Indeed, Gun Crime is much, much worse in those countries where guns are banned.
I wish people would stop repeating the fundamental misstatement that is "gun crime". Reducing "gun crime" does not indicate an actual improvement in our lives. Reducing violent crime and murder does. "Gun crime" is fallaciously subdividing the problem of violent crime in order to try to make it look like statistics show an improvement when, in fact, the problem is unchanged or worsened. The term originates from politicians trying to make it seem like their measures have been effective after analysis of the data showed no benefit.
Let me provide a rather extreme hypothetical case to explain:
Case A, a man goes into a home with an axe and attacks a family. A family member grabs a gun, shoots at him and misses, then chases him out into the street shooting at him and eventually killing him.
Case B, laws have gone into effect that prevent gun ownership. A man goes into a house and kills six unarmed people with an axe.
So, let's analyze the different cases with regard to violent crime. Violent crime is higher in B than in A because instead of one manslaughter, we have six homicides. Now Lets examine the same two cases with regard to "gun crime". "gun crime" is lower in B than A because one manslaughter did not happen. This misguided analysis supports the theory that we should promote laws that encourage case B instead of A.
Any analytical subdivision that can take the above hypothetical situation and show case B as a net positive for society over case A, is misstatement of the problem. Please people, ignore fallacious "gun crime" studies and please look at real numbers on murder and violent crime. Note, I'm not advocating for or against any particular gun control law here. I'm just asking that people please be careful when you look at the data. From what I've seen, there hasn't been any convincing scientific evidence that gun control laws have much correlation (let alone causation) with either reductions or increases in violent crime or murder, when normalized for other, known factors.
Fog of war? I mean, had there been opposing demonstrators you might have had a full blown fire fight.
I've seen this argument before, but it sure doesn't hold up when you look at the numbers. Armed citizens intervening in conflicts shoot an innocent bystander in the confusion much less often as a percentage of incidents as police responding to a shooting. Frankly, a nonprofessional that is there when things happen is a better judge of where to aim their firearm than a policeman coming into a situation with minimal information.
"Likely any alien would be so alien communication would be an even bigger issue than with dolphins.)"
If Dolphins had developed technology then it would be likely they would understand the language of maths.
Roman numerals? Hexadecimal? Math as abstracted visual images like some autistic savants? Math that is understood in a way that it is still completely incomprehensible to us? I don't understand why because they understand how to solve problems using it, means they can communicate it to us or us to them. Nor does it alleviate the problem of their intelligences being so alien that their motives and actions have no meaning or pattern that we can discern.
It's as though we expect the kids to grok the process from a demonstration of the results (generally pointing out the trivia about it) and a few cartoony diagrams.
Potentially because those teachers sometimes have to teach those uncomfortable facts and shouldn't have to worry about being fired. I'm not saying I'm in favor of tenure for most schools (I haven't studied the issue or really formed an opinion). I'm just saying there may well be a similar reason. I know several parents that would push to have teachers fired for teaching such "controversial" topics as evolution without disclaimers about how it isn't really true.
I support limited rights for animals, in parallel with the level of legal responsibility that they have. Already many animals are granted a few very limited rights, like the right to not be tortured (even if it is legal to kill them). This goes along with their legal responsibilities as in they are not held accountable for their actions like theft or even murder, which is the responsibility of the owner (if there is one). The courts might order an animal put down, but only as a protection for the community, not as a punishment.
So, what level of rights should dolphins be granted? What level or responsibility? Should we make it illegal to kill them? Should we convict them of murder if they kill another person or dolphin? Should it be illegal to confine them? Should they be held responsible if they steal fish from a net?
I suspect much of the problem is one of communication. Dolphin are simply so alien to humans that we may not think similarly enough to communicate richly enough to make sense of this sort of ethical issue. (It also shatters all those awesome sci-fi fantasies about meeting cool alien species who are similar to us, but different, but we communicate and get along. Likely any alien would be so alien communication would be an even bigger issue than with dolphins.)
You can make that argument because math is provable, but it is a semantic argument about the definition of 'fact". (I might note that I recall being taught to memorize the answers to multiplication problems.) Science, however, is a process and not a fact. It is not provable nor indisputably correct but rather is logically supported by inductive evidence derived from deductive experimentation. Either way, making people memorize the definition of the scientific method doesn't really get us anywhere either. We need to teach an understanding of the application and appreciation for the results.
It's as though we expect the kids to grok the process from a demonstration of the results (generally pointing out the trivia about it) and a few cartoony diagrams.
Rather, I suspect at this point most of the instructors don't grok the process either and so teach "the materials" as a sort of memorization exercise just like every other subject taught these days. I doubt instructors expect students to be able to understand and apply the scientific method because most of the instructors just aren't very bright themselves and don't understand it themselves, or the importance of understanding logic and science in everyday life, or in many cases that there is anything to understand beyond memorizing a phrase that describes "science".
The only electronics I need in my car is an easy to use, powerful interface to my smart phone. It has everything I need to play music, check email, message, surf the web, navigate, etc.
I'd settle for a law requiring all car stereos to have a $0.30 headphone in jack to cut the absurd waste of resources and money that goes into adapters, new stereos, and FM transmitters in order to get the audio output all of a foot and communicating with the stereo.
No, first time you run the mac app store, applications installed via other means (installer,copy,whatever) are detected and marked as already installed.
Marked installed? What does that mean? How is Apple identifying apps they've never even seen?
There are reports that applications are detected even when copied to other volumes.
What reports and how are people claiming to know this? Citation?
Until there is evidence to the contrary or apple explicitly denies it, assume that your apple id is associated with whatever you had in your hard drive.
Now that would be quite a trick since the App Store app doesn't autodetect your AppleID from iTunes and asks you to enter one. So how are they associating the ID with the apps on the first run, when the application doing the detecting does not yet know the ID?
All I've seen so far is some fairly spurious guessing on your part. Perhaps you're confused because the app does look for some Apple created apps for purposes of keeping those updated (since Apple ditched the other update mechinsms for them). I suppose we can test your hypothesis with a couple of Macs to see how the data transferred changes (in volume if nothing else) between a Mac with a lot of apps and one with none, but rather than make an assumption either way, maybe it's better to gather some actual facts?
