I had mod points, and as a Firebird user, I was going to mod up the grandparent -- Firebird is very often ignored despite a host of positive features.
But as you had a bad experience, and you link to your problem list, I thought I'd respond instead. Better to answer questions than just mod up friendlies.
- Database path: Yes, firebird supports aliases. Our app doesn't use them, but they're there. - ISQL: I hear Oracle's SQL*Plus isn't much better. However, I use isql every once in a while, and I have command-history, backspacing, etc. available to me. From what I hear, it's more of a function of the shell you use (around isql) than isql itself. If you set up your environment properly, isql and its ilk automatically get command-history. (That's what I'm told, note. Anyone who can explain this is welcome to. I'm not a sysadmin.) - Never seen it freeze. - Corruption: we've had exactly one database issue, where it seems a backup/restore script ran in the middle of the day, restoring the database to its state from 4 hours earlier. In 4 years of use, with somewhere around 60 users in a medical clinic/insurance/billing environment, we've had no corruption. Using forced-writes is important, however. The careful-write strategy is really, really reliable, but it still can't protect you from faulty hard drives or operating systems that refuse to send data to the disk in the order requested (cf. Windows). M1 Abrams tank story, anyone? - IBDataPump and other third-party tools exist for some of the other features you're interested in. I'm not sure I know how even I feel about some things only being offered by third-parties. Oracle's tools suck enough people buy other products... heck, why bother? Just develop a good RDBMS with a good API, and let others fight it out? (That's an open question.)
Feature-wise, and maybe target-audience-wise, Firebird and PostgreSQL are similar. Stored procedures, triggers, check constraints, MVCC (Postgresql seems to have copied MVCC off of Interbase, note), savepoints/nested (but not concurrent) sub-transactions, etc. It lacks a lot of the UDT (type) features of PostgreSQL (you can define domains, but not entirely new datatypes) -- note that Postgres was specifically designed with UDT's in mind. Firebird does support UDF (function) features though, and you can get some of the same flexibility that way if you're masochistic (save data in octet or blob fields and use UDF's to interpret the data). Pg also has neat SP language support, letting you write your SPs in a variety of languages -- Fb doesn't. Unlike Postgres, it's really easy to install, particularly on windows (that was a problem for Pg up until semi-recently) and it practically maintains itself. (Happily, the Pg team eventually got their vacuum, equivalent to Fb's sweep, to not take down the database, so Pg can now run 24/7 too.) Fyracle has been trying to make Firebird more Oracle-like in SP language support and some of Oracle's more interesting query abilities (CONNECT BY). Yes, I occasionally get feature-lust and look at other DBMS's. I don't need Oracle features, but Pg features would sometimes be nice. But I don't use Pg, so I don't know what annoyances it has that Pg users would be thinking about. Maybe it's all-around better, I don't know.
Both are really good projects, with their own strengths. I would say comparing Firebird and PostgreSQL is a much fairer comparison than Pg and MySQL or MySQL and Fb. Pg and Fb are more of a 'niche' comparison. MySQL has nowhere near the features of either of them, isn't nearly as safe, and just isn't designed with the same requirements in mind.
Every single experience I've had with MySQL has been one of "fixing" stuff for a MySQL user who just couldn't get things to work. Joins that wouldn't work (but should have), joins that were slow, data being eaten... And then there's reliability... ugh. MySQL just wasn't designed with data integrity in mind, while Pg and Fb were. "Foreign key constraints can be
When buying items from a store, is the contract not purely with the store? How does the manufacturer come into this? Even if the store has a contract with their supplier, wouldn't they then have to make clear (on their own) that by buying from them, you agree to terms they are bound by contract to enforce on you, from their supplier? (Honest question.)
Oddly, I see that as a closely-related field. Logic (a big chunk by itself), communication of ideas, language structure, context-sensitive language, syntax vs. semantics, truth / predicates... common themes to both fields. At least for programming languages and databases, it's a start; probably a much better start than, say, an english major.
Does how much impact it has on 'families and children' have any bearing on whether or not it should be restricted / censored / tagged / whatever? Obviously, having this data won't harm us, and I'm not complaining that you would ask, I'm just curious about the assumption behind the question. How would you (plural) react if the data came back conclusively proving that porn has an enormous negative impact? No impact? Positive impact?
(Speaking of assumptions, it took me a minute to even realize I should account for the possibility that porn might have a positive impact, because it's been so thoroughly demonized.)
Personally, I've found porn occasionally useful as a communication tool. (*Gasp!*) You're supposed to (in a healthy relationship, or so I hear) communicate with your partner about what turns you on, what doesn't, etc. -- sometimes that's easier with porn than words. I know, it sounds weird, but... if it leads to a better relationship...
Meta: I'm pretty sure this discussion isn't evolving toward a solution. And I propose that we move it offsite to email; without an audience, we're less likely to make bold claims and accusations. This isn't a matter of saving face; at this point, I doubt anyone's watching anyway. But at this point, we probably take up well over a page just to ourselves, and we don't deserve it.
Non-meta: You must have misunderstood me at some point, and I'll take the blame (if it makes it easier) by saying I must have been unclear. I'm not sure that repeating myself (nor quoting you) will really fix that, but if you insist...
Your assertion: "That's why we don't have gills, tails or fur anymore: they take energy and resources to make, so children that don't have them get by on less." My response: You may be right, but I highly doubt you have any historical record of the process. You're merely guessing, and I'm telling you it's only a guess. Every event has a nearly infinite number of possible causes, and you've picked one arbitrarily. I argue for a lack of assertion, you argue for an assertion. The burden of proof is on you. It's not even an important point, why not let it go?
Your falsehood_and_fallacy URL returns a page starting with: Sorry, but there's no such page. You may have followed an inaccurate or outdated link. Maybe you just sent the wrong URL? I'll read it again when you provide a better URL.
You must have taken the "scientists and creationists" thing to heart just a bit too much. I'm not painting you as a religious nut as you claim, I'm merely pointing out that many people, on all sides of the issue, tend to assume that things are 'perfect' (or near-perfect) simply because they are the way they are. But I certainly wasn't claiming you to be a religious nut simply because you seemed to see things the same way both some scientists and (all?) creationists do. That's a leap I won't take.
