MS and Apple change much more significant aspects of their UI almost regularly, so I call BS on your claim about domain lettering style being 'unlearnable'.
Users are uneducated and shouldn't know DNS even is. If it looks like chasebank.com in the address bar then they arrived at the right place. Another example of horrid disdain and elitism. Most people don't know what a domain is because techies like yourself have written them off. And of course, arguing for active babysitter-like controls certainly helps the job security of most people on this site. So we have a case of the fox guarding the henhouse, where a certain negative (even misanthropic) attitude is a cultivated kneejerk response out of self-interest.
Many new phishing tools poison the cache of the pc or host file so grandma wont know the difference since the address bar will exactly match the correct address. How do you fix this? This is what SSL certificates are for: YOU check that the domain you typed is spelled correctly, and the CA makes sure your browser is going where you want it to go. You only need to know that 1) the domain is crucial, 2) certificate warnings protect you from spoofed domains, 3) it is still up to you to judge if the site you want to communicate with is "good" or not.
So my question for you is: How often have you told people about this?
No one I have asked has ever received these very basic instructions for using the Web.
And adoption of GPLv3 will never reach the lions' share of projects.
Why? Because the 'split' the article talks about indicates that a significant set of coders on either side want different terms. And they are now both served by different versions of the GPL, instead of sticking with v2 and leaving a sizable chunk of hardware-freedom advocates without a common license. IOW, the addition of v3 has successfully expanded our options.
WRT to the owners, free hardware was all we had up until DRM and in fact the whole body of free software discussion starkly embodies that assumption. Now it can no longer be taken for granted and so freely re-programmable hardware must be asserted as a prerequisite for free software culture: The GPLv3 does just that.
Having it disabled by default is one saving grace, I will admit.
But we are certain to hear techies ramble on to their acquaintances about clicking that checkbox in preferences, instead of telling them to keep looking at the domain (the latter being the only way to truly safeguard yourself in the longrun anyway).
Any IT dept who pushes this is stupid, because they are leaking internal employee activities to an external site.
and some of those users are fundamentally un-educatable. Then you are part of the problem, and given your level of knowledge implied by your presence here, you share in the responsibility for that failure. I'd wager you like to push fancy, dramatic-sounding 'technologies' to people, instead of teaching fundamentals that save tons of resources (and freedom) down the line. It is because of disdainful people like you that we have millions of users operating computers who were never even shown the concepts of "file" and "pathname" yet never hear the end of malware removal and fascist-leaning Internet filtering schemes.
I will second that: Firefox 2 is far more stable than 1.5 (on Linux and OS X, don't know about Windows).
They have done a lot of good work and I consider FF very stable. But they have problems with CPU hogging from time to time. I just experienced a recurrence when Gnome components (and presumably GTK enhancements that affect Firefox) were installed onto my KDE system. Switching away from the default Firefox theme made the symptom go away.
What triggers the CPU hogging for me on Linux is the presence of enhanced GTK theming in the system.
For instance, after I installed Kubuntu and added Firefox, FF looked a little clunky but worked smoothly with no CPU hogging. But then I installed a Gnome app and bunch of Gnome stuff got installed along with it. From that point on, Firefox looked nicer but always used nearly 50% CPU after 20 minutes of brisk usage where closing web pages didn't help.
Workaround: Changing the Firefox theme to something other than 'default' (I eventually chose 'iFox Smooth') which also looked nice, but the idle CPU usage went below 1%. The default FF theme is more extensive than most, it seems, and the way you can tell is by comparing how the dropdown menus look. The FF default menus will look quite fancy with all of the Gnome stuff installed, while a user-installed theme will result in flat menus.
Because the people who put it in FF are acting like idiots by assuming average users are dumb and won't learn a couple of simple instructions. Hence, the idiots (i.e. many people in IT) don't even bother to suggest proper URL usage and instead concoct convoluted and invasive crap based on what a central authority considers socially acceptable for web browsing (and don't tell me the blacklist won't be expanded beyond suspected phishers-- you know it will).
The best thing they could do, IMO, is to render every URL in the address bar with the domain in red BOLD letters. Then, on first-use of Firefox the user gets a popup baloon coming out of the address bar advising them to always keep an eye on the domain field. This has the added benefit of making SSL certificates worthwhile, since certificates only work if you pay attention to the domain you are connecting to.