It points out the real problem with science education: we're not teaching the big facts and then delving into the intricacies, we're teaching the intricacies and hoping the big facts are obvious.
I disagree. The problem is we're concentrating on teaching facts instead of methods. "2+2=4" is a fact. Knowing how to add is a method. It's the process of forming opinions that is broken, because science students aren't learning to reason logically and follow a logical methodology when forming opinions. Another problem is our focus on having an answer even when we don't have data. We punish students for not guessing, which trains them to guess instead of honestly answer "I don't know" then follow a proper method to find out. It should be acceptable in an exam to answer, "I don't know but I'll find out with some reference materials or experiments". Worse yet, we train students to not only guess, but then to defend their guesses using rhetoric. It's not about coming to correct decisions, but about convincing others that you're right.
So here's the thing we take what six years of science classes at a minimum? And then many go on to college and get a degree. Yet I found myself last year at the bar talking to a guy with a Bachelors of Science degree and he couldn't even describe to me what the scientific method was, let alone how one would apply it to solve a simple problem. That is what the focus on memorizing facts has gotten us. Doubtless many of his peers could still dredge up a definition of "science" from memory, but how many could apply it? How many do apply it every day to make decisions rationally?
Screw our educational system. Teach kids math and informal logic and critical thinking and proper research skills. Screw standardized closed book tests. Lets reduce the number of tests kids take to prove how much they know and how good their teachers are and give a few hands on tests a year where they have to actually demonstrate a real understanding of how to do something or how something works or how and why something happened.
In a lot of other countries with a much better education system, teachers are recruited from the top of the graduating classes and are given incentives to go and teach.
Which countries are these?
Finland is a good example and among the best in the world. Their teachers are one of their highest paid professions and are government jobs. Less than 3% of grade school students go to private school. All education is free right up through doctoral degrees and includes free meals and healthcare.
Many other countries are starting to use them as a model for changes to their own educational system, including China who has been taking great interest in reforming their education along those lines. I don't expect it to ever happen in the US though, we're a bit too isolationist and the majority opinion is that we're "better" at education and many other things despite objective evidence to the contrary.
Since when was taking advantage of gaping exploits in software not hacking...
Since when is not implementing strict DRM an exploit? Quick OS X has a huge exploit and doesn't check for a valid serial number! Quick OpenOffice has a huge exploit, you can copy it without paying anyone!
The level of DRM a developer wants to implement is up to them. If they decide not to check or to check only for any valid account, that's up to them. They might make such a decision because they want to get to market faster and don't want to code and test it or because they actually don't mind people copying as they think that will promote more sales in the long run. Calling it an "exploit" seems a bit hyperbolic.
Apple simply wants a revenue stream where people can easily purchase and install licensed versions of software.
Like iTunes and the iPhone App Store, I suspect this is about selling hardware. Taking a 30% cut of app sales while providing the hosting and the credit card processing and while taking on the burden of hosting the lion's share of all the freeware in existence is unlikely to be a significant money maker. It certainly has not been on the IPhone. Rather, this is a way to make more people think Macs are easy to use by making getting apps easier, reducing crashes, and slightly mitigating security risks. The store is about selling hardware, just like their other stores.
Seriously, the whole story is that some apps aren't checking to see if the Mac in question has a receipt for that app. Most apps on OS X don't bother checking registration now. Heck, OS X doesn't even check to see if the user has a valid key. First, how is this news? Second, why the hell is apps not using DRM being spun on Slashdot as a BAD thing? Seriously, when did Slashdot become pro-DRM? Oh no apps are freely copyable and users can share them without DRM getting in the way, if the app developer made them that way! Seriously people?
Look I'm not going to go through all this again. You seem to be replying to only parts of my post, parts where I don't address the direct failures of the current system on most Linux distros. So lets simplify. For use cases where you need to move an application between two systems, non-bundled apps are harder to move. You must either re-download a copy from somewhere on the internet if it is available or you must gather up the binary and the resources, possibly zipped then extracted or whatever, and then copy them over into the correct location and make the binary executable. The latter is much, much more cumbersome and beyond the capabilities of many normal users.
For use cases where you're running an app remotely, OS X makes use of locally stored three tiered configuration files paired with an optional config file in the bundle. This means you can run an app stored on a flash or remote drive from computer A, and it will inherit the configuration changes made for that computer (A), group, and user as well as global changes for the app if the developer wants that. When you run the same app from the same location using computer B it will inherit the configuration changes for computer B, group, user, and app. These are persistent so replacing an app with a newer version or using apps where you have five floating copies on flash drives "just works" exactly as the user who has no idea what's happening in the background would expect.
To replicate the above using Linux, especially when Linux is running on computers with different architectures takes a great deal of work by a sys admin setting up some cool symlinks. It is way, way, way beyond the understanding of a normal user, brittle, and complex.
Lets be clear. Linux does not work "just fine" for these same cases. Linux falls down because it does not support bundled apps with proper hooks in the OS for FAT binaries and storing of preferences. Further, there is no standard for bundled apps so automatically mitigating the disk cost of FAT binaries becomes very difficult for systems with limited space and extraction of resources becomes a process of hunting around and hoping.
Now as to the opposite situation. Linux still has better more capable package managers that can integrate easily with multiple repositories and it is a more widely used mechanism on Linux. As a result a smaller subset of OS X software gains the benefits of centralized auto updating when compared to OS X.
In both cases the issue is solvable and the OS can be extended to handle the deficiency just as the other OS does. The difference is, I doubt Apple will add more repositories because they don't see it as a beneficial business case. I don't see application bundles being properly supported in Linux, because most Linux developers focus on the server and don't care and those that focus on the desktop are blindly resistant to admitting there could be any way in which Linux is not perfect and thus won't even consider improvements that already exist in OS X. (As a user of both I have to say at least Apple has been slowly copying the cool bits of Linux, I've seen very little of the cool bits of OS X pulled into Linux.)
