There was a time in this country, a long time ago, when reading wasn't just for fags, and neither was writing. People wrote books and movies. Movies that had stories, so you cared whose ass it was and why it was farting, and I believe that time can come again!
Nothing is disabled, turned off, removed, commented out, locked etc
And yet, OS X on the desktop supports a bluetooth keyboard. The iPhone supposedly has a stripped-down OS X. Therefore, they must have stripped out USB HID support.
I've not seen bluetooth keyboard in ANY other phone I've ever owned,
See other posts in this thread. It's not just the keyboard, and other phones -- even the ones that aren't "smartphones" -- do far better than the iPhone in that respect.
This post, in particular? Of course not. Your use of the internet to view television shows? Yes, you probably should, if you wish to think of yourself as completely honest.
Why? What is the difference?
You choose to watch television, without paying for it,
In what way am I "not paying for" my own bandwidth? Read my sig.
If you mean I'm "not paying for" the content, I'm not paying to read your content here, either.
If you're assuming piracy, you clearly didn't read. The Daily Show is actually free, as are several other shows.
A mutual agreement might be arrived at, whereby you pay 5 or 10 percent extra for internet, without being charged for the unused television service.
Paying to whom?
My ISP? That's a blatant violation of network neutrality.
A media corporation? That's both a blatant violation of network neutrality, and anticompetitive towards the other shows (Revision 3, anyone?) which aren't available on network TV, but are on the Internet.
Discussing your digital activity on the interwebz with the post office is so absurd that no rational person would consider it as more than a joke.
So is discussing it with people from my local TV company. Demonstrate the difference, if you can.
Better yet, tell me why talking to my ISP about it is any different than talking to my phone company about who I want to call, or talking to my local grocery store about just how I intend to eat that apple.
Fuck no. You provide me a good or a service, which I pay for. Then I use that for precisely what I want. If you don't want me to use it for that, don't provide the service.
Obviously, you haven't read the REST of my posts in this discussion.
Not much, no. If you like, feel free to point to any you feel are particularly relevant.
Where is this "imaginary non theiving world" to which you refer?
The fact that you refer to us as theives means you're either suggesting that it's possible for someone not to be a thief -- and I'm suggesting that by your criteria, this is impractical and not desirable -- or you're suggesting that everyone is a thief, kind of like the Christian view that everyone is a sinner.
Again: If I were to avoid being a thief, does that mean I have to discuss this post with my ISP and the local post office, and get approval from both? If not, how is posting this different than the activities you've described as "stealing"?
Opensource is as good as my own property, for the taking, because I can do ANYTHING with it.
Clearly, you are quite ignorant of Open Source, as defined by the OSI.
Free software is something that I am permitted to use AS LONG AS I use it in a way approved of by the authors/owners.
Actually, that is true of most Open Source as well. It's also far from the FSF's definition of "free software".
There isn't a lot of software which could be classified "open source", but not "free software", by your definition. I could only mention sqlite, and DJB's stuff (qmail, djbdns, etc) -- these are public domain.
No, the difference is actually quite an obvious one: Open Source only requires that the source code be available. Before DJB's work was open source, it was very much open source, in that everyone could freely acquire the software. However, it was not free software -- once you have the source code, you are not allowed to redistribute it with your own patches.
If you wanted to improve qmail, for instance, you had two choices: Either beg DJB to include your change, or distribute your patch directly to users -- they would have to then download the qmail sources, apply your patch, and compile.
Free Software, on the other hand, only requires that you be able to both obtain the software, and fork it to distribute your own version. It doesn't place any other restrictions, it doesn't even require "free as in beer".
By the FSF's definition, the official distribution of sqlite is very much Free Software. By the OSI's definition, it is very much Open Source. And it is completely public domain, meaning there is absolutely no restriction placed on its use.
Your confusion probably stems from the fact that the FSF advocates the GPL, which does impose quite a lot of restrictions. The GPL's purpose is to ensure that this software must continue to be Free Software -- compare to sqlite, which could be forked into a proprietary version.
But there's nothing about the idea of Free Software itself which requires the GPL, or anything like it, and it's quite ignorant of you to suggest that.
"OOOHHH! Google USED TO allow us to use their resources, now they don't! OH! EVIL GOOGLE!"
Nice strawman. Can you point to a single post with that sentiment?
It's possible you can, but I've been reading this thread for awhile, and I've found nothing of the sort. I've found a few people who suggest that Google should have considered the implications, and certainly some people are disappointed, but I haven't actually seen anyone say that Google is evil or wrong for doing that.
If your ISP offers both internet and television, and you purchase only one, then continue to make use of both,
Ah, but I'm not. I'm watching the Daily Show on a website -- this may or may not have anything to do with actual television. For that matter, if I were to watch Ask a Ninja, which isn't on any television network that I know of, is it still "stealing"?
YOU ARE CONSCIOUSLY ATTEMPTING TO ABUSE THE SYSTEM.
What you call "abuse", I call "use".
Did you read the TOS?
Yes. And while it was a bit unreasonable about bandwidth expectations, there were no restrictions on actual content beyond "keep it legal".
Did you discuss your plan with the ISP?
I don't know -- do I discuss every time I plan to use Slashdot with my ISP?
Wouldn't it kind of defeat the fucking point of the Internet if I had to discuss every website I visited with my ISP?
Were you completely open and honest with the ISP about your intentions?
My ISP never asked. And that is the honest truth -- they never once asked what I intended to use my Internet service for. Being reasonably savvy themselves, they had to assume that "Internet" includes audiovisual media.
Frankly, I don't see why that's any of their business, any more than the phone company has any business asking who I intend to call. In fact, I'm allowed to telephone their competitors! Fancy that!
Or, did you just sort of sneak around and put this thing together to your own satisfaction, hoping that no one would catch on and block you?
No, I did it pretty openly, and legally, hoping that the ISP would realize that this is precisely what the Internet is for, and why it is valuable -- to promote the free exchange of information. I was also pretty much hoping the ISP would be in favor of net neutrality, and not openly "block" anyone, as that would be blatant censorship.
Now, I am a bit lucky in that respect -- my own ISP, while they do sell a television service, they realize that they are for the moment primarily an Internet Service Provider. They also realize that if I use too much bandwidth, they can purchase more to cover it, and charge me more if they have to -- for the moment, they openly support network neutrality.
But that is how it's supposed to work.
Tell me, in your imaginary non-thieving world, should I have discussed this post with my ISP, and with the local post office, just to make sure I wasn't "stealing" by not sending it via snail mail? And just how is anyone supposed to get anything done on the Internet, if the ISP is meant to micromanage it so much?
It is abuse to take advantage of some free service,
Is it, now? Well, you better stop posting on Slashdot, then...
thereby circumventing the telecom's charges.
I don't really see how.
Put another way: If my ISP offers both Internet and TV, but I choose to only purchase Internet, and watch the Daily Show free online, is that "abuse", because I'm not consuming exactly the service they want with exactly the fees I want?
It seems to me it's simply a more efficient use of a service I am paying for. After all, that app presumably does use the telecom's own Internet service, meaning you are paying for that -- if that's abuse, the telecom clearly isn't charging enough for bandwidth. I would suggest that, instead, they are clearly charging too much for text messages.
If it's not using the telecom's own Internet service, but rather some local wifi, I see even less "stealing" going on, given that you're taking place in a communication completely outside of what the telecom is providing. Should it be considered "stealing" that I prefer instant messaging to text messaging, and actually don't send text messages?
If you are going to steal Google's (or anyone else's) services, at least admit that you are thieving.
You first. Or have you just purchased something you saw in a Slashdot ad?
Clearly, Slashdot makes enough money from all of us visiting this website to keep it running. That, or someone is being very, very generous -- but if it really is losing money, that's not "stealing", that's their own failed business model, and they could shut it down.
Similarly, this service Google has provided was shut down, because they didn't anticipate that much traffic, and couldn't figure out how to monetize that. That is Google's failing in creating a service which could handle it, not our failing in using it.
