaScript but we ran in multiple problems, that required almost always some kind of work around.
Don't get me wrong, I do think there's a LOT of potential in XUL. Specially, XBL has been wonderful to develop some data binded components that saved us a lot of time in the long run, and they were quite easy to develop.
However, theres a lack of maturity in some of it features that make them look as it they were an after thought, or just not really well designed. Templates have a lot of limitations, remote XUL just doesnt work right (or at least the docs that explain how to sign components are just too outdated), many errors result in segfaults (agree, the problem was in how we were using some components, but you shouldn't segfault in any case!!), theres no clear separation between components and services (I really don't understand why they don't share a common interface!), XBL lacks of obvious things (like a script tag, please). And honestly, javascript sucks to structure your code... we had to implement an include () directive to import other.js, there should be a way to do that in the language. And, there should be support for other datasources besides RDF.
And the worst part is that there are not many apps developed in XUL, so documentation (specially about XPCOM components) is really scarce. The reference is OK (most of the time), but you dont have a single example for most of the XPCOM components. A search in google usually returns C/IDL headers from the mozilla implementation or the reference at xulplanet;).
I really hope XUL development speeds up. Its really a good platform, and the separation between presentation/data/logic really shows up. But as it is, its darn difficult, and slow, to develop for it (at least till you know most of the work arounds).
By the way, has anyone developed both in Luxor and mozilla XUL and have some insights in their pros and cons? (besides what the luxor page says, which of course I wont believe until I try it:P).
In my experience, code is almost always on maintenance. Be it because someone (the client for instance) want to add new functionality, change existing one, or modify the environment under which the program is running .
New programming languages and paradigms (like oo, which is new in comparison to qbasic:)) aren't tools just for creating new software; they also make modifying and maintaining that software much easier. When your code grows and developes, i would like to see how you handle the beast using qbasic, and how much time will take you to do it.
Besides, the issue is not only with the client perception of the app, its with code development costs. Generally, it will be easier to code the app in modern languages than in qbasic.
Maybe it was because I was just a kid (well, a kid of 16 or so, but a kid nonetheless!:P), but I still have a vivid memory of the forest outside the hunter 's lodge at night, on day 6 of the game, entering the wolf's cave and finding the naked german covered in blood and chewing bones.... creepy. In fact the entire game was very creepy, with the flashbacks and all the black wulf legend thing.
I still consider GK2 one of the best games I played.
I didn't have a single problem installing NWN. Yes, the 'official' way to install it is using a separate windows installation. But the day after NWN for linux was released, a linux installer was posted for the latest NWN version that didn't need the windows installation.
Im a proud user of NWN, and I truly find that posting an article such as this is way to spread FUD, which clearly Bioware doesn't need nor deserve.
Kudos to Bioware, and thanks for taking the linux gaming community seriously!!!
Most people don't like anything that takes them out of their rutine. They give a damn about openess, and even size.
We have a multi-national client (a BIG one) which wanted to do a survey in our country. This survey its done in a lot of diferent countries, and each country can present the results anyway they like (in fact they even encourage the presentation to be diferent so it can accomodate to each situation) but all of them have to send the results to their headquarters in France for revision.
Our idea was to automatically generate the presentations from the survey data (which was obvious from looking at the reports), creating a really nice looking, and easy to use, document in HTML. The guy who where conducting the survey liked it a lot, so we developed the app. After it was done and we sent it for revision, they told us that they had some trouble reading them (we specially designed it so it could be opened with both netscape > 4 and ie > 4.1, so maybe they used lynx, i dont know) and why it was that we couldnt send the report in powerpoint as every other country did.
We tried to explained them how much easier it was to generate HTML than to generate a PPT, how much plataform independent, etc. etc.. Their answer was that they didnt care, the rest of the countries sent it in PPT, so we should too.
The bottom line is that we had to cut and paste most of the document into PPT slides (because PowerPoint has a lot of trouble importing HTML with many tables). Oh, and i forgot to mention it, the PPT document is almost 5MB (because we have to save it as a previous PPT version).
Even if its probably possible, its a bit hard to define a coding style for a language whose motto is "There's more than one way to do it". Probably the amount of things you should define to have code that looks similar its BIG (where should you use fors instead of maps? should i use $_ or assign it to something else? etc., etc., etc.)...
Its easy to define *your* coding style, but imposse it to your fellow coders might be harder (always speaking about perl). Anyway, you might like to read the "Common practices" chapter in Programming in Perl where Larry gives some good advices about coding style in perl....
My suggestion, if you think the code might start to get bigger... start slowly to switch to another language, 30K lines of perl code just gives me the creeps...
The big question is why we should try to improve current humans with artificial components?
When someone says that artificial intelligence is the next evolutionary step, and that machines will eventually make humans obsolete, people always say that then we should try to improve humans by adding machine components instead of making robots. I can't really say where's the difference, machine will probably be more efficient than biological components (correct me if im wrong), so why do humanity wants to stay so badly? Because we might loose control? Control to do what? We have to stop seeing humanity as the most important piece in the universe, and its final objective.