Now I think I've clearly stated specific use cases where the same task is harder or just doesn't work when using Linux. If you want to discuss those specific cases, please do, but lets not go off onto weird tangents about edge cases.
P.S. with regard to your question about memory usage and libraries; I don't want to spend a lot of time on it because as I said, memory is a minor part of the library issue. Rather faster propagation of fixes into libraries used by apps is the main benefit. If you really want to understand the issue you need to read up on both OS X's dynamic linking scheme and their pre-binding for libraries not shipped with OS X itself.
you kind of ignored the last two paragraphs of his reply.
Yup, because they were empty of coherent content.
But just for you:
Unix in general can do that too.
Can do what? What are you talking about? Do you even know what you're replying to?
That's probably where the original idea for this came from in NextStep to begin with.
The original idea for application bundles in NextStep came from what? Unix? Umm, NextStep was Unix. Again, WTF are you talking about? The idea for bundles came about to simplify and compartmentalize applications to make things cleaner, easier and better architected to solve the problems traditional Unix had with messy applications leaving bits everywhere and creating dependency hell.
Linux apps were using this sort of approach when MacOS was still in it's original form...
Gee, don't specify what approach you're talking about or anything. How does one graduate high school writing class without being able to compose a specific sentence?
In short, I suspect the previous poster did not even read the thread and was just writing some sort of nonsense in the hope of a reply or is really, really drunk and trying to defend Linux against some imagined threat or something. Really, it's incoherent and makes no specific mention of ANYTHING. I mean he fucking quoted my personal observation about a reply, not any argument made in the reply.
Actually, Linux relies upon dependency resolution at install time. OS X uses self contained packages with a dynamic linking scheme. That's the difference I was bringing up and what enables OS X to have more easily portable applications and better ability to use remote software.
Again, no, you seem to misunderstand what linux does and does not do. Both systems work in both of the ways you have described. See e.g. MATLAB for linux (no install time dependency resolution), or Fink/Macports which does install-time dependency resolution on OSX.
I think you're being intentionally obtuse. Yes, the Macports system uses the same mechanism as Linux, but that's an edge case. The vast majority of applications installed on OS X are installed as openstep bundles and that is quite different. As for Matlab, it does not, however, gain the benefits of a portable application bundle described in my previous post. Sure you can bring up edge cases like Linux distros based upon GNUStep, but those are rarities where they have addressed this deficiency. That is not the case in mainstream distros.
On OS X the executable(s) and resources are in the same directory along with the libraries that aren't standard on the OS.
That's exactly the same way that 3rd party self-contined rather than package-managed software works on Linux. And the standard Linux way is exactly the same way that third-party package-managed software works on OSX (e.g. Fink).
No, it is not. For the most part third party software on Linux is distributed as a binary installer, which then writes out binaries in/bin and resources elsewhere. While some Linux distros save binary installers so those can be used to reinstall elsewhere, that is less common and you still can't just grab the application you're running and let the system worry about moving it. Further, for remote applications it does not do a good job about handling multiple configuration setups unless you're using a mess of symlinks or a file server protocol for that specific purpose.
As for unable to share libraries, that's not true. They do share libraries dynamically linking to the most up to date within the stable line. You can literally install a singed package and your other apps will upgrade or fall back to their own copy as needed because multiple copies are stored (one per app that uses it).
Are you claiming that if two different.apps have the same.dylib buried in their directory somewhere, then when the two apps are running, only one copy of the.dylib will reside in RAM? If so, then [citation needed] because I've never heard of that happening before.
That is not exactly what I was talking about. Two different apps have different versions of the same library, but the prebinding will bind to the most up to date, compatible one. This does not translate directly to RAM savings it can result in cache savings. The main advantage, however, are security and stability and bug fixes being more rapidly propagated.
Of course the package manager doesn't manage non-packages. Much like the.app method doesn't help executables that aren't.apps. For non managed packages the install process is usually a case or running the installer executable, which is I will grant more awkward than using a.app on OSX (though plenty of OSX programs also seem to require installing, too). But not much, given that the majority of installed software is done through the package management system.
The difference is, most users on Linux, whether home users installing games or pro users installing dev tools do end up installing applications utside of te package manager. Very, very few users ever install software on OS X that is not a.app bundle.
How does an uninformative reply like this, that has no useful information or even argument get modded up? All you have is an opinion that Macs are not a "system" and everything is thrown together? Do you have any idea what the hell we're talking abut?
PC-BSD (based on FreeBSD) is doing this with their PBI system right now:
Interesting, although it looks like it is still placing multiple parts of the app in different locations and both must be copied in order to transfer an app (instead of a single package). Also, I'm not sure I see how this addresses running apps off a flash drive when they need to be run by multiple platforms and OS versions. Assuming you installed an app on the Flash drive, wouldn't you still need to create a local symlink on each machine so it could find the resources?
In general most software on Linux systems comes from the distro's repositories, so having a decent GUI is generally sufficient to allow for easy installation.
Look, I'm a Linux user as well as an OS X one. Yes, most of the software is in the repositories... but by no means all of it. I still end up having to manually move or re-install a lot of software because a LOT of software I need or want is just not there, especially commercial packages.
It also gives you a central location for updates, whereas on Mac OS X each app--along with the OS--has its own update mechanism.
It only gives you a central location for updates for software provided through it, not for all software. And as of today, that's basically the same situation on OS X.
Similarly if you add a third-party repository...
And here is where other repository systems are still ahead of OS X. That said, commercial vendors find setting up their own repositories too cumbersome compared to Web downloads and hosted repositories take distribution out of their own control too much. So far, only Ubuntu has really addressed the need for payware software developers to distribute through the central repository.
For the standard application distribution method they provide, it's largely unnecessary. They really do provide a mechanism that's simple and elegant and self-contained.
I actually disagree. For updates and app discovery, Linux is still ahead of OS X, likewise for the ability to handle multiple repositories. That said, Apple took a big step in the right direction today. Apple's application format, on the other hand, still provides significant benefits not yet matched by Linux formats. There is no reason for the two to be mutually exclusive.