Teaching basic URL awareness also assumes that people who don't bother to spell correctly (or blithely click 'OK' on certificate warnings) will get what they deserve.
This feature could make me switch away from Firefox.
Yeah, which is why they encourage their users to enter the email addresses of all their friends and family.
Did you know that I have a facebook profile, without ever previously visiting their site? So now I have to sign up myself to find out what this profile says about me!
Even while extolling the virtues of SSL certificate authorities here, I am also newly aware of a potential problem... a betrayal of sorts. VeriSign has 60% of the CA market and doubtless hold most of the keys. They have also entered a new market that 'synergizes' with their existing one: VeriSign is now "lawful intercept" subcontractor. Under the expanded scope of CALEA, they spy on both voice and data communications for the FBI and NSA. In the case of purely USA-domestic links, they presumably act only on a court warrant, but where any hint of the international is involved (very easy to construe on the Internet) there isn't even a need for a warrant.
VeriSign are capitalizing on what seems to be their unique ability to stage MITM attacks undetected.
I don't mind recommending certs in response to a topic about the explosion of garden-variety Internet crime. But there is that ultimate question of privacy to contend with; of countering widespread government surveillance.
Community-organized CAs could perhaps gain trust through PGP signing, for what are otherwise normal SSL CA services. Major Linux distros could start CAs based on their considerable PGP signing histories, since they already use public key identification for their repositories.
Perhaps the OS that was developed over the Internet could end up saving it.
This is a much more insightful comment than appears on the surface.
Windows is the standard-bearer of personal computing culture. That being the set of expectations that allow people to buy their own computer, then add 3rd-party drivers and applications at will. (And yes, I know that Apple had that basic formula before MS.) Without the PC business model, what we are left with is a mainframe culture: Central authority decides what you can run.
The PC plays a critical role in electronic freedom, and Windows' poor engineering (among other factors) is putting it all in jeopardy. If the Desktop Linux people get a clue, we can step in and fill the gap. But it will have to involve more than selling the idea of glorified thin clients.
Certificates never had anything to do with reputation or authenticating the motives and business practices of the site operators. And it works very well in situations where users are actively paying attention to the URLs they visit.
Don't want to be discerning about URLs? Then you don't deserve Internet security.
What you suggest with PGP is admirable and interesting, but is something entirely different. If turned on in browsers by default, it would attract 100x more politicking than simple certification of domain names; Wikipedia would be a comparative stroll in the park.
The point of certificates is that you are supposed to be paying attention to the domain.com address you are connecting to in the first place, not blindly trusting wherever the links take you. But many people these days do not even know what the address bar is for: they type "www.yahoo.com" into MSN search field, and consider the address field to be gibberish because no one introduced them to it.
Certificates mean that you can tell the "etrade.com" you are trying to reach is real, with no hijacked address or MITM eavesdropping. That is all.
And that is all that's necessary IF you pay attention to the domains that appear in your address bar. Implicitly trusting everyone with a certificate isn't necessary; all the certificate means is that the DOMAIN is what it says.
Beyond that, you still have to be discerning in which addresses you decide to connect with.
I wouldn't think that a call for flawless code would be the solution, either, though I do appreciate your general point of striving for safer implementations.
There is very much talk about safer implementation every time an exploit is discussed (one could drown in it). But IMO, software interfaces are obsessed over at the complete expense of the user interface. Its imbalanced to the point of ignoring all of the low-hanging fruit that can be reaped with a simple 3-point introduction.
Even if this high standard of engineering that you speak of were achieved, clicking on surreptitiously-placed links is still a big privacy risk, particularly WRT email.
I find it odd that people have become so intensely aware of the need for firewalls and securing wireless: These sound very technical and techies drill it into users' heads constantly. Techies love to ramble on about such geek-mystique stuff, always pointing out their router is bigger/shinier and more power-packed than the person s/he is talking to (or lording over). My theory of the security dynamic is that URLs on the status bar, and SSL dialogs, do not have the same status-setting, impressing potential as bloviating about fancy hardware, port scanners and active software "guardians" that tax the system. We need to embarrass such people with elegant security advice.
Its Panther that I use regularly (the status bar thing is limited to Tiger maybe?) But thanks, that's really interesting.
You're right: Remotely-delivered scripting should never be allowed to change anything in the browser windowframe.