I think that now that we can start thinking and talking about true AI we should consider that maybe that is our objective, that that is what we are here to do, and what happens after that with humanity is just a sideline in history. Why is what we are here to do? Well development of human level AI its the only event in which i can think of that would make humanity absolutely obsolete, its the only event in which we can say if this was our objective, then we have achieved it, and if there was some bigger one, then our "descendents" will be more fit to achieve it than us. If we are here to do something, then we should do it and stop throw away all the "i might loose control" thing, if we are not here to do something, but just to exists, then well, it doesnt matter what we do, it will be right (lewis carroll anyone?).
If you really think that hapiness can be an end in itself, i suggest you read Aldus Huxley's Brave New World. Even if what he describres will never happen by far, the most obvious conclusion of the story is that happiness can't be a goal in itself, because if the goal of humanity lies in humanity itself it just means that the problem that humanity is trying to solve exists only because humanity exists, so just wipe out humanity and you solved the problem. Its as saying that individual humans do have a purpose, making everyone happy, but humanity as itself doesn't. But if you believe in something greater, that goes beyond your own and humans existance, then humanity might have a purpose (and btw, that something greater might be the development of artificial intelligence, im not talking about a god here...). I belive humans should try to define what do they think, and why, its humanity objective in the universe (why its here), and then try to act to achieve that objective. If the object of humanity is humanity, then humanity doesn't exists outside humanity;);).
Of course, what we do is based on selfishness. It doesn't matter if you do it to feel good or help somebody else, you are always doing things because you have an objective that you want and need to achieve, once that objective is defined you act to get the best value in an utility function to get to that objective. The big question is what objectives are rational ones. Its ok if you think you have to act to help people, but you need to answer why you do it, and if that answer make sense at all. People are often trying to have something define they more basic moral rules so they can build their right/bad actions graph from that solid ground (what it is fact what most religions and societies do), but you can obtain the same results by defining whats is your goal, and building the graph backwards, from the goal you want to achieve, to the most basic moral rules. The problem is that people always have a hard time thinking about goals and objectives. Specially when it comes to define "the" goal. Sometimes its easy to think that our responsability its just with ourselves, so helping ourselves its the only goal. I believe that there are some greater truths, truthts of which humanity is just a small part of their actions.
People often say that moral and ethics are not rational. I think they are extremely rational, based in a deep understanding of tit-for-tat strategies, so deep understanding that we can see why its the best strategy and how we have to handle its rules to avoid getting forever locked into attack/attack situations.
The same thing happened when he came to Uruguay, and suggested us to change the UYLUG name. We discussed it, voted, and stayed with UYLUG (2 against 17).
However, i don't consider that it was out of place from his part, we all know, more or less, Stallman's views, and we probably would have been suprised if he didnt asked!
Stallman is not a politician, if he was one, he would try to be as social as he could, try to make everyone happy, and leave all that ethical issues out of the picture. He's just really convinced of his points of view, and you will never, never, see him contradict himself on those issues. And that is more than i can ask from a politician.
You can be with, or against Stallman's ideas, but you can't call him unethical. And i wouldn't call egoistical a man who dedicated more than 15 years of his life entirely to advocate and make Free Software a good alternative.
Its very funny, but probably because the pathetical extremists of the XX century we now consider that absolutely all extremes are bad, and in fact we have to be open minded and don't take sides on any issue and consider that all is relative. There are things you must be able to consider absolute, if you get by thinking and reflection to (for you) an unavoidable truth (as happened with Stallman), then you have to go for it. Of course, it might be false, you might be wrong, but until someone shows you you are actually wrong, you have to stay with that course of action, and that implies saying things that might disgust someone.
People always says that open source problem lies in that there are too many options, that too many GUIs confuse users and multiplicate the work. But having many options is the heart of the movement. What i hate most about microsoft products is that they turn a personal computer into an electrodomestic, theres only one way to do things, and that makes ms community (if theres such a thing) to be much more static and subjugated to ms decisions of "innovation".
I admit its hard to track of all the new versions of open source projects, keep updating the tools, but i think that is that constant evolution, branching, and merging of projects what its most promising of the open source model of development, because it makes me believe that progress is always on the way, and that is a question of time for a new revolutionary way to do things to appear.
And about the marketing things, well, i dont believe that the objective of open source its to get to the mainstream. Its objective is develop better and independent software, getting to the mainstream might help to do it, but its a medium, not the goal. If the mainstream decides not to use the awesome pieces of software produced by open source, its a problem of the mainstream, im not going to feel bad because people are stupid.
While we have the resources to keep evolving software for free, the open source movement is safe. If people believe more in marketing than in results, its their choice.
Software engeering is not occurring at the rate of Moore's Law, but technological milestones are happening at smaller and smaller intervals. And i dont know why you believe we can have perfect hardware but flawed software, if that was the case, then we just need to implement the AI seed on hardware and we would have solved the problem.
Why its more reasonable to devote energy in enhancing our biological hardware than in building a new kind of hardware? If you get a better specie, biological or silicon based, its the same. But there are a lot of benefits in the digital approach, probably more than in the biological one (practical and theorical). Just think what would happend if someone suggested to use some embryos to test some probably usefull mutations... or reprogramm some guy mind to test a new feature. Of course, there will be some machines right activists, but that will take longer;)
The singularity that these guys are talking about is far more important, mainly because in case of getting into it we have changed forever the position of mankind.