Can an Apple user send a current intel-only app to someone running OSX on PPC or iOS on a mobile device -- no!
If the app was built as Intel only, no they can't run it on both, but when there are multiple architectures in play, both running OS X, yes it can. We're talking about the capabilities of the OS application package structure. I can still access apps on a flash drive and run them from my legacy PPC machine, then plug them into a brand new Intel machine and run the very same app.
So picture this, ARM takes off and becomes a viable OS X platform going forward. So app developers add that checkbox when compiling apps and suddenly, with no effort on their part, all the apps being produced become portable between both types of machines. Meanwhile Linux users have to download separate packages for both systems.
Then we get to the question of why anybody would ever want to copy a raw binary instead of using a linux distros package manager?
Your question is flawed. Everyone has moved to using package managers now. The question becomes, what advantages of OS X's package format still brings benefits that current Linux package managers cannot replicate. One is migration. When I move to a new laptop, I suck all my apps across rather than re-downloading and reconfiguring them a process that takes considerable time and bandwidth. If my new laptop is a different architecture than my last, with Linux I have to re-download them all (or would if I wasn't already running my Linux desktop in a VM). And what about commercial apps? You do know not all apps are in repositories. I regularly have to download apps from Web pages, whether they are server components like zencart or commercial games. (My compliments to Ubuntu for starting to get some commercial apps in their repositories.) In my example before about IM'ing an app, it was actually a real world example. The app was no longer being offered by the developer and was nowhere to be found except on my machine. Now with some work and know how I could have replicated that feat on Linux on the same architecture it would not have been easy and it would have failed when it went between architectures.
Fat binaries were a temporary workaround and both Apple and MS include emulators.
FAT binaries are a forward thinking mechanism that promotes usability in a world where we are exposed to different architectures over the years. There's nothing temporary about the concept unless you think we've settled on the "one true architecture" now and forever.
Can I run a linux ARM binary via a distro under QEMU from another arch -- yes.
Yes... but not easily and not unless your emulator and OS supports all the architectures. And you take a serious performance and stability hit doing it instead of using a native binary.
Weird I thought you were going to argue the point too, but then you didn't address any of the real and practical application differences I pointed out. That's not much of an argument.
OK, you clearly do NOT understand how Linux works, or ironically OSX. The underlying mechanisms are then same in both operating systems that we may as well consider them the same for the purposes of discussion.
Actually, Linux relies upon dependency resolution at install time. OS X uses self contained packages with a dynamic linking scheme. That's the difference I was bringing up and what enables OS X to have more easily portable applications and better ability to use remote software.
Linux and OSX implement self-contained programs in exactly the same way. They ship around and archive containing the executable, the requires shared libraries and resources.
On OS X the executable(s) and resources are in the same directory along with the libraries that aren't standard on the OS. Further, they're in predetermined locations so it's easy to grab resources like music or images.
In your preferred system, every package would have to have all of its own libraries. Not only would that vastly increase the amount of disk space used, it would (more importantly) vastly increase the RAM usage since different executables would be unable to share the same libraries.
It does increase disk space usage, but not significantly as resources are actually the big part of apps these days. If you get really crunched on space you can prune binaries and unused libraries at the expense of losing portability going forward. As for unable to share libraries, that's not true. They do share libraries dynamically linking to the most up to date within the stable line. You can literally install a singed package and your other apps will upgrade or fall back to their own copy as needed because multiple copies are stored (one per app that uses it).
I like the package management based solution. It basically works like magic with absoloutely no effort.
It doesn't work as well, especially for apps installed not using the package manger (as a Linux user I'm sure you have to deal with these as well) and it falls down in the several, specific use cases I mentioned in my last post (and which you did not address).
You know you can't just copy the programs in/sw/bin to another mac, right?
You do know that basically no applications get stored in/sw/bin right? That's mostly for bad ports and legacy software. Even OpenOffice installs as a.app these days and it can be stored anywhere the user likes.
If "developers will hand over 30 percent of the purchase price to Apple," what will consumer prices be?
Have you ever worked in the end user software development business? 30% going to distribution, credit card processing, and managing updates isn't bad. When you add in the amount of publicity it generates by being in THE searchable software database for end users, well, likely prices will drop as advertising will drive more sales, more price competition, and larger volumes.
Really? Is that why I can move my home directory from one linux install to another and the programs will still run?
Please don't even argue this point. Linux is a bit behind the curve and the only people who would argue otherwise are people who don't use both OS's. Sure you can copy your home directory on Linux, or use the stored installer (if you are expert enough to know where they go) for an individual app (on some distros)... all provided you are running on the same architecture.
With OS X you can literally drag an application into a chat window to a friend, who is running a different version of your OS, running on a different chipset and that friend can double click the app and run it. It's a great deal more painless since all the apps are the installers and are self contained directories ending in.app. It's one of the things Apple got right and where no Linux distro has enough pull to push change, especially since it is not a big pain point for end users. Additionally, the OpenStep packages make running software off a network drive or flash drive or anywhere really, easier by allowing for multiple sets of preferences and multiple included binaries to get around the whole hack of symlinks or multiple copies for multiple architectures.
Linux is not ahead in every area, just as OS X and Windows are behind in other areas. Get over it.
Apple "innovates" again and re-invents the package manager Linux has had for ages...
Yeah, pretty much. I wish it had not taken them so long though. I wish they'd do a better job copying virtual desktops while they're at it. I notice Canonical innovated and added apps for sale to their own app repositories. Now I wish Linux distros would innovate and re-invent GNU-step style packages and required package signing with real credentials to get into the default repositories, and heck system services while they're at it.
I really, really like it when OS's copy the best parts of other OS's and my daily computing experience is made easier. I don't really care that someone else came up with something first.
TFA makes it sound a lot like apt or the package manager for any Linux, except here you have to pay for some of the apps, and without a community repo.
What do you mean you don't have to pay for any of the apps on Linux? Ubuntu has included paid apps in the USC for some time now.