That's a bit like suggesting that it's "stealing" for me to actually use the bandwidth I paid for, because I am slowing everyone else down. Well, no, I paid for my bandwidth -- if the ISP oversold it, and I'm now lagging someone else, that's really the ISP's fault, not mine. I'm just using what I paid for.
The need for driver's ed has nothing to do with the complexity of the car; rather it's about the motor skills, judgment, and understanding of the rules of the road.
It's not entirely divorced from the complexity of the car -- see oil changes, tire checks, etc.
But you're right. Training for a computer should be about logic skills, judgement, and understanding the rules of the Internet. As you said:
that has nothing to do with the complexity of the device; it's about a failure to understand the basic rules of effective and safe operation.
There are some basic rules for effective and safe operation of a computer, that many people with an IT department -- salespeople in general -- are willfully ignorant of. They almost seem to see it as somehow beneath them.
Proper design goes a long way to eliminating human error;
There is only so much you can do to avoid human error. Eventually, you have to train the humans.
Put another way, yes, an automatic transmission helps -- it's harder to stall the car by suddenly shifting to the wrong gear. However, if you suddenly shift to park while on the freeway, bad things happen. And at a very basic level, if you turn the wheel left, the car will go left.
It's possible you meant to go right, and there are very dangerous things to the left, but if you turn the wheel left, the car goes left.
Similarly, we could argue all day about how to build a better virus scanner, but ultimately, if a user tells their computer to run an executable file that they've downloaded from the Internet, then that's what the computer will do.
It's certainly possible to cripple the device. You can force people to ride the train instead, and they will have no ability to steer -- they'll literally be on rails. Similarly, you can remove the ability to run all but a predefined set of executables, or the ability to download any file types that aren't known to be safe, and otherwise lock the machine down to protect the user from themself.
But I don't think that basic level of training is asking too much, and I think it is ultimately more effective.
Too many people think computer training means understanding the OS and why things work like they do; rather than what do you need to do to effectively and safely use the machine.
I would say, some basic theory will help... I would lean heavily in favor of teaching more basic skills, rather than learning by rote.
For example, here are some basic things I teach:
- Try right-clicking on whatever you're trying to change.
- If it isn't obvious what a button does, hover over it.
- Look at the keyboard shortcuts next to that menu item.
- If you don't know where to start, explore. Wander through menus. Hover over buttons. Click on things -- if you're about to do something bad, you'll probably get an "are you sure?" message.
- When asking for help, phrase your question in terms of a goal, not a tool. The person you're asking might know of a better tool.
Compare to, "In order to add a header to the document, go to the Format menu, click 'Enable document headers' if it exists. If there is an item which says 'Disable document headers', leave it alone. Then go to the View menu, choose 'Display Mode', then 'Page Mode'. Then click the small rectangle at the top of the page and type your header."
That will be much quicker if a user simply wants to insert a header and be done with it. But it's a lot to remember, and I know I won't remember it by the next time I'm asked. It's much easier to teach people to look through the menus for something having to do with 'headers', or to check the documentation under the Help menu.
The problem is, teaching these general things, and their most important applications, can take a bit of time. This is why I compare it to driver's education -- I would expect it to take at least that long, for
I guess the biggest reason it is possible to have such a system, is because Linux itself is "free" as well as almost all software that runs under it.
Actually, that may have something to do with the current implementation. It has absolutely nothing to do with the potential of that to work.
After all, both Microsoft and Apple have centralized update systems -- they just haven't opened them up to third-party developers.
For that matter, Valve provides Steam, which provides reasonably secure (I assume) downloads through a centralized package manager, plus automatic updates -- and with some games, it can download only what's needed to launch a game, and download the rest on demand.
Yet almost none of the software distributed via Steam, nor Steam itself, is "free" in either sense of the word.
There is no technical reason you couldn't have a similar system, with less of the centralized control (and licensing cost), and more of the features we've talked about.
One reason why Linux is not making much headway on the desktop, is because there are a number of mutually incompatible flavors.
So use Ubuntu. This really isn't an issue anymore.
Consider, also: Windows now comes in some six or seven flavors, per release. Each has different features enabled or not. The "ultimate" versions cost significantly more than the others, if you really want everything.
Ubuntu comes in precisely one flavor, with everything. No need to worry about whether you can live with "home basic" or whether you need "home premium".
The user has to know which flavor is on the machine and then connect to the proper repository.
For most software, it will already be in the main repositories, which means you don't need to know what flavor you've got except "mine". Just open your distro's package manager, find the software you want, and click "install".
If you insist on getting it from third parties, yes, you do need to know what distro you're running. I'm really not sure why you think it's harder to remember "Ubuntu" than it is to remember "Windows" or "OS X", never mind "XP" or "Vista" or "Leopard".
Another reason is that hardware manufacturers are reluctant to put all their hardware eggs into the Linux basket because they have no official central place, such as Microsoft, to get software support.
How so? I'm really curious what support system Microsoft has for manufacturers...
End users who install Linux also don't get a single entity to squawk to, but have to find the right complaint Department for their particular distribution.
There are the Ubuntu Forums. There isn't a "department" -- you just post on the forums, and people come help.
Or, if you really want to compare apples to apples, buy a Dell with Ubuntu. You can then call Dell, just as you would with Windows.
Also, most people instinctively know that it is not cool to look a gift horse in the mouth. That makes people more reluctant to complain...
Have you ever actually done any open source work?
I can pretty much guarantee you are dead wrong here. People absolutely do complain, frequently, and are quick to blame anyone but themselves, no matter what the problem.
Linux is sort like one of these bicycles that comes in a compact box. It may be a great bicycle at a low price, but the user first has to get some tools and assemble the thing
Unless, again, you buy a Dell with Ubuntu preinstalled. Otherwise, you're not comparing apples to apples -- it's a much closer comparison to a user who bought a machine with Vista and now wants to install XP.
And that is usually more frustrating, by the way. It took far longer to find XP drivers for all the hardware in this laptop than it did to find Ubuntu drivers, since the thing was designed for Vista (and Ubuntu).
This is far from the only computer I've used on which Ubuntu was easier to install than Windows.
Since OSX and Linux share the underlying *NIX architecture, it should not be too difficult to implement the strengths of each on both of them.
That's a bit like saying, since men and women share the underlying Human architecture, it should not be too difficult to implement the strengths of each on both of them.
I don't know about you, but while I have nothing against hermaphrodites, I'm not particularly attracted to them.
So while I think it is possible to create something which has the most important strengths of each, you're always going to have someone who insists that a particular trait is a strength, where you see it as a weakness. For example, I really don't want to download binaries for some fifteen different platforms -- I'd rather just get the one x86_64 binary I need, and download others when I have another architecture that actually needs them.
Whereas you would rather waste the bandwidth and disk space, so you can easily transfer an application from one machine to another, without worrying about architecture. I'd much rather re-download it on the second machine.
Personally, I do not update most of my software, unless there is a really compelling reason to do so.
The problem is, if it's a security reason, you won't know that compelling reason until it's too late.
That is one advantage of the Debian and Ubuntu models -- you can stick to an old version until you need a new one.
For purposes of trying out new software, I usually create a special account for that purpose and then install all trial software in that account only, leaving the system entirely untouched.
I tend to trust the repository, if it's in there. I could as easily create a test account, install the software to the system, and try it out on the test account -- if I then want to remove all trace of it, I just have to uninstall with --purge (purge config files), and then nuke the test account.
For software that isn't available as a trusted package, I'm probably going to end up compiling from a source -- which I can do in a test account, and keep it completely isolated. If it is hardcoded to need root access and certain paths, I can chroot it and run it with "fakeroot", keeping it just as effectively sandboxed as if it was a virtual machine.
That's not as easy as it could be, and I'd rather see more isolation by default, out of the box, as well as making it more user-friendly. Most desktop accounts could easily allocate several UIDs per tab.
But at the very least, it's possible, even if it isn't often needed. Centralized updates, optionally automatic, aren't really available anywhere else.