Again, we are back to essential UI policy. Interfaces are a serious business, whether they have metal pins and sockets, binary data structures and chip registers, or colorful icons and flowing text. But the UI is not being handled today in anything like a sober fashion.
Wondered how long it would be before someone brought 'pros' up.
Yes indeedy... I keep running into Cisco-certified people, for instance, why are mystified by my advice and not knowing what a certificate is. "Taken aback" is just how I feel in these situations; thanks for expressing it. IBM tech support people: not a clue.
Based on this sort of evidence alone, one could conclude that the IT industry has jumped the shark. The trade press is also implicated, since they don't mention basic security techniques whenever they report on a crisis; instead they play-up this "Internet-II" idea with nebulous references and awe.
All this makes me think about starting a "security n00bs" blog, where people can learn they can make a big improvement in the security of mundane tasks in ways that many experts don't let on about.
On every... single.... page. It does sound a little painful, put that way. But I use NoScript and it really isn't painful.
And there's even room for smoothing the way: Allow scripting automatically for SSL mode. You want to get your code running easily on other peoples' computers? Then be prepared to carry the burden of trusted connections. It's simple ethics.
Really, whenever I have these discussions and putting out ideas I start thinking W3C, IETF and the like are out to lunch. Seems they are only able to contemplate "solutions" anymore that reek of convoluted centralized control.
In addition to Sancho's reply, I'd also like to point out that Unix policies also tend to limit a virus' ability to spread itself. The argument about the effects on an isolated individual is a red-herring.
OK I'll bite:
It may not be hard to teach, but how many of them want to learn. Honestly, its pretty hard to say how receptive people are with almost no one making it an issue. You seem to want to write them off though.
Once you lower the bar, there's no raising it back up again. Apple made personal computers easier to use, not the MS copycats. So by your logic, Macs ought to be havens for viruses and trojans.
I don't think a digital-elitist attitude is going to add anything to tackling the issue.
Also, no applications I know would save files to disk with the Executable flag set. Not even binaries and scripts.
Trojans are hard for avg. people to guard against because the UI does not adhere to a trojan-resistance policy. It is seemingly little details that would make trojans stick out like a sore thumb... if only MS would do something with the UI.
Average people check for counterfeits every minute of every hour at the cash register. It is not the ultimate in authentication, but then most web fraud is not the ultimate in user deception.
That job isn't paying enough. Let me know when it gets past $50 bucks an hour. Until then I've got paying work and when I'm not doing that I'd like to spend time with the family. You are a Web Consumer, not a citizen then. You all want services in the form of shiny things you can click on and pay for to grease the way. Well the address and status bars are the most important factors in web security, and they aren't linked to paid consumer service industries or other notions of boutique consumerism.
The car analogy (as is often the case) doesn't fit. PC culture has been driven by pros and enthusiasts alike who can informally make recommendations, and a large chunk of the population cultivate relationships with their "PC guy" type friends and relatives. The best anyone can do in this situation of fraud proliferation is to educate people on the most basic and effective measures, esp. since the service-based model of security if failing. In a culture with a growing market of "Geek Squad" and "Nerdmobile" techs administering virus scanners and such, we find that criminals increasingly run amok.
Since the issue is web surfing (driving), your analogy could only be saved by asserting that what people need are paid chauffeurs to do their web surfing for them.
I'd like to expand on my first post by pointing out a few ways for fighting malware that are the most freedom-friendly, encouraging users to make responsible decisions. These depend on OS vendors employing sane UI policies:
Do not engage in filename-mangling! If a file is named "apicture.jpg.exe" then it MUST be displayed that way and must not undergo any automatic alteration (falsification) that, for instance, makes an executable appear as data.
Additionally, all executable files are shown with a red warning flag whenever that filename is displayed by the desktop, file manager or file dialog. This is important, as Windows will execute files ending in ".com" and this suffix is a part of most websites the user trusts; clicking on a "monster.com" file is natural so another indicator is necessary to cut down on trojans.
Make web site scripting purely an opt-in affair by default. This goes for anything else the html engine is used for, like chat clients.
No more "Open this file" option in download dialogs. Period. If the user cannot manage opening the file themselves from the regular UI, then hopefully they will get stuck and sign up for an introductory computer class.
No other distros support full-disk encryption that I am aware.
Debian is the only one to offer FDE at install time.