Slashdot Mirror
Indeed, Gun Crime is much, much worse in those countries where guns are banned.
I wish people would stop repeating the fundamental misstatement that is "gun crime". Reducing "gun crime" does not indicate an actual improvement in our lives. Reducing violent crime and murder does. "Gun crime" is fallaciously subdividing the problem of violent crime in order to try to make it look like statistics show an improvement when, in fact, the problem is unchanged or worsened. The term originates from politicians trying to make it seem like their measures have been effective after analysis of the data showed no benefit.
Let me provide a rather extreme hypothetical case to explain:
Case A, a man goes into a home with an axe and attacks a family. A family member grabs a gun, shoots at him and misses, then chases him out into the street shooting at him and eventually killing him.
Case B, laws have gone into effect that prevent gun ownership. A man goes into a house and kills six unarmed people with an axe.
So, let's analyze the different cases with regard to violent crime. Violent crime is higher in B than in A because instead of one manslaughter, we have six homicides. Now Lets examine the same two cases with regard to "gun crime". "gun crime" is lower in B than A because one manslaughter did not happen. This misguided analysis supports the theory that we should promote laws that encourage case B instead of A.
Any analytical subdivision that can take the above hypothetical situation and show case B as a net positive for society over case A, is misstatement of the problem. Please people, ignore fallacious "gun crime" studies and please look at real numbers on murder and violent crime. Note, I'm not advocating for or against any particular gun control law here. I'm just asking that people please be careful when you look at the data. From what I've seen, there hasn't been any convincing scientific evidence that gun control laws have much correlation (let alone causation) with either reductions or increases in violent crime or murder, when normalized for other, known factors.
Fog of war? I mean, had there been opposing demonstrators you might have had a full blown fire fight.
I've seen this argument before, but it sure doesn't hold up when you look at the numbers. Armed citizens intervening in conflicts shoot an innocent bystander in the confusion much less often as a percentage of incidents as police responding to a shooting. Frankly, a nonprofessional that is there when things happen is a better judge of where to aim their firearm than a policeman coming into a situation with minimal information.
"Likely any alien would be so alien communication would be an even bigger issue than with dolphins.)"
If Dolphins had developed technology then it would be likely they would understand the language of maths.
Roman numerals? Hexadecimal? Math as abstracted visual images like some autistic savants? Math that is understood in a way that it is still completely incomprehensible to us? I don't understand why because they understand how to solve problems using it, means they can communicate it to us or us to them. Nor does it alleviate the problem of their intelligences being so alien that their motives and actions have no meaning or pattern that we can discern.
It's as though we expect the kids to grok the process from a demonstration of the results (generally pointing out the trivia about it) and a few cartoony diagrams.
Potentially because those teachers sometimes have to teach those uncomfortable facts and shouldn't have to worry about being fired. I'm not saying I'm in favor of tenure for most schools (I haven't studied the issue or really formed an opinion). I'm just saying there may well be a similar reason. I know several parents that would push to have teachers fired for teaching such "controversial" topics as evolution without disclaimers about how it isn't really true.
I support limited rights for animals, in parallel with the level of legal responsibility that they have. Already many animals are granted a few very limited rights, like the right to not be tortured (even if it is legal to kill them). This goes along with their legal responsibilities as in they are not held accountable for their actions like theft or even murder, which is the responsibility of the owner (if there is one). The courts might order an animal put down, but only as a protection for the community, not as a punishment.
So, what level of rights should dolphins be granted? What level or responsibility? Should we make it illegal to kill them? Should we convict them of murder if they kill another person or dolphin? Should it be illegal to confine them? Should they be held responsible if they steal fish from a net?
I suspect much of the problem is one of communication. Dolphin are simply so alien to humans that we may not think similarly enough to communicate richly enough to make sense of this sort of ethical issue. (It also shatters all those awesome sci-fi fantasies about meeting cool alien species who are similar to us, but different, but we communicate and get along. Likely any alien would be so alien communication would be an even bigger issue than with dolphins.)
The process for addition is also a fact.
You can make that argument because math is provable, but it is a semantic argument about the definition of 'fact". (I might note that I recall being taught to memorize the answers to multiplication problems.) Science, however, is a process and not a fact. It is not provable nor indisputably correct but rather is logically supported by inductive evidence derived from deductive experimentation. Either way, making people memorize the definition of the scientific method doesn't really get us anywhere either. We need to teach an understanding of the application and appreciation for the results.
It's as though we expect the kids to grok the process from a demonstration of the results (generally pointing out the trivia about it) and a few cartoony diagrams.
Rather, I suspect at this point most of the instructors don't grok the process either and so teach "the materials" as a sort of memorization exercise just like every other subject taught these days. I doubt instructors expect students to be able to understand and apply the scientific method because most of the instructors just aren't very bright themselves and don't understand it themselves, or the importance of understanding logic and science in everyday life, or in many cases that there is anything to understand beyond memorizing a phrase that describes "science".
The only electronics I need in my car is an easy to use, powerful interface to my smart phone. It has everything I need to play music, check email, message, surf the web, navigate, etc.
I'd settle for a law requiring all car stereos to have a $0.30 headphone in jack to cut the absurd waste of resources and money that goes into adapters, new stereos, and FM transmitters in order to get the audio output all of a foot and communicating with the stereo.
No, first time you run the mac app store, applications installed via other means (installer,copy,whatever) are detected and marked as already installed.
Marked installed? What does that mean? How is Apple identifying apps they've never even seen?
There are reports that applications are detected even when copied to other volumes.
What reports and how are people claiming to know this? Citation?
Until there is evidence to the contrary or apple explicitly denies it, assume that your apple id is associated with whatever you had in your hard drive.
Now that would be quite a trick since the App Store app doesn't autodetect your AppleID from iTunes and asks you to enter one. So how are they associating the ID with the apps on the first run, when the application doing the detecting does not yet know the ID?
All I've seen so far is some fairly spurious guessing on your part. Perhaps you're confused because the app does look for some Apple created apps for purposes of keeping those updated (since Apple ditched the other update mechinsms for them). I suppose we can test your hypothesis with a couple of Macs to see how the data transferred changes (in volume if nothing else) between a Mac with a lot of apps and one with none, but rather than make an assumption either way, maybe it's better to gather some actual facts?