Learning to drive has little to do with the complexity of the device- you can drive quite well and never have a clue about why the car works.
This is true.
You do need to know the basic principles, however. You don't need to know the physics of how hydroplaning works, but you do need to know to slow down when there's water on the road. You don't need to know exactly how shifting into reverse on the highway will tear your transmission to shreds -- actually, I don't have a clue, maybe the transmission is fine and the engine explodes -- you just need to know it is a Bad Idea.
But you haven't provided a case for why computer users shouldn't be required to attain at least a basic level of competency.
You seem to be hinting that a computer shouldn't require a user to know how it works, but really, it doesn't. You don't have to know why downloading programs is dangerous, any more than you have to know why unprotected sex is dangerous. Just some simple instructions: Use a condom, and don't try to install software unless you know what you're doing.
Even that "knowing what you're doing" isn't difficult -- for any given program, do some quick research online to find out whether you can trust it, favor open source, etc. If it's too much work to do that, ask a geek. You wouldn't make a major modification to your car without talking to a mechanic, would you?
A salesperson only needs to be able to understand how to produce the material they need to make sales; not what's under the keyboard.
But they do need to understand how to do their job effectively, which includes not screwing up what's under the keyboard.
I could, for example, slam on the brakes at every intersection, tires squealing, because it's the only way I know to stop. I could leave the key at the "start" position for a full two minutes because someone told me it was a good idea. I could leave my seatbelt unbuckled, because it really has nothing to do with getting where I'm going -- if anything, it's an annoyance, slowing me down when I need to reach for something. And I could be willfully ignorant of basic maintenance, never changing the oil.
Then, I would either be killed, or I would bring the car in for maintenance much more often, spending much more money on replacement parts that I destroyed through my own ignorance, and all the while complaining that it's somehow the manufacturer's fault -- if skidding is so bad, a car shouldn't let me do it! The ignition should just be an on-off switch -- or better yet, a button, like computers have! The oil should tell me it needs to be changed! That light is unintuitive -- it should tell me in plain english, like any good error message!
Understanding why something works is not necessary. But there is a basic level of competence that is generally required in every other industry. If someone behaved as I've described with a car, they'd be lucky to still have their license -- yet people behave this way every day with computers.
And keep in mind: There is an entire high school class called "Driver's Education". I see no reason there couldn't be a similar class on basic computer usage and maintenance. And I don't mean "this is a mouse" education -- everyone knows, no one holds up a steering wheel and says "This is a wheel, turn it left and you go left, turn it right and you go right." I mean all the stuff you really should know -- like when to accelerate instead of brake to avoid an accident, and why common sense with file extensions can be more effective than any antivirus.
I'm not asking that people be able to put a computer together from parts, or that they understand assembly language, or even that they be able to use a commandline. I'm asking that they take a little responsibility, and learn at least as much about computers as they do about other tools that are important to them.
Given the choice, I would much rather have a world in which some people attempt to keep their inventions secret -- but are vulnerable to reverse engineering -- than one in which I have to be careful what I invent, to be sure it's not already patented.
But the best compromise I see for patents would be to limit them to an incredibly short time -- perhaps one or two years. Plenty of time to get a first-mover's advantage, but not enough time to allow for patent trolls, or business models based solely on licensing patents. If you were to sue someone for infringing on your patent, they might just decide, "Oh well, we'll sell it next year."
I really can't say, though. This book makes a good point for removing patent protection entirely.
That argument used to be valid in the early days of computing when these things were very expensive.
Still is.
when 4GB of PC6400 DDR2 memory sell for less than $20
First: Can you get that in a netbook?
Second: It still has an effect on cache. That has been increasing, too, but not nearly as much.
Most modern computers come with disk drives of several hundred GB where storing 18 MB file or one of 30 MB makes no practical difference at all.
There are still a large number of people stuck on dialup, or with bandwidth caps. And we're talking about a single program. Two things to remember here:
First, there are quite literally thousands of packages in a typical Linux install, and that's what comes with it. OS X manages to hide most of this by including so much with the OS itself, including Ruby on Rails with every Leopard installation.
Second, while often increases in efficiency seem unnecessary, you eventually find some sudden innovation that relies on it. It's easiest to see this in gaming, where often things are "faked" -- for example, lighting used to be calculated ahead of time, and baked into the level. If a spinning fan cast a shadow, that's because the artists added the shadow when they added the fan.
Suddenly, with a certain amount of raw power, and the efficiency of Carmack's Reverse, we no longer have to fake those -- everything can be lit in real time. Perhaps the most obvious demonstration is the Doom 3 flashlight -- far from being a static, precalculated light, or the simple "make this area brighter" flashlight of things like Half-Life, it was a real light source the player could carry around, and it did cast shadows.
So, consider: If the initial download of a program is much smaller, due to many of the shared libraries and interpreters already being present, it's possible to get to a point where native applications can be installed nearly instantly. Provide a sandbox and a decent UI, and you have what Java tried to do with WebStart, and what AJAX is currently trying to do, but for non-web applications.
That is, no need to "install" unless you wish to mark some as being available offline, otherwise, you could see hundreds of programs in your program list, and the ones not already installed would be downloaded on first use.
To my knowledge, this has never caught on, but it also hasn't been tried on a large scale, and with the advantage of a package manager.
That is certainly not true for the vast majority of programs for OSX that I know about. Many programs do not even have or use a special installer program.
True. However, enough do, including important programs like photoshop, that this cannot be entirely dismissed.
Simply dragging the application icon into the trash is all that is necessary to completely get rid of ALL of a program.
Unless the program has installed a menu to the menu bar, like the built-in AirPort UI. For examples, see pre-Spaces attempts at virtual desktops, VPN clients, etc.
Or maybe it actually made a fundamental change to the system -- for example, a program to force any window fullscreen with a keystroke.
These are the kinds of programs for which installers and uninstallers make sense. Perhaps dragging to the trash is a good UI for uninstalling -- though I might debate that, in these cases -- but it should do more than simply move some files around.
In OSX a downloaded DMG file appears in the downloads folder. Also in Safari double-clicking on the filename in the downloads window immediately opens (mounts) it.
Right...
Dragging the program icon to the applications folder icon installs a program in the one place where all programs a user would generally require are present.
This is the step that isn't often done. What I usually find instead are people who ha
Linux is an operating system made by geeks for geeks. That sentiment has been expressed numerous times in this thread.
While that is true, the other side of it is false:
Linux is not for grandma, unless she has a Linux using grandson/daughter or other relative or friend.
With recent Linux distros, I've seen no reason Linux can't be for grandma. In fact, she'd probably be better off -- she only has a few things she needs to do, which she can do very well, like emailing her grandchildren, or receiving photos from them.
But, she's going to find it somewhat more difficult to install some random screensaver that someone emailed to her, which can only be a good thing.
Why should a program need to be installed on a computer? Why not simply copy a desired program from whatever source, online or a piece of plastic, to a well marked, established place in the computer?
The main problem I have with this are, programs are more than one piece. Common pieces -- libraries, interpreters, and the like -- may or may not already be installed. At the very least, you need a dependency system.
Mac programs can be "installed" by simply copying them from a source disk or download to the applications folder. Any initializing and setups are done when the new program is first run. Surely, it must be possible to do that sort of thing in Linux also.
Certainly. But it would be a bad idea.
First, consider: OS X is designed to run on two different processors, in 32-bit and 64-bit modes. Depending how much of the program is compiled, this means quite a lot must be included in a "universal binary", and most of it will sit there completely unused, wasting disk space, RAM, and bandwidth.
Case in point: VLC for OS X is available as Intel and PowerPC specific packages. These are a little over 18 megabytes each. It is also available as a Universal Binary, which is about 30 megabytes.
The only advantage to that waste is that you can copy an application from one machine to another without worrying about what their architectures are. But it's worth noting, when installing a package via Synaptic, you don't need to know your architecture either -- it will automatically download the package for you.
