If that's how people on this site define the word "sane", then yes I am.
If your reaction to being called a fanboy is to assume I'm that much biased in the other direction, you're making my point for me.
No, I do think you're mostly sane. But "sane" doesn't mean "agrees with me", and the same goes the other way. I do not have to be insane to disagree with you.
So the guy who yells "FUCKING KILL GOOGLE" (so you say) is so subtle that he'd try to auction the patents to some other firm?
Microsoft has been deliberately "subtle" in the past -- see MS Java, Dr DOS and Windows 3.1, "It ain't done till Lotus won't run", etc.
Again, the question here is: Is it so unbelievable that someone would both overtly rail at his enemies, and covertly attempt to undermine them? Ballmer wouldn't be the first to do this.
And yet so short-sighted that he didn't foresee the possibility that a firm that wants to defend Linux would buy the patents in the same auction?
Actually, that's not what I'm suggesting. What I'm suggesting is more paranoid -- that said firm could easily be a front, or could be tempted with sufficient amounts of money, so that it looks as though they've sold the patents to a Linux-friendly company.
It's much more likely that Microsoft auctioned the patents for some wholly un-related reason, and didn't care who purchased them.
Oh, I'll grant that, easily.
But keep in mind, this is the same Microsoft who sent astroturfers to a Linux convention, saying things like "It's all over, the suits are taking over," back when IBM started showing up at these conventions. This is the same Microsoft who essentially stole the GUI concept wholesale from a Macintosh prototype -- keep in mind, Apple did actually license it from Xerox. And this is the same Microsoft who continues to fund SCO.
It is not paranoid to expect a company which has been underhanded and "subtle" in the past to continue to do so in the future. It may not be the most likely possibility, but it's hardly "insane".
"Taskbar thumbnail previews" are something I remember having in Compiz long before Win7, and I don't remember seeing them on OS X, either. Want to be specific about what Mac things they're ripping off?
That is because Linux doesn't really have a drop in replacement for AD+Exchange+Outlook+Sharepoint+GPO.
The problem is firstly that people think they actually need all that -- do people actually use SharePoint? Anecdotes point to maybe.
But Exchange + Outlook? Plenty of groupware. AD/GPO? Samba is working on it, if it isn't released already, and Unix-centric solutions have existed for decades.
If I understand your argument, though, they're trying to concentrate on fighting Apple?
I think you just don't know what "public domain" means. It means there is no owner, no one can exert property rights in it, there's no legal way to exclude anyone from using it.
From the Wikipedia article:
The public domain can be defined in contrast to several forms of intellectual property; the public domain in contrast to copyrighted works is different from the public domain in contrast to trademarks or patented works.
Here's the problem: You were talking specifically about documentation, which would imply "public domain" as contrast to copyrighted works. And patenting an invention is specifically designed to ensure that all of the work behind it -- designs, schematics, papers, everything -- is released into the public domain.
I don't know that there is a way to release a patentable concept into the public domain, other than ensuring that it's patented by someone who won't abuse the patent.
I also think you don't really understand software patents, especially how Microsoft (and its ilk) use them to interfere with others competing with them, or even with interoperating with them.
Actually, I've been following the whole TomTom VFAT thing, not to mention the current issue with h.264 patents -- where essentially neither Apple nor Google will implement patent-less video formats, yet no other browser makers care to spend the money on licenses for h.264.
Covenants not to sue are transient. You'd have to sue them to enforce it, which makes it useless to small operations.
Or you simply violate their patent and dare them to break that covenant. When they come sue you, you have legal grounds to call them on it -- not to mention loads of bad press for them.
What's more a shame to me is just how much this seems to confirm the direction the OLPC has set, ever since they decided they had to support Windows, and then decided they were dropping Sugar.
Basically, they're moving it less in the direction of being a ridiculously cheap education appliance, and more in the direction of being just a sort of cheap netbook PC.
Not that there's anything wrong with that, but the netbook market is pretty saturated, and they had a number of things that used to set them apart. By moving in this direction, instead of, say, cheaper, more rugged, more reliable, and improving Sugar, it looks like they're going to be competing head-on with the major OEMs very soon, if they aren't already.
I mean... I don't know, maybe I'm jaded, but I haven't seen an encouraging OLPC story in awhile now. There just doesn't seem to be anything interesting left.
Well, it's not on all pages, nor is it fully featured, but they do have a tech demo.
Main missing features are fullscreen video and Theora.