It points out the real problem with science education: we're not teaching the big facts and then delving into the intricacies, we're teaching the intricacies and hoping the big facts are obvious.
I disagree. The problem is we're concentrating on teaching facts instead of methods. "2+2=4" is a fact. Knowing how to add is a method. It's the process of forming opinions that is broken, because science students aren't learning to reason logically and follow a logical methodology when forming opinions. Another problem is our focus on having an answer even when we don't have data. We punish students for not guessing, which trains them to guess instead of honestly answer "I don't know" then follow a proper method to find out. It should be acceptable in an exam to answer, "I don't know but I'll find out with some reference materials or experiments". Worse yet, we train students to not only guess, but then to defend their guesses using rhetoric. It's not about coming to correct decisions, but about convincing others that you're right.
So here's the thing we take what six years of science classes at a minimum? And then many go on to college and get a degree. Yet I found myself last year at the bar talking to a guy with a Bachelors of Science degree and he couldn't even describe to me what the scientific method was, let alone how one would apply it to solve a simple problem. That is what the focus on memorizing facts has gotten us. Doubtless many of his peers could still dredge up a definition of "science" from memory, but how many could apply it? How many do apply it every day to make decisions rationally?
Screw our educational system. Teach kids math and informal logic and critical thinking and proper research skills. Screw standardized closed book tests. Lets reduce the number of tests kids take to prove how much they know and how good their teachers are and give a few hands on tests a year where they have to actually demonstrate a real understanding of how to do something or how something works or how and why something happened.
In a lot of other countries with a much better education system, teachers are recruited from the top of the graduating classes and are given incentives to go and teach.
Which countries are these?
Finland is a good example and among the best in the world. Their teachers are one of their highest paid professions and are government jobs. Less than 3% of grade school students go to private school. All education is free right up through doctoral degrees and includes free meals and healthcare.
Many other countries are starting to use them as a model for changes to their own educational system, including China who has been taking great interest in reforming their education along those lines. I don't expect it to ever happen in the US though, we're a bit too isolationist and the majority opinion is that we're "better" at education and many other things despite objective evidence to the contrary.
Since when was taking advantage of gaping exploits in software not hacking...
Since when is not implementing strict DRM an exploit? Quick OS X has a huge exploit and doesn't check for a valid serial number! Quick OpenOffice has a huge exploit, you can copy it without paying anyone!
The level of DRM a developer wants to implement is up to them. If they decide not to check or to check only for any valid account, that's up to them. They might make such a decision because they want to get to market faster and don't want to code and test it or because they actually don't mind people copying as they think that will promote more sales in the long run. Calling it an "exploit" seems a bit hyperbolic.
Apple simply wants a revenue stream where people can easily purchase and install licensed versions of software.
Like iTunes and the iPhone App Store, I suspect this is about selling hardware. Taking a 30% cut of app sales while providing the hosting and the credit card processing and while taking on the burden of hosting the lion's share of all the freeware in existence is unlikely to be a significant money maker. It certainly has not been on the IPhone. Rather, this is a way to make more people think Macs are easy to use by making getting apps easier, reducing crashes, and slightly mitigating security risks. The store is about selling hardware, just like their other stores.
Seriously, the whole story is that some apps aren't checking to see if the Mac in question has a receipt for that app. Most apps on OS X don't bother checking registration now. Heck, OS X doesn't even check to see if the user has a valid key. First, how is this news? Second, why the hell is apps not using DRM being spun on Slashdot as a BAD thing? Seriously, when did Slashdot become pro-DRM? Oh no apps are freely copyable and users can share them without DRM getting in the way, if the app developer made them that way! Seriously people?
Look I'm not going to go through all this again. You seem to be replying to only parts of my post, parts where I don't address the direct failures of the current system on most Linux distros. So lets simplify. For use cases where you need to move an application between two systems, non-bundled apps are harder to move. You must either re-download a copy from somewhere on the internet if it is available or you must gather up the binary and the resources, possibly zipped then extracted or whatever, and then copy them over into the correct location and make the binary executable. The latter is much, much more cumbersome and beyond the capabilities of many normal users.
For use cases where you're running an app remotely, OS X makes use of locally stored three tiered configuration files paired with an optional config file in the bundle. This means you can run an app stored on a flash or remote drive from computer A, and it will inherit the configuration changes made for that computer (A), group, and user as well as global changes for the app if the developer wants that. When you run the same app from the same location using computer B it will inherit the configuration changes for computer B, group, user, and app. These are persistent so replacing an app with a newer version or using apps where you have five floating copies on flash drives "just works" exactly as the user who has no idea what's happening in the background would expect.
To replicate the above using Linux, especially when Linux is running on computers with different architectures takes a great deal of work by a sys admin setting up some cool symlinks. It is way, way, way beyond the understanding of a normal user, brittle, and complex.
Lets be clear. Linux does not work "just fine" for these same cases. Linux falls down because it does not support bundled apps with proper hooks in the OS for FAT binaries and storing of preferences. Further, there is no standard for bundled apps so automatically mitigating the disk cost of FAT binaries becomes very difficult for systems with limited space and extraction of resources becomes a process of hunting around and hoping.
Now as to the opposite situation. Linux still has better more capable package managers that can integrate easily with multiple repositories and it is a more widely used mechanism on Linux. As a result a smaller subset of OS X software gains the benefits of centralized auto updating when compared to OS X.
In both cases the issue is solvable and the OS can be extended to handle the deficiency just as the other OS does. The difference is, I doubt Apple will add more repositories because they don't see it as a beneficial business case. I don't see application bundles being properly supported in Linux, because most Linux developers focus on the server and don't care and those that focus on the desktop are blindly resistant to admitting there could be any way in which Linux is not perfect and thus won't even consider improvements that already exist in OS X. (As a user of both I have to say at least Apple has been slowly copying the cool bits of Linux, I've seen very little of the cool bits of OS X pulled into Linux.)
