You know, it's funny. I'm a bit fuzzy on the dates, but child pornography was only made illegal in the US about 40-or-so years ago.
Of course, you have to separate pornography into two categories: 1) a media work showing an explictly sexual act (masturbation, penetrative sex, oral sex, etc) 2) a media work of a prurient nature that does not explicitly show "sex."
The former was illegal (minor consent laws and all that), but the latter was kinda legal.
When I say "kinda" I'm not being wishy-washy, it's that we're coming up against anachronism: according to experts on this sort of thing[1] attitudes of the "man on the street" have shifted drastically concerning photographs/drawings of young children. What would be considered "cute" and "childlike" back in the 1950's would be considered "grotesque" and "unsuitable for public consumption" now. Most of this can be linked to the witchhunts regarding child pornography.
Even more amusing, since the enactment of child pornography laws the average age of actors involved in sex scenes and -- and this is very odd -- in just plain romantic kisses in movies has gone way down. Way, way down.
I think it was Ebert (or was it Stephen King? King wrote about this a bit as well) who commented that youngsters used to go to the movies to see oldsters in a sex scene. Now it's reversed.
Salon commented[2] that the rabid and far-reaching bans on child "pornography"[3] has caused us all to think like pedophiles. Reminds me of the joke about the sexaholic who goes into the psychiatrist, who gives him a rorschach test.
"What do you see?"
"Two people having sex. A naked woman. A threesome."
"Jane, I think you've got a fixation on sex."
"What? Doc, you're the one showing me all these dirty pictures!"
Ah well.
So my point is that it's tough to say what's child porn and what's not, with the fact that we've gotten hypersensitive about it recently. I don't really feel like having the morality police check out all of my mom's photo albums to see if they're kosher. Especially morality police from the MPC[4]. Though being a smart son who knows the power of embarassment blackmail from mothers, I have removed all of my naked baby pictures. Now if only I could get the picture of me hugging then penguin at Sea World when I was 12....
And to spare myself accusations: no. I'm not "into" kids. To be honest, only recently has my age group become somewhat appealing to me at all: I've always been after older women, which is a real drag as women are used to young guys being...well, young guys.
So, just something to think about.
[1] I'm thinking of the historians interviewed in Salon concerning the somewhat-recent Paul Ruebens case. Interesting article.
[2] Again, the Paul Ruebens case.
[3] bare-butt baby picture arrests, anyone? Traditional Brazilian breast feeding family photos getting your kids under the care of CPS, like just happened in Dallas about a year ago?
Personally I think being within 10% - 15% of the top end PC would have been fine.
My guess is that the amount of effort and documentation required to make a legitimate test look like a legitimate test for "X% Faster" (where X is what people interpret Apple as claiming) versus that required for "within 10%-15% of the top end PC" would be little when, if they believe that the tests are legitimate, they would be hurting their position.
Why? Because the plausibility is not in how the G5 compares to the Intel, it's in how the G5 compares to the G4. People don't trust huge leaps in performance within the same line of chips, and demand more documentation.
And rightfully so - if a company is going to spend a lot of money and develop an all-new architecture that they're going to market as being a massive move from the previous one requiring developer time to move things between the two architectures, then they damned well better be a lot better performing.
Also, people have this weird assumption that just because a chip speaks the same instruction set as the previous one, and is being released by the same company (effectively), then it must be the same stuff internally. We are big believers in the "ground up rebuild" (note how often it's used in marketing and is repeated by others. Marketing is designed to tell us what we want to hear). So if you didn't rebuild the entire chip, and you didn't because it works with the old one, then how could it be N times faster?
Also, marketing companies have been telling us for years "new, improved, works the same as the old one, 10 times faster than the old one" and we've stopped believing them. General loss of credibility.
So, given that sociological assumption, and (charitably) given that Apple believes the benchmarks actually reflect reality, then why not go for the gold? You'd have to weaken your position quite a bit to be considered much more plausible given the general performance and age of the G4 and G4 architecture[1].
Just one possible explanation.
[1] - here I'm talking about the problems we all know about: low bus speed, weak integer performance and a few other things. It really was a very nice chip and once developers started optimizing for it, it showed enough strength in certain areas to work around its deficiencies for a surprising amount.
You mean the new chips from Intel that were announced the same day as the G5s?
Yes, and I'm sure Intel would be so very helpful in getting one of their prime competitors pre-release sample versions of their chips.
Apple most likely did have to actually buy that Dell box just like the rest of us poor slobs (if us poor slobs were to buy a Dell, that is). Since they're a big competitor, I doubt they got any special favors about first ship or anything else.
"Pissing on motorcycles" bit was referring to people who don't respect the tools they use, ie motorcycles. Not saying you do that, agreeing that the people who do that are a problem.
As for being rewarded...I think the best reward we could get would be the respect of other motorists and a realization that the entire road is not just for cars (where I live the belief is that the entire road is for trucks, with cars being second-class citizens and motorcycles being flies that bother you).
Then - at my behest - the garage took a look at the black box in my car
It was your decision to look at the black box. That's cool. Not an invasion of privacy at all. On the other hand, if you had forced the other person to turn over his black box, then that would be an invasion of privacy - whether it would be acceptable would be another matter.
As a practical matter (as you well know), physics tells us that if total speed is known and *a* car's speed is known, we can know the other car's speed.
Hence it only takes one car having a black box and a willing owner to produce the results you need, no violation of privacy is necessary.
That's not what concerns us "privacy nuts."
It's when you might be forced to turn over YOUR black box. I don't want that to happen to you. First off, because I don't think that's right (though I'm willing to be convinced otherwise).
Second off, as the above basic physics indicates, there is no reason for you to have to turn over your black box for any situation involving two cars if those cars fall within the 98% of those out on the road.
Soooooo...why would the law want to be able to yank your black box if they don't need it for a two-car accident like that? Doesn't sound quite right to me. Maybe they're just not thinking the logic through.
I've read the actuarial summaries related to car accidents. The #1 influencing factor is not the skill of the driver, the state of the road, or even the absolute speed of the car, funny enough.
The biggest indicator of your likelihood to get in an accident is your relative speed related to the other drivers, regardless of the speed limit. If you are driving greater than or less than 10 mph relative to traffic, you are very likely to get in an accident.
Hence your idea sounds good but wouldn't work out too well, unfortunately. Wish it would, though.
One of the other big influencing factors is the number of occupants in a car, particularly where the average age is low (moms with many young children and teenagers in general).
I think we'd all do better to pay attention to the actuarial data and allow insurance companies to make much more specific policies regarding coverage rather than making more laws.
Just an idea.
PS: as a person who likes motorcycles, I agree with you that "cool" is the big reason that motorcycles have gottten a bad name: no helmets, no protective gear, more bike than you can handle, and (on a minor note) incredibly loud pipes. Of course, I don't want to see laws being made regarding any of this, but it's still frustrating to see something you like get pissed on.
