You'd have to run as super user to install apps under Ubuntu. So, if the kids can figure out the password for that, you're correct, they could trash it quite easily. What the grandparent is pointing out is that, by default, you don't actually need the root pasword to install software in Ubuntu. You only need your own password since, by default, all users have sudoer rights.
Of course, this is very easy to fix (e.g. by removing the users from the relevent group, or using visudo).
as for your comment about editing a file to configure monitors, the latest Ubuntu beta due later this month (finally) has a GUI tool to edit and configure monitors, as well as making it easy to plug in a second screen. It's required very little configuring out of the box, and I'm quite impressed, as a Linux user since 2003. For a start, the beta came out ages ago; it's the release candidate and RTM that are due later this month. More importantly however, the new xrandr front-end is very, very, very far for being a complete multiple monitor solution. For example, there's no way of specify the relative position of your screens. If it guesses wrong? Tough. Back to the command line.
On the other hand, at least xrandr on the command line does occasionally work, which is more than can be said for the old Screens and Graphics utility. Which I have never seen do anything, ever, except trash xorg.conf to the point where X won't start.
It is time more people acknowledged that X11 is a piece of crap and is the reason why Linux doesn't have more market share. You still hear people trumpeting about X.org's backwards compatibility and how an xorg.conf that worked in 1980 will still work today. It's a windowing system; we don's need backward compatibility with several decades ago, we need to to work now!
Rights are properly configured on Linux by default. Your hypothetical kids in the library won't be able to touch anything system related, or anything not owned by the user. There is no configuration required to enforce this. Minor point, but the one thing you *don't* in this situation is to leave user rights management how Ubuntu configures it "by default"; since by default all users have sudoer rights, so can get admin access whether they want by typing in their own password!
Of course, this is very easy to fix; but trumpeting "There is no configuration required to enforce this!" when there actually is, however minor (removing the users from the 'admin' group, or using visudo), might be counterproductive.
Saying the results are wrong because they don't come out to the expected results doesn't help the code as an argument to convince someone who isn't already convinced of the results. The only point I was making was that, whichever way you're arguing, the results must still sum to one. If you think that it doesn't matter whether you switch, the result would be (50%, 50%); if you think it does, the result would be (33.3..%, 66.6..%). A result which doesn't sum to one will convince nobody of anything, since it is contradictory.
As far as looking over the code goes, I'm afraid that it's a quarter to three in the morning where I am, and I'm in no condition to peruse code in a language I'm not familiar with just before I finally collapse from Slashdot into bed;)
Car is behind 1 -I choose 1, -Host chooses 2 -Stay = won -Switch = lost -Host chooses 3 -Stay = won -Switch = lost -Host CAN'T choose 1 -I choose 2 -Host chooses 3 -Stay = lost -Switch = won -Host CAN'T choose 1 -Host CAN'T choose 2 -I choose 3 -Host chooses 2 -Stay = lost -Switch = won -Host CAN'T choose 1 -Host CAN'T choose 3
Count: 4 wins, 4 loses True. But each of those four wins and losses are not equally likely.
You have two (equally probable) possibilities under case "I choose 1" -- "Host chooses 2" and "Host chooses 3". The probability of "I choose 1" is a third, so each of the two possibilities have probability a sixth.
But each of the other cases -- I choose 2 or 3 -- only have one possibility each. So since the probability of choosing each of the cases is a third, each of the possibilities have probability a third.
Count up the probabilities for each outcome, and you get the actual result.
I'th thought of another argument to convince the ItDoesn'tMatterWhetherYouSwitch-ists.
When you first choose, the probability of there being a car behind the door you choose is 1/3. Now imagine that as soon as you've chosen, you stick your fingers in your ears, shut your eyes, ignore anything Monty Hall does, and sing very loudly, "I'm Sticking With This One, I'm Sticking With This One..." over and over again.
Clearly, if at any time during your singing you were to open the door, there is always a probability of 1/3 there being a car behind it. No matter what Monty Hall does.
