I agree with your analysis (ie: that criminals are criminals and that they will just move on to the next thing). But just like drug legalization would remove the demand for illegal drugs, opening the borders would remove the demand for human trafficking. Just food for thought.
How exactly do you propose growing the Coca plant in the US? Hint: there's a reason it is trafficked all the way from Columbia.
A prohibition on all exports and imports of drugs would be as effective as a similar measure on produce. While we grow tomatoes here, we also import them from and export them to Mexico. Market supplies are global, and that isn't going to change. Tariffs and prohibitions only create artificial scarcity, which of course leads to black markets and inflated prices.
Uh, no. First of all, raw milk can be legally sold in more than half of the states in the US. Second, in the states where it is illegal, it is not criminal. It is merely prohibited, which means if you are caught you will face civil penalties (fines, suspension of food license, etc).
The money made by the drug cartels is in cocaine and heroine as the GP stated. Other drugs, like pot, ecstasy, and meth are much easier to obtain in the US. With the high risk involved, there is no sense in trafficking it. The profitability is too low. Also, a few states have already decriminalized possession of marijuana. I don't expect all of them to any time soon, but more will eventually. So for the recreational user it won't be hard to obtain.
Well, sure, but what do you propose? Retailers either have to absorb risk, or pay for mandated security. As far as absorbing risk is concerned, they do some. Verification of signatures for transactions is required to show due diligence, for example. But they mostly pass on the cost of fraudulent transactions. In cases where retailers do absorb the entire cost of fraud (ex: checks), they just drop it as a form of payment. So, no, I don't think there is a realistic scenario where small retailers can afford to absorb the costs of fraud because, as you say, profit margins are usually low already. So that leaves paying the (possibly high) cost for good security....
A ton of breaches happen at the small retail level. That's why the bankcard industry is requiring increased security. They aren't lucrative breaches, so you don't hear about it on the news, but they are significant. If an organized group can steal $10 from 10,000 accounts, that's $100,000. It adds up. And if it's a credit transaction, Visa et al are left absorbing the damages. So they have an incentive to tighten security.
Uh, you do realize those "man made rules" are based on observation. Extensive and thorough observation. Basically, the need for oxygen, carbon, nitrogen, hydrogen, sulfur, and phosphorus to sustain life has been demonstrated not just at the organismal level, but also at the molecular level, with predictable and repeating patterns in every life form ever studied. That is the definition of a "Law" in science. Something that is assumed to be a fact because exceptions have never been observed. It, of course, doesn't mean there is no possibility of a life form that "breaks the Law", but the probability is low and such claims are rightfully met with a healthy dose of skepticism. As long as the right controls are done, though, and alternative explanations ruled out, it will eventually be accepted by the scientific community, most likely with a lot of enthusiasm. Findings such as these typically raise more questions than they answer, and a lot of scientists will surely be interested in the underlying enzymology and such, if it holds up to the current scrutiny.
Ok, fair enough, you are right. I see what you are arguing for. You can't prove it is theoretically impossible to do what you are describing. However, I like to think of progress in science, and in particular the success of projects like this, as proceeding in two ways. The first is the revolutionary approach, where completely new and groundbreaking ways of doing things are discovered, or made economical. The second is constrained by practical considerations, and therefore takes a more incremental approach. Both approaches have played a role in the development and progress of science.
Let's take superconductors as an example. Room temperature superconductors are not theoretically impossible, and many would believe that they are very possible. If we succeed in developing them, we could revolutionize the use of techniques like MRI, NMR, and FT-MS. We could make instruments more portable, more energy efficient, more economical. We could envision uses of these devices that aren't possible now because they currently need environmentally-controlled facilities and a full-time trained staff to operate them. Despite all of this, however, we still have a ways to go before we get there. On the other hand, we have had success in making superconductors that can operate at warmer temperatures. So there may be ways to use this to make incremental improvements in these devices (perhaps there already are, I'm not sure).
So in the context of a contest like this, which approach will you choose: the revolutionary approach, or the incremental approach? I guess it depends on your preference, but I think the incremental approach is more likely to yield results in the short term. We currently have the ability to identify pathogens very accurately. And it isn't very expensive, either. It just requires a lab. So the incremental approach would be to try to miniaturize the lab. Like a lab-on-a-chip sort of thing. Something that could be operated anywhere, by minimally-trained staff. I think it's possible and practical, and there are groups already working on things like this. George Whitesides has made a career in investing in ideas like this. And it works. He has been very successful.
So, yeah, the revolutionary approach is interesting and exciting. It shouldn't be abandoned. It absolutely has a place in the advancement of science. Without it, we would never try to achieve new things. But in the interest of practicality and development of an actual product, sometimes it's best to just leverage current technology. To figure out how to make it better, rather than how to replace it with something else.