This is true for some demographics, particularly the "budget" ones - ie the areas of insurance mandated by the government and such.
It is possible to buy "no fault" insurance, that pays no matter whose fault it was. If there was demand for it, I can see there being "no criminal fault" insurance - if you are not assigned the fault by a court, then the insurance pays out.
Why would the insurance companies do this? Twofold: they don't make much money on just liability insurance. In states that mandate liability insurance, people whose actuarial classification is in the "Danger! Danger, Will Robinson!" area get placed into a state pool where the insurance companies do a lottery to see who has to insure them. This pool costs them money, which they then make up from the next highest bracket. Ever wonder why it takes a long time for your payments to come in line with your actuarial classification? That's why.
Also, most people who get dicked over by insurance companies go for minimum coverage. This is like coach class in an airplane: don't expect great service.
Again, the problem is that people are not willing to pay for what they want. Of course insurance companies are going to try to reduce their payouts to customers who shop around all the time trying to find the lowest rate.
I predict that if we do get insurance companies getting persnickity with the monitoring equipment, then there will be additional grades of coverage. There is competition in the profitable brackets of insurance.
My favorite was the part where the RIAA FSM claimed that RIAA members wished there was greater variety on the radio.
Gosh, wouldn't that be nice.
Of course, the fact that the RIAA members have been focusing on Billboard and mega-hits would have nothing to do with the dearth of radio content. Nothing at all.
Oh, and the fact that payola is alive and well (aided and abetted by the RIAA members) would also have nothing to do with it. Nope. Nothing at all.
But have they actually done this? [...]Can you give me an actual example where WalMart.. not the studio but WalMart.. ordered the editing of the movie?
Pardon, I misspoke. What I intended to convey was that I do not like bowdlerizing movies, and by extension the carrying of those movies as the selection in a store. It does not matter to me who is doing it. Again, I don't think that movies should be edited until they are appropriate, they should be appropriate to begin with, or not.
However, I'm making it sound like I wish for this to be forbidden, and nothing could be further from the truth. I am merely stating why I do not buy movies from Wal-Mart. That's all.
My earlier post was an attempt to summarize the problems people have with Wal-Mart in order to show that it is not the censorship that's the issue at heart (witness my statement "You might as well complain at the smaller stores for not experimenting a bit (they don't have much to lose if Wal-Mart is in town), or the locals for not supporting alternatives.").
This leads me to your next point:
Censorship means it is removed from you.. you cannot get it. you most certainly can still get that version of the movie (if it exists) somewhere else. Walmart just chooses not to carry it. there is a huge difference from one to the other.[split for comment -ed]
Fine by me. I very much support stores being more than just warehouses --- Wal-Mart has an identity: people know that if they buy movies there that it won't be porn. That's great. I still wish they wouldn't sell or request the making of a censored version, not because I think it's wrong, but because I think it makes the point less "Family Friendly" and more "Family Friendly Because We've Cut Out Everything You Might Not Like" (much like when USA showed Animal House a few years ago. The movie was much shorter and there were a lot of lips moving where different voices were coming out. It didn't stop it from being a raunchy film, it just made it a superficially clean one. Bleaching, if you will). But I will by no means try to get a law passed keeping Wal-Mart from doing it.
[resume -ed]After all, it is the studio that is making the changes to the film.. not walmart.. and they might well be doing it at walmarts request.. but it is still the person making the product that is making the changes.
Yes, ultimately we must take responsibility for what we do; However, it is still Wal-Mart's decision that a recut movie is acceptable while the original is not.
And as much hyperbole as you want to call it, I have a 10 year old, and another child on the way. I try to listen to what my son listens to, and pay attention to it and discuss it with him.
I am sorry, it seems to be my day for giving the wrong impression. I'll make another attempt: what I want is for you to have good and accurate information on the media out there and the ability to permit your child to see what you think is ok.
For the first, I think the current rating systems are just plain awful. They convey next to nothing. I gave the example of ScreenIt as a good (and free, no less) service
that does this, and I hope that it catches on and similar things come out. My problem with most family rating systems --- that I have seen --- is that they do not give you the information to make your own decision. Rather, they give age ranges or pulpit pounding. That's okay if that's what you want, but I really don't think there's enough good information out there right now.
Of course, the best thing is to keep track of what's going on yourself, as you are, so that's great. But it's not always possible.
Secondly, I dislike laws and regulations that set blanket denials on minors without regard to what the parents think is ok. Again, my problem is with the lack
I believe that the frustration most people
feel towards Wal-Mart lies within two realms:
Only available local store
Many people in smaller towns are frustrated by the homogeneity of local selection. Wal-Mart is the biggest example of that around, so it's easy to vent at them. You might as well complain at the smaller stores for not experimenting a bit (they don't have much to lose if Wal-Mart is in town), or the locals for not supporting alternatives.
Bowdlerizing
This is the camp that I fall into. I believe that a movie should be in line with your standards concerning watchability by itself, not after editing; therefore, I do not approve of removing content from movies[1]. Wal-Mart does this (at least, they did last time I checked. Blockbuster is bad about this as well). In addition to this annoyance, often times the fact that the movie is edited is not apparent on the box (mostly to not mess with the artwork), so I never know if I'm seeing the movie or not.
Finally, Wal-Mart has been known to carry enough weight to cause censorship in movies for the theatrical release version (it's expensive to make two versions and doesn't really please anyone, and Wal-Mart likes to have a large selection of movies to sell). That bothers me.
So, it's not really a problem with Wal-Mart's decision of what to sell and not sell, and I salute their intent if not their actions.
Also, I would like to add that those of us who decry censorship do not advocate "streaming porn on the projector at elementary schools." I know your statement was hyperbole, but it reflects upon an important issue: it's not that we want universal availability of all media to minors (after all, that would be trampling upon the rights of their guardians, who are supposed to be raising the kids, after all). For myself, I would like more in-depth information on movies (ScreenIt does a wonderful job here, but she's only one gal and there's a lot of movies out there), and I would like for parents to be able to give kids the freedom to access more at the parents' discretion.
I'd be interested in hearing your opinion on this, as you have a differing one from mine.
[1] I also believe in the creation of an artist (no matter how bad), and that the only good reason not to watch all of a piece of art is if you're not enjoying it.
We don't like to control what people do. Honestly, if AIM does it for them, they're welcome to keep doing it.
My issue is mostly that a) AIM is unprofessional, and tends to get your home life mixed with your work life in an annoying way, and b) some people don't want to sign up for AIM.
These could both be solved by Jabber, but I wouldn't really be serving people if I blocked AIM.
We're a pretty hands-off shop. We get more done that way.