That thing about the future, well, there are a lot of things that we can't stop because we just dont know exactly what produces them. A lot of the emergent phenomena that we generate as a specie probably respond to laws that we can't yet explain, and, for that reason, can't yet avoid. Why you said the thing about "as long real people are in control we can't go wrong" really amazes me. Humanity went wrong a LOT of times, the least that we can expect is that somone more intelligent will at least go wrong less.
The fifth one sounded like FUD, anti-humans? are they anti-themselves? well, i think they just believe that there might be something more important that humans. In fact i dont see where is the arrogance in saying that humans are just a small piece in a bigger mechanism, its sounds more arrogant to me to think that the objective of humanity lies in humanity.
The other arguments that the manifesto gives, like that since software is so buggy we can't think about developing a human level AI, well, its just a temporary one, and it doesnt goes against the fact that sometime maybe it will not be that buggy. And its important to see if people want to go that way, or prefer to avoid all that branch of science (like Joy).
Take in account that in Kurzweil's view, the singularity is not only a matter of better hardware, but of development in all the areas of research. So, according to his point of view, hardware and software will get caught in that singularity.
I do believe that the singularity actually has a more specific name, the development of artficial human level intelligence. If sometime humans get to that, then we can say that probably the objective of humanity have been acomplished, in developing something that is better than us in that that makes us different, and superior, from the rest of the species, and therefore leaving the world in a better situation that when we found it;) (so making BP happy). We have improved ourselves. If we do get to human level AI, then humanity is obsolete, and more than being afraid we should be proud (in spanish we could say that is "el fin de la humanidad", but "el fin" in both senses).
Humans being are very inefficient in a lot of computation tasks, but that can be improved much faster if we have a digital AI than our biological brains. We could try some optimization, and if suddenly we go nuts we could restore the backup;).
People are extremely afraid of being superated, they would prefer (like Joy) to don't develop some areas of research because that may make humans an obsolete technology, well, in my opinion we are just some very small part of a higher scheme (by who? who knows, maybe evolution), and if we can make something that is actually better and more fit than humans, please do it.
I dont think that it's a misconception at all. Perl is easy to learn. You dont have many constructs, in fact just three different types of data to do everything, and the C like loop constructs and operands. If you want to program in perl in a similar way than in C you can do it with few modifications. Obviously, you would be using a chainsaw as a scissor....
"Programming Perl" isn't a language reference. I learned perl reading it and IMHO its the best way to grasp the idea of what perl is about. Why? Because it keeps you moving, changing, showing you diferent things at the same time, it helps you to start thinking "perl". A more orderly book would probably produce programmers that wouldn't get the idea of what TMTOWTDI is about.
I know a good number of languages, but i didnt find one that gaves me such an expressive power as perl. I dont always use it anyway, because sometimes i do belive that i need more control on my code. But i can't understand why some people dont like perl. Its the only programming language in wich i can say more things in a sentence that in a full page of natural language...
Exolab's Castor JDO is not the same as Sun JDOm which seems to be a really usefull initiative... anyway Castor scheme its quite good, but i dont like the idea of doing absolutely everything on runtime (like deciding which tables to use), and probably the idea of that your base classes are the ones that know how to persist themselves is much more appealing that castor's database object...
Besides, Sun's JDO will be usefull to persist to any local storage system, from indexed files to complex OORDBM. Of course, rigth not its not more than a review draft;).
BTW, what about object relational databases, where do they fall short? do they?. Systems like Oracle and PostgreSQL are o/r, but some of the people here consider them OO, are they as good as pure OO or not??
I know this is a bit old, but i enjoy the discussion:):).
In the turing's test the judge knows that one of the terminals is being used by a computer. In the eliza case they didnt. If i make a bot that sends messages saying "uh uh i need pics!!" theres no reason for someone to think that it's a computer;). The turing's test relies on that human intelligence is the only intelligence that we accept to exists without any doubt. Therefore the only thing against we can compare a new system to see if we can call it intelligent is human intelligence. Again, if a computer pass a "fair" turing's test, then we should consider it intelligent.
Im not pulling buzz phrases;) (but however, yes im just a newbie in the study of this things;), fuzzy logic is uncertain logic. That means that accepts propositions, and its able to make inferences, that can't be classified as true or false. I dont know why do you say that its "the study of case statements with (or sometimes without) knowledge of one or more previous cases".
As if its new or not... well i would say its quite a new concept (not multi-valued logic in itself, but the special case of fuzzy logic as proposed by Zadeh, 1965 if im right). In fact, neural networks follow the same rules that any other program, and the fact that you can make a neural network on a digital computer means that you dont need another kind of computer for making them!:) You can probalby make a more efficient computer to handle nn specifically, but it would be equals in capacity to the previous one but with more processing power:). If theres a way to implement a mechanical AI, then it can be done by actual pc. If theres not, then well, thats the important discussion anyway!:)
Of course you dont need "only" processing power. What i meant is that with current techniques and enough processing power something similar to intelligence can be achieved (the comment was regarding what you said, that for obtaining stable structures you would need to pass countless generations on genetic algoritms). And in fact Minsky is a defender of "strong AI", and probably would agree with the idea that is possible to achieve human level intelligence on computers.