I'm not sure there's currently any sort of spec for making an HTML5 video go fullscreen. I can see it being restricted somehow -- to a native control, or maybe in the same way pop-ups are, because otherwise, any script could throw up a fullscreen porn ad -- but it seems like a kind of glaring omission to have no support for it other than calling "save as" on the video (which isn't always easy).
Chances are, they'll never add Theora, which means if they switch over, Firefox may never support it -- or at least, I get the impression that there are a few idealistically blinded people who want to remove all hint of a possibility of a way to play HTML5 h.264 in Firefox, not even as a third-party plugin.
Interesting little fact, though: The video tag does support data urls. That means a clip can be embedded into the HTML page itself, base64-encoded. I have absolutely no idea why this is useful, but I thought it was cool.
It's not paranoid when they actually are out to get you.
And really -- I know you're often pro-Microsoft, borderline fanboy, but even you should be able to see that Steve "FUCKING KILL GOOGLE" Ballmer would love a chance to cut Linux off at the knees.
take all the documentation of the patent, and put it in the public domain.
That's what a patent does in the first place!
It just provides monopoly protection for actually using said patent.
In fact, this was the whole point of patents. Say I invented lemonade. Without a patent, I'd keep it my secret family recipe for generations, and anyone who wanted to make lemonade would have to reverse engineer it -- but if someone did, I wouldn't be able to say much.
With a patent, I would publish the documentation ("It's just sugar, water, and lemon juice.") and then only I can make lemonade. Or I can license the recipe to others -- it's not like they don't know how to make it now, it's that they legally can't unless I let them.
If they were really serious about merely protecting the community, they'd give up the patent control entirely.
I'm not sure it's legally possible to do that, is the problem. Moreover, having a process patented provides clear documentation that you patented it first, thus putting the burden on anyone's infringing patent to prove that they invented it before you did.
No, my big problems with this are not that I think the result is bad, but because I think it should be unnecessary -- I highly doubt Microsoft has any stunning invention that Linux "stole" for which prior art doesn't exist a thousandfold, and even if there were, I'm not sure software patents should exist at all.
But if I'm going to accept that they exist, and that someone has to hold them, I'd much rather that someone not be Microsoft, no matter how legally binding their "covenant not to sue" is.
Though it would be pretty slimy if this new organization doesn't have some sort of "covenant not to sue." Maybe that's the motive? Blech, now I have to go wash the evil from my brain...
Well, it made no damned sense, as there was no authentication that I know of (WIDE open, so long as you're on Windows), and ethernet worked just fine. For what it's worth, udhcpc is a very minimalist DHCP client -- designed for embedded systems, as I understand it.
So it seems like this is either a bug in two separate DHCP clients, or in the server itself, but not necessarily a reflection of how much better (or worse) one client is than another. For example, it could very well be that dhclient is behaving entirely within the spec, but confusing the DHCP server (whatever it is).
Also, there was nothing about "persistence", really -- udhcpc got a lease pretty much instantly. Like, less than a second to associate with the LAN, run DHCP, and get a lease, whereas dhclient would time out after more like ten, twenty, thirty seconds.
Legitimate use is easy and non-annoying. In other words, if you purchase the product on iTunes or Steam or any service implementing this protocol, you can use the product where ever and whenever you want, on whatever devices.
So long as you transfer it every fucking time. No thanks.
To give another example: Steam is nice because if my original Windows installation is completely annihilated, I can just re-download games onto a new machine. This is worse than Steam, and most people don't even like Steam.
There's no "Kindle 1984" scenario looming
Actually, that's assuming they don't do that anyway.
no need to buy special "DRM-ready HDMI" cables.
Actually, that's highly likely to be needed anyway, as HDCP isn't going away just because the content is protected in some other way than Blu-Ray's AACS or BD+.
Infringing use is easier to prove/disprove. This assumes that the files are watermarked with your account identifier and digitally signed in some fashion (not difficult -- iTunes does something like this already).
You are aware that it's also trivial to change this watermark in iTunes? It's no more "proof" than an IP address.
OTOH, though, if the FBI seizes your hard drive, and all the files on it are properly watermarked and digitally signed, they have no case.
Which sets a really bad precedent for those of us who prefer unencumbered media. Do I have to dig up the actual CDs I ripped my FLACs from to prove my innocence?
I shouldn't have to prove my innocence. It's "innocent before proven guilty", right?
Obviously, some hackers will find a way to crack the file format pretty much the day it is announced, and the BitTorrents will continue. That's OK; a little piracy never hurt anybody.