Just my $0.02. I'll be interested to see how this pans out. It certainly encourages people to have new ideas and to develop them. So let's see what they come up with.
I like your enthusiasm, but I think you are really overestimating the feasibility of your approach. How exactly do you "IR for a long, complex chain molecule", as you replied to another poster (hint, there are millions of them in biological systems)? The point he was making is that all of the common spectroscopy techniques are useless: IR, NMR, and UV/Vis. Even a short pure DNA strand can't be sequenced with IR, NMR, or UV/Vis...that is, unless someone invents a miraculous way to deconvolute the spectrum. It is far more complicated than a dictionary search, useful as those have proven to be in other areas.
There is a reason why we use the techniques we do. But of course that doesn't mean we can't do it better. DNA sequencing has been around since the 70's, but improvements over the last decade have really revolutionized its use in biological research. We can now feasibly talk about sequencing the genome of an organism as a routine thing, rather than a years long and millions of dollars effort. And single-molecule sequencing is also starting to have some seriously promising results. So maybe the right thing to do is figure out how to efficiently retrieve and prepare samples from patients, so that sequencing for pathogenic DNA can be done quickly and cheaply. It's not quite like a tricorder, but it has the potential to be automated in the way the contest seems to want.
Honestly, I'm not sure how helpful something like this is, though. We can already do diagnostic screening very accurately and quickly. Making it not require a lab would certainly be a nice development. But, as another poster already replied, screening for pathogens is a very small part of medicine these days, even in developing countries. Things like x-ray scanners and MRIs are becoming a lot cheaper and more portable, though. For MRI what they really need is a breakthrough in the magnet technology, but in the meantime there are companies specializing in things like small hand MRIs, which can have some useful applications. One "diagnosis machine" is still very much in sci-fi land, though.
Well, this is an interesting idea, but the major cost being left out of this calculation is the operating cost. No nuclear reactor will be licensed to operate in the US without a full staff complement. At a minimum, this usually entails two full-time licensed nuclear reactor operators, security, and auditing and regular safety review. Don't be fooled. This is a power plant, even if it is a small one. It's not going to be economical for a random business. The website says it would work for things like large government and hospital complexes. In these cases it might be conceivable to hire staff to run an in-house power generation facility, but I doubt it would work out for an office building downtown.
True or not, the perception is that after a certain point... that highly specialized degree doesn't really translate into any knowledge that anybody needs,
I really don't buy this argument at all. My only experience is with the life sciences. Maybe some of the other fields are different. But in the sciences your research has to be funded by a grant. The granting process is extremely competitive. You're not going to get funding without convincing others that your research is significant, useful, and that you are productive. I don't know where people get the idea that the research is so highly specialized it is useless. Highly specialized, maybe, but not useless. Unless you think basic research related to cancer and infectious disease is useless.
Then again, maybe the author is his own example. He is so specialized in his own field that he doesn't know what happens in the academic world outside of it.
Why should people have to change at Ubuntu's pace to continue to get their work done?
Um, they don't. You don't have to upgrade to the latest version. I'm still running 9.10 on one of my desktops because I needed it to be a stable machine to write my thesis on. I didn't have time to upgrade and I didn't want to spend a lot of time doing backups, repartitioning, etc. So I just left it, and it still works great. Seriously, if you don't want/need the new stuff, don't upgrade.
Well, doctors tend to be a conservative bunch for a number of very good reasons. New ideas take time before they get traction, and there is enough experience to make them confident recommending it to their patients. Some doctors are more comfortable recommending experimental techniques...I'm not sure if that makes them better doctors. I would prefer established practices unless there was a compelling reason to try something different (as in the article's case).
As for the OB comment, I just had a baby daughter born yesterday. Throughout the whole process, I never heard anybody from the medical profession refer to mid-wifery as "potentially costing my baby her life." However, from a personal viewpoint, I am much more comfortable having had the delivery in a hospital. There were no complications, but if there had been, an emergency c-section would have been available. The epidural for my wife was also available, which is very good considering the amount of pain she was in after her 12th hour of labor.
Wait, so somebody in your lab did an experiment better than you (has prettier data), and the only reasoning you can come up with is they fudged the data...interesting. The more humble among us might at least acknowledge the possibility that we aren't as good at doing the experiments, especially if we are new to the field and don't have the same level of experience.
Well, after reading the article and number of these rebuttals, I have to say this is an interesting topic to discuss. I think you will find some scientists are more willing to accept the uncertainties in their results (testing for publication bias, etc) than others. I tend to be very skeptical with my results and want to be very sure before publishing and I've met others who are the same, but I've also met scientists who just want to "get the data out there" and let peer-review sort out problems with methodology and such. I think there are two points to the article that the rebuttals are missing because they are focusing on the scientific method question, which most scientists will agree is only as good as the practitioner who follows it.