I'm not seeing the slander, but okay. Nor the red herring. This may be the result of a simple misunderstanding; perhaps we agreed, you thought we disagreed, so you stated my position to be the opposite of itself so you could state that you agreed with my position while disagreeing with me? I'm a wee bit confused how that went. You say If you don't know statistics, you may not understand what that means, but it's quite poison to this notion of perfection of yours. As I was trying to point out, the process has no concept of perfection. So... I think we wound up agreeing. The point of this was that without perfection, the above assumption (about things being perfect because they exist) cannot be true. (Which, skipping a bit, implies that just because you see a way things could improve, doesn't mean they will. Why is this important? Only because you stated the appendix -will- disappear in a certain number of generations, because you see no reason for it to stick around.)
I'm no longer sure which question you think I'm avoiding (I'm replying in order, mind you)... but that section wasn't an ad hominem. There was exactly one of those (at least intended) -- and that was the 'troll' bit (for which I'll apologize farther down.)
I'm not ignoring the data. I'm saying the data proves nothing, it merely supports for the time being. Absence of evidence is not evidence of absence -- other, undiscovered data (or, as some claim, a lot of data we already have) could make natural selection seem as silly of a claim as spontaneous generation. But I'm not saying it will, nor that it has. Merely that supportive data proves nothing. You haven't disagreed with this, and I don't disagree with you. (Except when you say That said, over the course of the fossil record, and recently in life, we have evidence of this sort of thing happening hundreds of thousands of times. "Happening" must refer to natural selection; although I don't disagree with this being likely the case, I think we could take the same data and say it was evidence that God had make lots of
Err... yes. Most languages get used both carefully and not. Even the best keep-you-from-shooting-yourself-in-the-foot languages won't prevent you from writing incorrect programs; I'm really not sure why people bother. Buffer overflows aren't hard to prevent, and if a programmer gets caught writing one once, will probably never do so again. SQL injection is an issue in any language because people insist (note, I do this too) on writing their SQL as strings and injecting values into those strings, as opposed to using prepared statements with placeholders (which is much safer, but I'm not sure all API's support it.) It's a matter of using libraries correctly. But hey, let's complain about divide-by-zero errors and blame the language rather than the programmer.
Buffer overflows and SQL injection can be prevented, if you just learn a little bit. Those responsible are either chronically irresponsible and shouldn't be coding at all, or haven't learned enough yet, and should be taught. Irresponsibility and ignorance aren't language issues, they're people issues. Language choice is much more about preference/habit/comfort than language features or available libraries.
Gills: you never proved you knew why we lack them, only that you have an idea for one possible reason. I'm not the one with the burden of proof here.
Fallacy: logical error. Yes, you took a wrong turn in your reasoning, inferring a goal from a trend (see below.) And your link is useless, to boot. Congrats.
Religious nut: no, actually, that's poisoning the well, by attributing an evil purpose to my method in order to attack the method. And it doesn't work. But then, as we've learned from political ads, being loud is more useful than being right.
Reversal: I specifically deny perfection, so your notion that statistics trump perfection really only serves my point. But thanks. (A friend of mine calls this "rephrase and riposte", and we think it's cute. http://www.pseudotheos.com/view_object.php?object_ id=804.)
Evidence: no, actually, we don't have evidence of this happening hundreds of thousands of times. What we have is evidence of variations that happen to be congruent with a particular set of theories, thus failing to disprove those theories. Finding theories that explain a lot, seem plausible, fit the data, and aren't ugly hacks is all well and good -- but the data should not be equated with the theory. You have data currently most-usefully explained by natural selection. You do not have a historical record of it actually happening.
Citations: this one's funny. It's your first citation of any project name, and you're whining that I'm not providing citations? Ooh, and a straw man, attributing to me the accusation that these methods are incapable of producing useful results, too. Nice. Yes, actually, I have read some of Darwin's work -- no, not every word of it. Despite your questioning, you failed to assert the same of yourself, so to be even, have you read Darwin's work? And yes, I have built a primitive evolving system. No, it wasn't successful. No, that's not proof that I don't know what I'm talking about, but I expect you'll assert otherwise. And no, I don't need to cite works exemplifying failures in the AI world to show, logically, that the system can be skewed. It's intrinsic to the process. This isn't the same as saying it'll always fail, merely that you have no guarantees. At this point, you may feel free to cite statistics showing that evolving systems succeed 99% of the time, thus proving my point. Or you can try to logically prove that it's impossible for an evolving system to get stuck in a rut, but I predict that you'll fail.
Your reversi example (if we can call it that -- you don't cite any publications of your own, so we just have to trust that you're not bluffing) does not disprove that these systems can get terribly unstable -- merely that your one example did not seem to, in your own opinion. You didn't disprove the logic, you merely provided a data point. Exaggerated.
Skipping vacuous blabbering. You'll whine, nobody cares. (Apparent tactic: use short questions or assertions that will take forever to answer properly, making any response either pedantic or vague. Possibly effective. Nice work.)
To come back to the original point, no, you cannot assert that the emergent properties of a system demonstrate a goal inherent to the system. A trend is not a purpose. The individual goals making up the process are merely cogs in the machine. Bacteria (this is a stretch) may desire to individually maximize their ability to gather food, but this does not imply that natural selection desires it.
At least you've quietly stopped asserting that the appendix will disappear in 10k generations, simply because natural selection always does what you think it should do. (Paraphrase.)
I'm not sure how you got the karma level you have, considering your public message history. You're not wrong just because you're a troll, but a troll, you are. I actually deserve to be moderated down for feeding a troll, but I elect to put you in your place.
AT-Robots or some other assembly-like language to teach them how to program with registers, conditional jumps, interrupts, and all that. They'll later realize they hate goto from this. But mostly it's really stinkin' fun, and they'll hardly realize they're programming. Instead they'll be thinking about the problem they're solving, trying to beat each other in competition.
Then a language like C, to learn about functions, loops of various sorts, pointers, maybe even function pointers. This is when it gets boring.
Then a few other languages to learn about OO, dynamic typing, etc.
They really should be exposed to both interpreted and compiled languages, strongly typed and otherwise, with and without OO,... learning to program isn't about learning to program -something- in particular.
I would object to teaching them specifically "Ruby on Rails" -- that's like teaching someone how to hammer together a bird-house, as opposed to teaching them to be a carpenter. Also make sure they know the difference between a language and a library, if you can make that distinction.
You're a little late, but you needn't make up for it with rudeness.
No. Theory only predicts that there is likely to be a 'selection' if there is a means for the selection (variation) and a reason for the selection (preference). An organ may be useless, but if there's no variation, or if its cost is minimal compared to other selection factors, then selection isn't expected to occur. It might (if there's variation) by entirely random chance, particularly in small groups, taking the species in an effectively random direction. But that doesn't help your point.