I'm glad we agree. So why do we need it?
The idea is to protect regular people -- folks who just want to buy or rent a song or movie and play it without a big hassle and without giving control of their computers over to some other company
How the fuck does this "protect" them more than "no DRM at all"?
and to help the big publishers feel comfortable about moving towards digital distribution.
If that's its entire purpose, then it can rot in hell.
Do you think the average college helpdesk is prepared to answer random Linux questions?
At my community college, they were prepared for the specific Linux question I had. Unfortunately, the only answer they had was "We don't know." They were glad to have a solution once I solved it, though.
Due to a few bad decisions in college and the economy imploding at work, I am now going to Indian Hills Community College, trying to at least keep it together for a semester or two before I invest in a four-year college again.
I have no idea whether it's required, but there is a Unix/Linux class. But then, there's also a Visual Basic class.
Aside from the fact that a few things (VB?) will require Windows -- though they at least have the decency to have an MSDNAA license, meaning free copies of Windows and nearly everything needed -- there's also the fact that Macs sort of get a passing reference saying "We hope it works, but we can't support you," and Linux gets no mention.
The wireless fails out of the box with Ubuntu, yet works with Windows. Talking to the help desk, they basically said "We don't support that, we don't know much, but our vendor assures us that the problem is not with our equipment, but with Linux."
To get it working again, I had to switch DHCP clients. Neither dhclient nor dhcpcd worked, but udhcpc did. It's worth mentioning, this is not a common problem -- I used NetworkManager's point and click interface pretty much everywhere, and it worked pretty much everywhere, from hotel rooms to hospitals to crappy little Linksys routers -- it even worked if I plugged into ethernet in school -- everywhere except the school wireless.
The conclusion to this story? I mailed the helpdesk again with my findings, and with the little script I wrote to disable NetworkManager, bring up wlan0 manually, and run udhcpc. They seemed very glad to have a solution.
So, I'm not really sure what to make of it. On the one hand, it was obviously a priority, and I was pretty much left to fend for myself. On the other hand, no one actually has a problem with me using Linux, most of the time.
I realize that doesn't answer your questions about printing or VPNs -- I haven't had to do either yet. Printing, I've only done from lab computers (all Windows, naturally), and they don't require a VPN, though my personal VPN works fine from the school wireless. Their website is an abomination, but it mostly works fine in Chrome, with only one place so far which requires Firefox, and I haven't had to use IE yet, except on lab computers.
Just for fun, another anecdote: Iowa State University, when I was there, had a lab full of top-of-the-line Linux workstations. In the classes I was taking, they were used mainly to run rdesktop, which seems profoundly retarded, but I never had a problem due to running Linux or OS X. This was around 2005-2006.
Erm... seems like you might've responded to the wrong comment, or blatantly misread mine? I'm the atheist, responding to "God-boy", who never mentioned that particular argument (nor did I).
I admire your goals, but being a dick about it isn't going to change minds.
Because sometimes it works better? Even in Unix world, there's Emacs as a testament
...to the fact that some people would much rather use vim.
Moreover, Emacs isn't monolithic. Aside from the fact that it's just a set of Lisp scripts for a much simpler editor (TECO), there's the fact that most of the more interesting things you might want to do with it are also going to call some external program at some point.
One obvious example: Version control. I very much doubt anyone's actually written bindings for Emacs for all the various version control systems it supports. Specifically, I know it supports Git, and I know Git has no simple library or API beyond its commandline interface -- in other words, if you use Git with Emacs, that means Emacs is calling git commands.
Small programs, doing one thing well. One unified interface acting as a frontend for these small programs. Sounds exactly like the Unix philosophy to me -- indeed, the best of both worlds.
I cn not install software on those systems. I can not add anything on those systems. Further not all logins are weblogins and some that are only work on very locked down IE machines
Look at shvytejima's reply -- that is, PasswordHasher is available both as an extension and as a standalone HTML page -- thus, easy to host yourself, if you don't trust someone else to. Visit with your locked down IE, type your password and the base domain, copy/paste the hash into the page (or program) you needed.
So by increasing the security on their side by doing all the things that are possible, they actually have decreased it in the end.
Which is exactly why the important thing here is to allow tighter security, not require it. The hardware key is a great example -- I'd stick it on my keychain and be happy to have it. Most people would probably physically attach it to the machine it's used from. Make it optional and we can both be happy.