1) While scientists tend to recognize and acknowledge the weaknesses in their scientific models (ie: they are just models subject to change) and the ever-continuing refinement of our understanding as new knowledge is gained and methodological flaws are identified in past experiments, the non-scientist public generally does not. This extends from the popular science writers, to the science news articles, to the moms who are deciding what is safe for their babies, and eventually to the policy-makers. There is a large impact on the general public that novel scientific advancements have, especially trendy ones. So an effort to educate the general public about some of the inherent uncertainties in science (this is a New Yorker article after all, not a Science paper), is important and useful.
2) The real underlying thread of the article is the amount of time and money spent trying to reproduce bad experiments and chasing results that are statistical anomalies. It's not an argument about the scientific process itself, but one of practicality and an efficient use of research time and funding. If you really think about it, this is something that most scientists should really agree on. If we can better spend our time doing more of the "right experiments" and less tail-chasing, the progress of science will benefit and the public understanding of science will not face as much skepticism (especially when you want to convince a large number of people to make significant lifestyle or societal changes). In other words, we need to be more systematic as scientists. We need to have more discipline to carefully design our experiments and analyze the data. We shouldn't be afraid to publish when we lack certainty, but we should openly acknowledge any doubts we may have about our data and our process (and be honest with ourselves about it too). We shouldn't be wasting the time of the scientific community with bad experiments just because the peer-review process will eventually sort it out. There are better things we can be doing with the limited research funding that is available.
Yeah, that's true with this particular implementation. But I think the concept can be extended. Since the essence of the patch is that it automatically creates process groups based on a criteria (in this case the TTY number), I think it could work something like this: the desktop environment runs a daemon in the background that manages at least two process groups--a background process group and an everything else group. Things like dvd rippers, media encoders, matlab simulations, whatever go in the background process group, and the kernel handles the rest.
The question for scheduling always is "what needs the highest priority?" This patch groups process by tty because its a natural way of separating processes. Obviously the ones that should have the least latency are the ones you are actively using (ie: in the active tty). To do this with a modern gui, the desktop environment has to get involved somehow with the process grouping. Because otherwise there is no way for the kernel to know which processes are being actively used and which ones are in the background (it knows which ones are using the most cpu, but that's about it). Even for the desktop environment it will be tricky, but it has a lot more information to work with than the kernel alone--which windows are on which workspace, which workspace is visible, which windows are on top, which are minimized, when was the last time an Xinput event was in a window, is the application in the window constantly updating the framebuffer, etc....
And Debian updater out-of-the-box is worse than Windows. You have to be root to apply updates,
Well, it depends on your perspective. You may not want it to update without your approval. But there are ways to do unattended updates. The easiest is to probably use the Ubuntu update manager, which I think is also available on Debian.
So, if I want to install the package for "testing" on "stable", it and all of its dependencies that are not already on the system get installed somewhere where they do not affect the rest of the system, even if that means having two copies of some library but named differently.
That's not a bad idea. I wonder if it could be easily implemented....
Once upon a time autopackage made some inroads into the cross-distribution packaging problem. They aren't around anymore. I really think it's just a matter of having a good way to map library names and locations across distributions. It requires the cooperation of all of the distribution maintainers, though, so it will probably not happen any time soon.
It's not really about the app wanting to use a newer version. That's an easy problem, and unless the versions conflict in some horrible way (like libc6), you can easily have multiple versions of a library installed. The problem is when you install from different distributions (testing and stable are effectively separate distributions from stable), the apps are linked against a different set of libraries (it may be the same version, but a different name or whatever). That's why cross-distribution packaging is such a pain. Most distributions use more or less the same library versions, but they name them differently, and version control (the package itself, not the library) them differently, so the binary package isn't cross-compatible, but the source is. The solution so far, for things like Debian stable, is to have a backports repository where newer software versions are compiled against the libraries in stable. Then you just have to upgrade the single package and not everything in your distribution. But yeah, cross-distribution packaging would be nice....
Otherwise we have to go Windows-style and have every software package ship with it's own installer and no knowledge of anything else that may be on the system, no way to centrally update them all, and no way to universally apply security patches to every package that uses a given library (ie: every package has to have its own crappy update utility running in the system tray bogging down everything and irritating you with annoying popups all the time).
For now I can leave it like this (also, when installing Firefox from Mozilla's site, I didn't have to upgrade any package at all) - automatic updates work etc.