No. We don't lack gills simply because they take energy to grow and therefore children without gills somehow survive more easily. Or rather, you'll have trouble proving what you've asserted. It's a possibility, yes. But then, it's also possible our ancestors simply thought gills were ugly, abandoned their children, killed and ate them, or refused to mate with any creature with gills. Or maybe you were there and can give us a history lesson? Reverse-engineering the process from the result is difficult work.
You fall prey to the same fallacy as everyone else, assuming that because an organ is useless in your eyes, it will be useless in the virtual eye of natural selection. Natural selection doesn't come with any guarantees, no matter how much time you give it -- 10,000 generations or not. If it did come with guarantees, that might almost be proof of a Creator, who would have designed the process with a goal in mind.
In simulations, the result is only as good as the selection process. Our ecology is based on all of us having been part of the selection process together. If you evolve AIs in a lab, and then have them meet real-world gamers, you're introducing an entirely new, foreign constraint to them. Furthermore, if you try to hurry things up in the lab by keeping only the top z% of the agents (cream of the cream), it's entirely possible that your population will skew in a random direction -- butterfly effect, essentially. Without continuous outside testing and some way of rectifying these anomalies, you may in the end have AIs that are really good against each other, but terrible against a real-world player. All the iterations in the world won't have helped you then. So no, I'm not assuming one-at-a-time testing, but I am assuming a good bit of outside control, if you're growing AIs with a particular result in mind. If you have no end result in mind (as is, apparently, the case of our planet) then leaving things to run their course is entirely appropriate (inasmuch as anything is 'appropriate' or 'good' in the absence of a goal.)
But thanks for your insight. On that subject -- you should be complaining about 'informative' moderations, not 'insightful', and even then it's a stretch. In our pomo world, what's the point of a 'truth' moderation? I'd love to see the politics section then...
At least in artificial evolutionary systems, yes. ("Why" is an odd question from a purely naturalistic point of view. Did nature -decide- to inject mutations?) However, mutations only get you so far. From somewhere near a local maximum, you're basically reaching out trying to hit another local maximum. If you hit something less-than-as-good-as-you-already-have, even if it's on the side of a better curve, you'll probably not get selected. Or at least, probabilistically not. If the same mutation keeps happening, yes, maybe one of them will stick around long enough to start climbing the newly-reached curve. But that requires, again, a large population. It seems like the chances of true 'innovation' go down as the (different) species reach their local maxima -- each will improve on itself less and less, then stagnate. In nature, you rarely get breeding between two completely random species (the general rule seems to be 'never'), so even if you have several specialized species, you're not going to have them naturally breeding and forming a third, new species that can start its own evolutionary path. But we do have that advantage in simulations. We can take a bird and an ant, merge them, and see what happens. Nature tends to avoid that (possibly with good effects) but it'd be a good way to re-introduce "proven" mutations into a species and see if they can be useful elsewhere. (Much like we pull useful genes out of random plants and stick them in other plants or animals. The mutation is common in one species but not another, both are successful, so we merge them to see what happens. Neither would naturally evolve into the other.)
Err... 'evolution' results, if all goes well, in a local-maximum. Nothing about the process of evolution, real or simulation, says you'll get "the best" anything. Only that, on average, things will tend to improve in a way that matches the particular constraints at the time, according to available variations. If the constraints change, or there were several ways around a problem ("hack",) or there wasn't sufficient diversity, or a bad trait just happened to get rewarded along with a good trait, you may wind up with a terribly bad result.
This is why I get so annoyed when scientists (and creationists) ask "what is this organ useful for?" expecting that every animal's every organ is entirely well-suited to its environment -- because either evolution or the hand of god made it perfect. That's not what the theory(!) of evolution predicts. Narwhals wound up with a long tooth, and sure, maybe they use it to impress the females now -- but is that why they have it in the first place, do they maybe only use it because they have it? Could it be that somewhere along the line, some freak just happened to survive an accident when others didn't, and passed on the freak gene causing this tooth to be a horn?
Weird results from genetic algorithms are even more likely in small-population scenarios like games. You can only send so many 'test' enemies at the player before he gets bored. Particularly considering that in most games, either the player surives or the computer survives, I'm not seeing at what point you can reward the AI by letting it reproduce except when the player loses (at least in an FPS setting.) In that case, the game only gets better if the player loses a lot -- and most games try to make sure the player doesn't lose too much, but is instead always on the brink of losing (to keep him hopeful but challenged.) On the other hand, if you train them in the studio, you'll have to be careful to not train them to be good only against the testers. You don't want to release a game in which the AI is really good -- so long as you don't lure it into getting stuck in a corner, just because no tester thought to do that often enough to breed it out.
But genetic algorithms are certainly not guaranteed to produce good results. They merely might.
Lemme get this straight -- they'll require a backdoor on those little boxes you can buy with commercial service, but not from random software you can download and use with a microphone and speakers?
Criminals and terrorists can get said software if they want to. Heck, they can probably get it for free, legally. They can use that to talk to each other over the internet; granted, it won't be as convenient as phone-like service with standard phone numbers, but it'll work. Meanwhile, the government retains the ability to wiretap citizens who are up to no evil, or are, but were too stupid to find a way not to get wiretapped.
This helps us how? And is anyone x-raying all USPS mail, OCR'ing it, and then analyzing the text content to discover hidden messages being sent by mail? Do they wiretap your mailbox?
French science fiction -- Verne, Barjavel... wait, planet of the apes was french (Boulle)? Woah. Sadly, wikipedia has no equivalent "Spanish science fiction" page, nor, that I can find, a page listing sci-fi by language. Sorry.
As they note, french sci-fi tends to be 'different'. I can't quite put my finger on it, despite having watched plenty of made-in-france sci-fi animated stuff as a kid ("Il etait une fois... l'espace", "Mysterieuses Cites d'Or", "Ulysse 31") and read french sci-fi (like Barjavel's). It seems obsessed with the words "infinite", "time", and with ancient civilizations... and a sort of steam-punk anachronism of those civilizations. I know it's not uniquely french, but it just seems much more present in french sci-fi than in, say, american. And a lot less with war (at least not told in the present tense) and more with the consequences of war, ancient history, etc. I guess we can attribute this to WW2, they were the ones being invaded, and dealing with that legacy, while we were the ones saving their butts, and dealing with -that- legacy (more with the concepts of prelude to war, justifying war, engaging in war, deciding which side to be on, etc.) And although it's again not literature, Jeunet's "Delicatessen" amused the hell out of me. (Freaked everyone else out -- again a story of ordinary life after a war.)