That said, the process for Ubuntu isn't hard. K/Ubuntu's graphical splash screen does interfere with the passphrase prompt, so you will have to keep removing "splash" from the/boot/grub/menu.lst 'kernel' line each time that/boot gets re-written (i.e. the kernel is updated).
Indeed, the only possible "success" from the whitelist idea is that the Internet morphs into television (shudder).
Q: Where has the Internet failed?
A: Its main proponents and enthusiasts ignored Drivers' Ed for the info-superhighway. They didn't teach people how to use web browser and email programs, didn't show how to read a URL and pay attention to the protocol and domain, nor instill the habit of mousing-over links to see where they go beforehand. Teaching people about the padlock symbol should have also included how to deal with SSL certificate alerts.
The result of this neglect is that people cannot recognize authenticity on the Internet, so the value of the Internet's "currency" is spoiling. Imagine if people weren't clued-in on how to authenticate a $20 bill: Over time only certain government and corporate entities would be trusted to handle currency to prevent spoiling by counterfeiters.
Our job as Internet cognoscenti is to keep correcting the people around you on the right way to use Web and email. Granted, this is not a cure-all given the other major factor here (Windows malware) but its several steps in the right direction. This stuff is not hard.
The alternative is an Internet-II re-worked around big corporations and government sites through a whitelist enforced by Trusted Computing remote attestation. Don't think they won't be opportunistic enough to scare the public into that corner.
Dmix has not solved the problem, though. Linux apps have switched to ALSA and they are STILL!! getting exclusive access to audio output without even trying. I have Kubuntu 7.10 (kernel 2.6.22) and checked that my audio device is indeed being run by ALSA, and the situation is little different than it ever was: there is perhaps 30% less audio blockage than before.
Your statement on ABIs assumes that they must be constantly fluid in order for progress to occur. But a responsible OS project would freeze them, planning for change during major revisions. So the problem is based on their development culture.
YOU still have a choice to use a proprietary driver, as long as vendor cares about updating it for newer kernels. How are HW vendors going to make sure the update takes place? It doesn't happen automatically now. Where has this design goal been all these years?
PM indeed is one of biggest problems for Linux currently. You say that only because it is the most intractable problem for techies like us. To a typical desktop/laptop user, all of the issues I mentioned equally resemble a big brick wall. Your perception that PM hasn't been a problem on desktops until recently is odd, since I have never been able to suspend a desktop on Linux and HDs never spin down... so where is the power savings?? Your answer indicates that it has become a problem in your mind only recently, pointing to the problem being a cultural (not technical) phenomenon within the Linux community.
Yep, there could be a "linux-compatible" trademark as a reward for companies which have drivers in the vanilla kernel. Yeah, they could even do the professional thing and have a test suite for 3rd-party drivers. Lets face it: No smooth 3rd-party driver accommodation means you do not meet essential expectations of the personal computing experience. I think a big underlying cause here is that (unlike Apple) the Linux people have adopted and are pushing the Unix server/thin client model which is actually anti-PC.
A real personal computer supports easy installation and management of 3rd-party components (HW, drivers and applications). Linux and about 90% of the GNU world are constructed to badly mimic the formerly-big names in Unix, however, which shun those use-cases.
Slashdot Mirror
So my question for you is: How often have you told people about this?
No one I have asked has ever received these very basic instructions for using the Web.
And adoption of GPLv3 will never reach the lions' share of projects.
Why? Because the 'split' the article talks about indicates that a significant set of coders on either side want different terms. And they are now both served by different versions of the GPL, instead of sticking with v2 and leaving a sizable chunk of hardware-freedom advocates without a common license. IOW, the addition of v3 has successfully expanded our options.
WRT to the owners, free hardware was all we had up until DRM and in fact the whole body of free software discussion starkly embodies that assumption. Now it can no longer be taken for granted and so freely re-programmable hardware must be asserted as a prerequisite for free software culture: The GPLv3 does just that.
But we are certain to hear techies ramble on to their acquaintances about clicking that checkbox in preferences, instead of telling them to keep looking at the domain (the latter being the only way to truly safeguard yourself in the longrun anyway).