Second, consider the libraries. VLC on OS X is 18 megabytes, and that's the platform-specific version. On Ubuntu, it's about 3.6 megabytes. The rest is in libraries, which you may or may not have installed -- but you don't have to care, Synaptic will handle it for you.
This is pretty consistently true, throughout the OS.
Now, what's the advantage of the OS X method? It's probably easier to install from a CD, maybe. But from the Internet? You must be joking, unless something's improved. Last I checked, the typical method was:
- Download a DMG file.
- Find the downloaded file, and double-click.
- Find the newly-mounted drive.
- Click and drag the app to the Applications folder.
- Unmount the drive.
- Throw away the DMG.
Compare this with Windows:
- Download an EXE
- Find the downloaded file, and double-click.
- Follow instructions on screen.
- Throw away the EXE.
Now compare with Linux:
- Open Synaptic
- Find the program you want
- Check it
- click "apply"
Looking here, it looks as though the Linux version is just as easy as Windows, and quicker, but Windows and Linux both beat OS X. Also keep in mind, the Linux version will pull from the official repositories, meaning the program will be signed -- on Windows and OS X, the closest you'll get is a security prompt, which basically says "This was downloaded from the Internet!"
What's more, OS X doesn't always do things that way -- there's often a.mpkg involved, which is really an installer. Here, on Linux, I can go to Synaptic,
Right now if something is patented, you need to figure out another way to do the same thing. Sometimes the new method is even better than the original. THAT IS THE [IMPLIED] GOAL.
Actually, the implied goal is to provide an incentive for the original invention -- the theory being that no one will create anything without it being protected.
Without a patent, everyone would just use the 1st method and nobody would want to improve upon it.
Bullshit. The incentive is still the same: Build a better mousetrap.
Do you honestly think that Apple created the iPhone because existing physical keypads for cell phones were patented?
Sure, there are exceptions -- PNG was created because Gif was copyrighted. But for every example like that, I'm sure I can find ten examples of inventions that would have been created without patent law, and twenty of innovation that was held back significantly by patent law, going back to the first steam engines.
Linux is now, and has been for awhile, far easier to learn and use than Windows. OS X might be easier, but I would argue there are places Linux does better, even from the "clueless user" perspective.
The problem is, as this article identifies, it's not enough to be better. You have to balance a tightrope of being exactly like Windows, so Windows users don't have a learning curve, and yet so incredibly better than Windows that people are willing to ditch Windows (and all their Windows apps) to give it a try -- even if there was no learning curve, it can be a hard sell.
This is not Linux' fault. It's the essential problem that with a single OS dominating the market, it's nearly impossible for a third party to gain any traction -- which is why we were suing Microsoft for antitrust in the first place.
In other words: Linux has pretty much no hope, but it's not through any fault of Linux, or because anything else is easier to use.
They aren't competent because they have no incentive to be -- if they screw up their computers, that's IT's problem. If it suddenly became their problem, they might see things a little differently.
Just for fun, here's a car analogy: A car is a rather complex piece of machinery, and takes a lot of training -- typically an entire class of driver's education. While some people go on to master it and become stunt drivers, or simply improve their skills and get a truck license, etc, most are content to at least reach some level of competence.
But if you never bother to reach that much, you end up driving into a tree, or a telephone pole, or another person, and it's generally your fault.
Aside from the fact that cars are actually dangerous, and can cause bodily harm, I'll go with the fact that it is entirely the responsibility of the driver to be properly licensed and at least competent, and if they can't do that, it's entirely on their own head, both literally and financially.
Now, granted, many corporations don't like the idea of having to fire their best salesmen because said salesmen are morons about computers. But that only perpetuates the myth that it's somehow hard to attain some level of competence, and allows the salesmen to continue to see computer knowledge as somehow beneath them.
Re:My experiences so far
on
The Age of Steam
·
· Score: 2, Interesting
I hear you on the credentials...
I forgot my steam password, once, when setting up a new machine. The old one wasn't exactly functioning, and I'd simply copied all of the files off the hard drive, assuming that would be enough for anything I needed. Fortunately, I also had a disk image, which i convinced to run in a virtual machine, then downloaded and ran a small utility to find my password (from a very isolated virtual machine, so no way it could send it to anyone).
I'm sure it's possible to do it without actually running the Windows in question, but there was no ready-made script, and it was way more than I wanted to learn about Windows crypto.
But yes, after that, there is no way I'm forgetting that password.
Perhaps, but really, ads on seeks is going too far.
Do I see an ad when I fastforward a VCR tape? No. How about previews. Are they skippable? Yes.
Do I see an ad when I fastforward or arbitrarily seek a DVD? No. Are the ads skippable? On most of them, yes. In fact, most move the ads to a "special features" area.
Do I see an ad when opening a recorded show on a DVR, or seeking within it? No. Are the ads skippable? Hell yes, that's half the point of a DVR in the first place!
Do I see an ad anywhere in an mkv I downloaded via BitTorrent? No.
Do I see ads when seeking or otherwise fastforwarding in Hulu? Yes. Are they skippable? No.
So, even one ad on seek -- and considering it's unskippable, as are the ads at the beginning and every 15 minutes or so -- makes Hulu the single most obnoxious platform out of all of the above.
Again, I kind of want to like it. Digital delivery, on demand, using just the Web (and, ok, Flash)... But no HD, US-only, slow to release, and almost more ads than I'd get watching it on TV.
The only thing you can do is make the process so painful and bothersome that they decide it's not worth it.
And make it so long and time-consuming that they will be caught by coworkers/cameras and fired for doing something clearly against company policy.
Speak loudly about that company policy, by the way. After all, the job market sucks right now. Make the employee ask themselves whether that after-hours torrent is really worth their job.
Treating adults like adults is good in theory, but when you have 300+ people trying to beat "the man", you want to take away as much temptation as possible.
Has it occurred to you that if you treat them like adults in the first place, there will be less desire to try to beat "the man"?
The dirt-simple solution, which I never, ever seem to see implemented, is:
- Let employees be responsible for the well-being of their own machines
- Inform them that files are only safe on the fileservers, and make frequent incremental backups
- If they FUBAR the machine, re-image it at the touch of a button
- Apply security at the network level -- block outbound port 25 and anything else likely to get you in trouble, monitor for badness, etc. But this is an easy job when users are responsible for what they (or their computers) do.
Actually, I lied. I'm told this is the way it's done at Google -- at least the last part.
By the way, no real training needed. Just give them a quick presentation on how very alone they are, and let them know that if they don't feel they're computer-literate enough to do their own IT, the real IT department will be happy to remove root access, and give them a list of guidelines to follow (like Don't Install Software, Ever) -- which still takes fewer IT resources than locking them down against their will, and playing a cat and mouse game trying to keep users from doing what they clearly want to.
I've been fortunate enough to always work in that kind of environment -- the company provides the hardware (usually), I maintain the software. So long as I'm productive, and don't cause problems, I just saved the company some IT time.
due to scalability requirements it will eventually turn into a wireless mesh system.
I would guess that'll happen because of the threat of censorship, and the relative cheapness, more than anything else. Fiber is pretty scalable.
As networks grow very large, they _must_ become increasingly decentralized and therefore increasingly resilient to attacks of the kind that net neutrality seeks to prevent.
Keep in mind, the Internet currently is very centralized in other ways as well.
For example: How do we find anything on the Internet? Google. How does eBay allow individuals to become sellers? By routing them through the corporate hub of, well, eBay. Who decides how to allocate DNS and IP? The IANA.
And yet, when you completely decentralize it, you open yourself up to spam. That is, if everything is defined by a consensus of peers, all someone has to do is control a large number of those peers, either by infecting real peers, or by fabricating them.
I don't have a good solution, and I have no idea what a good solution would look like, unless it went entirely peer-to-peer. But then we'd have to set about building a web of trust that spans the planet, and any one entity might still not have a good path to trust another entity.
Slashdot Mirror
What makes you think that they use the same Bluetooth stack on your Mac and on your iPhone?