Yes, that's because the libraries are statically-linked...it doesn't use any of your system libraries, but includes it's own. There's a huge debate over which approach is better. I personally think dynamic-linking is better, for security, maintenance, and resource-usage reasons. But then Apple takes the opposite stance with their "App-folders" where every application bundles its own libraries. There are never any dependency issues, but at any given time you might have 10 different copies of a font library (or something like that) loaded and running because each application uses its own and none of them share.
And I would have used it if I knew that if I added "unstable" or "testing" repository,
Or use the lenny-backports repository I mentioned above. Incidentally, I don't run Debian anymore, and they've changed the repositories around (you used to have to add a different repository to get things like Flash which you don't have to do anymore). But the first hit when you Google Debian Flash is the Debian Flash wiki, which redirects you to the Debian FlashPlayer package, which tells you where you can get it. And if you do a search on packages.debian.org for Flash, you can see which repositories it is in. You can do that generally for any package you might want to install.
But if I only want the newest version of FF (IW) but leave the rest of the system as is?
I assume if I add the "unstable" repository the updater would see that half of my system is out of date and would update it.
Mmmm...I knew you were going to ask that question. You can install only a single package (with the required dependencies) from the unstable or testing distributions, but this is definitely a more advanced configuration. You have to use apt-pinning. Here is a simple tutorial, http://jaqque.sbih.org/kplug/apt-pinning.html
This is one thing that is nice about Debian. It allows you to do things like this even when it's not recommended. Keep in mind, the larger packages like Firefox and OpenOffice touch the system in a lot of places, so you will probably find it upgrading a number of packages when you install them (things like gnome libraries and glibc versions). You won't get a clean upgrade of just a single package. But it won't upgrade things it doesn't need to. I suggest setting it up to pull newer packages from just the testing repository first. You will be able to get most of the things you want in that distribution. You only need to add unstable if you really want to be on the bleeding edge.
When I installed Debian (Lenny), Iceweasel was lower version (3.0.x IIRC) than FF and didn't have the nice features of Firefox >3.6.x
Yeah, that's why you need to use testing or unstable if you want the latest software. The stable branch is really stable, which necessarily means the software is usually at least one if not a few versions behind the latest. Don't worry, you can use testing or unstable safely. It's just the way Debian does it's versioning...it's not going to break horribly if you use it.
I downloaded flash from adobe's site, that wasn't a problem, the problem was that I had to hunt down the "free" flash plugin because for some reason it was used first if both were installed.
Yes, but if you use the package, you can take advantage of the features the packaging offers, like automatically updating your plugin registry to use it by default instead of the open source version. There's a reason why package managers are so favored in the linux community. You should try to use it (dpkg/apt) as much as possible when installing/removing/updating/configuring software on your machine. It's there because it's good.
Yes, that is certainly a lot more work. Thankfully, it is unnecessary. There was a Firefox rebranding issue with Debian, so they started calling it Iceweasel instead. So try apt-get install iceweasel and see if that works. I'm kind of surprised it didn't pop up a message saying there was a replacement package for firefox. Usually they are pretty good about putting those notes in the repository.
For the flash plugin, you need: apt-get install flashplugin-nonfree
According to the database it is in both unstable and testing. If you are running stable (Lenny), you need to add the backports repository to get it.
Ever tried to install the actual Firefox on Debian?
Yes. apt-get install firefox Debian doesn't have the nice "Ubuntu Software Center" because it isn't a desktop-oriented distribution. But I would say it is easy to install software.
It's harder than just downloading an installer binary and running it.
Maybe. It depends on whether you already know what to do or not. If you are learning it for the first time, then yes, it is probably more difficult, but only because it is different. Also, Mozilla has that big download button on their first Firefox page that automatically detects the version and everything for you. That's nice. But most software distributed for Windows is not that easy.
Can you install Linux software from CD or flash drive?
Yes. How do you think your distribution gets installed in the first place...usually from a CD/DVD.
Will it work with the regular package manager?
Depends. If it's packaged then yes.
Let's say i have a PC without internet connection. I can download the software and record it on a CD on my regular PC then go there and install it, Can I do the same with Linux software that is not part of the install CDs?
Yes.
Linux way of installing software is great until you want to install a program that "the community" does not like for some reason. Maybe the software is closed source. Maybe it uses patented code (even though my country does not recognize software patents and I do not care about them).
I use at least four proprietary (read: closed source) applications on my linux desktop on a regular basis. I had no trouble installing those. They didn't use the package manager, yes, but that's mainly because they would have to tailor it to individual distributions, which is too much of a pain. So they just shipped it with their own binary installer, and it works just fine.
I would definitely like to see better cross-distribution package management. It's a problem a number of projects have tried to address. But it's difficult. And I guess the benefit is not great enough to persevere.
Slashdot Mirror
I agree with your analysis (ie: that criminals are criminals and that they will just move on to the next thing). But just like drug legalization would remove the demand for illegal drugs, opening the borders would remove the demand for human trafficking. Just food for thought.