The right way to solve the problem users have with their passwords is educating those users, not writing off a service as "insecure" because the people using it are. After all, social engineering is always possible.
So if after say 20 years of intensive computer usage by non-geeks what we do now does not work, I would suggest we should start looking for something else.
Depends what you mean by "what we do now".
Everything you've just mentioned is fairly stupid design, where the systems are entirely outside your control as a user, and you are "forced" to have a level of security they will accept, leading to you doing things like writing passwords down. About all I could suggest from your side is, that text file should be encrypted with a password you can remember.
But we weren't talking about that. We were talking about webmail -- and most webmail is either smart enough to use OpenID, or not smart enough to do more than a straight username/password. In either case, it's possible to be both reasonably secure (seemingly-random password) and convenient (hashing, save password, etc).
It's certainly possible to be at least secure enough that no one's going to pull it directly off a hard drive or a sticky note, nor guess it based on your pet's name, favorite movie, dictionary attack, etc.
So I think I'm justified in not feeling a lot of sympathy for someone who had their webmail account cracked by guessing a password. I can sort of see someone not knowing to force https, at least when using wireless (hint: https://mail.google.com/, and Gmail will remember it once you login), but it's not that difficult to deal with a secure password, especially in that context.
I prefer a browser I can easily extend myself to one that has "everything" built in, where "everything" might not include something Opera hasn't thought of.
Then you have users who don't even know "Open in new tab exists", so they keep clicking on the left mouse button, thereby using only one-two tabs.
That's why performance of a single tab was also measured.
Other users (like me for example) got used to clicking the middle mouse button so all new pages are opened in new tab.
So opening eight tabs is still a useful measure of how quickly it can open a new tab.
Scientifically, it would be relatively easy to write a software that would
So write it.
I think your definition of "relatively" is a bit odd, considering we're talking about the difference between that and just telling the browser to open eight tabs, and we're talking about a tech reporter, not necessarily a programmer, doing the measurement.
Yeah, I'm on Jaunty 64-bit. And I'm using the Chromium nightly PPA -- Google for it -- means I have to download maybe 10-20 megs per day, but it also means I don't have to compile anything.
I am not allowed to save them. I must memorize them.
Nonsense. While Chrome doesn't seem to have this yet, Firefox and Konqueror come with encrypted password stores out of the box.
That is, you enter one master password, and it then remembers all your passwords for you.
I also have a friend who wrote a Firefox extension, which I'm seriously considering replicating (or finding, if he ever published it), which would take one master password that he'd remember, combine it with the domain, and computer a hash. Thus, nothing is ever stored, but there's still only one password to remember.
This scheme prevents a breach at one website from compromising others, so long as your master password and/or local password store is safe. And it's a lot easier to try to keep that safe than to try to create and memorize tons of random passwords.
Finally, there is the option to use client-side certificates and/or OpenID, with services that support them. This would allow you to choose whatever means of authentication you like, passwords or otherwise.
The point is, you're not allowed to save them somewhere obvious in plain text, or especially, taped to your monitor.
It is a known fact that people are stupid. If you make something that proves that fact, then the problem is not the moron users, but the designers.
But trying to idiot-proof it is the wrong approach, or at least, should not be a priority. As the saying goes, they'll always build a better idiot.
No, the right approach is to increase the ease with which someone could use the system properly, and how far "properly" extends. After you've done that -- in this case, after OpenID is ubiquitous -- then you can worry about how to dumb it down to where an idiot can use it.
But if you design the system for an idiot in the first place, you're both creating more idiots, and in this case (using passwords and "pet's name" security questions), making the system less secure and/or less convenient for experienced users.
Indeed it is, but it works well enough, except for this one issue.
I suppose it's possible that this is one thing which hasn't been ported to 64-bit, and I should try a 32-bit Flash 10. But it is kind of frustrating that this "alpha" works flawlessly for everything I throw at it, including an "alpha" version of 64-bit Chromium, except for fullscreen.
In other words, that XKCD is still relevant.
Especially because it's not even Linux's fault. It's whatever asshat they have working for Adobe, who somehow claims using XVideo (what mplayer uses to play fullscreen 1080p flawlessly while Flash struggles with 480) would be impossible because of some bullshit like color spaces. Which might've been a valid argument, if Gnash hadn't already implemented XVideo support.
Also, no it's not. Even if I ran the slowest and most unstable distro ever made, it doesn't explain why Flash 9 would work, and Flash 10 would turn into a slideshow.
Slashdot Mirror
If that's how people on this site define the word "sane", then yes I am.