Also, you can tie your current authentication systems into the server. jabberd, the open source server, can easily be extended with a script to use PAM (and there are several out there to do that), and there are also several scripts out to authenticate against a domain.
There is a server for windows that allows you to authenticate directly to the domain.
Has anyone here succesfully rolled out Jabber in a corporate environment, how succesfull was it and how well did the users react to it?
There have been quite a few, according to the Jabber mailing lists. We're working on one right now, and it has done rather poorly - most of the users are not very interested in keeping in touch. That and the clients are pretty bad. Oh, and the ones who do do IM already run on client, and are highly resistant to running another one. In other words: inertia is a big factor.
We've had some people start to use it, mostly to communicate with me, personally.
We saw some usage in the first week, then it's like everyone forgot about it. I view it as a lost cause, but I'm going to keep it available as I use it and I'm moving our team over to it.
On the good side, every time AOL's server becomes inaccessible, people hop on our Jabber server. They usually stay on until the end of the day.
Would you skydive with a parachute pre-packed and sold to you by some anonymous guy in a back alley?
For the same reasons no one should use an insulin pump that isn't guaranteed by a company.
If no guarantee is required to sell insulin pumps and a fellow chooses an insulin pump that is not guaranteed and it fails, he has only himself to blame (financial matters aside).
The issue is that guaranteeing something is an expensive process, just like getting an SLA for service quality you already have is most likely going to cost quite a bit. The UCITA would have you believe that a guarantee is free: it comes with every piece of software. This is not --- and should not be --- the case: I can't afford to guarantee my work, but I do promise to label my releases by how confident I am in them.
Unless, of course, you were to pay me. A lot of money. Then I could hire someone to do an outside code audit as well as do my own internal beefing up of a version I am confident in. And I could build a testing environment for it. And prove the algorithms correct. And get a custom machine built for it. That's what it takes to "guarantee" software.
This "big media monopoly" is such a myth. The networks, newspapers, internet sites compete viciously against each other
Perhaps it is.
In this case, the article is discussing the issues that can result from a group that pools its patents, creating a "virtual monopoly" - more like an oligopoly, but not really.
As to the myth of the "big media monopoly", I believe the "monopoly" is more of a statement of frustration regarding the fact that most people get their news from organizations that toe one of two party lines: you have the somewhere-in-the-socialist-field CNN, NYT and company. Then you have the somewhere-in-the-authoritarian-field FoxNews, and company.
This is probably because most people only accept the first story they hear regarding an issue that does not immediately contradict their preconceptions, and if they hear anything else to the contrary, it's filed under B for Bullshit.
This leads to homogenous news organizations.
Also, there are about two major cliques in the news media, and journalists seek approval from one of them. This results in a dichotomous but homogenous sub-culture that pervades news organizations. This culture picks who gets to be a reporter, and who doesn't.
So, is the "media monopoly" a myth? Yes. There is no one select group that controls and commands the news media. But there certainly is a behavioral system that regulates the news media.
To clarify what I mean by this, and to show how meaningless yet meaningful the conclusion is, let me compare this situation to Asshole Drivers(tm).
Asshole Drivers are everywhere. They seem to be a bit different from city to city, but they are everywhere. They cut you off, they slow down traffic because they refuse to merge, they speed up to keep you from merging. We know them by sight.
What causes there to be so many of them? Hard to say, but I think it's pretty easy to say that a combination of environment and human nature combines to create the self-centered bastards. I don't think they are employed to be that way by a business, nor do I believe that it is a form of religious worship (though some days I have my doubts....).
If this is true, then Asshole Drivers are not created or controlled by a monopoly, but there certainly is a behavioral system that produces them.
Now, I'll make this last part quick. Your comment about "I see plenty of choices on tv, radio, and the Internet than ever before."
Yes, they do compete, but the movie and music companies (who are the ones at issue here) not only compete with one another, but they gang up to destroy smaller companies. You can do that without being controlled by one person or board of directors.
There are numerous cases (that I wish I had links to so that I could cite them) showing collusion amongst RIAA and MPAA members to oust independents. Strong-arming distribution companies, prevention of advertisement from stores that wish to keep their "special deals." Etc, etc, etc.
Now, having said all this, let me tell you that I do not believe "all big corporations are evil." I believe that most problems with society right now come about because entities are not equal before the law and law-makers. I do not know how to solve these problems, but I do know they are a problem, and I'd like to hear solutions.
Thanks for the heads up, I wasn't aware it was so bad.
Luckily, we do not use Excel's statistical routines, we use an Excel plugin that utilizes Matlab. Once new methods are developed from our informatics person's tweaking the data and ideas, everything is then done in a combination of Matlab and Excel. No calculations are done with Excel, however, except for some very trivial ones (addition, subtraction).
My point was mostly about the interfaces being used, and the nature of "open" development in most of science: you can see what it's doing at the conceptual level, and if you invest enough time (Matlab), you can control some more. It is not, however, Open.
A research lab, especially a public one, is first and foremost obligated to do science. To clarify what I mean by this, let me say what I mean:
The experiment must
Have a clear question in mind, and a method for discerning success from failure (not as easy at it sounds, when your result is not binary).
Be repeatable by your lab.
Be repeatable by other labs.
In a physical experiment, it is perfectly acceptable to use a proprietary kit, provided that you can:
Show that the results are replicable
Give other labs access to said kit (for a price) and have the ability to produce more
Show that the results have relevance to what it is that you are trying to achieve
That is true for physical experiments. For manipulation of the data of an experiment, however, the procedure must follow a published or publication-pending method of analysis if you intend to have your research be considered legitimate.
It is coming to pass that algorithms are becoming complex enough and analysists savvy enough, that it is often more practical to produce clear, well-documented source code in addition to your paper than it is to go over and over again the fine points of your method with every interested party.
This leads me to my point: in physical interaction, proprietary and closed methods will most likely remain prevalent for many years to come.
In information manipulation, however, open methods are becoming a dominating trend, if only for the clarity they afford.
In my lab we do bioinformatics research, and we could not do research on the scale we are doing at the pace we are doing it if we were depending upon proprietary software: the proprietary companies cannot compete on customization, new development (after all, we're the ones creating the method), user interface (we're improving all the time, most proprietary packages have ancient user interfaces that are clunky and just plain awful, read: GeneSpring and friends), and cross-lab communication and auditing.
These proprietary solutions are forcing many theoreticians to use software that is "open" in another way: Excel. Yes, I kid you not. Quite a bit of bioinformatic analysis development is done in Excel because the proprietary solutions are just too closed.
The only proprietary companies that are on the right track, in my opinion, are the ones that allow you to use the app as a hub for many other componentized programs.
I'm not saying I know what the answer is, I'm just arguing that easily modifiable engines that hang around for five+ years is certainly not it.
I think people said (and say) the same thing about Flash: most things produced with it are awful. Sites that use it to the exclusion of normal navigation are even worse.