How exactly do you propose growing the Coca plant in the US? Hint: there's a reason it is trafficked all the way from Columbia.
A prohibition on all exports and imports of drugs would be as effective as a similar measure on produce. While we grow tomatoes here, we also import them from and export them to Mexico. Market supplies are global, and that isn't going to change. Tariffs and prohibitions only create artificial scarcity, which of course leads to black markets and inflated prices.
Uh, no. First of all, raw milk can be legally sold in more than half of the states in the US. Second, in the states where it is illegal, it is not criminal. It is merely prohibited, which means if you are caught you will face civil penalties (fines, suspension of food license, etc).
The money made by the drug cartels is in cocaine and heroine as the GP stated. Other drugs, like pot, ecstasy, and meth are much easier to obtain in the US. With the high risk involved, there is no sense in trafficking it. The profitability is too low. Also, a few states have already decriminalized possession of marijuana. I don't expect all of them to any time soon, but more will eventually. So for the recreational user it won't be hard to obtain.
Well, sure, but what do you propose? Retailers either have to absorb risk, or pay for mandated security. As far as absorbing risk is concerned, they do some. Verification of signatures for transactions is required to show due diligence, for example. But they mostly pass on the cost of fraudulent transactions. In cases where retailers do absorb the entire cost of fraud (ex: checks), they just drop it as a form of payment. So, no, I don't think there is a realistic scenario where small retailers can afford to absorb the costs of fraud because, as you say, profit margins are usually low already. So that leaves paying the (possibly high) cost for good security....
Of course the two features that are truly innovative have been left off the list: Versions and AirDrop. Hmm....
A ton of breaches happen at the small retail level. That's why the bankcard industry is requiring increased security. They aren't lucrative breaches, so you don't hear about it on the news, but they are significant. If an organized group can steal $10 from 10,000 accounts, that's $100,000. It adds up. And if it's a credit transaction, Visa et al are left absorbing the damages. So they have an incentive to tighten security.
Uh, you do realize those "man made rules" are based on observation. Extensive and thorough observation. Basically, the need for oxygen, carbon, nitrogen, hydrogen, sulfur, and phosphorus to sustain life has been demonstrated not just at the organismal level, but also at the molecular level, with predictable and repeating patterns in every life form ever studied. That is the definition of a "Law" in science. Something that is assumed to be a fact because exceptions have never been observed. It, of course, doesn't mean there is no possibility of a life form that "breaks the Law", but the probability is low and such claims are rightfully met with a healthy dose of skepticism. As long as the right controls are done, though, and alternative explanations ruled out, it will eventually be accepted by the scientific community, most likely with a lot of enthusiasm. Findings such as these typically raise more questions than they answer, and a lot of scientists will surely be interested in the underlying enzymology and such, if it holds up to the current scrutiny.
Ok, fair enough, you are right. I see what you are arguing for. You can't prove it is theoretically impossible to do what you are describing. However, I like to think of progress in science, and in particular the success of projects like this, as proceeding in two ways. The first is the revolutionary approach, where completely new and groundbreaking ways of doing things are discovered, or made economical. The second is constrained by practical considerations, and therefore takes a more incremental approach. Both approaches have played a role in the development and progress of science.
Let's take superconductors as an example. Room temperature superconductors are not theoretically impossible, and many would believe that they are very possible. If we succeed in developing them, we could revolutionize the use of techniques like MRI, NMR, and FT-MS. We could make instruments more portable, more energy efficient, more economical. We could envision uses of these devices that aren't possible now because they currently need environmentally-controlled facilities and a full-time trained staff to operate them. Despite all of this, however, we still have a ways to go before we get there. On the other hand, we have had success in making superconductors that can operate at warmer temperatures. So there may be ways to use this to make incremental improvements in these devices (perhaps there already are, I'm not sure).
So in the context of a contest like this, which approach will you choose: the revolutionary approach, or the incremental approach? I guess it depends on your preference, but I think the incremental approach is more likely to yield results in the short term. We currently have the ability to identify pathogens very accurately. And it isn't very expensive, either. It just requires a lab. So the incremental approach would be to try to miniaturize the lab. Like a lab-on-a-chip sort of thing. Something that could be operated anywhere, by minimally-trained staff. I think it's possible and practical, and there are groups already working on things like this. George Whitesides has made a career in investing in ideas like this. And it works. He has been very successful.
So, yeah, the revolutionary approach is interesting and exciting. It shouldn't be abandoned. It absolutely has a place in the advancement of science. Without it, we would never try to achieve new things. But in the interest of practicality and development of an actual product, sometimes it's best to just leverage current technology. To figure out how to make it better, rather than how to replace it with something else.