Now I think I've clearly stated specific use cases where the same task is harder or just doesn't work when using Linux. If you want to discuss those specific cases, please do, but lets not go off onto weird tangents about edge cases.
P.S. with regard to your question about memory usage and libraries; I don't want to spend a lot of time on it because as I said, memory is a minor part of the library issue. Rather faster propagation of fixes into libraries used by apps is the main benefit. If you really want to understand the issue you need to read up on both OS X's dynamic linking scheme and their pre-binding for libraries not shipped with OS X itself.
you kind of ignored the last two paragraphs of his reply.
Yup, because they were empty of coherent content.
But just for you:
Unix in general can do that too.
Can do what? What are you talking about? Do you even know what you're replying to?
That's probably where the original idea for this came from in NextStep to begin with.
The original idea for application bundles in NextStep came from what? Unix? Umm, NextStep was Unix. Again, WTF are you talking about? The idea for bundles came about to simplify and compartmentalize applications to make things cleaner, easier and better architected to solve the problems traditional Unix had with messy applications leaving bits everywhere and creating dependency hell.
Linux apps were using this sort of approach when MacOS was still in it's original form...
Gee, don't specify what approach you're talking about or anything. How does one graduate high school writing class without being able to compose a specific sentence?
In short, I suspect the previous poster did not even read the thread and was just writing some sort of nonsense in the hope of a reply or is really, really drunk and trying to defend Linux against some imagined threat or something. Really, it's incoherent and makes no specific mention of ANYTHING. I mean he fucking quoted my personal observation about a reply, not any argument made in the reply.
Actually, Linux relies upon dependency resolution at install time. OS X uses self contained packages with a dynamic linking scheme. That's the difference I was bringing up and what enables OS X to have more easily portable applications and better ability to use remote software.
Again, no, you seem to misunderstand what linux does and does not do. Both systems work in both of the ways you have described. See e.g. MATLAB for linux (no install time dependency resolution), or Fink/Macports which does install-time dependency resolution on OSX.
I think you're being intentionally obtuse. Yes, the Macports system uses the same mechanism as Linux, but that's an edge case. The vast majority of applications installed on OS X are installed as openstep bundles and that is quite different. As for Matlab, it does not, however, gain the benefits of a portable application bundle described in my previous post. Sure you can bring up edge cases like Linux distros based upon GNUStep, but those are rarities where they have addressed this deficiency. That is not the case in mainstream distros.
On OS X the executable(s) and resources are in the same directory along with the libraries that aren't standard on the OS.
That's exactly the same way that 3rd party self-contined rather than package-managed software works on Linux. And the standard Linux way is exactly the same way that third-party package-managed software works on OSX (e.g. Fink).
No, it is not. For the most part third party software on Linux is distributed as a binary installer, which then writes out binaries in /bin and resources elsewhere. While some Linux distros save binary installers so those can be used to reinstall elsewhere, that is less common and you still can't just grab the application you're running and let the system worry about moving it. Further, for remote applications it does not do a good job about handling multiple configuration setups unless you're using a mess of symlinks or a file server protocol for that specific purpose.
As for unable to share libraries, that's not true. They do share libraries dynamically linking to the most up to date within the stable line. You can literally install a singed package and your other apps will upgrade or fall back to their own copy as needed because multiple copies are stored (one per app that uses it).
Are you claiming that if two different .apps have the same .dylib buried in their directory somewhere, then when the two apps are running, only one copy of the .dylib will reside in RAM? If so, then [citation needed] because I've never heard of that happening before.
That is not exactly what I was talking about. Two different apps have different versions of the same library, but the prebinding will bind to the most up to date, compatible one. This does not translate directly to RAM savings it can result in cache savings. The main advantage, however, are security and stability and bug fixes being more rapidly propagated.
Of course the package manager doesn't manage non-packages. Much like the .app method doesn't help executables that aren't .apps. For non managed packages the install process is usually a case or running the installer executable, which is I will grant more awkward than using a .app on OSX (though plenty of OSX programs also seem to require installing, too). But not much, given that the majority of installed software is done through the package management system.
The difference is, most users on Linux, whether home users installing games or pro users installing dev tools do end up installing applications utside of te package manager. Very, very few users ever install software on OS X that is not a .app bundle.
How does an uninformative reply like this, that has no useful information or even argument get modded up? All you have is an opinion that Macs are not a "system" and everything is thrown together? Do you have any idea what the hell we're talking abut?
PC-BSD (based on FreeBSD) is doing this with their PBI system right now:
Interesting, although it looks like it is still placing multiple parts of the app in different locations and both must be copied in order to transfer an app (instead of a single package). Also, I'm not sure I see how this addresses running apps off a flash drive when they need to be run by multiple platforms and OS versions. Assuming you installed an app on the Flash drive, wouldn't you still need to create a local symlink on each machine so it could find the resources?
In general most software on Linux systems comes from the distro's repositories, so having a decent GUI is generally sufficient to allow for easy installation.
Look, I'm a Linux user as well as an OS X one. Yes, most of the software is in the repositories... but by no means all of it. I still end up having to manually move or re-install a lot of software because a LOT of software I need or want is just not there, especially commercial packages.
It also gives you a central location for updates, whereas on Mac OS X each app--along with the OS--has its own update mechanism.
It only gives you a central location for updates for software provided through it, not for all software. And as of today, that's basically the same situation on OS X.
Similarly if you add a third-party repository...
And here is where other repository systems are still ahead of OS X. That said, commercial vendors find setting up their own repositories too cumbersome compared to Web downloads and hosted repositories take distribution out of their own control too much. So far, only Ubuntu has really addressed the need for payware software developers to distribute through the central repository.
For the standard application distribution method they provide, it's largely unnecessary. They really do provide a mechanism that's simple and elegant and self-contained.
I actually disagree. For updates and app discovery, Linux is still ahead of OS X, likewise for the ability to handle multiple repositories. That said, Apple took a big step in the right direction today. Apple's application format, on the other hand, still provides significant benefits not yet matched by Linux formats. There is no reason for the two to be mutually exclusive.