Now consider that probabilities for mutually exclusive and exaustive options must sum to 1. Now, if you hadn't stuck your finger in your ear, you would have had an option to switch. Since the car is either behind your door or the door you could switch to, and the probability it's behind your door is 1/3 (probabilities don't change just because you've not stuck your fingers in your ears), the probability that it's behind the other door must be 2/3; so you should switch.
Won when switching %: 0.5
Won when wasn't switching %: 0.333333333333333 Well, I'm afraid you must have done something wrong, since whatever you think of the problem, those two probabilities should sum to one!
You're correct, of course, but you're preaching to the converted: the parent was making a point about the grandparent, who had started a truth table analysis with:
Choose door 1...
Choose door 1...
Choose door 2...
Choose door 3... for the first stage of the problem, when you first choose a door. Of course, since the probability of choosing each door initially is, wlog, a third, if you list "Choose door 1" twice since it has two possible consequences (door 1 had the car in this analysis) then you must apply a weighting of 1/2 to them. Once this is done, the method accordingly spits out the correct solution (of course, 2/3 probability of winning if you switch, 1/3 if you don't switch, so you should switch).
Choose door 1...
Choose door 1...
Choose door 2...
Choose door 3... When you initially choose a door, you pick each door with a one third probability. So in your analysis there, since you've listed "Choose door 1" twice, you need to apply a weighting of 1/2 to all the results you get for that. The sum of the probabilities at each stage must equal one, remember.
Once you do that, you get the same result the grandparent had: 2/3 probability of winning if you switch, 1/3 if you don't switch, so you should switch.
The result? No matter how competent Firefox developers are, IE works faster than Firefox or consumes lesser resources Oh, come on; don't be ridiculous. Opera is a hell of a lot faster than either Firefox or IE, and they don't have any access to secret APIs.
- The reason IE is quite fast is because it's just the bog-standard Windows HTML rendering component wrapped in a barebones toolbar.
- The reason Firefox is slow is because it wasn't designed to be superfast, it was designed to be easy to develop for and easy to port (the user interface is done in interpreted XML, for god's sake).
- The reason Opera is fast is because it's designed to be minimally resource intensive because the same codebase is used to develop the mobile versions of Opera (for smartphones, Wii, PocketPC etc.).
There are more than enough blatantly obvious reasons without having to resort to conspiracy theories.
Given that the leaked source code was from Windows 2000, and Windows 2000 is listed as being not affected by that vulnerability, I very much doubt it!
If I use Windows 98... Windows 2000, what should I do?
By default, these operating systems do not natively provide a version of the vulnerable component and are not affected. Just because A occurs chronologically after event B, doesn't mean that B caused A.
No-one's suggesting that this was a deliberate policy decision by HP; the suggestion is that it was a disgruntled worker or somesuch that did it deliberately for some unknown ends.
Unfortunately, you can't use the Group Policy Editor on Windows XP Home Edition. Who in their right mind would have XP Home edition installed on an HP ProLiant Server?
The main purpose for having floppies in servers is because Windows requires them to install mass storage drivers during installation on hardware such as RAID arrays and SATA drives IIRC, Only on Windows installers that use WinPE 1.x. That includes 2003 Server, but not 2008 Server (which uses WinPE 2.1). So hopefully now floppies should actually become a thing of the past.
Not, of course, that that in any way absolves MS -- it's still shocking that floppies were sometimes needed for a server OS released a mere half decade ago! Although at least you could always install remotely over a network using RIS or WDS and avoid the issue entirely, which is I suppose what most enterprises probably do anyway.
I doubt fonts that come only in one, fixed, non-antialiasable, microscopically tiny size would be a good idea for *reducing* eyestrain. I'd suggest that the precise font doesn't matter* as long as it's at a generous point size (at least 12), on a decently sized monitor, that you're not sitting too close too.
* I'm quite partial to Consolas myself, although freetype's bytecode hinting interpreter does make a bit of a hash of it at 12 point and below.
When Microsoft actually starts releasing code under a real open-source license, then we can start talking. A "real" open source license? Presumably, then, you think that licenses such as the Microsoft Reciprocal License and Microsoft Public License are fake open-source licenses. Don't you think you'd better call the Free Software Foundation and tell them they've been hoodwinked into certifying both these fake licenses as Free Software licenses?