Just my $0.02. I'll be interested to see how this pans out. It certainly encourages people to have new ideas and to develop them. So let's see what they come up with.
I like your enthusiasm, but I think you are really overestimating the feasibility of your approach. How exactly do you "IR for a long, complex chain molecule", as you replied to another poster (hint, there are millions of them in biological systems)? The point he was making is that all of the common spectroscopy techniques are useless: IR, NMR, and UV/Vis. Even a short pure DNA strand can't be sequenced with IR, NMR, or UV/Vis...that is, unless someone invents a miraculous way to deconvolute the spectrum. It is far more complicated than a dictionary search, useful as those have proven to be in other areas.
There is a reason why we use the techniques we do. But of course that doesn't mean we can't do it better. DNA sequencing has been around since the 70's, but improvements over the last decade have really revolutionized its use in biological research. We can now feasibly talk about sequencing the genome of an organism as a routine thing, rather than a years long and millions of dollars effort. And single-molecule sequencing is also starting to have some seriously promising results. So maybe the right thing to do is figure out how to efficiently retrieve and prepare samples from patients, so that sequencing for pathogenic DNA can be done quickly and cheaply. It's not quite like a tricorder, but it has the potential to be automated in the way the contest seems to want.
Honestly, I'm not sure how helpful something like this is, though. We can already do diagnostic screening very accurately and quickly. Making it not require a lab would certainly be a nice development. But, as another poster already replied, screening for pathogens is a very small part of medicine these days, even in developing countries. Things like x-ray scanners and MRIs are becoming a lot cheaper and more portable, though. For MRI what they really need is a breakthrough in the magnet technology, but in the meantime there are companies specializing in things like small hand MRIs, which can have some useful applications. One "diagnosis machine" is still very much in sci-fi land, though.
Well, this is an interesting idea, but the major cost being left out of this calculation is the operating cost. No nuclear reactor will be licensed to operate in the US without a full staff complement. At a minimum, this usually entails two full-time licensed nuclear reactor operators, security, and auditing and regular safety review.
Don't be fooled. This is a power plant, even if it is a small one. It's not going to be economical for a random business. The website says it would work for things like large government and hospital complexes. In these cases it might be conceivable to hire staff to run an in-house power generation facility, but I doubt it would work out for an office building downtown.
True or not, the perception is that after a certain point ... that highly specialized degree doesn't really translate into any knowledge that anybody needs,
I really don't buy this argument at all. My only experience is with the life sciences. Maybe some of the other fields are different. But in the sciences your research has to be funded by a grant. The granting process is extremely competitive. You're not going to get funding without convincing others that your research is significant, useful, and that you are productive. I don't know where people get the idea that the research is so highly specialized it is useless. Highly specialized, maybe, but not useless. Unless you think basic research related to cancer and infectious disease is useless.
Then again, maybe the author is his own example. He is so specialized in his own field that he doesn't know what happens in the academic world outside of it.
Why should people have to change at Ubuntu's pace to continue to get their work done?
Um, they don't. You don't have to upgrade to the latest version. I'm still running 9.10 on one of my desktops because I needed it to be a stable machine to write my thesis on. I didn't have time to upgrade and I didn't want to spend a lot of time doing backups, repartitioning, etc. So I just left it, and it still works great. Seriously, if you don't want/need the new stuff, don't upgrade.
Well, doctors tend to be a conservative bunch for a number of very good reasons. New ideas take time before they get traction, and there is enough experience to make them confident recommending it to their patients. Some doctors are more comfortable recommending experimental techniques...I'm not sure if that makes them better doctors. I would prefer established practices unless there was a compelling reason to try something different (as in the article's case).
As for the OB comment, I just had a baby daughter born yesterday. Throughout the whole process, I never heard anybody from the medical profession refer to mid-wifery as "potentially costing my baby her life." However, from a personal viewpoint, I am much more comfortable having had the delivery in a hospital. There were no complications, but if there had been, an emergency c-section would have been available. The epidural for my wife was also available, which is very good considering the amount of pain she was in after her 12th hour of labor.
Wait, so somebody in your lab did an experiment better than you (has prettier data), and the only reasoning you can come up with is they fudged the data...interesting. The more humble among us might at least acknowledge the possibility that we aren't as good at doing the experiments, especially if we are new to the field and don't have the same level of experience.
Well, after reading the article and number of these rebuttals, I have to say this is an interesting topic to discuss. I think you will find some scientists are more willing to accept the uncertainties in their results (testing for publication bias, etc) than others. I tend to be very skeptical with my results and want to be very sure before publishing and I've met others who are the same, but I've also met scientists who just want to "get the data out there" and let peer-review sort out problems with methodology and such. I think there are two points to the article that the rebuttals are missing because they are focusing on the scientific method question, which most scientists will agree is only as good as the practitioner who follows it.