As to the Hugo question, I haven't a clue (and don't feel like checking.)
At this point, do we even care if porn-producing companies stay in business? There's so much of it -- is it even physically possible for any given human to see all the porn in the world (that is of any interest to him) in the course of a lifetime? Once the product hits the public domain, in a hundred (million) years... it'll be gratis *and* libre, and plentiful enough to keep everyone distracted. Forever. (Wait, isn't that a bad thing?)
(I wish our COBOL class hadn't restricted us to such an ancient dialect. We might have almost come to like it, if we could have used newer features -- or at least we could have had an updated dislike of the language. But no, our profs believed such modern COBOL to be heretical... my last assignment was therefore 666 lines long. Yes, Dr. H., that was on purpose. Yes, I swear I'll never use the sort function against an array.)
Last time I tested speed (very simple timing of a large increment loop),... compared to some simple C/C++ (at this point, the difference is moot)... perl, php, and python were all in the range of 1/10 to 1/14 the speed of C (pretty much equal.) Ruby was 1/40, which... sucked. But that was... let's see... four years ago? And a really icky way to test speed anyway?
But really, it doesn't matter. You can write good and bad code in any of them, and raw looping speed rarely matters. Most of our apps are just glue, talking to databases, networks, files -- all bottlenecks. The CPU isn't our problem. And where speed does matter, pretty much all of these languages rely on C libraries that -are- speedy. Most of them seem to compile to some form of bytecode at the very least, and will cache that bytecode to avoid a recurring cost.
These days, we pick languages more on features and libraries than speed. And most people just pick by features, because they're not interested in building -new- architecture where features matter. I still am, and C++ does that for me. But for web apps? I'm not coding architecture with really neat stuff going on... I just want to grab from a db and spit it out. PHP does that for me, so it's fine. Perl would too. So would python, and ruby, and probably COBOL if I were willing to ever do that again. I think the only reason I went with PHP was the db function set, and that's not much to go by, considering that most of these languages rely on similar/same C libraries, so it's just a matter of time before they all have the same library features -- just the time required to write wrappers to make C functions available in the scripted language itself.
If you do write web-apps in C++... apache modules aren't that bad, but there's a certain amount of "getting started" cost. Find yourself a good set of CGI functions if you're doing it the CGI way -- C++ doesn't hand you your $_POST stuff neatly the way PHP does. Beyond that... meh? It's pretty much all the same.
I'm surprised this is the only comment pointing that out -- I... I couldn't finish reading the summary after that one. Joysticks, cockpits... our world is full of fun references nobody pays attention to anymore.
Who said they had to have a valid plan for using this information? I see this more as a cover-your-ass thing. Imagine if they didn't ask for this ability, and the public were to find out there was something, at some point, that they could have done "better" (by their own, uninformed logic)... suddenly we get yet another intelligence-agency re-org for the sake of making things look like they're changing, and... again, nothing really happens. But it's a lot like politicians and lawyers, using every trick even remotely at their disposal to get things 'done' -- if you don't, and someone sees you missed an opportunity, you're in trouble. If the public is convinced that security is attainable, that it requires technology, and that our intelligence agencies are capable of and responsible for doing everything in their power to use technology to make us safe... then they'll grill anyone who doesn't try. But nobody said it had to work or be useful in practice -- it just has to look good to the public.
While we're pointing out all the holes that haven't been plugged, making the whole effort rather moot, the public just hears that there are holes that could be plugged and aren't, and get scared, and angry, and then they start writing letters... and then random holes get plugged that don't matter.
The religion gets stripped out, but if you look at the Homeworld RTS games, they're pulling pretty strongly from middle-eastern religious themes (and music) -- the jewish diaspora, the hebrew/arab relationship, the trinity (christianity thrown in?), the struggle to find a home... but it's not really approached from a religious perspective. It's much more "the story of the jews, but without god, and in space".
Besides -- you don't 'game' religion. Nothing about religion is predictable from a scientific point of view. If it were, people would be using prayer tactically to their advantage. Coding a game in which no results are ever guaranteed, nor even terribly predictable (don't even want to introduce the concept of probability that your prayer will be answered vs. the cost of praying)... there's just nothing left. Random background noise, probability-wise? You can't "play" that -- there's no technique, no challenge. If anything, the game would teach you that you can do just fine without religion helping you. Oops.
So it winds up always being story-oriented. And you're not very free to change the story. So you wind up with games like "go find the animals for Noah's ark"... which really isn't about religion, it's more like slapping a theme pack on top of a game like Tetris -- the artwork can be religious, but the game isn't.
Not so many centuries ago, the concept of 'hour' was flexible, depending on the season. 12 day hours, 12 night hours, regardless of the ratio between them. Back to the old?
There's evidence that temporal behaviour is again critical, that the timing of pulses from neuron to neuron may be the information storage for short-term memory, and that the information is not 'stored' anywhere apart from in the pulse-train.
Is that anything at all like 'storing' information on a network, by sending the data stream to bounce off an echo port at the other end of the internet, constantly cycling data back out into tcp/ip-land?
I happened to rant about that the other day. We -should- redesign our UI toolkits, even if not for deaf/blind/etc. It's just smart from an app-design standpoint. As it is, it's really hard to make an app future-proof in terms of visuals (will it work with future theme support?) and cross-platform compatible UI-wise (it's getting easier by using more wrappers like wxWidgets, but those in turn are still vulnerable to OS changes.)
What you're asking about is called cross-compiling. Linux is mostly C/C++ code, which is processor-agnostic (mostly). Most of the time, compilers translate that code to whatever the current processor is in our machines -- but given extra information, you can compile code on one machine that's meant for another machine (entirely different processor). It won't run where it's compiled, but if you can figure some way to move it over to the target machine, it'll run. But as the other poster points out, it's not just a matter of recompiling existing code, you sometimes have to write new code (drivers) as well for specific hardware that's never been supported before. But once you get all the basic stuff working again (text, graphics, sound, input, etc.) then the already-written applications can be cross-compiled as well -- everything they depend on will be ready for them. So long as software libraries are sufficiently decoupled from each other... but yeah. Google around for more on cross-compiling. Use an x86 PC to compile linux for your Mac, etc.
Slashdot Mirror
I had mod points, and as a Firebird user, I was going to mod up the grandparent -- Firebird is very often ignored despite a host of positive features.
... heck, why bother? Just develop a good RDBMS with a good API, and let others fight it out? (That's an open question.)