But at the same time, Flash has caused an explosion of amateur animator's work to be available online. A lot of it is awful, but some of it is good.
As the tools get better, less and less focus is made upon the technical ability with the tool (though we will always appreciate excellence in design). This allows the right people to do something.
Consider this: most of the best stories in our culture were at one time oral tradition: It was the only medium accessible to your average storyteller. Print was out of the picture.
Over time, advances have made it so that, now, any damned fool can write a story that could be viewable by the whole world. This lower technical barrier to entry has resulted in more crap, but it has also resulted in more good stuff being available.
My point is that once the rendering aspect of an engine stops being the selling point (Carmack believes this will be true Real Soon Now, and I am inclined to agree with him), the focus will shift to making the engine a tool instead of the centerpiece. We are seeing the vestiges of this right now.
One of the few truly innovative mods I've seen for FPS games has been Natural Selection. Not 100% original, but certainly quite a bit different from your average mod. It really tweaks with the team dynamics: something I haven't seen done successfully in any mod to date.
When short, playable, proof-of-concept games can be cranked out about as fast as a rough draft of a short story, we will see great innovation in games (note: and be of about the same quality, depth, and length as said short story).
Also, I am interested in the techno system of creativity: one person puts out something that's "pretty good." Other people come upon it, play with it, and one or two will come up with something much, much better. This willingness to play together gives us quite a bit in creativity.
The problem right now is that, to play with a game you pretty much have to entirely recreate it or be very familiar with the coding style of the programmer involved - and that's if you have access to their code and can use the engine that they use.
Text-based adventure games had some elements of "quick to crank out" and "can play with another's code", and that was without as big of a following and without the internet (until modern times, and some of the stuff coming out now is quite good - though I have never had the knack for text adventure games, sadly enough).
These are just some thoughts, let me know what you think.
Something about this article really set me off. Read if you're interested.
First off, the only way to solve the "too much crufty information" is to implement a central planning system of some sort. This doesn't have to be a "central committee" or anything like that. It can be as simple as marketing types deciding what's cool and what's not. Then you'd have to deny people the ability to put up (or at least have linked to) what you consider "uncool."
You'd have to do this to reduce the amount of cruft.
Of course, we've seen the success of these restrictive planning systems.
The funny thing is that the internet does this, after a fashion. Things that people aren't interested in are not linked to, and thus are far less likely to be found by search engines. If you wanted to eliminate cruft, you might just not pay attention to anything below a certain linking threshhold. It wouldn't be perfect, but it wouldn't be bad.
Also, a note about receiving crufty emails. Yes, it happens. People send you the strangest, most useless crap all the time. We have a situation where I work where two or three secrataries dominate my inbox by inundating it with "word tips," "lost and found notices," and seminar announcements. It makes up about 1/4 of my non-automated email every day. Can't delete it because important stuff is sometimes in there.
The problem is because the sender thinks that these things are important. Most likely, in her little microcosm, it is. If you ask a secretary the most important things about running an office, you'll probably get answers like "office supplies and appointments," or "maintaining a friendly atmosphere" (which they interpret to mean everyone helping everyone else).
Of course, she's not alone. All of us have different priorities for running the business. I think that keeping the main fileserver and code repository running is a big thing. Surprise, surprise, that's my main job.
I think the solution to many of these problems is not to filter the emails (though that certainly helps), nor is it to disallow the emails. I think the best solution is to provide an appropriate forum. A private usenet server is an excellent place to post seminar announcements. A web page or wiki is a great place to have end-users help end-users. Might also be a good spot to have that lost-and-found list.
We've put up a wiki, and I'm introducing people to it, especially the serious email bombers (or maybe they're strafers?). We'll see how it works out.
Another problem, of course, is making all of these tools available and usable in the right way without the tools consuming all of your work time. I don't know the solution to this, but I do think that proper tools and proper integration are the future path of the information world.
If we accept as given that most people code open source to "scratch an itch" not "replace something that works" (and that does seem to be true. Most "replacements" are replacements that do something more/different), then I don't think that this is true. Well, it is and it isn't. Hear me out.
A programmer working on a package to get it to do something that he needs done will work on it to the point that it does what he wants. A bug that blocks him from doing what he wants means that he's not done yet, hence bugs will get fixed as well. A counter-argument might be made that the programmer is less likely to write good code to fix a bug than to write good code to implement a feature, but, honestly, this is not a big issue. Fix the bug to where it's fixed, and next time you implement a feature, re-organize the code. That seems to be what most people do anyways.
But why, then, do we have buggy programs and no one willing to fix those bugs before implementing a feature? Well, very few people get up in the morning and thing "I'm going to knock out some bugs that don't really affect me today." Unless that person is the maintainer, of course, then he might think that.
People are goal oriented, not task oriented. I think that's one of the reasons why so few of us have the ability to do very good time management, yet most of us have the ability to get something done.
Anyways, that's my theory, let me know what you think.
What I've found in my experience of writing software is that you end up writing your own libraries for practically everything because the existing library is missing a feature, has bugs, is too slow, isn't being supported anymore, or just does things in strange ways.
Forgive some sarcasm, but I often find that I rewrite libc and qt. getopt and perl are re-implemented all the time, and gcc is a horrid hack compared to a hand-tuned compiler for my specific application. make is a total hack, and I write my build scripts in assembly.
Seriously, this bit of "writing your own because the others aren't good" is the exact problem facing us today. Here we go:
"Missing a feature." Well, if it's missing a feature, why not add it and get the benefits of the other person's bugfixes and reduces the amount of virgin code introduced.
"has bugs" why not fix them? Fix the devil you know, and avoid inviting in the devil you don't.
"isn't supported anymore" the reason so many of these libraries aren't supported anymore is because people write their own. Very, very, very few libraries can stand the test of 5 years of upgrades to software without any maintenance and upkeep. Your unwillingness to learn a new system and fix its problems, rather to forge ahead with your own problems and lack of maintenance for the future bespeaks a lack of understanding as to where the problem lies, IMHO.
As far as being dependent upon another for core code...I have no problem with that as long as I modularize my code sufficiently that I can switch to another or my own system. If the other person's code isn't good enough, I'll put up with it until my system is good enough that the other's library is the critical bug, then I will swap out their code for another, or perhaps my own.
Note that many larger programs are often written using buggy subcomponents that are later swapped out for better ones rather than sitting in the citadel and crafting the Perfect System From Scratch.
Also, you run into the issue that your code may work fine for you (works for you, no way to tell if it's broken), but it breaks when you give it to other people.
Anyways, sorry for jumping on you, I might be wrong about all of this (though I doubt it), but your post is along the lines of some ideas I've had to deal with recently.
I think that some of the research being done into programs that walk code and look for certain things to be true (a subset of code checkers) will produce some good things, hopefully soon.