1) While scientists tend to recognize and acknowledge the weaknesses in their scientific models (ie: they are just models subject to change) and the ever-continuing refinement of our understanding as new knowledge is gained and methodological flaws are identified in past experiments, the non-scientist public generally does not. This extends from the popular science writers, to the science news articles, to the moms who are deciding what is safe for their babies, and eventually to the policy-makers. There is a large impact on the general public that novel scientific advancements have, especially trendy ones. So an effort to educate the general public about some of the inherent uncertainties in science (this is a New Yorker article after all, not a Science paper), is important and useful.
2) The real underlying thread of the article is the amount of time and money spent trying to reproduce bad experiments and chasing results that are statistical anomalies. It's not an argument about the scientific process itself, but one of practicality and an efficient use of research time and funding. If you really think about it, this is something that most scientists should really agree on. If we can better spend our time doing more of the "right experiments" and less tail-chasing, the progress of science will benefit and the public understanding of science will not face as much skepticism (especially when you want to convince a large number of people to make significant lifestyle or societal changes). In other words, we need to be more systematic as scientists. We need to have more discipline to carefully design our experiments and analyze the data. We shouldn't be afraid to publish when we lack certainty, but we should openly acknowledge any doubts we may have about our data and our process (and be honest with ourselves about it too). We shouldn't be wasting the time of the scientific community with bad experiments just because the peer-review process will eventually sort it out. There are better things we can be doing with the limited research funding that is available.
Yeah, that's true with this particular implementation. But I think the concept can be extended. Since the essence of the patch is that it automatically creates process groups based on a criteria (in this case the TTY number), I think it could work something like this: the desktop environment runs a daemon in the background that manages at least two process groups--a background process group and an everything else group. Things like dvd rippers, media encoders, matlab simulations, whatever go in the background process group, and the kernel handles the rest.
The question for scheduling always is "what needs the highest priority?" This patch groups process by tty because its a natural way of separating processes. Obviously the ones that should have the least latency are the ones you are actively using (ie: in the active tty). To do this with a modern gui, the desktop environment has to get involved somehow with the process grouping. Because otherwise there is no way for the kernel to know which processes are being actively used and which ones are in the background (it knows which ones are using the most cpu, but that's about it). Even for the desktop environment it will be tricky, but it has a lot more information to work with than the kernel alone--which windows are on which workspace, which workspace is visible, which windows are on top, which are minimized, when was the last time an Xinput event was in a window, is the application in the window constantly updating the framebuffer, etc....
Just found this,
http://listaller.nlinux.org/
And Debian updater out-of-the-box is worse than Windows. You have to be root to apply updates,
Well, it depends on your perspective. You may not want it to update without your approval. But there are ways to do unattended updates. The easiest is to probably use the Ubuntu update manager, which I think is also available on Debian.
So, if I want to install the package for "testing" on "stable", it and all of its dependencies that are not already on the system get installed somewhere where they do not affect the rest of the system, even if that means having two copies of some library but named differently.
That's not a bad idea. I wonder if it could be easily implemented....
Once upon a time autopackage made some inroads into the cross-distribution packaging problem. They aren't around anymore. I really think it's just a matter of having a good way to map library names and locations across distributions. It requires the cooperation of all of the distribution maintainers, though, so it will probably not happen any time soon.
Just checked...the newer Firefox (err, Iceweasel) is in lenny-backports. So there's another way to get the latest version without apt-pinning.
It's not really about the app wanting to use a newer version. That's an easy problem, and unless the versions conflict in some horrible way (like libc6), you can easily have multiple versions of a library installed. The problem is when you install from different distributions (testing and stable are effectively separate distributions from stable), the apps are linked against a different set of libraries (it may be the same version, but a different name or whatever). That's why cross-distribution packaging is such a pain. Most distributions use more or less the same library versions, but they name them differently, and version control (the package itself, not the library) them differently, so the binary package isn't cross-compatible, but the source is. The solution so far, for things like Debian stable, is to have a backports repository where newer software versions are compiled against the libraries in stable. Then you just have to upgrade the single package and not everything in your distribution. But yeah, cross-distribution packaging would be nice....
Otherwise we have to go Windows-style and have every software package ship with it's own installer and no knowledge of anything else that may be on the system, no way to centrally update them all, and no way to universally apply security patches to every package that uses a given library (ie: every package has to have its own crappy update utility running in the system tray bogging down everything and irritating you with annoying popups all the time).
For now I can leave it like this (also, when installing Firefox from Mozilla's site, I didn't have to upgrade any package at all) - automatic updates work etc.