I disagree. You have to know where you're going, when you're alone in a vast desert looking for the one oasis. You have a small chance of stumbling on it, but I would not bet on it.
Thats one way to look at it. The other is that intelligence is the sand, but you ignore that intelligence is the sand because you think that the only thing that can be called intelligence is what you have;).
I believe that there are diferent levels of intelligence. If what we are talking here is about getting to the level of human intelligence, then we do have a goal.
What we need then is a way to check when we have reached human intelligence. I agree that turing test its not a very good test, in the sense that is a sufficient test but not a necesary one. If a machine can pass turing test, well, then probably we could say that it has a similar level of intelligence to that of a human.
I talked about quantum computers because they are order of magnitude faster than standard computers (if you have enough qbits at least). You can emulate non-boolean logic with standard computers, and probably it will be necesary to use non boolean logic (e.g. fuzzy logic) to create an AI.
For me its a problem of processing power. Of course, you might need more processing power than that of a human brain to create a human brain, but it all comes to processing. And if the growth of processing speed continues to go as fast as kurzweill says, then we probably will be alive when a computer reaches human intelligence.
About the unstability of recursion in intelligence, i dont know what do you mean exactly by unstability, but in the case you mean unpredictability, well, that one of the things that an intelligent being should have, at least in some way. In the case you mean that, becuase of its complexity, it could get in an infinite recursion loop, well, we humans do have the same problem, but we usually get tired before ending the recursion;);) (in fact, our recursion is so bad that we have enormous problems in defining what are we and what we call human intelligence, just kidding:)).
1. Creating a machine with the "prcessing power of the human brain" (whatever that is) does not imply that it will be intelligent.
True, but you also have the technology called "intrusive/non-intrusive brain scanning", which, according to ray, would allow us to "download" our brain structure into computers. Yeah, I considered the idea a bit weird too when i read it.
2. We have yet to define where the line between intelligence and sentience is.
You dont have to define the line to cross it. Maybe we already crossed it and we dont know we did it;) (but i really doubt it...).
3. We don't yet know how to produce even intelligence (say, on the order of a cat).
Well, if you take a look at some of the existing robots (i dont remember the name of the one im talking about...) you already have dog-robots which, in my opinion, act very similar to organic ones. Of course they follow rules programmed into them, but those rules have an enormous complexity which dont allow you to predict what would be their next action.
4. It's quite possible that sentience requires a very different kind of computer than we have (e.g. non-Von Neuman architectures).
This could be true, but quantum computers are on the way:):).
This problem doesn't have anything to do with the three hat problem, and it's just a case of bad reasoning.
The three guys paid 10 dollars each, and the girl gives one dollar back to each one, so you have 27 dollars, what is exactly what they paid. 25 to the hotel owner and two more for the woman. The other three dollars are in posesion of the guys, so there's no problem. The error lies in adding the 2 dollars to the 27, cos they are already included in those 27.
Slashdot Mirror
"One of the few popular languages with first-class functions"? Allow me to disagree but almost every dynamic language I know of has first-class functions.
Other languages, like C, C++, C#, also allow you to use functions as a data type.
I agree that, in some of them, their syntax does not make it easy to define functions as, for example, an argument, but you can define the function first, and pass it as an argument later if needed.
Closures are another thing altogether, but they are supported on many dynamic languages.
You mean the Douglas Crockford that wrote http://www.crockford.com/javascript/private.html?
When I talk about an object oriented programming language I'm referring to a language that allows you to use the concepts of OOP in a *natural* and *homogeneous* way. I don't want to write a library and helper methods to write an OO program, I want to use the language.
It's OK if it doesn't has classes, and therefore inheritance does not have a place in Javascript, just stop trying to force it to be something that it was not meant to be (that is a general purpose language to write medium to large scale applications).
We already have a good number of established scripting languages that fill the niche. What does Javascript brings into this world that makes it interesting to work again on compilers, module platforms, optimizers, etc.?
The arguments posted in the article about what changed to consider moving javascript out of the browser, are in my opinion, pretty weak:
* We discovered AJAX: besides doesn't having anything to do with the argument, we might say that more than AJAX, browsers started to be a *little* more standard compliant, so designing complex HTML application became less painful. AJAX is really a so simple thing that I really don't believe is the responsible for our buzzed web 2.0 (besides... we always had iframes). Heavy support for CSS, fixing of layout issues, etc., that's what probably brought our web interfaces as we know them today.
* Its included on every consumer computer: Yes, in the browser. You will not use the browser to run these javascript programs, since there are limitations, and for a reason, on what you can do from the browser. Probably, you will need to download a javascript runtime to execute this new javascript programs anyway.
* Designers know how to program javascript: And that is why it's in the browser and not running on your server or free on your computer. Have you ever looked at the average javascript source code? People program as if they needed to save every byte on their source code, avoiding white space and having tons of a,b,c variables. Have they never heard of minimizers? And if size is such a problem, standarize an optimized intermediate representation instead.