... And then there's reliability ... ugh. MySQL just wasn't designed with data integrity in mind, while Pg and Fb were. "Foreign key constraints can be
But as you had a bad experience, and you link to your problem list, I thought I'd respond instead. Better to answer questions than just mod up friendlies.
- Database path: Yes, firebird supports aliases. Our app doesn't use them, but they're there.
- ISQL: I hear Oracle's SQL*Plus isn't much better. However, I use isql every once in a while, and I have command-history, backspacing, etc. available to me. From what I hear, it's more of a function of the shell you use (around isql) than isql itself. If you set up your environment properly, isql and its ilk automatically get command-history. (That's what I'm told, note. Anyone who can explain this is welcome to. I'm not a sysadmin.)
- Never seen it freeze.
- Corruption: we've had exactly one database issue, where it seems a backup/restore script ran in the middle of the day, restoring the database to its state from 4 hours earlier. In 4 years of use, with somewhere around 60 users in a medical clinic/insurance/billing environment, we've had no corruption. Using forced-writes is important, however. The careful-write strategy is really, really reliable, but it still can't protect you from faulty hard drives or operating systems that refuse to send data to the disk in the order requested (cf. Windows). M1 Abrams tank story, anyone?
- IBDataPump and other third-party tools exist for some of the other features you're interested in. I'm not sure I know how even I feel about some things only being offered by third-parties. Oracle's tools suck enough people buy other products
Feature-wise, and maybe target-audience-wise, Firebird and PostgreSQL are similar. Stored procedures, triggers, check constraints, MVCC (Postgresql seems to have copied MVCC off of Interbase, note), savepoints/nested (but not concurrent) sub-transactions, etc. It lacks a lot of the UDT (type) features of PostgreSQL (you can define domains, but not entirely new datatypes) -- note that Postgres was specifically designed with UDT's in mind. Firebird does support UDF (function) features though, and you can get some of the same flexibility that way if you're masochistic (save data in octet or blob fields and use UDF's to interpret the data). Pg also has neat SP language support, letting you write your SPs in a variety of languages -- Fb doesn't. Unlike Postgres, it's really easy to install, particularly on windows (that was a problem for Pg up until semi-recently) and it practically maintains itself. (Happily, the Pg team eventually got their vacuum, equivalent to Fb's sweep, to not take down the database, so Pg can now run 24/7 too.) Fyracle has been trying to make Firebird more Oracle-like in SP language support and some of Oracle's more interesting query abilities (CONNECT BY). Yes, I occasionally get feature-lust and look at other DBMS's. I don't need Oracle features, but Pg features would sometimes be nice. But I don't use Pg, so I don't know what annoyances it has that Pg users would be thinking about. Maybe it's all-around better, I don't know.
Both are really good projects, with their own strengths. I would say comparing Firebird and PostgreSQL is a much fairer comparison than Pg and MySQL or MySQL and Fb. Pg and Fb are more of a 'niche' comparison. MySQL has nowhere near the features of either of them, isn't nearly as safe, and just isn't designed with the same requirements in mind.
Every single experience I've had with MySQL has been one of "fixing" stuff for a MySQL user who just couldn't get things to work. Joins that wouldn't work (but should have), joins that were slow, data being eaten
When buying items from a store, is the contract not purely with the store? How does the manufacturer come into this? Even if the store has a contract with their supplier, wouldn't they then have to make clear (on their own) that by buying from them, you agree to terms they are bound by contract to enforce on you, from their supplier? (Honest question.)
Oddly, I see that as a closely-related field. Logic (a big chunk by itself), communication of ideas, language structure, context-sensitive language, syntax vs. semantics, truth / predicates ... common themes to both fields. At least for programming languages and databases, it's a start; probably a much better start than, say, an english major.
I don't get it. The ess-queue-el to the bourne shell? I guess it rhymes, but ...
Does how much impact it has on 'families and children' have any bearing on whether or not it should be restricted / censored / tagged / whatever? Obviously, having this data won't harm us, and I'm not complaining that you would ask, I'm just curious about the assumption behind the question. How would you (plural) react if the data came back conclusively proving that porn has an enormous negative impact? No impact? Positive impact?
... if it leads to a better relationship ...
(Speaking of assumptions, it took me a minute to even realize I should account for the possibility that porn might have a positive impact, because it's been so thoroughly demonized.)
Personally, I've found porn occasionally useful as a communication tool. (*Gasp!*) You're supposed to (in a healthy relationship, or so I hear) communicate with your partner about what turns you on, what doesn't, etc. -- sometimes that's easier with porn than words. I know, it sounds weird, but
Meta: I'm pretty sure this discussion isn't evolving toward a solution. And I propose that we move it offsite to email; without an audience, we're less likely to make bold claims and accusations. This isn't a matter of saving face; at this point, I doubt anyone's watching anyway. But at this point, we probably take up well over a page just to ourselves, and we don't deserve it.
... I think we wound up agreeing. The point of this was that without perfection, the above assumption (about things being perfect because they exist) cannot be true. (Which, skipping a bit, implies that just because you see a way things could improve, doesn't mean they will. Why is this important? Only because you stated the appendix -will- disappear in a certain number of generations, because you see no reason for it to stick around.)
... but that section wasn't an ad hominem. There was exactly one of those (at least intended) -- and that was the 'troll' bit (for which I'll apologize farther down.)
Non-meta: You must have misunderstood me at some point, and I'll take the blame (if it makes it easier) by saying I must have been unclear. I'm not sure that repeating myself (nor quoting you) will really fix that, but if you insist...
Your assertion: "That's why we don't have gills, tails or fur anymore: they take energy and resources to make, so children that don't have them get by on less."
My response: You may be right, but I highly doubt you have any historical record of the process. You're merely guessing, and I'm telling you it's only a guess. Every event has a nearly infinite number of possible causes, and you've picked one arbitrarily. I argue for a lack of assertion, you argue for an assertion. The burden of proof is on you. It's not even an important point, why not let it go?
Your falsehood_and_fallacy URL returns a page starting with: Sorry, but there's no such page. You may have followed an inaccurate or outdated link. Maybe you just sent the wrong URL? I'll read it again when you provide a better URL.
You must have taken the "scientists and creationists" thing to heart just a bit too much. I'm not painting you as a religious nut as you claim, I'm merely pointing out that many people, on all sides of the issue, tend to assume that things are 'perfect' (or near-perfect) simply because they are the way they are. But I certainly wasn't claiming you to be a religious nut simply because you seemed to see things the same way both some scientists and (all?) creationists do. That's a leap I won't take.