Any IT dept who pushes this is stupid, because they are leaking internal employee activities to an external site. and some of those users are fundamentally un-educatable. Then you are part of the problem, and given your level of knowledge implied by your presence here, you share in the responsibility for that failure. I'd wager you like to push fancy, dramatic-sounding 'technologies' to people, instead of teaching fundamentals that save tons of resources (and freedom) down the line. It is because of disdainful people like you that we have millions of users operating computers who were never even shown the concepts of "file" and "pathname" yet never hear the end of malware removal and fascist-leaning Internet filtering schemes.
I will second that: Firefox 2 is far more stable than 1.5 (on Linux and OS X, don't know about Windows).
They have done a lot of good work and I consider FF very stable. But they have problems with CPU hogging from time to time. I just experienced a recurrence when Gnome components (and presumably GTK enhancements that affect Firefox) were installed onto my KDE system. Switching away from the default Firefox theme made the symptom go away.
It doesn't matter to me whether a users' privacy is invaded before or after they give in and sign up.
What triggers the CPU hogging for me on Linux is the presence of enhanced GTK theming in the system.
For instance, after I installed Kubuntu and added Firefox, FF looked a little clunky but worked smoothly with no CPU hogging. But then I installed a Gnome app and bunch of Gnome stuff got installed along with it. From that point on, Firefox looked nicer but always used nearly 50% CPU after 20 minutes of brisk usage where closing web pages didn't help.
Workaround: Changing the Firefox theme to something other than 'default' (I eventually chose 'iFox Smooth') which also looked nice, but the idle CPU usage went below 1%. The default FF theme is more extensive than most, it seems, and the way you can tell is by comparing how the dropdown menus look. The FF default menus will look quite fancy with all of the Gnome stuff installed, while a user-installed theme will result in flat menus.
Because the people who put it in FF are acting like idiots by assuming average users are dumb and won't learn a couple of simple instructions. Hence, the idiots (i.e. many people in IT) don't even bother to suggest proper URL usage and instead concoct convoluted and invasive crap based on what a central authority considers socially acceptable for web browsing (and don't tell me the blacklist won't be expanded beyond suspected phishers-- you know it will).
The best thing they could do, IMO, is to render every URL in the address bar with the domain in red BOLD letters. Then, on first-use of Firefox the user gets a popup baloon coming out of the address bar advising them to always keep an eye on the domain field. This has the added benefit of making SSL certificates worthwhile, since certificates only work if you pay attention to the domain you are connecting to.
Teaching basic URL awareness also assumes that people who don't bother to spell correctly (or blithely click 'OK' on certificate warnings) will get what they deserve.
This feature could make me switch away from Firefox.
Yeah, which is why they encourage their users to enter the email addresses of all their friends and family.
Did you know that I have a facebook profile, without ever previously visiting their site? So now I have to sign up myself to find out what this profile says about me!
It's like automated blackmail.
Let me admit something:
Even while extolling the virtues of SSL certificate authorities here, I am also newly aware of a potential problem... a betrayal of sorts. VeriSign has 60% of the CA market and doubtless hold most of the keys. They have also entered a new market that 'synergizes' with their existing one: VeriSign is now "lawful intercept" subcontractor. Under the expanded scope of CALEA, they spy on both voice and data communications for the FBI and NSA. In the case of purely USA-domestic links, they presumably act only on a court warrant, but where any hint of the international is involved (very easy to construe on the Internet) there isn't even a need for a warrant.
VeriSign are capitalizing on what seems to be their unique ability to stage MITM attacks undetected.
I don't mind recommending certs in response to a topic about the explosion of garden-variety Internet crime. But there is that ultimate question of privacy to contend with; of countering widespread government surveillance.
Community-organized CAs could perhaps gain trust through PGP signing, for what are otherwise normal SSL CA services. Major Linux distros could start CAs based on their considerable PGP signing histories, since they already use public key identification for their repositories.
Perhaps the OS that was developed over the Internet could end up saving it.
Just a thought...
This is a much more insightful comment than appears on the surface.
Windows is the standard-bearer of personal computing culture. That being the set of expectations that allow people to buy their own computer, then add 3rd-party drivers and applications at will. (And yes, I know that Apple had that basic formula before MS.) Without the PC business model, what we are left with is a mainframe culture: Central authority decides what you can run.
The PC plays a critical role in electronic freedom, and Windows' poor engineering (among other factors) is putting it all in jeopardy. If the Desktop Linux people get a clue, we can step in and fill the gap. But it will have to involve more than selling the idea of glorified thin clients.