What reason would I have to believe they'd throw out a perfectly good desktop Bluetooth stack, and rewrite it from scratch for the iPhone?
There was a time in this country, a long time ago, when reading wasn't just for fags, and neither was writing. People wrote books and movies. Movies that had stories, so you cared whose ass it was and why it was farting, and I believe that time can come again!
There go my mod points, but I had to...
Nothing is disabled, turned off, removed, commented out, locked etc
And yet, OS X on the desktop supports a bluetooth keyboard. The iPhone supposedly has a stripped-down OS X. Therefore, they must have stripped out USB HID support.
I've not seen bluetooth keyboard in ANY other phone I've ever owned,
See other posts in this thread. It's not just the keyboard, and other phones -- even the ones that aren't "smartphones" -- do far better than the iPhone in that respect.
This post, in particular? Of course not. Your use of the internet to view television shows? Yes, you probably should, if you wish to think of yourself as completely honest.
Why? What is the difference?
You choose to watch television, without paying for it,
In what way am I "not paying for" my own bandwidth? Read my sig.
If you mean I'm "not paying for" the content, I'm not paying to read your content here, either.
If you're assuming piracy, you clearly didn't read. The Daily Show is actually free, as are several other shows.
A mutual agreement might be arrived at, whereby you pay 5 or 10 percent extra for internet, without being charged for the unused television service.
Paying to whom?
My ISP? That's a blatant violation of network neutrality.
A media corporation? That's both a blatant violation of network neutrality, and anticompetitive towards the other shows (Revision 3, anyone?) which aren't available on network TV, but are on the Internet.
Discussing your digital activity on the interwebz with the post office is so absurd that no rational person would consider it as more than a joke.
So is discussing it with people from my local TV company. Demonstrate the difference, if you can.
Better yet, tell me why talking to my ISP about it is any different than talking to my phone company about who I want to call, or talking to my local grocery store about just how I intend to eat that apple.
Fuck no. You provide me a good or a service, which I pay for. Then I use that for precisely what I want. If you don't want me to use it for that, don't provide the service.
Obviously, you haven't read the REST of my posts in this discussion.
Not much, no. If you like, feel free to point to any you feel are particularly relevant.
Where is this "imaginary non theiving world" to which you refer?
The fact that you refer to us as theives means you're either suggesting that it's possible for someone not to be a thief -- and I'm suggesting that by your criteria, this is impractical and not desirable -- or you're suggesting that everyone is a thief, kind of like the Christian view that everyone is a sinner.
Again: If I were to avoid being a thief, does that mean I have to discuss this post with my ISP and the local post office, and get approval from both? If not, how is posting this different than the activities you've described as "stealing"?
Opensource is as good as my own property, for the taking, because I can do ANYTHING with it.
Clearly, you are quite ignorant of Open Source, as defined by the OSI.
Free software is something that I am permitted to use AS LONG AS I use it in a way approved of by the authors/owners.
Actually, that is true of most Open Source as well. It's also far from the FSF's definition of "free software".
There isn't a lot of software which could be classified "open source", but not "free software", by your definition. I could only mention sqlite, and DJB's stuff (qmail, djbdns, etc) -- these are public domain.
No, the difference is actually quite an obvious one: Open Source only requires that the source code be available. Before DJB's work was open source, it was very much open source, in that everyone could freely acquire the software. However, it was not free software -- once you have the source code, you are not allowed to redistribute it with your own patches.
If you wanted to improve qmail, for instance, you had two choices: Either beg DJB to include your change, or distribute your patch directly to users -- they would have to then download the qmail sources, apply your patch, and compile.
Free Software, on the other hand, only requires that you be able to both obtain the software, and fork it to distribute your own version. It doesn't place any other restrictions, it doesn't even require "free as in beer".
By the FSF's definition, the official distribution of sqlite is very much Free Software. By the OSI's definition, it is very much Open Source. And it is completely public domain, meaning there is absolutely no restriction placed on its use.
Your confusion probably stems from the fact that the FSF advocates the GPL, which does impose quite a lot of restrictions. The GPL's purpose is to ensure that this software must continue to be Free Software -- compare to sqlite, which could be forked into a proprietary version.
But there's nothing about the idea of Free Software itself which requires the GPL, or anything like it, and it's quite ignorant of you to suggest that.
"OOOHHH! Google USED TO allow us to use their resources, now they don't! OH! EVIL GOOGLE!"
Nice strawman. Can you point to a single post with that sentiment?
It's possible you can, but I've been reading this thread for awhile, and I've found nothing of the sort. I've found a few people who suggest that Google should have considered the implications, and certainly some people are disappointed, but I haven't actually seen anyone say that Google is evil or wrong for doing that.
If your ISP offers both internet and television, and you purchase only one, then continue to make use of both,
Ah, but I'm not. I'm watching the Daily Show on a website -- this may or may not have anything to do with actual television. For that matter, if I were to watch Ask a Ninja, which isn't on any television network that I know of, is it still "stealing"?
YOU ARE CONSCIOUSLY ATTEMPTING TO ABUSE THE SYSTEM.
What you call "abuse", I call "use".
Did you read the TOS?
Yes. And while it was a bit unreasonable about bandwidth expectations, there were no restrictions on actual content beyond "keep it legal".
Did you discuss your plan with the ISP?
I don't know -- do I discuss every time I plan to use Slashdot with my ISP?
Wouldn't it kind of defeat the fucking point of the Internet if I had to discuss every website I visited with my ISP?
Were you completely open and honest with the ISP about your intentions?
My ISP never asked. And that is the honest truth -- they never once asked what I intended to use my Internet service for. Being reasonably savvy themselves, they had to assume that "Internet" includes audiovisual media.
Frankly, I don't see why that's any of their business, any more than the phone company has any business asking who I intend to call. In fact, I'm allowed to telephone their competitors! Fancy that!
Or, did you just sort of sneak around and put this thing together to your own satisfaction, hoping that no one would catch on and block you?
No, I did it pretty openly, and legally, hoping that the ISP would realize that this is precisely what the Internet is for, and why it is valuable -- to promote the free exchange of information. I was also pretty much hoping the ISP would be in favor of net neutrality, and not openly "block" anyone, as that would be blatant censorship.
Now, I am a bit lucky in that respect -- my own ISP, while they do sell a television service, they realize that they are for the moment primarily an Internet Service Provider. They also realize that if I use too much bandwidth, they can purchase more to cover it, and charge me more if they have to -- for the moment, they openly support network neutrality.
But that is how it's supposed to work.
Tell me, in your imaginary non-thieving world, should I have discussed this post with my ISP, and with the local post office, just to make sure I wasn't "stealing" by not sending it via snail mail? And just how is anyone supposed to get anything done on the Internet, if the ISP is meant to micromanage it so much?
It is abuse to take advantage of some free service,
Is it, now? Well, you better stop posting on Slashdot, then...
thereby circumventing the telecom's charges.
I don't really see how.
Put another way: If my ISP offers both Internet and TV, but I choose to only purchase Internet, and watch the Daily Show free online, is that "abuse", because I'm not consuming exactly the service they want with exactly the fees I want?
It seems to me it's simply a more efficient use of a service I am paying for. After all, that app presumably does use the telecom's own Internet service, meaning you are paying for that -- if that's abuse, the telecom clearly isn't charging enough for bandwidth. I would suggest that, instead, they are clearly charging too much for text messages.
If it's not using the telecom's own Internet service, but rather some local wifi, I see even less "stealing" going on, given that you're taking place in a communication completely outside of what the telecom is providing. Should it be considered "stealing" that I prefer instant messaging to text messaging, and actually don't send text messages?
If you are going to steal Google's (or anyone else's) services, at least admit that you are thieving.
You first. Or have you just purchased something you saw in a Slashdot ad?
Clearly, Slashdot makes enough money from all of us visiting this website to keep it running. That, or someone is being very, very generous -- but if it really is losing money, that's not "stealing", that's their own failed business model, and they could shut it down.