Don't worry though, Ubuntu 8.04 will have "Bulletproof X" which will make things loads easier I believe. Better support for multiple monitors and video cards too I believe. Bulletproof X doesn't make getting the setup you want any easier; it just automatically drops you back to VESA drivers if whatever configuration you've just tried doesn't work. Unless you're happy to stick with 640*480 and 256 colours indefinitely, the only advantage is that you can try to fix whatever went wrong in gedit rather than pico. Kinda like Windows safe mode.
Regarding the better support for multiple monitors in Hardy, it comes in the form of an xrandr front end applet. Having had a rather bad time trying to get dual monitors set up in Gutsy, I tried this when the beta came out (using a vanilla auto-configured xorg.conf). Looked pretty good at first; it shows the two monitors side by side, showing the one I hadn't been using with a screen resolution set to 'off'. I set that to 1152x864, and pressed 'apply': Lo and behold, it turned on and showed my desktop at that resolution -- except that the monitor I had been using before was now set to 'off'. I used the applet to turn that monitor on, it did so -- and turned my secondary monitor back off. Not impressed.
The old 'Screens and Graphics' manager is still installed, only it's been moved over to the 'applications' menu for some reason. It still works identically to how it worked in Gutsy. By which I mean: not at all.
However, even though the graphical front end to xrandr doesn't work, xrandr itself *does*. After a lot of experimentation, most of which is due to the fact that Compiz Fusion fails miserably when you try to enable dual screens, I found that I am now able to get dual screens working (with metacity only), using "xrandr --output DVI-0 --left-of VGA-0".
Under Windows you'll find data in Program Files, C:\ (my sister uses that for example *shudder*) and in the registry. The programs which use the 'correct' folder are somewhat rare. I addressed all this in my post (first Windows paragraph, third bullet point). Yes, it certainly is possible to store data anywhere you like if you're logged in as admin (root in Linux), and most people usually were admin in XP and prior. The obvious answer is that if you actively choose to store your data somewhere other than your userfolder, in any operating system, then back that place up as well as your userfolder! You claim that it is rare for programs to use the correct folder *by default* is not true in my experience: very few programs go to the bother of writing their own file dialogues; most just use the Windows common dialogues, which default to My Documents unless the program explicitly specifies some other parameter (which few do).
90% of programs store all their config data in [the registry] and its impossible to backup properly. xcopy %UserProfile%\NTUser.dat D:\backup\/h ?
I dont really care about new features Vista has. I'm telling it how it happens in reality and not what 'should' happen. ...That you don't care about the way programs in recent versions of Windows store files and settings is a slightly strange statement, given that we are engaged in a discussion about the way programs in Windows and Linux store files and settings...
(Oh btw you forgot about ~ on linux.) Using $HOME is more reliable than ~, since tilde epansion is a feature of the specific shell you're using rather than Linux itself; and some shells (e.g. ASH) don't support it. $HOME will always work.
Doing that on Windows will copy a fraction of your settings, some of your files and a huge amount of crap.
Linux gets every single file of yours, every single setting and very little crap.
Slight difference dont you think? Well, yes, there would be... if your first two sentences were accurate. Sadly, however, repeating stating the same thing without evidence doesn't make it true.
Let's take a quick look at the differences between Windows' \Users and Linux's/home, for both files and settings. Files first.
Linux:
- Save As dialogues almost universally default to the current user's home directory, which itself can be placed anywhere (e.g. on a different partition). - The Home directory's location can be easily found programatically though the environment variable $HOME. - Saving documents in places in the filesystem other than the home directory is strongly discouraged, a policy enforced through permissions (root access is needed).
- However, users do usually have permissions to save files anywhere on any other mounted hard drives (adjustable using user/nouser flags in fstab).