What remains is that javascript is cool. That's probably right if you feel cool when you write ugly hacks to make things work.
And regarding problems with javascript to be used on large applications and not as glue code, I would say prototype programming is one of my main concerns. Weakly typed languages already have the disadvantage of lacking compile time type checking, and the difficulties to perform automated refactoring since you don't know to what a variable will refer. But with prototype languages you also add the difficulty to know what's the structure of an object.. In other dynamic languages you can also do it (ie. changing the structure of a class during runtime), but, being there and doing that, it's a probable road to "WTF is going on" (with exceptions, of course).
An instant classic! It kind of reminds me of Dragon Strike, without the skeletons of course.
After reading a thread about why C++ wasn't used for developing git (basically, apparently it's used by bad programmers), I decided to take a look at git source code, first file I check and I find this: :P.
http://gitorious.org/git/mainline/blobs/master/commit.c#line144
I suppose that one wasn't written by Linus, otherwise I would tend to say that even wiz programmers tend to misuse goto
Agreed, every time I have to develop complex web based interfaces I feel like I have to hack my way into something that wasn't designed for what I need to do. Truly hate it. It seldom feels right and clean, and I do consider myself a clean programmer.
I don't think the issue is if you can do it with HTML or not, but wether you should. Its entretaining to push a technology to its limits and think you are a wizard, but its also a waste of time. I really hope XUL kicks off (its great, but its still hackish). It doesn't make sense to keep trying to fix HTML by adding new out of the hat components that try to make a file cabinet work as a database. They are just another hack.
well, not exactly our experience here. We have been trying to develop a system using XUL and JavaScript but we ran in multiple problems, that required almost always some kind of work around.
Don't get me wrong, I do think there's a LOT of potential in XUL. Specially, XBL has been wonderful to develop some data binded components that saved us a lot of time in the long run, and they were quite easy to develop.
However, theres a lack of maturity in some of it features that make them look as it they were an after thought, or just not really well designed. Templates have a lot of limitations, remote XUL just doesnt work right (or at least the docs that explain how to sign components are just too outdated), many errors result in segfaults (agree, the problem was in how we were using some components, but you shouldn't segfault in any case!!), theres no clear separation between components and services (I really don't understand why they don't share a common interface!), XBL lacks of obvious things (like a script tag, please). And honestly, javascript sucks to structure your code... we had to implement an include () directive to import other .js, there should be a way to do that in the language. And, there should be support for other datasources besides RDF.
And the worst part is that there are not many apps developed in XUL, so documentation (specially about XPCOM components) is really scarce. The reference is OK (most of the time), but you dont have a single example for most of the XPCOM components. A search in google usually returns C/IDL headers from the mozilla implementation or the reference at xulplanet ;).
I really hope XUL development speeds up. Its really a good platform, and the separation between presentation/data/logic really shows up. But as it is, its darn difficult, and slow, to develop for it (at least till you know most of the work arounds).
By the way, has anyone developed both in Luxor and mozilla XUL and have some insights in their pros and cons? (besides what the luxor page says, which of course I wont believe until I try it :P).
In my experience, code is almost always on maintenance. Be it because someone (the client for instance) want to add new functionality, change existing one, or modify the environment under which the program is running . New programming languages and paradigms (like oo, which is new in comparison to qbasic :)) aren't tools just for creating new software; they also make modifying and maintaining that software much easier. When your code grows and developes, i would like to see how you handle the beast using qbasic, and how much time will take you to do it.
Besides, the issue is not only with the client perception of the app, its with code development costs. Generally, it will be easier to code the app in modern languages than in qbasic.
Maybe it was because I was just a kid (well, a kid of 16 or so, but a kid nonetheless! :P), but I still have a vivid memory of the forest outside the hunter 's lodge at night, on day 6 of the game, entering the wolf's cave and finding the naked german covered in blood and chewing bones.... creepy. In fact the entire game was very creepy, with the flashbacks and all the black wulf legend thing.
I still consider GK2 one of the best games I played.
I didn't have a single problem installing NWN. Yes, the 'official' way to install it is using a separate windows installation. But the day after NWN for linux was released, a linux installer was posted for the latest NWN version that didn't need the windows installation.
Im a proud user of NWN, and I truly find that posting an article such as this is way to spread FUD, which clearly Bioware doesn't need nor deserve.
Kudos to Bioware, and thanks for taking the linux gaming community seriously!!!
Its amazing what a single comma might do...
Most people don't like anything that takes them out of their rutine. They give a damn about openess, and even size.
We have a multi-national client (a BIG one) which wanted to do a survey in our country. This survey its done in a lot of diferent countries, and each country can present the results anyway they like (in fact they even encourage the presentation to be diferent so it can accomodate to each situation) but all of them have to send the results to their headquarters in France for revision.
Our idea was to automatically generate the presentations from the survey data (which was obvious from looking at the reports), creating a really nice looking, and easy to use, document in HTML. The guy who where conducting the survey liked it a lot, so we developed the app. After it was done and we sent it for revision, they told us that they had some trouble reading them (we specially designed it so it could be opened with both netscape > 4 and ie > 4.1, so maybe they used lynx, i dont know) and why it was that we couldnt send the report in powerpoint as every other country did.