Can an Apple user send a current intel-only app to someone running OSX on PPC or iOS on a mobile device -- no!
If the app was built as Intel only, no they can't run it on both, but when there are multiple architectures in play, both running OS X, yes it can. We're talking about the capabilities of the OS application package structure. I can still access apps on a flash drive and run them from my legacy PPC machine, then plug them into a brand new Intel machine and run the very same app.
So picture this, ARM takes off and becomes a viable OS X platform going forward. So app developers add that checkbox when compiling apps and suddenly, with no effort on their part, all the apps being produced become portable between both types of machines. Meanwhile Linux users have to download separate packages for both systems.
Then we get to the question of why anybody would ever want to copy a raw binary instead of using a linux distros package manager?
Your question is flawed. Everyone has moved to using package managers now. The question becomes, what advantages of OS X's package format still brings benefits that current Linux package managers cannot replicate. One is migration. When I move to a new laptop, I suck all my apps across rather than re-downloading and reconfiguring them a process that takes considerable time and bandwidth. If my new laptop is a different architecture than my last, with Linux I have to re-download them all (or would if I wasn't already running my Linux desktop in a VM). And what about commercial apps? You do know not all apps are in repositories. I regularly have to download apps from Web pages, whether they are server components like zencart or commercial games. (My compliments to Ubuntu for starting to get some commercial apps in their repositories.) In my example before about IM'ing an app, it was actually a real world example. The app was no longer being offered by the developer and was nowhere to be found except on my machine. Now with some work and know how I could have replicated that feat on Linux on the same architecture it would not have been easy and it would have failed when it went between architectures.
Fat binaries were a temporary workaround and both Apple and MS include emulators.
FAT binaries are a forward thinking mechanism that promotes usability in a world where we are exposed to different architectures over the years. There's nothing temporary about the concept unless you think we've settled on the "one true architecture" now and forever.
Can I run a linux ARM binary via a distro under QEMU from another arch -- yes.
Yes... but not easily and not unless your emulator and OS supports all the architectures. And you take a serious performance and stability hit doing it instead of using a native binary.
Please don't even argue this point.
Of course I will.
Weird I thought you were going to argue the point too, but then you didn't address any of the real and practical application differences I pointed out. That's not much of an argument.
OK, you clearly do NOT understand how Linux works, or ironically OSX. The underlying mechanisms are then same in both operating systems that we may as well consider them the same for the purposes of discussion.
Actually, Linux relies upon dependency resolution at install time. OS X uses self contained packages with a dynamic linking scheme. That's the difference I was bringing up and what enables OS X to have more easily portable applications and better ability to use remote software.
Linux and OSX implement self-contained programs in exactly the same way. They ship around and archive containing the executable, the requires shared libraries and resources.
On OS X the executable(s) and resources are in the same directory along with the libraries that aren't standard on the OS. Further, they're in predetermined locations so it's easy to grab resources like music or images.
In your preferred system, every package would have to have all of its own libraries. Not only would that vastly increase the amount of disk space used, it would (more importantly) vastly increase the RAM usage since different executables would be unable to share the same libraries.
It does increase disk space usage, but not significantly as resources are actually the big part of apps these days. If you get really crunched on space you can prune binaries and unused libraries at the expense of losing portability going forward. As for unable to share libraries, that's not true. They do share libraries dynamically linking to the most up to date within the stable line. You can literally install a singed package and your other apps will upgrade or fall back to their own copy as needed because multiple copies are stored (one per app that uses it).
I like the package management based solution. It basically works like magic with absoloutely no effort.
It doesn't work as well, especially for apps installed not using the package manger (as a Linux user I'm sure you have to deal with these as well) and it falls down in the several, specific use cases I mentioned in my last post (and which you did not address).
You know you can't just copy the programs in /sw/bin to another mac, right?
You do know that basically no applications get stored in /sw/bin right? That's mostly for bad ports and legacy software. Even OpenOffice installs as a .app these days and it can be stored anywhere the user likes.
If "developers will hand over 30 percent of the purchase price to Apple," what will consumer prices be?
Have you ever worked in the end user software development business? 30% going to distribution, credit card processing, and managing updates isn't bad. When you add in the amount of publicity it generates by being in THE searchable software database for end users, well, likely prices will drop as advertising will drive more sales, more price competition, and larger volumes.
Really? Is that why I can move my home directory from one linux install to another and the programs will still run?
Please don't even argue this point. Linux is a bit behind the curve and the only people who would argue otherwise are people who don't use both OS's. Sure you can copy your home directory on Linux, or use the stored installer (if you are expert enough to know where they go) for an individual app (on some distros)... all provided you are running on the same architecture.
With OS X you can literally drag an application into a chat window to a friend, who is running a different version of your OS, running on a different chipset and that friend can double click the app and run it. It's a great deal more painless since all the apps are the installers and are self contained directories ending in .app. It's one of the things Apple got right and where no Linux distro has enough pull to push change, especially since it is not a big pain point for end users. Additionally, the OpenStep packages make running software off a network drive or flash drive or anywhere really, easier by allowing for multiple sets of preferences and multiple included binaries to get around the whole hack of symlinks or multiple copies for multiple architectures.
Linux is not ahead in every area, just as OS X and Windows are behind in other areas. Get over it.
Apple "innovates" again and re-invents the package manager Linux has had for ages...
Yeah, pretty much. I wish it had not taken them so long though. I wish they'd do a better job copying virtual desktops while they're at it. I notice Canonical innovated and added apps for sale to their own app repositories. Now I wish Linux distros would innovate and re-invent GNU-step style packages and required package signing with real credentials to get into the default repositories, and heck system services while they're at it.
I really, really like it when OS's copy the best parts of other OS's and my daily computing experience is made easier. I don't really care that someone else came up with something first.
TFA makes it sound a lot like apt or the package manager for any Linux, except here you have to pay for some of the apps, and without a community repo.
What do you mean you don't have to pay for any of the apps on Linux? Ubuntu has included paid apps in the USC for some time now.