Windows:
- Save As dialogues usually default to the relevent subfolder of the current user's userfolder depending on the application; such as the Documents subfolder. The userfolder can can be placed anywhere (e.g. on a different partition). - The user folder's location can be easily found programatically though the environment variable %UserProfile%, and various subfolders through variables such as %Documents%. - Saving documents on the hard drive other than the userfolder is discouraged, a policy enforced through permissions in more recent versions of Windows (admin priviliges needed), but not enforced by default in older versions.
- Users do usually have permissions to save files anywhere on any other mounted hard drives (adjustable using ACLs).
Looks pretty similar so far, doesn't it? A differently named environment variable here, an s/root/admin there, but they're broadly similar. Let's look at the picture for settings.
Linux:
- All user-specific settings stored in hidden subdirectories of the home folder.
- Settings can usually be copied to a different computer without problems.
Windows:
- User-specific settings stored in \appdata subfolder of the userfolder, and in the HKEY_CURRENT_USER section of the registry. The latter is stored as a hive in two files in the userfolder. In older versions of Windows, many programs tended to store settings in the program's folder in Program Files rather than the userfolder; in more recent versions, such writes are silently redirected to the appropriate subfolder of the userfolder through file and registry virtualisation.
- User-specific application data and registry hives are divided up into "Local" and "Roaming". Application settings stored in "Roaming" (including the registry hive NTUser.dat) must be copiable to a different computer without problems, which I assume is the basis for the Roaming user profiles in Active Directory. However, application settings stored in "Local" (including the registry hive Usrclass.dat) will not work when copied to a different computer, as the name suggests (e.g. this is where web browsers might store their cache, and other computer-specific crap).
So, a much larger difference between the OSes in the way they store settings than files. And Linux certainly is the simpler one, and has the definite advantage that you can just copy and paste the entire home folder when restoring without having to worry aboutg the difference between Local and Roaming.
But the evidence for your claims of huge differences in the way the OSes store user data is notably lacking.
But on Linux, backing everything up is far simpler than windows. cp -rf ~/backup does the job.
Compare it to Windows... Is xcopy %UserProfile% D:\backup/s/i/h really that much more effort to type than cp -rf ~/backup? I mean, I know the latter is a bit shorter, but it's not exactly "far simpler".
Actualy, changing window managers has been depreciated and unsupported for quite a while. What? Ubuntu, to pick one example, even lets you choose between window managers on one of the tabs in its 'appearance' dialogue box (and has done for the last two versions)! Admitted the choice is limited to only Metacity or Compix Fusion, but the capability is certainly there.
Persistent per-application window-manager settings are a big one for me. Some apps should always be visible on all desktops, some should always be on top, etc. That's a window manager thing. GNOME doesn't force you to use one particular window manager, as KDE does with KWin; so if Metacity (the default window manager) doesn't do what you want, use another, such as Compiz -- I'm pretty sure there's a Compiz Fusion plugin that does what you want.
Your argument would make sense if GNOME did not let you change the fact that there was a panel at the bottom. But it does. Gnome panels can be placed on one, more, or none arbitrary screen edges, with arbitrary content on each one. The point which you seem to have misseds that including the ability to configure and change the defaults does not remove an obligation for the designer to think hard about -- to design -- sensible and thoughtful defaults in the first place.
Slashdot Mirror
Of course, this is very easy to fix (e.g. by removing the users from the relevent group, or using visudo).
On the other hand, at least xrandr on the command line does occasionally work, which is more than can be said for the old Screens and Graphics utility. Which I have never seen do anything, ever, except trash xorg.conf to the point where X won't start.
It is time more people acknowledged that X11 is a piece of crap and is the reason why Linux doesn't have more market share. You still hear people trumpeting about X.org's backwards compatibility and how an xorg.conf that worked in 1980 will still work today. It's a windowing system; we don's need backward compatibility with several decades ago, we need to to work now!
Bollocks. http://en.wikipedia.org/wiki/Usage_share_of_desktop_operating_systems lists five different sources, and none of them show anything like the change you claim has occured in the last 18 months.
Of course, this is very easy to fix; but trumpeting "There is no configuration required to enforce this!" when there actually is, however minor (removing the users from the 'admin' group, or using visudo), might be counterproductive.