We tried to explained them how much easier it was to generate HTML than to generate a PPT, how much plataform independent, etc. etc.. Their answer was that they didnt care, the rest of the countries sent it in PPT, so we should too.
The bottom line is that we had to cut and paste most of the document into PPT slides (because PowerPoint has a lot of trouble importing HTML with many tables). Oh, and i forgot to mention it, the PPT document is almost 5MB (because we have to save it as a previous PPT version).
bwaaaahahahaha!!!
Even if its probably possible, its a bit hard to define a coding style for a language whose motto is "There's more than one way to do it". Probably the amount of things you should define to have code that looks similar its BIG (where should you use fors instead of maps? should i use $_ or assign it to something else? etc., etc., etc.)...
Its easy to define *your* coding style, but imposse it to your fellow coders might be harder (always speaking about perl). Anyway, you might like to read the "Common practices" chapter in Programming in Perl where Larry gives some good advices about coding style in perl....
My suggestion, if you think the code might start to get bigger... start slowly to switch to another language, 30K lines of perl code just gives me the creeps...
The big question is why we should try to improve current humans with artificial components?
When someone says that artificial intelligence is the next evolutionary step, and that machines will eventually make humans obsolete, people always say that then we should try to improve humans by adding machine components instead of making robots. I can't really say where's the difference, machine will probably be more efficient than biological components (correct me if im wrong), so why do humanity wants to stay so badly? Because we might loose control? Control to do what? We have to stop seeing humanity as the most important piece in the universe, and its final objective.
I think that now that we can start thinking and talking about true AI we should consider that maybe that is our objective, that that is what we are here to do, and what happens after that with humanity is just a sideline in history. Why is what we are here to do? Well development of human level AI its the only event in which i can think of that would make humanity absolutely obsolete, its the only event in which we can say if this was our objective, then we have achieved it, and if there was some bigger one, then our "descendents" will be more fit to achieve it than us. If we are here to do something, then we should do it and stop throw away all the "i might loose control" thing, if we are not here to do something, but just to exists, then well, it doesnt matter what we do, it will be right (lewis carroll anyone?).
Santiago Aguiar
If you really think that hapiness can be an end in itself, i suggest you read Aldus Huxley's Brave New World. Even if what he describres will never happen by far, the most obvious conclusion of the story is that happiness can't be a goal in itself, because if the goal of humanity lies in humanity itself it just means that the problem that humanity is trying to solve exists only because humanity exists, so just wipe out humanity and you solved the problem. Its as saying that individual humans do have a purpose, making everyone happy, but humanity as itself doesn't. But if you believe in something greater, that goes beyond your own and humans existance, then humanity might have a purpose (and btw, that something greater might be the development of artificial intelligence, im not talking about a god here...). I belive humans should try to define what do they think, and why, its humanity objective in the universe (why its here), and then try to act to achieve that objective. If the object of humanity is humanity, then humanity doesn't exists outside humanity ;);).
Of course, what we do is based on selfishness. It doesn't matter if you do it to feel good or help somebody else, you are always doing things because you have an objective that you want and need to achieve, once that objective is defined you act to get the best value in an utility function to get to that objective. The big question is what objectives are rational ones. Its ok if you think you have to act to help people, but you need to answer why you do it, and if that answer make sense at all. People are often trying to have something define they more basic moral rules so they can build their right/bad actions graph from that solid ground (what it is fact what most religions and societies do), but you can obtain the same results by defining whats is your goal, and building the graph backwards, from the goal you want to achieve, to the most basic moral rules. The problem is that people always have a hard time thinking about goals and objectives. Specially when it comes to define "the" goal. Sometimes its easy to think that our responsability its just with ourselves, so helping ourselves its the only goal. I believe that there are some greater truths, truthts of which humanity is just a small part of their actions.
People often say that moral and ethics are not rational. I think they are extremely rational, based in a deep understanding of tit-for-tat strategies, so deep understanding that we can see why its the best strategy and how we have to handle its rules to avoid getting forever locked into attack/attack situations.
Santiago Aguiar
The same thing happened when he came to Uruguay, and suggested us to change the UYLUG name. We discussed it, voted, and stayed with UYLUG (2 against 17).
However, i don't consider that it was out of place from his part, we all know, more or less, Stallman's views, and we probably would have been suprised if he didnt asked!
Stallman is not a politician, if he was one, he would try to be as social as he could, try to make everyone happy, and leave all that ethical issues out of the picture. He's just really convinced of his points of view, and you will never, never, see him contradict himself on those issues. And that is more than i can ask from a politician.
You can be with, or against Stallman's ideas, but you can't call him unethical. And i wouldn't call egoistical a man who dedicated more than 15 years of his life entirely to advocate and make Free Software a good alternative.