Similarly, this service Google has provided was shut down, because they didn't anticipate that much traffic, and couldn't figure out how to monetize that. That is Google's failing in creating a service which could handle it, not our failing in using it.
That's a bit like suggesting that it's "stealing" for me to actually use the bandwidth I paid for, because I am slowing everyone else down. Well, no, I paid for my bandwidth -- if the ISP oversold it, and I'm now lagging someone else, that's really the ISP's fault, not mine. I'm just using what I paid for.
The need for driver's ed has nothing to do with the complexity of the car; rather it's about the motor skills, judgment, and understanding of the rules of the road.
It's not entirely divorced from the complexity of the car -- see oil changes, tire checks, etc.
But you're right. Training for a computer should be about logic skills, judgement, and understanding the rules of the Internet. As you said:
that has nothing to do with the complexity of the device; it's about a failure to understand the basic rules of effective and safe operation.
There are some basic rules for effective and safe operation of a computer, that many people with an IT department -- salespeople in general -- are willfully ignorant of. They almost seem to see it as somehow beneath them.
Proper design goes a long way to eliminating human error;
There is only so much you can do to avoid human error. Eventually, you have to train the humans.
Put another way, yes, an automatic transmission helps -- it's harder to stall the car by suddenly shifting to the wrong gear. However, if you suddenly shift to park while on the freeway, bad things happen. And at a very basic level, if you turn the wheel left, the car will go left.
It's possible you meant to go right, and there are very dangerous things to the left, but if you turn the wheel left, the car goes left.
Similarly, we could argue all day about how to build a better virus scanner, but ultimately, if a user tells their computer to run an executable file that they've downloaded from the Internet, then that's what the computer will do.
It's certainly possible to cripple the device. You can force people to ride the train instead, and they will have no ability to steer -- they'll literally be on rails. Similarly, you can remove the ability to run all but a predefined set of executables, or the ability to download any file types that aren't known to be safe, and otherwise lock the machine down to protect the user from themself.
But I don't think that basic level of training is asking too much, and I think it is ultimately more effective.
Too many people think computer training means understanding the OS and why things work like they do; rather than what do you need to do to effectively and safely use the machine.
I would say, some basic theory will help... I would lean heavily in favor of teaching more basic skills, rather than learning by rote.
For example, here are some basic things I teach:
- Try right-clicking on whatever you're trying to change.
- If it isn't obvious what a button does, hover over it.
- Look at the keyboard shortcuts next to that menu item.
- If you don't know where to start, explore. Wander through menus. Hover over buttons. Click on things -- if you're about to do something bad, you'll probably get an "are you sure?" message.
- When asking for help, phrase your question in terms of a goal, not a tool. The person you're asking might know of a better tool.
Compare to, "In order to add a header to the document, go to the Format menu, click 'Enable document headers' if it exists. If there is an item which says 'Disable document headers', leave it alone. Then go to the View menu, choose 'Display Mode', then 'Page Mode'. Then click the small rectangle at the top of the page and type your header."
That will be much quicker if a user simply wants to insert a header and be done with it. But it's a lot to remember, and I know I won't remember it by the next time I'm asked. It's much easier to teach people to look through the menus for something having to do with 'headers', or to check the documentation under the Help menu.
The problem is, teaching these general things, and their most important applications, can take a bit of time. This is why I compare it to driver's education -- I would expect it to take at least that long, for
I guess the biggest reason it is possible to have such a system, is because Linux itself is "free" as well as almost all software that runs under it.
Actually, that may have something to do with the current implementation. It has absolutely nothing to do with the potential of that to work.
After all, both Microsoft and Apple have centralized update systems -- they just haven't opened them up to third-party developers.
For that matter, Valve provides Steam, which provides reasonably secure (I assume) downloads through a centralized package manager, plus automatic updates -- and with some games, it can download only what's needed to launch a game, and download the rest on demand.
Yet almost none of the software distributed via Steam, nor Steam itself, is "free" in either sense of the word.
There is no technical reason you couldn't have a similar system, with less of the centralized control (and licensing cost), and more of the features we've talked about.
One reason why Linux is not making much headway on the desktop, is because there are a number of mutually incompatible flavors.
So use Ubuntu. This really isn't an issue anymore.
Consider, also: Windows now comes in some six or seven flavors, per release. Each has different features enabled or not. The "ultimate" versions cost significantly more than the others, if you really want everything.
Ubuntu comes in precisely one flavor, with everything. No need to worry about whether you can live with "home basic" or whether you need "home premium".
The user has to know which flavor is on the machine and then connect to the proper repository.
For most software, it will already be in the main repositories, which means you don't need to know what flavor you've got except "mine". Just open your distro's package manager, find the software you want, and click "install".
If you insist on getting it from third parties, yes, you do need to know what distro you're running. I'm really not sure why you think it's harder to remember "Ubuntu" than it is to remember "Windows" or "OS X", never mind "XP" or "Vista" or "Leopard".
Another reason is that hardware manufacturers are reluctant to put all their hardware eggs into the Linux basket because they have no official central place, such as Microsoft, to get software support.
How so? I'm really curious what support system Microsoft has for manufacturers...
End users who install Linux also don't get a single entity to squawk to, but have to find the right complaint Department for their particular distribution.
There are the Ubuntu Forums. There isn't a "department" -- you just post on the forums, and people come help.
Or, if you really want to compare apples to apples, buy a Dell with Ubuntu. You can then call Dell, just as you would with Windows.
Also, most people instinctively know that it is not cool to look a gift horse in the mouth. That makes people more reluctant to complain...
Have you ever actually done any open source work?
I can pretty much guarantee you are dead wrong here. People absolutely do complain, frequently, and are quick to blame anyone but themselves, no matter what the problem.
Linux is sort like one of these bicycles that comes in a compact box. It may be a great bicycle at a low price, but the user first has to get some tools and assemble the thing
Unless, again, you buy a Dell with Ubuntu preinstalled. Otherwise, you're not comparing apples to apples -- it's a much closer comparison to a user who bought a machine with Vista and now wants to install XP.
And that is usually more frustrating, by the way. It took far longer to find XP drivers for all the hardware in this laptop than it did to find Ubuntu drivers, since the thing was designed for Vista (and Ubuntu).
This is far from the only computer I've used on which Ubuntu was easier to install than Windows.
Since OSX and Linux share the underlying *NIX architecture, it should not be too difficult to implement the strengths of each on both of them.
That's a bit like saying, since men and women share the underlying Human architecture, it should not be too difficult to implement the strengths of each on both of them.
I don't know about you, but while I have nothing against hermaphrodites, I'm not particularly attracted to them.
So while I think it is possible to create something which has the most important strengths of each, you're always going to have someone who insists that a particular trait is a strength, where you see it as a weakness. For example, I really don't want to download binaries for some fifteen different platforms -- I'd rather just get the one x86_64 binary I need, and download others when I have another architecture that actually needs them.
Whereas you would rather waste the bandwidth and disk space, so you can easily transfer an application from one machine to another, without worrying about architecture. I'd much rather re-download it on the second machine.
Personally, I do not update most of my software, unless there is a really compelling reason to do so.
The problem is, if it's a security reason, you won't know that compelling reason until it's too late.
That is one advantage of the Debian and Ubuntu models -- you can stick to an old version until you need a new one.
For purposes of trying out new software, I usually create a special account for that purpose and then install all trial software in that account only, leaving the system entirely untouched.
I tend to trust the repository, if it's in there. I could as easily create a test account, install the software to the system, and try it out on the test account -- if I then want to remove all trace of it, I just have to uninstall with --purge (purge config files), and then nuke the test account.
For software that isn't available as a trusted package, I'm probably going to end up compiling from a source -- which I can do in a test account, and keep it completely isolated. If it is hardcoded to need root access and certain paths, I can chroot it and run it with "fakeroot", keeping it just as effectively sandboxed as if it was a virtual machine.
That's not as easy as it could be, and I'd rather see more isolation by default, out of the box, as well as making it more user-friendly. Most desktop accounts could easily allocate several UIDs per tab.