As far as looking over the code goes, I'm afraid that it's a quarter to three in the morning where I am, and I'm in no condition to peruse code in a language I'm not familiar with just before I finally collapse from Slashdot into bed
You have two (equally probable) possibilities under case "I choose 1" -- "Host chooses 2" and "Host chooses 3". The probability of "I choose 1" is a third, so each of the two possibilities have probability a sixth.
But each of the other cases -- I choose 2 or 3 -- only have one possibility each. So since the probability of choosing each of the cases is a third, each of the possibilities have probability a third.
Count up the probabilities for each outcome, and you get the actual result.
I'th thought of another argument to convince the ItDoesn'tMatterWhetherYouSwitch-ists.
When you first choose, the probability of there being a car behind the door you choose is 1/3. Now imagine that as soon as you've chosen, you stick your fingers in your ears, shut your eyes, ignore anything Monty Hall does, and sing very loudly, "I'm Sticking With This One, I'm Sticking With This One..." over and over again.
Clearly, if at any time during your singing you were to open the door, there is always a probability of 1/3 there being a car behind it. No matter what Monty Hall does.
Now consider that probabilities for mutually exclusive and exaustive options must sum to 1. Now, if you hadn't stuck your finger in your ear, you would have had an option to switch. Since the car is either behind your door or the door you could switch to, and the probability it's behind your door is 1/3 (probabilities don't change just because you've not stuck your fingers in your ears), the probability that it's behind the other door must be 2/3; so you should switch.
Clear?
Won when wasn't switching %: 0.333333333333333 Well, I'm afraid you must have done something wrong, since whatever you think of the problem, those two probabilities should sum to one!
In fact, %Won when switching should be 2/3.
Choose door 1...
Choose door 2...
Choose door 3... for the first stage of the problem, when you first choose a door. Of course, since the probability of choosing each door initially is, wlog, a third, if you list "Choose door 1" twice since it has two possible consequences (door 1 had the car in this analysis) then you must apply a weighting of 1/2 to them. Once this is done, the method accordingly spits out the correct solution (of course, 2/3 probability of winning if you switch, 1/3 if you don't switch, so you should switch).
Choose door 1...
Choose door 2...
Choose door 3... When you initially choose a door, you pick each door with a one third probability. So in your analysis there, since you've listed "Choose door 1" twice, you need to apply a weighting of 1/2 to all the results you get for that. The sum of the probabilities at each stage must equal one, remember.
Once you do that, you get the same result the grandparent had: 2/3 probability of winning if you switch, 1/3 if you don't switch, so you should switch.
- The reason IE is quite fast is because it's just the bog-standard Windows HTML rendering component wrapped in a barebones toolbar.
- The reason Firefox is slow is because it wasn't designed to be superfast, it was designed to be easy to develop for and easy to port (the user interface is done in interpreted XML, for god's sake).
- The reason Opera is fast is because it's designed to be minimally resource intensive because the same codebase is used to develop the mobile versions of Opera (for smartphones, Wii, PocketPC etc.).
There are more than enough blatantly obvious reasons without having to resort to conspiracy theories.
If I use Windows 98
By default, these operating systems do not natively provide a version of the vulnerable component and are not affected. Just because A occurs chronologically after event B, doesn't mean that B caused A.
No-one's suggesting that this was a deliberate policy decision by HP; the suggestion is that it was a disgruntled worker or somesuch that did it deliberately for some unknown ends.
Not, of course, that that in any way absolves MS -- it's still shocking that floppies were sometimes needed for a server OS released a mere half decade ago! Although at least you could always install remotely over a network using RIS or WDS and avoid the issue entirely, which is I suppose what most enterprises probably do anyway.
I doubt fonts that come only in one, fixed, non-antialiasable, microscopically tiny size would be a good idea for *reducing* eyestrain. I'd suggest that the precise font doesn't matter* as long as it's at a generous point size (at least 12), on a decently sized monitor, that you're not sitting too close too.
* I'm quite partial to Consolas myself, although freetype's bytecode hinting interpreter does make a bit of a hash of it at 12 point and below.