I think the ability for programmers to write tests for programming errors in a general enough way will be wonderful. Being able to write a test to ensure that, no matter what, that specific coding error will never happen again (and as you discover more similar ones, your net will widen, and as you get false hits your accuracy will increase) will be a great boon for programmers.
Part of the problem with systems like Purify is that they do not allow the easy addition of tests to the battery (do-able in less than 5-10 minutes for a basic test).
But I think they are on the right track. We won't catch everything, but we can catch many things.
Oh, and as a corollary: I think that one of the most useful roles of a good IT person in a company is the role of liason between the user and the developers. You can train a liason to give and get good bug data, and get a good relationship going with them. Much harder to do with all of the individual users.
Also, some liasons are willing and able to write code. A lot of fink and Debian is done this way, I'm sure.
Slashdot Mirror
Damned if I can't find it. Must not've been in Salon. Weird. Ah well, I remember reading it.
You know, it's funny. I'm a bit fuzzy on the dates, but child pornography was only made illegal in the US about 40-or-so years ago.
Of course, you have to separate pornography into two categories: 1) a media work showing an explictly sexual act (masturbation, penetrative sex, oral sex, etc) 2) a media work of a prurient nature that does not explicitly show "sex."
The former was illegal (minor consent laws and all that), but the latter was kinda legal.
When I say "kinda" I'm not being wishy-washy, it's that we're coming up against anachronism: according to experts on this sort of thing[1] attitudes of the "man on the street" have shifted drastically concerning photographs/drawings of young children. What would be considered "cute" and "childlike" back in the 1950's would be considered "grotesque" and "unsuitable for public consumption" now. Most of this can be linked to the witchhunts regarding child pornography.
Even more amusing, since the enactment of child pornography laws the average age of actors involved in sex scenes and -- and this is very odd -- in just plain romantic kisses in movies has gone way down. Way, way down.
I think it was Ebert (or was it Stephen King? King wrote about this a bit as well) who commented that youngsters used to go to the movies to see oldsters in a sex scene. Now it's reversed.
Salon commented[2] that the rabid and far-reaching bans on child "pornography"[3] has caused us all to think like pedophiles. Reminds me of the joke about the sexaholic who goes into the psychiatrist, who gives him a rorschach test.
Ah well.
So my point is that it's tough to say what's child porn and what's not, with the fact that we've gotten hypersensitive about it recently. I don't really feel like having the morality police check out all of my mom's photo albums to see if they're kosher. Especially morality police from the MPC[4]. Though being a smart son who knows the power of embarassment blackmail from mothers, I have removed all of my naked baby pictures. Now if only I could get the picture of me hugging then penguin at Sea World when I was 12....
And to spare myself accusations: no. I'm not "into" kids. To be honest, only recently has my age group become somewhat appealing to me at all: I've always been after older women, which is a real drag as women are used to young guys being...well, young guys.
So, just something to think about.
[1] I'm thinking of the historians interviewed in Salon concerning the somewhat-recent Paul Ruebens case. Interesting article.
[2] Again, the Paul Ruebens case.
[3] bare-butt baby picture arrests, anyone? Traditional Brazilian breast feeding family photos getting your kids under the care of CPS, like just happened in Dallas about a year ago?
[4] Most Persnickity Country
Ok, you had me going until the "girlfriend" bit.
My guess is that the amount of effort and documentation required to make a legitimate test look like a legitimate test for "X% Faster" (where X is what people interpret Apple as claiming) versus that required for "within 10%-15% of the top end PC" would be little when, if they believe that the tests are legitimate, they would be hurting their position.
Why? Because the plausibility is not in how the G5 compares to the Intel, it's in how the G5 compares to the G4. People don't trust huge leaps in performance within the same line of chips, and demand more documentation.
And rightfully so - if a company is going to spend a lot of money and develop an all-new architecture that they're going to market as being a massive move from the previous one requiring developer time to move things between the two architectures, then they damned well better be a lot better performing.
Also, people have this weird assumption that just because a chip speaks the same instruction set as the previous one, and is being released by the same company (effectively), then it must be the same stuff internally. We are big believers in the "ground up rebuild" (note how often it's used in marketing and is repeated by others. Marketing is designed to tell us what we want to hear). So if you didn't rebuild the entire chip, and you didn't because it works with the old one, then how could it be N times faster?
Also, marketing companies have been telling us for years "new, improved, works the same as the old one, 10 times faster than the old one" and we've stopped believing them. General loss of credibility.
So, given that sociological assumption, and (charitably) given that Apple believes the benchmarks actually reflect reality, then why not go for the gold? You'd have to weaken your position quite a bit to be considered much more plausible given the general performance and age of the G4 and G4 architecture[1].
Just one possible explanation.
[1] - here I'm talking about the problems we all know about: low bus speed, weak integer performance and a few other things. It really was a very nice chip and once developers started optimizing for it, it showed enough strength in certain areas to work around its deficiencies for a surprising amount.
Yes, and I'm sure Intel would be so very helpful in getting one of their prime competitors pre-release sample versions of their chips.
Apple most likely did have to actually buy that Dell box just like the rest of us poor slobs (if us poor slobs were to buy a Dell, that is). Since they're a big competitor, I doubt they got any special favors about first ship or anything else.
"Pissing on motorcycles" bit was referring to people who don't respect the tools they use, ie motorcycles. Not saying you do that, agreeing that the people who do that are a problem.
As for being rewarded...I think the best reward we could get would be the respect of other motorists and a realization that the entire road is not just for cars (where I live the belief is that the entire road is for trucks, with cars being second-class citizens and motorcycles being flies that bother you).
It was your decision to look at the black box. That's cool. Not an invasion of privacy at all. On the other hand, if you had forced the other person to turn over his black box, then that would be an invasion of privacy - whether it would be acceptable would be another matter.
As a practical matter (as you well know), physics tells us that if total speed is known and *a* car's speed is known, we can know the other car's speed.
Hence it only takes one car having a black box and a willing owner to produce the results you need, no violation of privacy is necessary.
That's not what concerns us "privacy nuts."
It's when you might be forced to turn over YOUR black box. I don't want that to happen to you. First off, because I don't think that's right (though I'm willing to be convinced otherwise).
Second off, as the above basic physics indicates, there is no reason for you to have to turn over your black box for any situation involving two cars if those cars fall within the 98% of those out on the road.
Soooooo...why would the law want to be able to yank your black box if they don't need it for a two-car accident like that? Doesn't sound quite right to me. Maybe they're just not thinking the logic through.
I've read the actuarial summaries related to car accidents. The #1 influencing factor is not the skill of the driver, the state of the road, or even the absolute speed of the car, funny enough.
The biggest indicator of your likelihood to get in an accident is your relative speed related to the other drivers, regardless of the speed limit. If you are driving greater than or less than 10 mph relative to traffic, you are very likely to get in an accident.