But at the very least, it's possible, even if it isn't often needed. Centralized updates, optionally automatic, aren't really available anywhere else.
Learning to drive has little to do with the complexity of the device- you can drive quite well and never have a clue about why the car works.
This is true.
You do need to know the basic principles, however. You don't need to know the physics of how hydroplaning works, but you do need to know to slow down when there's water on the road. You don't need to know exactly how shifting into reverse on the highway will tear your transmission to shreds -- actually, I don't have a clue, maybe the transmission is fine and the engine explodes -- you just need to know it is a Bad Idea.
But you haven't provided a case for why computer users shouldn't be required to attain at least a basic level of competency.
You seem to be hinting that a computer shouldn't require a user to know how it works, but really, it doesn't. You don't have to know why downloading programs is dangerous, any more than you have to know why unprotected sex is dangerous. Just some simple instructions: Use a condom, and don't try to install software unless you know what you're doing.
Even that "knowing what you're doing" isn't difficult -- for any given program, do some quick research online to find out whether you can trust it, favor open source, etc. If it's too much work to do that, ask a geek. You wouldn't make a major modification to your car without talking to a mechanic, would you?
A salesperson only needs to be able to understand how to produce the material they need to make sales; not what's under the keyboard.
But they do need to understand how to do their job effectively, which includes not screwing up what's under the keyboard.
I could, for example, slam on the brakes at every intersection, tires squealing, because it's the only way I know to stop. I could leave the key at the "start" position for a full two minutes because someone told me it was a good idea. I could leave my seatbelt unbuckled, because it really has nothing to do with getting where I'm going -- if anything, it's an annoyance, slowing me down when I need to reach for something. And I could be willfully ignorant of basic maintenance, never changing the oil.
Then, I would either be killed, or I would bring the car in for maintenance much more often, spending much more money on replacement parts that I destroyed through my own ignorance, and all the while complaining that it's somehow the manufacturer's fault -- if skidding is so bad, a car shouldn't let me do it! The ignition should just be an on-off switch -- or better yet, a button, like computers have! The oil should tell me it needs to be changed! That light is unintuitive -- it should tell me in plain english, like any good error message!
Understanding why something works is not necessary. But there is a basic level of competence that is generally required in every other industry. If someone behaved as I've described with a car, they'd be lucky to still have their license -- yet people behave this way every day with computers.
And keep in mind: There is an entire high school class called "Driver's Education". I see no reason there couldn't be a similar class on basic computer usage and maintenance. And I don't mean "this is a mouse" education -- everyone knows, no one holds up a steering wheel and says "This is a wheel, turn it left and you go left, turn it right and you go right." I mean all the stuff you really should know -- like when to accelerate instead of brake to avoid an accident, and why common sense with file extensions can be more effective than any antivirus.
I'm not asking that people be able to put a computer together from parts, or that they understand assembly language, or even that they be able to use a commandline. I'm asking that they take a little responsibility, and learn at least as much about computers as they do about other tools that are important to them.
Given the choice, I would much rather have a world in which some people attempt to keep their inventions secret -- but are vulnerable to reverse engineering -- than one in which I have to be careful what I invent, to be sure it's not already patented.
But the best compromise I see for patents would be to limit them to an incredibly short time -- perhaps one or two years. Plenty of time to get a first-mover's advantage, but not enough time to allow for patent trolls, or business models based solely on licensing patents. If you were to sue someone for infringing on your patent, they might just decide, "Oh well, we'll sell it next year."
I really can't say, though. This book makes a good point for removing patent protection entirely.
That argument used to be valid in the early days of computing when these things were very expensive.
Still is.
when 4GB of PC6400 DDR2 memory sell for less than $20
First: Can you get that in a netbook?
Second: It still has an effect on cache. That has been increasing, too, but not nearly as much.
Most modern computers come with disk drives of several hundred GB where storing 18 MB file or one of 30 MB makes no practical difference at all.
There are still a large number of people stuck on dialup, or with bandwidth caps. And we're talking about a single program. Two things to remember here:
First, there are quite literally thousands of packages in a typical Linux install, and that's what comes with it. OS X manages to hide most of this by including so much with the OS itself, including Ruby on Rails with every Leopard installation.
Second, while often increases in efficiency seem unnecessary, you eventually find some sudden innovation that relies on it. It's easiest to see this in gaming, where often things are "faked" -- for example, lighting used to be calculated ahead of time, and baked into the level. If a spinning fan cast a shadow, that's because the artists added the shadow when they added the fan.
Suddenly, with a certain amount of raw power, and the efficiency of Carmack's Reverse, we no longer have to fake those -- everything can be lit in real time. Perhaps the most obvious demonstration is the Doom 3 flashlight -- far from being a static, precalculated light, or the simple "make this area brighter" flashlight of things like Half-Life, it was a real light source the player could carry around, and it did cast shadows.
So, consider: If the initial download of a program is much smaller, due to many of the shared libraries and interpreters already being present, it's possible to get to a point where native applications can be installed nearly instantly. Provide a sandbox and a decent UI, and you have what Java tried to do with WebStart, and what AJAX is currently trying to do, but for non-web applications.
That is, no need to "install" unless you wish to mark some as being available offline, otherwise, you could see hundreds of programs in your program list, and the ones not already installed would be downloaded on first use.
To my knowledge, this has never caught on, but it also hasn't been tried on a large scale, and with the advantage of a package manager.
That is certainly not true for the vast majority of programs for OSX that I know about. Many programs do not even have or use a special installer program.
True. However, enough do, including important programs like photoshop, that this cannot be entirely dismissed.
Simply dragging the application icon into the trash is all that is necessary to completely get rid of ALL of a program.
Unless the program has installed a menu to the menu bar, like the built-in AirPort UI. For examples, see pre-Spaces attempts at virtual desktops, VPN clients, etc.
Or maybe it actually made a fundamental change to the system -- for example, a program to force any window fullscreen with a keystroke.
These are the kinds of programs for which installers and uninstallers make sense. Perhaps dragging to the trash is a good UI for uninstalling -- though I might debate that, in these cases -- but it should do more than simply move some files around.
In OSX a downloaded DMG file appears in the downloads folder. Also in Safari double-clicking on the filename in the downloads window immediately opens (mounts) it.
Right...
Dragging the program icon to the applications folder icon installs a program in the one place where all programs a user would generally require are present.
This is the step that isn't often done. What I usually find instead are people who ha
Linux is an operating system made by geeks for geeks. That sentiment has been expressed numerous times in this thread.
While that is true, the other side of it is false:
Linux is not for grandma, unless she has a Linux using grandson/daughter or other relative or friend.
With recent Linux distros, I've seen no reason Linux can't be for grandma. In fact, she'd probably be better off -- she only has a few things she needs to do, which she can do very well, like emailing her grandchildren, or receiving photos from them.
But, she's going to find it somewhat more difficult to install some random screensaver that someone emailed to her, which can only be a good thing.
Why should a program need to be installed on a computer? Why not simply copy a desired program from whatever source, online or a piece of plastic, to a well marked, established place in the computer?
The main problem I have with this are, programs are more than one piece. Common pieces -- libraries, interpreters, and the like -- may or may not already be installed. At the very least, you need a dependency system.
Mac programs can be "installed" by simply copying them from a source disk or download to the applications folder. Any initializing and setups are done when the new program is first run. Surely, it must be possible to do that sort of thing in Linux also.
Certainly. But it would be a bad idea.
First, consider: OS X is designed to run on two different processors, in 32-bit and 64-bit modes. Depending how much of the program is compiled, this means quite a lot must be included in a "universal binary", and most of it will sit there completely unused, wasting disk space, RAM, and bandwidth.
Case in point: VLC for OS X is available as Intel and PowerPC specific packages. These are a little over 18 megabytes each. It is also available as a Universal Binary, which is about 30 megabytes.
The only advantage to that waste is that you can copy an application from one machine to another without worrying about what their architectures are. But it's worth noting, when installing a package via Synaptic, you don't need to know your architecture either -- it will automatically download the package for you.