Its very funny, but probably because the pathetical extremists of the XX century we now consider that absolutely all extremes are bad, and in fact we have to be open minded and don't take sides on any issue and consider that all is relative. There are things you must be able to consider absolute, if you get by thinking and reflection to (for you) an unavoidable truth (as happened with Stallman), then you have to go for it. Of course, it might be false, you might be wrong, but until someone shows you you are actually wrong, you have to stay with that course of action, and that implies saying things that might disgust someone.
People always says that open source problem lies in that there are too many options, that too many GUIs confuse users and multiplicate the work. But having many options is the heart of the movement. What i hate most about microsoft products is that they turn a personal computer into an electrodomestic, theres only one way to do things, and that makes ms community (if theres such a thing) to be much more static and subjugated to ms decisions of "innovation".
I admit its hard to track of all the new versions of open source projects, keep updating the tools, but i think that is that constant evolution, branching, and merging of projects what its most promising of the open source model of development, because it makes me believe that progress is always on the way, and that is a question of time for a new revolutionary way to do things to appear.
And about the marketing things, well, i dont believe that the objective of open source its to get to the mainstream. Its objective is develop better and independent software, getting to the mainstream might help to do it, but its a medium, not the goal. If the mainstream decides not to use the awesome pieces of software produced by open source, its a problem of the mainstream, im not going to feel bad because people are stupid.
While we have the resources to keep evolving software for free, the open source movement is safe. If people believe more in marketing than in results, its their choice.
Software engeering is not occurring at the rate of Moore's Law, but technological milestones are happening at smaller and smaller intervals. And i dont know why you believe we can have perfect hardware but flawed software, if that was the case, then we just need to implement the AI seed on hardware and we would have solved the problem.
;)
Why its more reasonable to devote energy in enhancing our biological hardware than in building a new kind of hardware? If you get a better specie, biological or silicon based, its the same. But there are a lot of benefits in the digital approach, probably more than in the biological one (practical and theorical). Just think what would happend if someone suggested to use some embryos to test some probably usefull mutations... or reprogramm some guy mind to test a new feature. Of course, there will be some machines right activists, but that will take longer
The singularity that these guys are talking about is far more important, mainly because in case of getting into it we have changed forever the position of mankind.
That thing about the future, well, there are a lot of things that we can't stop because we just dont know exactly what produces them. A lot of the emergent phenomena that we generate as a specie probably respond to laws that we can't yet explain, and, for that reason, can't yet avoid. Why you said the thing about "as long real people are in control we can't go wrong" really amazes me. Humanity went wrong a LOT of times, the least that we can expect is that somone more intelligent will at least go wrong less.
The fifth one sounded like FUD, anti-humans? are they anti-themselves? well, i think they just believe that there might be something more important that humans. In fact i dont see where is the arrogance in saying that humans are just a small piece in a bigger mechanism, its sounds more arrogant to me to think that the objective of humanity lies in humanity.
The other arguments that the manifesto gives, like that since software is so buggy we can't think about developing a human level AI, well, its just a temporary one, and it doesnt goes against the fact that sometime maybe it will not be that buggy. And its important to see if people want to go that way, or prefer to avoid all that branch of science (like Joy).
Santiago
-
Take in account that in Kurzweil's view, the singularity is not only a matter of better hardware, but of development in all the areas of research. So, according to his point of view, hardware and software will get caught in that singularity.
;) (so making BP happy). We have improved ourselves. If we do get to human level AI, then humanity is obsolete, and more than being afraid we should be proud (in spanish we could say that is "el fin de la humanidad", but "el fin" in both senses).
;).
I do believe that the singularity actually has a more specific name, the development of artficial human level intelligence. If sometime humans get to that, then we can say that probably the objective of humanity have been acomplished, in developing something that is better than us in that that makes us different, and superior, from the rest of the species, and therefore leaving the world in a better situation that when we found it
Humans being are very inefficient in a lot of computation tasks, but that can be improved much faster if we have a digital AI than our biological brains. We could try some optimization, and if suddenly we go nuts we could restore the backup
People are extremely afraid of being superated, they would prefer (like Joy) to don't develop some areas of research because that may make humans an obsolete technology, well, in my opinion we are just some very small part of a higher scheme (by who? who knows, maybe evolution), and if we can make something that is actually better and more fit than humans, please do it.
Santiago
-
I dont think that it's a misconception at all. Perl is easy to learn. You dont have many constructs, in fact just three different types of data to do everything, and the C like loop constructs and operands. If you want to program in perl in a similar way than in C you can do it with few modifications. Obviously, you would be using a chainsaw as a scissor....
"Programming Perl" isn't a language reference. I learned perl reading it and IMHO its the best way to grasp the idea of what perl is about. Why? Because it keeps you moving, changing, showing you diferent things at the same time, it helps you to start thinking "perl". A more orderly book would probably produce programmers that wouldn't get the idea of what TMTOWTDI is about.
I know a good number of languages, but i didnt find one that gaves me such an expressive power as perl. I dont always use it anyway, because sometimes i do belive that i need more control on my code. But i can't understand why some people dont like perl. Its the only programming language in wich i can say more things in a sentence that in a full page of natural language...
Santiago
-
Exolab's Castor JDO is not the same as Sun JDOm which seems to be a really usefull initiative... anyway Castor scheme its quite good, but i dont like the idea of doing absolutely everything on runtime (like deciding which tables to use), and probably the idea of that your base classes are the ones that know how to persist themselves is much more appealing that castor's database object...