Hence your idea sounds good but wouldn't work out too well, unfortunately. Wish it would, though.
One of the other big influencing factors is the number of occupants in a car, particularly where the average age is low (moms with many young children and teenagers in general).
I think we'd all do better to pay attention to the actuarial data and allow insurance companies to make much more specific policies regarding coverage rather than making more laws.
Just an idea.
PS: as a person who likes motorcycles, I agree with you that "cool" is the big reason that motorcycles have gottten a bad name: no helmets, no protective gear, more bike than you can handle, and (on a minor note) incredibly loud pipes. Of course, I don't want to see laws being made regarding any of this, but it's still frustrating to see something you like get pissed on.
This is true for some demographics, particularly the "budget" ones - ie the areas of insurance mandated by the government and such.
It is possible to buy "no fault" insurance, that pays no matter whose fault it was. If there was demand for it, I can see there being "no criminal fault" insurance - if you are not assigned the fault by a court, then the insurance pays out.
Why would the insurance companies do this? Twofold: they don't make much money on just liability insurance. In states that mandate liability insurance, people whose actuarial classification is in the "Danger! Danger, Will Robinson!" area get placed into a state pool where the insurance companies do a lottery to see who has to insure them. This pool costs them money, which they then make up from the next highest bracket. Ever wonder why it takes a long time for your payments to come in line with your actuarial classification? That's why.
Also, most people who get dicked over by insurance companies go for minimum coverage. This is like coach class in an airplane: don't expect great service.
Again, the problem is that people are not willing to pay for what they want. Of course insurance companies are going to try to reduce their payouts to customers who shop around all the time trying to find the lowest rate.
I predict that if we do get insurance companies getting persnickity with the monitoring equipment, then there will be additional grades of coverage. There is competition in the profitable brackets of insurance.
My favorite was the part where the RIAA FSM claimed that RIAA members wished there was greater variety on the radio.
Gosh, wouldn't that be nice.
Of course, the fact that the RIAA members have been focusing on Billboard and mega-hits would have nothing to do with the dearth of radio content. Nothing at all.
Oh, and the fact that payola is alive and well (aided and abetted by the RIAA members) would also have nothing to do with it. Nope. Nothing at all.
Pardon, I misspoke. What I intended to convey was that I do not like bowdlerizing movies, and by extension the carrying of those movies as the selection in a store. It does not matter to me who is doing it. Again, I don't think that movies should be edited until they are appropriate, they should be appropriate to begin with, or not.
However, I'm making it sound like I wish for this to be forbidden, and nothing could be further from the truth. I am merely stating why I do not buy movies from Wal-Mart. That's all.
My earlier post was an attempt to summarize the problems people have with Wal-Mart in order to show that it is not the censorship that's the issue at heart (witness my statement "You might as well complain at the smaller stores for not experimenting a bit (they don't have much to lose if Wal-Mart is in town), or the locals for not supporting alternatives.").
This leads me to your next point:
Fine by me. I very much support stores being more than just warehouses --- Wal-Mart has an identity: people know that if they buy movies there that it won't be porn. That's great. I still wish they wouldn't sell or request the making of a censored version, not because I think it's wrong, but because I think it makes the point less "Family Friendly" and more "Family Friendly Because We've Cut Out Everything You Might Not Like" (much like when USA showed Animal House a few years ago. The movie was much shorter and there were a lot of lips moving where different voices were coming out. It didn't stop it from being a raunchy film, it just made it a superficially clean one. Bleaching, if you will). But I will by no means try to get a law passed keeping Wal-Mart from doing it.
Yes, ultimately we must take responsibility for what we do; However, it is still Wal-Mart's decision that a recut movie is acceptable while the original is not.
I am sorry, it seems to be my day for giving the wrong impression. I'll make another attempt: what I want is for you to have good and accurate information on the media out there and the ability to permit your child to see what you think is ok.
For the first, I think the current rating systems are just plain awful. They convey next to nothing. I gave the example of ScreenIt as a good (and free, no less) service that does this, and I hope that it catches on and similar things come out. My problem with most family rating systems --- that I have seen --- is that they do not give you the information to make your own decision. Rather, they give age ranges or pulpit pounding. That's okay if that's what you want, but I really don't think there's enough good information out there right now.
Of course, the best thing is to keep track of what's going on yourself, as you are, so that's great. But it's not always possible.
Secondly, I dislike laws and regulations that set blanket denials on minors without regard to what the parents think is ok. Again, my problem is with the lack
I believe that the frustration most people feel towards Wal-Mart lies within two realms:
Many people in smaller towns are frustrated by the homogeneity of local selection. Wal-Mart is the biggest example of that around, so it's easy to vent at them. You might as well complain at the smaller stores for not experimenting a bit (they don't have much to lose if Wal-Mart is in town), or the locals for not supporting alternatives.
This is the camp that I fall into. I believe that a movie should be in line with your standards concerning watchability by itself, not after editing; therefore, I do not approve of removing content from movies[1]. Wal-Mart does this (at least, they did last time I checked. Blockbuster is bad about this as well). In addition to this annoyance, often times the fact that the movie is edited is not apparent on the box (mostly to not mess with the artwork), so I never know if I'm seeing the movie or not.
Finally, Wal-Mart has been known to carry enough weight to cause censorship in movies for the theatrical release version (it's expensive to make two versions and doesn't really please anyone, and Wal-Mart likes to have a large selection of movies to sell). That bothers me.
So, it's not really a problem with Wal-Mart's decision of what to sell and not sell, and I salute their intent if not their actions.
Also, I would like to add that those of us who decry censorship do not advocate "streaming porn on the projector at elementary schools." I know your statement was hyperbole, but it reflects upon an important issue: it's not that we want universal availability of all media to minors (after all, that would be trampling upon the rights of their guardians, who are supposed to be raising the kids, after all). For myself, I would like more in-depth information on movies (ScreenIt does a wonderful job here, but she's only one gal and there's a lot of movies out there), and I would like for parents to be able to give kids the freedom to access more at the parents' discretion.
I'd be interested in hearing your opinion on this, as you have a differing one from mine.
[1] I also believe in the creation of an artist (no matter how bad), and that the only good reason not to watch all of a piece of art is if you're not enjoying it.
We don't like to control what people do. Honestly, if AIM does it for them, they're welcome to keep doing it.
My issue is mostly that a) AIM is unprofessional, and tends to get your home life mixed with your work life in an annoying way, and b) some people don't want to sign up for AIM.
These could both be solved by Jabber, but I wouldn't really be serving people if I blocked AIM.
We're a pretty hands-off shop. We get more done that way.
Also, you can tie your current authentication systems into the server. jabberd, the open source server, can easily be extended with a script to use PAM (and there are several out there to do that), and there are also several scripts out to authenticate against a domain.