I'm not seeing the slander, but okay. Nor the red herring. This may be the result of a simple misunderstanding; perhaps we agreed, you thought we disagreed, so you stated my position to be the opposite of itself so you could state that you agreed with my position while disagreeing with me? I'm a wee bit confused how that went. You say If you don't know statistics, you may not understand what that means, but it's quite poison to this notion of perfection of yours. As I was trying to point out, the process has no concept of perfection. So
I'm no longer sure which question you think I'm avoiding (I'm replying in order, mind you)
I'm not ignoring the data. I'm saying the data proves nothing, it merely supports for the time being. Absence of evidence is not evidence of absence -- other, undiscovered data (or, as some claim, a lot of data we already have) could make natural selection seem as silly of a claim as spontaneous generation. But I'm not saying it will, nor that it has. Merely that supportive data proves nothing. You haven't disagreed with this, and I don't disagree with you. (Except when you say That said, over the course of the fossil record, and recently in life, we have evidence of this sort of thing happening hundreds of thousands of times. "Happening" must refer to natural selection; although I don't disagree with this being likely the case, I think we could take the same data and say it was evidence that God had make lots of
Err ... yes. Most languages get used both carefully and not. Even the best keep-you-from-shooting-yourself-in-the-foot languages won't prevent you from writing incorrect programs; I'm really not sure why people bother. Buffer overflows aren't hard to prevent, and if a programmer gets caught writing one once, will probably never do so again. SQL injection is an issue in any language because people insist (note, I do this too) on writing their SQL as strings and injecting values into those strings, as opposed to using prepared statements with placeholders (which is much safer, but I'm not sure all API's support it.) It's a matter of using libraries correctly. But hey, let's complain about divide-by-zero errors and blame the language rather than the programmer.
Buffer overflows and SQL injection can be prevented, if you just learn a little bit. Those responsible are either chronically irresponsible and shouldn't be coding at all, or haven't learned enough yet, and should be taught. Irresponsibility and ignorance aren't language issues, they're people issues. Language choice is much more about preference/habit/comfort than language features or available libraries.
Gills: you never proved you knew why we lack them, only that you have an idea for one possible reason. I'm not the one with the burden of proof here.
_ id=804.)
Fallacy: logical error. Yes, you took a wrong turn in your reasoning, inferring a goal from a trend (see below.) And your link is useless, to boot. Congrats.
Religious nut: no, actually, that's poisoning the well, by attributing an evil purpose to my method in order to attack the method. And it doesn't work. But then, as we've learned from political ads, being loud is more useful than being right.
Reversal: I specifically deny perfection, so your notion that statistics trump perfection really only serves my point. But thanks. (A friend of mine calls this "rephrase and riposte", and we think it's cute. http://www.pseudotheos.com/view_object.php?object
Evidence: no, actually, we don't have evidence of this happening hundreds of thousands of times. What we have is evidence of variations that happen to be congruent with a particular set of theories, thus failing to disprove those theories. Finding theories that explain a lot, seem plausible, fit the data, and aren't ugly hacks is all well and good -- but the data should not be equated with the theory. You have data currently most-usefully explained by natural selection. You do not have a historical record of it actually happening.
Citations: this one's funny. It's your first citation of any project name, and you're whining that I'm not providing citations? Ooh, and a straw man, attributing to me the accusation that these methods are incapable of producing useful results, too. Nice. Yes, actually, I have read some of Darwin's work -- no, not every word of it. Despite your questioning, you failed to assert the same of yourself, so to be even, have you read Darwin's work? And yes, I have built a primitive evolving system. No, it wasn't successful. No, that's not proof that I don't know what I'm talking about, but I expect you'll assert otherwise. And no, I don't need to cite works exemplifying failures in the AI world to show, logically, that the system can be skewed. It's intrinsic to the process. This isn't the same as saying it'll always fail, merely that you have no guarantees. At this point, you may feel free to cite statistics showing that evolving systems succeed 99% of the time, thus proving my point. Or you can try to logically prove that it's impossible for an evolving system to get stuck in a rut, but I predict that you'll fail.
Your reversi example (if we can call it that -- you don't cite any publications of your own, so we just have to trust that you're not bluffing) does not disprove that these systems can get terribly unstable -- merely that your one example did not seem to, in your own opinion. You didn't disprove the logic, you merely provided a data point. Exaggerated.
Skipping vacuous blabbering. You'll whine, nobody cares. (Apparent tactic: use short questions or assertions that will take forever to answer properly, making any response either pedantic or vague. Possibly effective. Nice work.)
To come back to the original point, no, you cannot assert that the emergent properties of a system demonstrate a goal inherent to the system. A trend is not a purpose. The individual goals making up the process are merely cogs in the machine. Bacteria (this is a stretch) may desire to individually maximize their ability to gather food, but this does not imply that natural selection desires it.
At least you've quietly stopped asserting that the appendix will disappear in 10k generations, simply because natural selection always does what you think it should do. (Paraphrase.)
I'm not sure how you got the karma level you have, considering your public message history. You're not wrong just because you're a troll, but a troll, you are. I actually deserve to be moderated down for feeding a troll, but I elect to put you in your place.
Seriously.
... learning to program isn't about learning to program -something- in particular.
AT-Robots or some other assembly-like language to teach them how to program with registers, conditional jumps, interrupts, and all that. They'll later realize they hate goto from this. But mostly it's really stinkin' fun, and they'll hardly realize they're programming. Instead they'll be thinking about the problem they're solving, trying to beat each other in competition.
Then a language like C, to learn about functions, loops of various sorts, pointers, maybe even function pointers. This is when it gets boring.
Then a few other languages to learn about OO, dynamic typing, etc.
They really should be exposed to both interpreted and compiled languages, strongly typed and otherwise, with and without OO,
I would object to teaching them specifically "Ruby on Rails" -- that's like teaching someone how to hammer together a bird-house, as opposed to teaching them to be a carpenter. Also make sure they know the difference between a language and a library, if you can make that distinction.
You're a little late, but you needn't make up for it with rudeness.
...
No. Theory only predicts that there is likely to be a 'selection' if there is a means for the selection (variation) and a reason for the selection (preference). An organ may be useless, but if there's no variation, or if its cost is minimal compared to other selection factors, then selection isn't expected to occur. It might (if there's variation) by entirely random chance, particularly in small groups, taking the species in an effectively random direction. But that doesn't help your point.