Second, consider the libraries. VLC on OS X is 18 megabytes, and that's the platform-specific version. On Ubuntu, it's about 3.6 megabytes. The rest is in libraries, which you may or may not have installed -- but you don't have to care, Synaptic will handle it for you.
This is pretty consistently true, throughout the OS.
Now, what's the advantage of the OS X method? It's probably easier to install from a CD, maybe. But from the Internet? You must be joking, unless something's improved. Last I checked, the typical method was:
- Download a DMG file.
- Find the downloaded file, and double-click.
- Find the newly-mounted drive.
- Click and drag the app to the Applications folder.
- Unmount the drive.
- Throw away the DMG.
Compare this with Windows:
- Download an EXE
- Find the downloaded file, and double-click.
- Follow instructions on screen.
- Throw away the EXE.
Now compare with Linux:
- Open Synaptic
- Find the program you want
- Check it
- click "apply"
Looking here, it looks as though the Linux version is just as easy as Windows, and quicker, but Windows and Linux both beat OS X. Also keep in mind, the Linux version will pull from the official repositories, meaning the program will be signed -- on Windows and OS X, the closest you'll get is a security prompt, which basically says "This was downloaded from the Internet!"
What's more, OS X doesn't always do things that way -- there's often a .mpkg involved, which is really an installer. Here, on Linux, I can go to Synaptic,
PNG was created because Gif was copyrighted
Sorry, should be because Gif was patented.
Right now if something is patented, you need to figure out another way to do the same thing. Sometimes the new method is even better than the original. THAT IS THE [IMPLIED] GOAL.
Actually, the implied goal is to provide an incentive for the original invention -- the theory being that no one will create anything without it being protected.
Without a patent, everyone would just use the 1st method and nobody would want to improve upon it.
Bullshit. The incentive is still the same: Build a better mousetrap.
Do you honestly think that Apple created the iPhone because existing physical keypads for cell phones were patented?
Sure, there are exceptions -- PNG was created because Gif was copyrighted. But for every example like that, I'm sure I can find ten examples of inventions that would have been created without patent law, and twenty of innovation that was held back significantly by patent law, going back to the first steam engines.
Linux is now, and has been for awhile, far easier to learn and use than Windows. OS X might be easier, but I would argue there are places Linux does better, even from the "clueless user" perspective.
The problem is, as this article identifies, it's not enough to be better. You have to balance a tightrope of being exactly like Windows, so Windows users don't have a learning curve, and yet so incredibly better than Windows that people are willing to ditch Windows (and all their Windows apps) to give it a try -- even if there was no learning curve, it can be a hard sell.
This is not Linux' fault. It's the essential problem that with a single OS dominating the market, it's nearly impossible for a third party to gain any traction -- which is why we were suing Microsoft for antitrust in the first place.
In other words: Linux has pretty much no hope, but it's not through any fault of Linux, or because anything else is easier to use.
They aren't competent because they have no incentive to be -- if they screw up their computers, that's IT's problem. If it suddenly became their problem, they might see things a little differently.
Just for fun, here's a car analogy: A car is a rather complex piece of machinery, and takes a lot of training -- typically an entire class of driver's education. While some people go on to master it and become stunt drivers, or simply improve their skills and get a truck license, etc, most are content to at least reach some level of competence.
But if you never bother to reach that much, you end up driving into a tree, or a telephone pole, or another person, and it's generally your fault.
Aside from the fact that cars are actually dangerous, and can cause bodily harm, I'll go with the fact that it is entirely the responsibility of the driver to be properly licensed and at least competent, and if they can't do that, it's entirely on their own head, both literally and financially.
Now, granted, many corporations don't like the idea of having to fire their best salesmen because said salesmen are morons about computers. But that only perpetuates the myth that it's somehow hard to attain some level of competence, and allows the salesmen to continue to see computer knowledge as somehow beneath them.
I hear you on the credentials...
I forgot my steam password, once, when setting up a new machine. The old one wasn't exactly functioning, and I'd simply copied all of the files off the hard drive, assuming that would be enough for anything I needed. Fortunately, I also had a disk image, which i convinced to run in a virtual machine, then downloaded and ran a small utility to find my password (from a very isolated virtual machine, so no way it could send it to anyone).
I'm sure it's possible to do it without actually running the Windows in question, but there was no ready-made script, and it was way more than I wanted to learn about Windows crypto.
But yes, after that, there is no way I'm forgetting that password.
Perhaps, but really, ads on seeks is going too far.
Do I see an ad when I fastforward a VCR tape? No.
How about previews. Are they skippable? Yes.
Do I see an ad when I fastforward or arbitrarily seek a DVD? No.
Are the ads skippable? On most of them, yes. In fact, most move the ads to a "special features" area.
Do I see an ad when opening a recorded show on a DVR, or seeking within it? No.
Are the ads skippable? Hell yes, that's half the point of a DVR in the first place!
Do I see an ad anywhere in an mkv I downloaded via BitTorrent? No.
Do I see ads when seeking or otherwise fastforwarding in Hulu? Yes.
Are they skippable? No.
So, even one ad on seek -- and considering it's unskippable, as are the ads at the beginning and every 15 minutes or so -- makes Hulu the single most obnoxious platform out of all of the above.
Again, I kind of want to like it. Digital delivery, on demand, using just the Web (and, ok, Flash)... But no HD, US-only, slow to release, and almost more ads than I'd get watching it on TV.
The only thing you can do is make the process so painful and bothersome that they decide it's not worth it.
And make it so long and time-consuming that they will be caught by coworkers/cameras and fired for doing something clearly against company policy.
Speak loudly about that company policy, by the way. After all, the job market sucks right now. Make the employee ask themselves whether that after-hours torrent is really worth their job.
Treating adults like adults is good in theory, but when you have 300+ people trying to beat "the man", you want to take away as much temptation as possible.
Has it occurred to you that if you treat them like adults in the first place, there will be less desire to try to beat "the man"?
The dirt-simple solution, which I never, ever seem to see implemented, is:
- Let employees be responsible for the well-being of their own machines
- Inform them that files are only safe on the fileservers, and make frequent incremental backups
- If they FUBAR the machine, re-image it at the touch of a button
- Apply security at the network level -- block outbound port 25 and anything else likely to get you in trouble, monitor for badness, etc. But this is an easy job when users are responsible for what they (or their computers) do.
Actually, I lied. I'm told this is the way it's done at Google -- at least the last part.
By the way, no real training needed. Just give them a quick presentation on how very alone they are, and let them know that if they don't feel they're computer-literate enough to do their own IT, the real IT department will be happy to remove root access, and give them a list of guidelines to follow (like Don't Install Software, Ever) -- which still takes fewer IT resources than locking them down against their will, and playing a cat and mouse game trying to keep users from doing what they clearly want to.
I've been fortunate enough to always work in that kind of environment -- the company provides the hardware (usually), I maintain the software. So long as I'm productive, and don't cause problems, I just saved the company some IT time.
due to scalability requirements it will eventually turn into a wireless mesh system.
I would guess that'll happen because of the threat of censorship, and the relative cheapness, more than anything else. Fiber is pretty scalable.
As networks grow very large, they _must_ become increasingly decentralized and therefore increasingly resilient to attacks of the kind that net neutrality seeks to prevent.
Keep in mind, the Internet currently is very centralized in other ways as well.
For example: How do we find anything on the Internet? Google. How does eBay allow individuals to become sellers? By routing them through the corporate hub of, well, eBay. Who decides how to allocate DNS and IP? The IANA.
And yet, when you completely decentralize it, you open yourself up to spam. That is, if everything is defined by a consensus of peers, all someone has to do is control a large number of those peers, either by infecting real peers, or by fabricating them.
I don't have a good solution, and I have no idea what a good solution would look like, unless it went entirely peer-to-peer. But then we'd have to set about building a web of trust that spans the planet, and any one entity might still not have a good path to trust another entity.
Fairfield.