There is a server for windows that allows you to authenticate directly to the domain.
There have been quite a few, according to the Jabber mailing lists. We're working on one right now, and it has done rather poorly - most of the users are not very interested in keeping in touch. That and the clients are pretty bad. Oh, and the ones who do do IM already run on client, and are highly resistant to running another one. In other words: inertia is a big factor.
We've had some people start to use it, mostly to communicate with me, personally.
We saw some usage in the first week, then it's like everyone forgot about it. I view it as a lost cause, but I'm going to keep it available as I use it and I'm moving our team over to it.
On the good side, every time AOL's server becomes inaccessible, people hop on our Jabber server. They usually stay on until the end of the day.
Would you skydive with a parachute pre-packed and sold to you by some anonymous guy in a back alley?
For the same reasons no one should use an insulin pump that isn't guaranteed by a company.
If no guarantee is required to sell insulin pumps and a fellow chooses an insulin pump that is not guaranteed and it fails, he has only himself to blame (financial matters aside).
The issue is that guaranteeing something is an expensive process, just like getting an SLA for service quality you already have is most likely going to cost quite a bit. The UCITA would have you believe that a guarantee is free: it comes with every piece of software. This is not --- and should not be --- the case: I can't afford to guarantee my work, but I do promise to label my releases by how confident I am in them.
Unless, of course, you were to pay me. A lot of money. Then I could hire someone to do an outside code audit as well as do my own internal beefing up of a version I am confident in. And I could build a testing environment for it. And prove the algorithms correct. And get a custom machine built for it. That's what it takes to "guarantee" software.
Perhaps it is.
In this case, the article is discussing the issues that can result from a group that pools its patents, creating a "virtual monopoly" - more like an oligopoly, but not really.
As to the myth of the "big media monopoly", I believe the "monopoly" is more of a statement of frustration regarding the fact that most people get their news from organizations that toe one of two party lines: you have the somewhere-in-the-socialist-field CNN, NYT and company. Then you have the somewhere-in-the-authoritarian-field FoxNews, and company.
This is probably because most people only accept the first story they hear regarding an issue that does not immediately contradict their preconceptions, and if they hear anything else to the contrary, it's filed under B for Bullshit.
This leads to homogenous news organizations.
Also, there are about two major cliques in the news media, and journalists seek approval from one of them. This results in a dichotomous but homogenous sub-culture that pervades news organizations. This culture picks who gets to be a reporter, and who doesn't.
So, is the "media monopoly" a myth? Yes. There is no one select group that controls and commands the news media. But there certainly is a behavioral system that regulates the news media.
To clarify what I mean by this, and to show how meaningless yet meaningful the conclusion is, let me compare this situation to Asshole Drivers(tm).
Asshole Drivers are everywhere. They seem to be a bit different from city to city, but they are everywhere. They cut you off, they slow down traffic because they refuse to merge, they speed up to keep you from merging. We know them by sight.
What causes there to be so many of them? Hard to say, but I think it's pretty easy to say that a combination of environment and human nature combines to create the self-centered bastards. I don't think they are employed to be that way by a business, nor do I believe that it is a form of religious worship (though some days I have my doubts....).
If this is true, then Asshole Drivers are not created or controlled by a monopoly, but there certainly is a behavioral system that produces them.
Now, I'll make this last part quick. Your comment about "I see plenty of choices on tv, radio, and the Internet than ever before."
Yes, they do compete, but the movie and music companies (who are the ones at issue here) not only compete with one another, but they gang up to destroy smaller companies. You can do that without being controlled by one person or board of directors.
There are numerous cases (that I wish I had links to so that I could cite them) showing collusion amongst RIAA and MPAA members to oust independents. Strong-arming distribution companies, prevention of advertisement from stores that wish to keep their "special deals." Etc, etc, etc.
Now, having said all this, let me tell you that I do not believe "all big corporations are evil." I believe that most problems with society right now come about because entities are not equal before the law and law-makers. I do not know how to solve these problems, but I do know they are a problem, and I'd like to hear solutions.
Thanks for the heads up, I wasn't aware it was so bad.
Luckily, we do not use Excel's statistical routines, we use an Excel plugin that utilizes Matlab. Once new methods are developed from our informatics person's tweaking the data and ideas, everything is then done in a combination of Matlab and Excel. No calculations are done with Excel, however, except for some very trivial ones (addition, subtraction).
My point was mostly about the interfaces being used, and the nature of "open" development in most of science: you can see what it's doing at the conceptual level, and if you invest enough time (Matlab), you can control some more. It is not, however, Open.
Yes and no.
A research lab, especially a public one, is first and foremost obligated to do science. To clarify what I mean by this, let me say what I mean:
The experiment must
In a physical experiment, it is perfectly acceptable to use a proprietary kit, provided that you can:
That is true for physical experiments. For manipulation of the data of an experiment, however, the procedure must follow a published or publication-pending method of analysis if you intend to have your research be considered legitimate.
It is coming to pass that algorithms are becoming complex enough and analysists savvy enough, that it is often more practical to produce clear, well-documented source code in addition to your paper than it is to go over and over again the fine points of your method with every interested party.
This leads me to my point: in physical interaction, proprietary and closed methods will most likely remain prevalent for many years to come.
In information manipulation, however, open methods are becoming a dominating trend, if only for the clarity they afford.
In my lab we do bioinformatics research, and we could not do research on the scale we are doing at the pace we are doing it if we were depending upon proprietary software: the proprietary companies cannot compete on customization, new development (after all, we're the ones creating the method), user interface (we're improving all the time, most proprietary packages have ancient user interfaces that are clunky and just plain awful, read: GeneSpring and friends), and cross-lab communication and auditing.
These proprietary solutions are forcing many theoreticians to use software that is "open" in another way: Excel. Yes, I kid you not. Quite a bit of bioinformatic analysis development is done in Excel because the proprietary solutions are just too closed.
The only proprietary companies that are on the right track, in my opinion, are the ones that allow you to use the app as a hub for many other componentized programs.
I think people said (and say) the same thing about Flash: most things produced with it are awful. Sites that use it to the exclusion of normal navigation are even worse.
But at the same time, Flash has caused an explosion of amateur animator's work to be available online. A lot of it is awful, but some of it is good.
As the tools get better, less and less focus is made upon the technical ability with the tool (though we will always appreciate excellence in design). This allows the right people to do something.
Consider this: most of the best stories in our culture were at one time oral tradition: It was the only medium accessible to your average storyteller. Print was out of the picture.
Over time, advances have made it so that, now, any damned fool can write a story that could be viewable by the whole world. This lower technical barrier to entry has resulted in more crap, but it has also resulted in more good stuff being available.