Yes, that's because the libraries are statically-linked...it doesn't use any of your system libraries, but includes it's own. There's a huge debate over which approach is better. I personally think dynamic-linking is better, for security, maintenance, and resource-usage reasons. But then Apple takes the opposite stance with their "App-folders" where every application bundles its own libraries. There are never any dependency issues, but at any given time you might have 10 different copies of a font library (or something like that) loaded and running because each application uses its own and none of them share.
And I would have used it if I knew that if I added "unstable" or "testing" repository,
Or use the lenny-backports repository I mentioned above. Incidentally, I don't run Debian anymore, and they've changed the repositories around (you used to have to add a different repository to get things like Flash which you don't have to do anymore). But the first hit when you Google Debian Flash is the Debian Flash wiki, which redirects you to the Debian FlashPlayer package, which tells you where you can get it. And if you do a search on packages.debian.org for Flash, you can see which repositories it is in. You can do that generally for any package you might want to install.
But if I only want the newest version of FF (IW) but leave the rest of the system as is?
I assume if I add the "unstable" repository the updater would see that half of my system is out of date and would update it.
Mmmm...I knew you were going to ask that question. You can install only a single package (with the required dependencies) from the unstable or testing distributions, but this is definitely a more advanced configuration. You have to use apt-pinning. Here is a simple tutorial,
http://jaqque.sbih.org/kplug/apt-pinning.html
This is one thing that is nice about Debian. It allows you to do things like this even when it's not recommended. Keep in mind, the larger packages like Firefox and OpenOffice touch the system in a lot of places, so you will probably find it upgrading a number of packages when you install them (things like gnome libraries and glibc versions). You won't get a clean upgrade of just a single package. But it won't upgrade things it doesn't need to. I suggest setting it up to pull newer packages from just the testing repository first. You will be able to get most of the things you want in that distribution. You only need to add unstable if you really want to be on the bleeding edge.
When I installed Debian (Lenny), Iceweasel was lower version (3.0.x IIRC) than FF and didn't have the nice features of Firefox >3.6.x
Yeah, that's why you need to use testing or unstable if you want the latest software. The stable branch is really stable, which necessarily means the software is usually at least one if not a few versions behind the latest. Don't worry, you can use testing or unstable safely. It's just the way Debian does it's versioning...it's not going to break horribly if you use it.
I downloaded flash from adobe's site, that wasn't a problem, the problem was that I had to hunt down the "free" flash plugin because for some reason it was used first if both were installed.
Yes, but if you use the package, you can take advantage of the features the packaging offers, like automatically updating your plugin registry to use it by default instead of the open source version. There's a reason why package managers are so favored in the linux community. You should try to use it (dpkg/apt) as much as possible when installing/removing/updating/configuring software on your machine. It's there because it's good.
Yes, that is certainly a lot more work. Thankfully, it is unnecessary. There was a Firefox rebranding issue with Debian, so they started calling it Iceweasel instead. So try apt-get install iceweasel and see if that works. I'm kind of surprised it didn't pop up a message saying there was a replacement package for firefox. Usually they are pretty good about putting those notes in the repository.
For the flash plugin, you need:
apt-get install flashplugin-nonfree
According to the database it is in both unstable and testing. If you are running stable (Lenny), you need to add the backports repository to get it.
Ever tried to install the actual Firefox on Debian?
Yes. apt-get install firefox
Debian doesn't have the nice "Ubuntu Software Center" because it isn't a desktop-oriented distribution. But I would say it is easy to install software.
It's harder than just downloading an installer binary and running it.
Maybe. It depends on whether you already know what to do or not. If you are learning it for the first time, then yes, it is probably more difficult, but only because it is different. Also, Mozilla has that big download button on their first Firefox page that automatically detects the version and everything for you. That's nice. But most software distributed for Windows is not that easy.
Can you install Linux software from CD or flash drive?
Yes. How do you think your distribution gets installed in the first place...usually from a CD/DVD.
Will it work with the regular package manager?
Depends. If it's packaged then yes.
Let's say i have a PC without internet connection. I can download the software and record it on a CD on my regular PC then go there and install it, Can I do the same with Linux software that is not part of the install CDs?
Yes.
Linux way of installing software is great until you want to install a program that "the community" does not like for some reason. Maybe the software is closed source. Maybe it uses patented code (even though my country does not recognize software patents and I do not care about them).
I use at least four proprietary (read: closed source) applications on my linux desktop on a regular basis. I had no trouble installing those. They didn't use the package manager, yes, but that's mainly because they would have to tailor it to individual distributions, which is too much of a pain. So they just shipped it with their own binary installer, and it works just fine.
I would definitely like to see better cross-distribution package management. It's a problem a number of projects have tried to address. But it's difficult. And I guess the benefit is not great enough to persevere.