Certificates never had anything to do with reputation or authenticating the motives and business practices of the site operators. And it works very well in situations where users are actively paying attention to the URLs they visit.
Don't want to be discerning about URLs? Then you don't deserve Internet security.
What you suggest with PGP is admirable and interesting, but is something entirely different. If turned on in browsers by default, it would attract 100x more politicking than simple certification of domain names; Wikipedia would be a comparative stroll in the park.
The point of certificates is that you are supposed to be paying attention to the domain.com address you are connecting to in the first place, not blindly trusting wherever the links take you. But many people these days do not even know what the address bar is for: they type "www.yahoo.com" into MSN search field, and consider the address field to be gibberish because no one introduced them to it.
Certificates mean that you can tell the "etrade.com" you are trying to reach is real, with no hijacked address or MITM eavesdropping. That is all.
And that is all that's necessary IF you pay attention to the domains that appear in your address bar. Implicitly trusting everyone with a certificate isn't necessary; all the certificate means is that the DOMAIN is what it says.
Beyond that, you still have to be discerning in which addresses you decide to connect with.
I wouldn't think that a call for flawless code would be the solution, either, though I do appreciate your general point of striving for safer implementations.
There is very much talk about safer implementation every time an exploit is discussed (one could drown in it). But IMO, software interfaces are obsessed over at the complete expense of the user interface. Its imbalanced to the point of ignoring all of the low-hanging fruit that can be reaped with a simple 3-point introduction.
Even if this high standard of engineering that you speak of were achieved, clicking on surreptitiously-placed links is still a big privacy risk, particularly WRT email.
I find it odd that people have become so intensely aware of the need for firewalls and securing wireless: These sound very technical and techies drill it into users' heads constantly. Techies love to ramble on about such geek-mystique stuff, always pointing out their router is bigger/shinier and more power-packed than the person s/he is talking to (or lording over). My theory of the security dynamic is that URLs on the status bar, and SSL dialogs, do not have the same status-setting, impressing potential as bloviating about fancy hardware, port scanners and active software "guardians" that tax the system. We need to embarrass such people with elegant security advice.
Its Panther that I use regularly (the status bar thing is limited to Tiger maybe?) But thanks, that's really interesting.
You're right: Remotely-delivered scripting should never be allowed to change anything in the browser windowframe.
Again, we are back to essential UI policy. Interfaces are a serious business, whether they have metal pins and sockets, binary data structures and chip registers, or colorful icons and flowing text. But the UI is not being handled today in anything like a sober fashion.
Wondered how long it would be before someone brought 'pros' up.
Yes indeedy... I keep running into Cisco-certified people, for instance, why are mystified by my advice and not knowing what a certificate is. "Taken aback" is just how I feel in these situations; thanks for expressing it. IBM tech support people: not a clue.
Based on this sort of evidence alone, one could conclude that the IT industry has jumped the shark. The trade press is also implicated, since they don't mention basic security techniques whenever they report on a crisis; instead they play-up this "Internet-II" idea with nebulous references and awe.
All this makes me think about starting a "security n00bs" blog, where people can learn they can make a big improvement in the security of mundane tasks in ways that many experts don't let on about.
And there's even room for smoothing the way: Allow scripting automatically for SSL mode. You want to get your code running easily on other peoples' computers? Then be prepared to carry the burden of trusted connections. It's simple ethics.
Really, whenever I have these discussions and putting out ideas I start thinking W3C, IETF and the like are out to lunch. Seems they are only able to contemplate "solutions" anymore that reek of convoluted centralized control.
In addition to Sancho's reply, I'd also like to point out that Unix policies also tend to limit a virus' ability to spread itself. The argument about the effects on an isolated individual is a red-herring.
I don't think a digital-elitist attitude is going to add anything to tackling the issue.
Also, no applications I know would save files to disk with the Executable flag set. Not even binaries and scripts.
Trojans are hard for avg. people to guard against because the UI does not adhere to a trojan-resistance policy. It is seemingly little details that would make trojans stick out like a sore thumb... if only MS would do something with the UI.
The car analogy (as is often the case) doesn't fit. PC culture has been driven by pros and enthusiasts alike who can informally make recommendations, and a large chunk of the population cultivate relationships with their "PC guy" type friends and relatives. The best anyone can do in this situation of fraud proliferation is to educate people on the most basic and effective measures, esp. since the service-based model of security if failing. In a culture with a growing market of "Geek Squad" and "Nerdmobile" techs administering virus scanners and such, we find that criminals increasingly run amok.