No. We don't lack gills simply because they take energy to grow and therefore children without gills somehow survive more easily. Or rather, you'll have trouble proving what you've asserted. It's a possibility, yes. But then, it's also possible our ancestors simply thought gills were ugly, abandoned their children, killed and ate them, or refused to mate with any creature with gills. Or maybe you were there and can give us a history lesson? Reverse-engineering the process from the result is difficult work.
You fall prey to the same fallacy as everyone else, assuming that because an organ is useless in your eyes, it will be useless in the virtual eye of natural selection. Natural selection doesn't come with any guarantees, no matter how much time you give it -- 10,000 generations or not. If it did come with guarantees, that might almost be proof of a Creator, who would have designed the process with a goal in mind.
In simulations, the result is only as good as the selection process. Our ecology is based on all of us having been part of the selection process together. If you evolve AIs in a lab, and then have them meet real-world gamers, you're introducing an entirely new, foreign constraint to them. Furthermore, if you try to hurry things up in the lab by keeping only the top z% of the agents (cream of the cream), it's entirely possible that your population will skew in a random direction -- butterfly effect, essentially. Without continuous outside testing and some way of rectifying these anomalies, you may in the end have AIs that are really good against each other, but terrible against a real-world player. All the iterations in the world won't have helped you then. So no, I'm not assuming one-at-a-time testing, but I am assuming a good bit of outside control, if you're growing AIs with a particular result in mind. If you have no end result in mind (as is, apparently, the case of our planet) then leaving things to run their course is entirely appropriate (inasmuch as anything is 'appropriate' or 'good' in the absence of a goal.)
But thanks for your insight. On that subject -- you should be complaining about 'informative' moderations, not 'insightful', and even then it's a stretch. In our pomo world, what's the point of a 'truth' moderation? I'd love to see the politics section then
At least in artificial evolutionary systems, yes. ("Why" is an odd question from a purely naturalistic point of view. Did nature -decide- to inject mutations?) However, mutations only get you so far. From somewhere near a local maximum, you're basically reaching out trying to hit another local maximum. If you hit something less-than-as-good-as-you-already-have, even if it's on the side of a better curve, you'll probably not get selected. Or at least, probabilistically not. If the same mutation keeps happening, yes, maybe one of them will stick around long enough to start climbing the newly-reached curve. But that requires, again, a large population. It seems like the chances of true 'innovation' go down as the (different) species reach their local maxima -- each will improve on itself less and less, then stagnate. In nature, you rarely get breeding between two completely random species (the general rule seems to be 'never'), so even if you have several specialized species, you're not going to have them naturally breeding and forming a third, new species that can start its own evolutionary path. But we do have that advantage in simulations. We can take a bird and an ant, merge them, and see what happens. Nature tends to avoid that (possibly with good effects) but it'd be a good way to re-introduce "proven" mutations into a species and see if they can be useful elsewhere. (Much like we pull useful genes out of random plants and stick them in other plants or animals. The mutation is common in one species but not another, both are successful, so we merge them to see what happens. Neither would naturally evolve into the other.)
Err ... 'evolution' results, if all goes well, in a local-maximum. Nothing about the process of evolution, real or simulation, says you'll get "the best" anything. Only that, on average, things will tend to improve in a way that matches the particular constraints at the time, according to available variations. If the constraints change, or there were several ways around a problem ("hack",) or there wasn't sufficient diversity, or a bad trait just happened to get rewarded along with a good trait, you may wind up with a terribly bad result.
This is why I get so annoyed when scientists (and creationists) ask "what is this organ useful for?" expecting that every animal's every organ is entirely well-suited to its environment -- because either evolution or the hand of god made it perfect. That's not what the theory(!) of evolution predicts. Narwhals wound up with a long tooth, and sure, maybe they use it to impress the females now -- but is that why they have it in the first place, do they maybe only use it because they have it? Could it be that somewhere along the line, some freak just happened to survive an accident when others didn't, and passed on the freak gene causing this tooth to be a horn?
Weird results from genetic algorithms are even more likely in small-population scenarios like games. You can only send so many 'test' enemies at the player before he gets bored. Particularly considering that in most games, either the player surives or the computer survives, I'm not seeing at what point you can reward the AI by letting it reproduce except when the player loses (at least in an FPS setting.) In that case, the game only gets better if the player loses a lot -- and most games try to make sure the player doesn't lose too much, but is instead always on the brink of losing (to keep him hopeful but challenged.) On the other hand, if you train them in the studio, you'll have to be careful to not train them to be good only against the testers. You don't want to release a game in which the AI is really good -- so long as you don't lure it into getting stuck in a corner, just because no tester thought to do that often enough to breed it out.
But genetic algorithms are certainly not guaranteed to produce good results. They merely might.
Lemme get this straight -- they'll require a backdoor on those little boxes you can buy with commercial service, but not from random software you can download and use with a microphone and speakers?
Criminals and terrorists can get said software if they want to. Heck, they can probably get it for free, legally. They can use that to talk to each other over the internet; granted, it won't be as convenient as phone-like service with standard phone numbers, but it'll work. Meanwhile, the government retains the ability to wiretap citizens who are up to no evil, or are, but were too stupid to find a way not to get wiretapped.
This helps us how? And is anyone x-raying all USPS mail, OCR'ing it, and then analyzing the text content to discover hidden messages being sent by mail? Do they wiretap your mailbox?
French science fiction -- Verne, Barjavel ... wait, planet of the apes was french (Boulle)? Woah. Sadly, wikipedia has no equivalent "Spanish science fiction" page, nor, that I can find, a page listing sci-fi by language. Sorry.
... l'espace", "Mysterieuses Cites d'Or", "Ulysse 31") and read french sci-fi (like Barjavel's). It seems obsessed with the words "infinite", "time", and with ancient civilizations ... and a sort of steam-punk anachronism of those civilizations. I know it's not uniquely french, but it just seems much more present in french sci-fi than in, say, american. And a lot less with war (at least not told in the present tense) and more with the consequences of war, ancient history, etc. I guess we can attribute this to WW2, they were the ones being invaded, and dealing with that legacy, while we were the ones saving their butts, and dealing with -that- legacy (more with the concepts of prelude to war, justifying war, engaging in war, deciding which side to be on, etc.) And although it's again not literature, Jeunet's "Delicatessen" amused the hell out of me. (Freaked everyone else out -- again a story of ordinary life after a war.)