Regarding the better support for multiple monitors in Hardy, it comes in the form of an xrandr front end applet. Having had a rather bad time trying to get dual monitors set up in Gutsy, I tried this when the beta came out (using a vanilla auto-configured xorg.conf). Looked pretty good at first; it shows the two monitors side by side, showing the one I hadn't been using with a screen resolution set to 'off'. I set that to 1152x864, and pressed 'apply': Lo and behold, it turned on and showed my desktop at that resolution -- except that the monitor I had been using before was now set to 'off'. I used the applet to turn that monitor on, it did so -- and turned my secondary monitor back off. Not impressed.
The old 'Screens and Graphics' manager is still installed, only it's been moved over to the 'applications' menu for some reason. It still works identically to how it worked in Gutsy. By which I mean: not at all.
However, even though the graphical front end to xrandr doesn't work, xrandr itself *does*. After a lot of experimentation, most of which is due to the fact that Compiz Fusion fails miserably when you try to enable dual screens, I found that I am now able to get dual screens working (with metacity only), using "xrandr --output DVI-0 --left-of VGA-0".
So, yay!
Slight difference dont you think? Well, yes, there would be... if your first two sentences were accurate. Sadly, however, repeating stating the same thing without evidence doesn't make it true.
Let's take a quick look at the differences between Windows' \Users and Linux's
Linux:
- Save As dialogues almost universally default to the current user's home directory, which itself can be placed anywhere (e.g. on a different partition).
- The Home directory's location can be easily found programatically though the environment variable $HOME.
- Saving documents in places in the filesystem other than the home directory is strongly discouraged, a policy enforced through permissions (root access is needed).
- However, users do usually have permissions to save files anywhere on any other mounted hard drives (adjustable using user/nouser flags in fstab).
Windows:
- Save As dialogues usually default to the relevent subfolder of the current user's userfolder depending on the application; such as the Documents subfolder. The userfolder can can be placed anywhere (e.g. on a different partition).
- The user folder's location can be easily found programatically though the environment variable %UserProfile%, and various subfolders through variables such as %Documents%.
- Saving documents on the hard drive other than the userfolder is discouraged, a policy enforced through permissions in more recent versions of Windows (admin priviliges needed), but not enforced by default in older versions.
- Users do usually have permissions to save files anywhere on any other mounted hard drives (adjustable using ACLs).
Looks pretty similar so far, doesn't it? A differently named environment variable here, an s/root/admin there, but they're broadly similar. Let's look at the picture for settings.
Linux:
- All user-specific settings stored in hidden subdirectories of the home folder.
- Settings can usually be copied to a different computer without problems.
Windows:
- User-specific settings stored in \appdata subfolder of the userfolder, and in the HKEY_CURRENT_USER section of the registry. The latter is stored as a hive in two files in the userfolder. In older versions of Windows, many programs tended to store settings in the program's folder in Program Files rather than the userfolder; in more recent versions, such writes are silently redirected to the appropriate subfolder of the userfolder through file and registry virtualisation.
- User-specific application data and registry hives are divided up into "Local" and "Roaming". Application settings stored in "Roaming" (including the registry hive NTUser.dat) must be copiable to a different computer without problems, which I assume is the basis for the Roaming user profiles in Active Directory. However, application settings stored in "Local" (including the registry hive Usrclass.dat) will not work when copied to a different computer, as the name suggests (e.g. this is where web browsers might store their cache, and other computer-specific crap).
So, a much larger difference between the OSes in the way they store settings than files. And Linux certainly is the simpler one, and has the definite advantage that you can just copy and paste the entire home folder when restoring without having to worry aboutg the difference between Local and Roaming.
But the evidence for your claims of huge differences in the way the OSes store user data is notably lacking.
Compare it to Windows... Is xcopy %UserProfile% D:\backup
Your argument would make sense if GNOME did not let you change the fact that there was a panel at the bottom. But it does. Gnome panels can be placed on one, more, or none arbitrary screen edges, with arbitrary content on each one. The point which you seem to have misseds that including the ability to configure and change the defaults does not remove an obligation for the designer to think hard about -- to design -- sensible and thoughtful defaults in the first place.