Since the issue is web surfing (driving), your analogy could only be saved by asserting that what people need are paid chauffeurs to do their web surfing for them.
I'd like to expand on my first post by pointing out a few ways for fighting malware that are the most freedom-friendly, encouraging users to make responsible decisions. These depend on OS vendors employing sane UI policies:
Do not engage in filename-mangling! If a file is named "apicture.jpg.exe" then it MUST be displayed that way and must not undergo any automatic alteration (falsification) that, for instance, makes an executable appear as data.
Additionally, all executable files are shown with a red warning flag whenever that filename is displayed by the desktop, file manager or file dialog. This is important, as Windows will execute files ending in ".com" and this suffix is a part of most websites the user trusts; clicking on a "monster.com" file is natural so another indicator is necessary to cut down on trojans.
Make web site scripting purely an opt-in affair by default. This goes for anything else the html engine is used for, like chat clients.
No more "Open this file" option in download dialogs. Period. If the user cannot manage opening the file themselves from the regular UI, then hopefully they will get stuck and sign up for an introductory computer class.
You should only need to remove "splash" once if you do it from the "# defoptions" line in menu.lst!
Leaving "quiet" in the options seems to actually help the passphrase prompt, so I leave it in.
No other distros support full-disk encryption that I am aware.
/boot/grub/menu.lst 'kernel' line each time that /boot gets re-written (i.e. the kernel is updated).
Debian is the only one to offer FDE at install time.
That said, the process for Ubuntu isn't hard. K/Ubuntu's graphical splash screen does interfere with the passphrase prompt, so you will have to keep removing "splash" from the
Indeed, the only possible "success" from the whitelist idea is that the Internet morphs into television (shudder).
Q: Where has the Internet failed?
A: Its main proponents and enthusiasts ignored Drivers' Ed for the info-superhighway. They didn't teach people how to use web browser and email programs, didn't show how to read a URL and pay attention to the protocol and domain, nor instill the habit of mousing-over links to see where they go beforehand. Teaching people about the padlock symbol should have also included how to deal with SSL certificate alerts.
The result of this neglect is that people cannot recognize authenticity on the Internet, so the value of the Internet's "currency" is spoiling. Imagine if people weren't clued-in on how to authenticate a $20 bill: Over time only certain government and corporate entities would be trusted to handle currency to prevent spoiling by counterfeiters.
Our job as Internet cognoscenti is to keep correcting the people around you on the right way to use Web and email. Granted, this is not a cure-all given the other major factor here (Windows malware) but its several steps in the right direction. This stuff is not hard.
The alternative is an Internet-II re-worked around big corporations and government sites through a whitelist enforced by Trusted Computing remote attestation. Don't think they won't be opportunistic enough to scare the public into that corner.
Your statement on ABIs assumes that they must be constantly fluid in order for progress to occur. But a responsible OS project would freeze them, planning for change during major revisions. So the problem is based on their development culture. YOU still have a choice to use a proprietary driver, as long as vendor cares about updating it for newer kernels. How are HW vendors going to make sure the update takes place? It doesn't happen automatically now. Where has this design goal been all these years? PM indeed is one of biggest problems for Linux currently. You say that only because it is the most intractable problem for techies like us. To a typical desktop/laptop user, all of the issues I mentioned equally resemble a big brick wall. Your perception that PM hasn't been a problem on desktops until recently is odd, since I have never been able to suspend a desktop on Linux and HDs never spin down... so where is the power savings?? Your answer indicates that it has become a problem in your mind only recently, pointing to the problem being a cultural (not technical) phenomenon within the Linux community. Yep, there could be a "linux-compatible" trademark as a reward for companies which have drivers in the vanilla kernel. Yeah, they could even do the professional thing and have a test suite for 3rd-party drivers. Lets face it: No smooth 3rd-party driver accommodation means you do not meet essential expectations of the personal computing experience. I think a big underlying cause here is that (unlike Apple) the Linux people have adopted and are pushing the Unix server/thin client model which is actually anti-PC.
A real personal computer supports easy installation and management of 3rd-party components (HW, drivers and applications). Linux and about 90% of the GNU world are constructed to badly mimic the formerly-big names in Unix, however, which shun those use-cases.