As they note, french sci-fi tends to be 'different'. I can't quite put my finger on it, despite having watched plenty of made-in-france sci-fi animated stuff as a kid ("Il etait une fois
As to the Hugo question, I haven't a clue (and don't feel like checking.)
At this point, do we even care if porn-producing companies stay in business? There's so much of it -- is it even physically possible for any given human to see all the porn in the world (that is of any interest to him) in the course of a lifetime? Once the product hits the public domain, in a hundred (million) years ... it'll be gratis *and* libre, and plentiful enough to keep everyone distracted. Forever. (Wait, isn't that a bad thing?)
Indeed. But LISP'ers will be quick to remind you it's nothing to be proud of, they've had it forever by accident.
COBOL
... my last assignment was therefore 666 lines long. Yes, Dr. H., that was on purpose. Yes, I swear I'll never use the sort function against an array.)
There's no reason for obscenities, either.
(I wish our COBOL class hadn't restricted us to such an ancient dialect. We might have almost come to like it, if we could have used newer features -- or at least we could have had an updated dislike of the language. But no, our profs believed such modern COBOL to be heretical
Last time I tested speed (very simple timing of a large increment loop), ... compared to some simple C/C++ (at this point, the difference is moot) ... perl, php, and python were all in the range of 1/10 to 1/14 the speed of C (pretty much equal.) Ruby was 1/40, which ... sucked. But that was ... let's see ... four years ago? And a really icky way to test speed anyway?
... I just want to grab from a db and spit it out. PHP does that for me, so it's fine. Perl would too. So would python, and ruby, and probably COBOL if I were willing to ever do that again. I think the only reason I went with PHP was the db function set, and that's not much to go by, considering that most of these languages rely on similar/same C libraries, so it's just a matter of time before they all have the same library features -- just the time required to write wrappers to make C functions available in the scripted language itself.
... apache modules aren't that bad, but there's a certain amount of "getting started" cost. Find yourself a good set of CGI functions if you're doing it the CGI way -- C++ doesn't hand you your $_POST stuff neatly the way PHP does. Beyond that ... meh? It's pretty much all the same.
But really, it doesn't matter. You can write good and bad code in any of them, and raw looping speed rarely matters. Most of our apps are just glue, talking to databases, networks, files -- all bottlenecks. The CPU isn't our problem. And where speed does matter, pretty much all of these languages rely on C libraries that -are- speedy. Most of them seem to compile to some form of bytecode at the very least, and will cache that bytecode to avoid a recurring cost.
These days, we pick languages more on features and libraries than speed. And most people just pick by features, because they're not interested in building -new- architecture where features matter. I still am, and C++ does that for me. But for web apps? I'm not coding architecture with really neat stuff going on
If you do write web-apps in C++
I'm surprised this is the only comment pointing that out -- I ... I couldn't finish reading the summary after that one. Joysticks, cockpits ... our world is full of fun references nobody pays attention to anymore.
Who said they had to have a valid plan for using this information? I see this more as a cover-your-ass thing. Imagine if they didn't ask for this ability, and the public were to find out there was something, at some point, that they could have done "better" (by their own, uninformed logic) ... suddenly we get yet another intelligence-agency re-org for the sake of making things look like they're changing, and ... again, nothing really happens. But it's a lot like politicians and lawyers, using every trick even remotely at their disposal to get things 'done' -- if you don't, and someone sees you missed an opportunity, you're in trouble. If the public is convinced that security is attainable, that it requires technology, and that our intelligence agencies are capable of and responsible for doing everything in their power to use technology to make us safe ... then they'll grill anyone who doesn't try. But nobody said it had to work or be useful in practice -- it just has to look good to the public.
... and then random holes get plugged that don't matter.
While we're pointing out all the holes that haven't been plugged, making the whole effort rather moot, the public just hears that there are holes that could be plugged and aren't, and get scared, and angry, and then they start writing letters
The religion gets stripped out, but if you look at the Homeworld RTS games, they're pulling pretty strongly from middle-eastern religious themes (and music) -- the jewish diaspora, the hebrew/arab relationship, the trinity (christianity thrown in?), the struggle to find a home ... but it's not really approached from a religious perspective. It's much more "the story of the jews, but without god, and in space".
... there's just nothing left. Random background noise, probability-wise? You can't "play" that -- there's no technique, no challenge. If anything, the game would teach you that you can do just fine without religion helping you. Oops.
... which really isn't about religion, it's more like slapping a theme pack on top of a game like Tetris -- the artwork can be religious, but the game isn't.
Besides -- you don't 'game' religion. Nothing about religion is predictable from a scientific point of view. If it were, people would be using prayer tactically to their advantage. Coding a game in which no results are ever guaranteed, nor even terribly predictable (don't even want to introduce the concept of probability that your prayer will be answered vs. the cost of praying)
So it winds up always being story-oriented. And you're not very free to change the story. So you wind up with games like "go find the animals for Noah's ark"
Not so many centuries ago, the concept of 'hour' was flexible, depending on the season. 12 day hours, 12 night hours, regardless of the ratio between them. Back to the old?
There's evidence that temporal behaviour is again critical, that the timing of pulses from neuron to neuron may be the information storage for short-term memory, and that the information is not 'stored' anywhere apart from in the pulse-train.
Is that anything at all like 'storing' information on a network, by sending the data stream to bounce off an echo port at the other end of the internet, constantly cycling data back out into tcp/ip-land?
http://www.pseudotheos.com/view_object.php?object_ id=1190
I happened to rant about that the other day. We -should- redesign our UI toolkits, even if not for deaf/blind/etc. It's just smart from an app-design standpoint. As it is, it's really hard to make an app future-proof in terms of visuals (will it work with future theme support?) and cross-platform compatible UI-wise (it's getting easier by using more wrappers like wxWidgets, but those in turn are still vulnerable to OS changes.)
What you're asking about is called cross-compiling. Linux is mostly C/C++ code, which is processor-agnostic (mostly). Most of the time, compilers translate that code to whatever the current processor is in our machines -- but given extra information, you can compile code on one machine that's meant for another machine (entirely different processor). It won't run where it's compiled, but if you can figure some way to move it over to the target machine, it'll run. But as the other poster points out, it's not just a matter of recompiling existing code, you sometimes have to write new code (drivers) as well for specific hardware that's never been supported before. But once you get all the basic stuff working again (text, graphics, sound, input, etc.) then the already-written applications can be cross-compiled as well -- everything they depend on will be ready for them. So long as software libraries are sufficiently decoupled from each other ... but yeah. Google around for more on cross-compiling. Use an x86 PC to compile linux for your Mac, etc.