If your reaction to being called a fanboy is to assume I'm that much biased in the other direction, you're making my point for me.
No, I do think you're mostly sane. But "sane" doesn't mean "agrees with me", and the same goes the other way. I do not have to be insane to disagree with you.
So the guy who yells "FUCKING KILL GOOGLE" (so you say) is so subtle that he'd try to auction the patents to some other firm?
Microsoft has been deliberately "subtle" in the past -- see MS Java, Dr DOS and Windows 3.1, "It ain't done till Lotus won't run", etc.
Again, the question here is: Is it so unbelievable that someone would both overtly rail at his enemies, and covertly attempt to undermine them? Ballmer wouldn't be the first to do this.
And yet so short-sighted that he didn't foresee the possibility that a firm that wants to defend Linux would buy the patents in the same auction?
Actually, that's not what I'm suggesting. What I'm suggesting is more paranoid -- that said firm could easily be a front, or could be tempted with sufficient amounts of money, so that it looks as though they've sold the patents to a Linux-friendly company.
It's much more likely that Microsoft auctioned the patents for some wholly un-related reason, and didn't care who purchased them.
Oh, I'll grant that, easily.
But keep in mind, this is the same Microsoft who sent astroturfers to a Linux convention, saying things like "It's all over, the suits are taking over," back when IBM started showing up at these conventions. This is the same Microsoft who essentially stole the GUI concept wholesale from a Macintosh prototype -- keep in mind, Apple did actually license it from Xerox. And this is the same Microsoft who continues to fund SCO.
It is not paranoid to expect a company which has been underhanded and "subtle" in the past to continue to do so in the future. It may not be the most likely possibility, but it's hardly "insane".
That ain't Linux they are ripping off
"Taskbar thumbnail previews" are something I remember having in Compiz long before Win7, and I don't remember seeing them on OS X, either. Want to be specific about what Mac things they're ripping off?
That is because Linux doesn't really have a drop in replacement for AD+Exchange+Outlook+Sharepoint+GPO.
The problem is firstly that people think they actually need all that -- do people actually use SharePoint? Anecdotes point to maybe.
But Exchange + Outlook? Plenty of groupware. AD/GPO? Samba is working on it, if it isn't released already, and Unix-centric solutions have existed for decades.
If I understand your argument, though, they're trying to concentrate on fighting Apple?
I think you just don't know what "public domain" means. It means there is no owner, no one can exert property rights in it, there's no legal way to exclude anyone from using it.
From the Wikipedia article:
The public domain can be defined in contrast to several forms of intellectual property; the public domain in contrast to copyrighted works is different from the public domain in contrast to trademarks or patented works.
Here's the problem: You were talking specifically about documentation, which would imply "public domain" as contrast to copyrighted works. And patenting an invention is specifically designed to ensure that all of the work behind it -- designs, schematics, papers, everything -- is released into the public domain.
I don't know that there is a way to release a patentable concept into the public domain, other than ensuring that it's patented by someone who won't abuse the patent.
I also think you don't really understand software patents, especially how Microsoft (and its ilk) use them to interfere with others competing with them, or even with interoperating with them.
Actually, I've been following the whole TomTom VFAT thing, not to mention the current issue with h.264 patents -- where essentially neither Apple nor Google will implement patent-less video formats, yet no other browser makers care to spend the money on licenses for h.264.
Covenants not to sue are transient. You'd have to sue them to enforce it, which makes it useless to small operations.
Or you simply violate their patent and dare them to break that covenant. When they come sue you, you have legal grounds to call them on it -- not to mention loads of bad press for them.
For that matter, take Theora:
On2, which owns patents that apply to the technical foundations of Theora, granted an unrevocable free license regarding those patents.
How is that better than a covenant not to sue? How is it different?
What's more a shame to me is just how much this seems to confirm the direction the OLPC has set, ever since they decided they had to support Windows, and then decided they were dropping Sugar.
Basically, they're moving it less in the direction of being a ridiculously cheap education appliance, and more in the direction of being just a sort of cheap netbook PC.
Not that there's anything wrong with that, but the netbook market is pretty saturated, and they had a number of things that used to set them apart. By moving in this direction, instead of, say, cheaper, more rugged, more reliable, and improving Sugar, it looks like they're going to be competing head-on with the major OEMs very soon, if they aren't already.
I mean... I don't know, maybe I'm jaded, but I haven't seen an encouraging OLPC story in awhile now. There just doesn't seem to be anything interesting left.