My point is that once the rendering aspect of an engine stops being the selling point (Carmack believes this will be true Real Soon Now, and I am inclined to agree with him), the focus will shift to making the engine a tool instead of the centerpiece. We are seeing the vestiges of this right now.
One of the few truly innovative mods I've seen for FPS games has been Natural Selection. Not 100% original, but certainly quite a bit different from your average mod. It really tweaks with the team dynamics: something I haven't seen done successfully in any mod to date.
When short, playable, proof-of-concept games can be cranked out about as fast as a rough draft of a short story, we will see great innovation in games (note: and be of about the same quality, depth, and length as said short story).
Also, I am interested in the techno system of creativity: one person puts out something that's "pretty good." Other people come upon it, play with it, and one or two will come up with something much, much better. This willingness to play together gives us quite a bit in creativity.
The problem right now is that, to play with a game you pretty much have to entirely recreate it or be very familiar with the coding style of the programmer involved - and that's if you have access to their code and can use the engine that they use.
Text-based adventure games had some elements of "quick to crank out" and "can play with another's code", and that was without as big of a following and without the internet (until modern times, and some of the stuff coming out now is quite good - though I have never had the knack for text adventure games, sadly enough).
These are just some thoughts, let me know what you think.
Something about this article really set me off. Read if you're interested.
First off, the only way to solve the "too much crufty information" is to implement a central planning system of some sort. This doesn't have to be a "central committee" or anything like that. It can be as simple as marketing types deciding what's cool and what's not. Then you'd have to deny people the ability to put up (or at least have linked to) what you consider "uncool."
You'd have to do this to reduce the amount of cruft.
Of course, we've seen the success of these restrictive planning systems.
The funny thing is that the internet does this, after a fashion. Things that people aren't interested in are not linked to, and thus are far less likely to be found by search engines. If you wanted to eliminate cruft, you might just not pay attention to anything below a certain linking threshhold. It wouldn't be perfect, but it wouldn't be bad.
Also, a note about receiving crufty emails. Yes, it happens. People send you the strangest, most useless crap all the time. We have a situation where I work where two or three secrataries dominate my inbox by inundating it with "word tips," "lost and found notices," and seminar announcements. It makes up about 1/4 of my non-automated email every day. Can't delete it because important stuff is sometimes in there.
The problem is because the sender thinks that these things are important. Most likely, in her little microcosm, it is. If you ask a secretary the most important things about running an office, you'll probably get answers like "office supplies and appointments," or "maintaining a friendly atmosphere" (which they interpret to mean everyone helping everyone else).
Of course, she's not alone. All of us have different priorities for running the business. I think that keeping the main fileserver and code repository running is a big thing. Surprise, surprise, that's my main job.
I think the solution to many of these problems is not to filter the emails (though that certainly helps), nor is it to disallow the emails. I think the best solution is to provide an appropriate forum. A private usenet server is an excellent place to post seminar announcements. A web page or wiki is a great place to have end-users help end-users. Might also be a good spot to have that lost-and-found list.
We've put up a wiki, and I'm introducing people to it, especially the serious email bombers (or maybe they're strafers?). We'll see how it works out.
Another problem, of course, is making all of these tools available and usable in the right way without the tools consuming all of your work time. I don't know the solution to this, but I do think that proper tools and proper integration are the future path of the information world.
Anyways, let me know what you think.
If we accept as given that most people code open source to "scratch an itch" not "replace something that works" (and that does seem to be true. Most "replacements" are replacements that do something more/different), then I don't think that this is true. Well, it is and it isn't. Hear me out.
A programmer working on a package to get it to do something that he needs done will work on it to the point that it does what he wants. A bug that blocks him from doing what he wants means that he's not done yet, hence bugs will get fixed as well. A counter-argument might be made that the programmer is less likely to write good code to fix a bug than to write good code to implement a feature, but, honestly, this is not a big issue. Fix the bug to where it's fixed, and next time you implement a feature, re-organize the code. That seems to be what most people do anyways.
But why, then, do we have buggy programs and no one willing to fix those bugs before implementing a feature? Well, very few people get up in the morning and thing "I'm going to knock out some bugs that don't really affect me today." Unless that person is the maintainer, of course, then he might think that.
People are goal oriented, not task oriented. I think that's one of the reasons why so few of us have the ability to do very good time management, yet most of us have the ability to get something done.
Anyways, that's my theory, let me know what you think.
Forgive some sarcasm, but I often find that I rewrite libc and qt. getopt and perl are re-implemented all the time, and gcc is a horrid hack compared to a hand-tuned compiler for my specific application. make is a total hack, and I write my build scripts in assembly.
Seriously, this bit of "writing your own because the others aren't good" is the exact problem facing us today. Here we go:
"Missing a feature." Well, if it's missing a feature, why not add it and get the benefits of the other person's bugfixes and reduces the amount of virgin code introduced.
"has bugs" why not fix them? Fix the devil you know, and avoid inviting in the devil you don't.
"isn't supported anymore" the reason so many of these libraries aren't supported anymore is because people write their own. Very, very, very few libraries can stand the test of 5 years of upgrades to software without any maintenance and upkeep. Your unwillingness to learn a new system and fix its problems, rather to forge ahead with your own problems and lack of maintenance for the future bespeaks a lack of understanding as to where the problem lies, IMHO.
As far as being dependent upon another for core code...I have no problem with that as long as I modularize my code sufficiently that I can switch to another or my own system. If the other person's code isn't good enough, I'll put up with it until my system is good enough that the other's library is the critical bug, then I will swap out their code for another, or perhaps my own.
Note that many larger programs are often written using buggy subcomponents that are later swapped out for better ones rather than sitting in the citadel and crafting the Perfect System From Scratch.
Also, you run into the issue that your code may work fine for you (works for you, no way to tell if it's broken), but it breaks when you give it to other people.
Anyways, sorry for jumping on you, I might be wrong about all of this (though I doubt it), but your post is along the lines of some ideas I've had to deal with recently.
I think that some of the research being done into programs that walk code and look for certain things to be true (a subset of code checkers) will produce some good things, hopefully soon.
I think the ability for programmers to write tests for programming errors in a general enough way will be wonderful. Being able to write a test to ensure that, no matter what, that specific coding error will never happen again (and as you discover more similar ones, your net will widen, and as you get false hits your accuracy will increase) will be a great boon for programmers.
Part of the problem with systems like Purify is that they do not allow the easy addition of tests to the battery (do-able in less than 5-10 minutes for a basic test).
But I think they are on the right track. We won't catch everything, but we can catch many things.
Oh, and as a corollary: I think that one of the most useful roles of a good IT person in a company is the role of liason between the user and the developers. You can train a liason to give and get good bug data, and get a good relationship going with them. Much harder to do with all of the individual users.
Also, some liasons are willing and able to write code. A lot of fink and Debian is done this way, I'm sure.