;).
Besides, Sun's JDO will be usefull to persist to any local storage system, from indexed files to complex OORDBM. Of course, rigth not its not more than a review draft
BTW, what about object relational databases, where do they fall short? do they?. Systems like Oracle and PostgreSQL are o/r, but some of the people here consider them OO, are they as good as pure OO or not??
Santiago
-
I know this is a bit old, but i enjoy the discussion :):).
;). The turing's test relies on that human intelligence is the only intelligence that we accept to exists without any doubt. Therefore the only thing against we can compare a new system to see if we can call it intelligent is human intelligence. Again, if a computer pass a "fair" turing's test, then we should consider it intelligent.
;) (but however, yes im just a newbie in the study of this things ;), fuzzy logic is uncertain logic. That means that accepts propositions, and its able to make inferences, that can't be classified as true or false. I dont know why do you say that its "the study of case statements with (or sometimes without) knowledge of one or more previous cases".
:) You can probalby make a more efficient computer to handle nn specifically, but it would be equals in capacity to the previous one but with more processing power :). If theres a way to implement a mechanical AI, then it can be done by actual pc. If theres not, then well, thats the important discussion anyway! :)
In the turing's test the judge knows that one of the terminals is being used by a computer. In the eliza case they didnt. If i make a bot that sends messages saying "uh uh i need pics!!" theres no reason for someone to think that it's a computer
Im not pulling buzz phrases
As if its new or not... well i would say its quite a new concept (not multi-valued logic in itself, but the special case of fuzzy logic as proposed by Zadeh, 1965 if im right). In fact, neural networks follow the same rules that any other program, and the fact that you can make a neural network on a digital computer means that you dont need another kind of computer for making them!
Of course you dont need "only" processing power. What i meant is that with current techniques and enough processing power something similar to intelligence can be achieved (the comment was regarding what you said, that for obtaining stable structures you would need to pass countless generations on genetic algoritms). And in fact Minsky is a defender of "strong AI", and probably would agree with the idea that is possible to achieve human level intelligence on computers.
Santiago
-
I disagree. You have to know where you're going, when you're alone in a vast desert looking for the one oasis. You have a small chance of stumbling on it, but I would not bet on it.
;).
;);) (in fact, our recursion is so bad that we have enormous problems in defining what are we and what we call human intelligence, just kidding :)).
Thats one way to look at it. The other is that intelligence is the sand, but you ignore that intelligence is the sand because you think that the only thing that can be called intelligence is what you have
I believe that there are diferent levels of intelligence. If what we are talking here is about getting to the level of human intelligence, then we do have a goal.
What we need then is a way to check when we have reached human intelligence. I agree that turing test its not a very good test, in the sense that is a sufficient test but not a necesary one. If a machine can pass turing test, well, then probably we could say that it has a similar level of intelligence to that of a human.
I talked about quantum computers because they are order of magnitude faster than standard computers (if you have enough qbits at least). You can emulate non-boolean logic with standard computers, and probably it will be necesary to use non boolean logic (e.g. fuzzy logic) to create an AI.
For me its a problem of processing power. Of course, you might need more processing power than that of a human brain to create a human brain, but it all comes to processing. And if the growth of processing speed continues to go as fast as kurzweill says, then we probably will be alive when a computer reaches human intelligence.
About the unstability of recursion in intelligence, i dont know what do you mean exactly by unstability, but in the case you mean unpredictability, well, that one of the things that an intelligent being should have, at least in some way. In the case you mean that, becuase of its complexity, it could get in an infinite recursion loop, well, we humans do have the same problem, but we usually get tired before ending the recursion
Santiago
-
1. Creating a machine with the "prcessing power of the human brain" (whatever that is) does not imply that it will be intelligent.
;) (but i really doubt it...).
:):).
True, but you also have the technology called "intrusive/non-intrusive brain scanning", which, according to ray, would allow us to "download" our brain structure into computers. Yeah, I considered the idea a bit weird too when i read it.
2. We have yet to define where the line between intelligence and sentience is.
You dont have to define the line to cross it. Maybe we already crossed it and we dont know we did it
3. We don't yet know how to produce even intelligence (say, on the order of a cat).
Well, if you take a look at some of the existing robots (i dont remember the name of the one im talking about...) you already have dog-robots which, in my opinion, act very similar to organic ones. Of course they follow rules programmed into them, but those rules have an enormous complexity which dont allow you to predict what would be their next action.
4. It's quite possible that sentience requires a very different kind of computer than we have (e.g. non-Von Neuman architectures).
This could be true, but quantum computers are on the way
Santiago
-
This problem doesn't have anything to do with the three hat problem, and it's just a case of bad reasoning.
The three guys paid 10 dollars each, and the girl gives one dollar back to each one, so you have 27 dollars, what is exactly what they paid. 25 to the hotel owner and two more for the woman. The other three dollars are in posesion of the guys, so there's no problem. The error lies in adding the 2 dollars to the 27, cos they are already included in those 27.
Santiago
-