Well, it's not on all pages, nor is it fully featured, but they do have a tech demo.
Main missing features are fullscreen video and Theora.
I'm not sure there's currently any sort of spec for making an HTML5 video go fullscreen. I can see it being restricted somehow -- to a native control, or maybe in the same way pop-ups are, because otherwise, any script could throw up a fullscreen porn ad -- but it seems like a kind of glaring omission to have no support for it other than calling "save as" on the video (which isn't always easy).
Chances are, they'll never add Theora, which means if they switch over, Firefox may never support it -- or at least, I get the impression that there are a few idealistically blinded people who want to remove all hint of a possibility of a way to play HTML5 h.264 in Firefox, not even as a third-party plugin.
Interesting little fact, though: The video tag does support data urls. That means a clip can be embedded into the HTML page itself, base64-encoded. I have absolutely no idea why this is useful, but I thought it was cool.
It's not paranoid when they actually are out to get you.
And really -- I know you're often pro-Microsoft, borderline fanboy, but even you should be able to see that Steve "FUCKING KILL GOOGLE" Ballmer would love a chance to cut Linux off at the knees.
take all the documentation of the patent, and put it in the public domain.
That's what a patent does in the first place!
It just provides monopoly protection for actually using said patent.
In fact, this was the whole point of patents. Say I invented lemonade. Without a patent, I'd keep it my secret family recipe for generations, and anyone who wanted to make lemonade would have to reverse engineer it -- but if someone did, I wouldn't be able to say much.
With a patent, I would publish the documentation ("It's just sugar, water, and lemon juice.") and then only I can make lemonade. Or I can license the recipe to others -- it's not like they don't know how to make it now, it's that they legally can't unless I let them.
If they were really serious about merely protecting the community, they'd give up the patent control entirely.
I'm not sure it's legally possible to do that, is the problem. Moreover, having a process patented provides clear documentation that you patented it first, thus putting the burden on anyone's infringing patent to prove that they invented it before you did.
No, my big problems with this are not that I think the result is bad, but because I think it should be unnecessary -- I highly doubt Microsoft has any stunning invention that Linux "stole" for which prior art doesn't exist a thousandfold, and even if there were, I'm not sure software patents should exist at all.
But if I'm going to accept that they exist, and that someone has to hold them, I'd much rather that someone not be Microsoft, no matter how legally binding their "covenant not to sue" is.
Though it would be pretty slimy if this new organization doesn't have some sort of "covenant not to sue." Maybe that's the motive? Blech, now I have to go wash the evil from my brain...
Well, it made no damned sense, as there was no authentication that I know of (WIDE open, so long as you're on Windows), and ethernet worked just fine. For what it's worth, udhcpc is a very minimalist DHCP client -- designed for embedded systems, as I understand it.
So it seems like this is either a bug in two separate DHCP clients, or in the server itself, but not necessarily a reflection of how much better (or worse) one client is than another. For example, it could very well be that dhclient is behaving entirely within the spec, but confusing the DHCP server (whatever it is).
Also, there was nothing about "persistence", really -- udhcpc got a lease pretty much instantly. Like, less than a second to associate with the LAN, run DHCP, and get a lease, whereas dhclient would time out after more like ten, twenty, thirty seconds.
Legitimate use is easy and non-annoying. In other words, if you purchase the product on iTunes or Steam or any service implementing this protocol, you can use the product where ever and whenever you want, on whatever devices.
So long as you transfer it every fucking time. No thanks.
To give another example: Steam is nice because if my original Windows installation is completely annihilated, I can just re-download games onto a new machine. This is worse than Steam, and most people don't even like Steam.
There's no "Kindle 1984" scenario looming
Actually, that's assuming they don't do that anyway.
no need to buy special "DRM-ready HDMI" cables.
Actually, that's highly likely to be needed anyway, as HDCP isn't going away just because the content is protected in some other way than Blu-Ray's AACS or BD+.
Infringing use is easier to prove/disprove. This assumes that the files are watermarked with your account identifier and digitally signed in some fashion (not difficult -- iTunes does something like this already).
You are aware that it's also trivial to change this watermark in iTunes? It's no more "proof" than an IP address.
OTOH, though, if the FBI seizes your hard drive, and all the files on it are properly watermarked and digitally signed, they have no case.
Which sets a really bad precedent for those of us who prefer unencumbered media. Do I have to dig up the actual CDs I ripped my FLACs from to prove my innocence?
I shouldn't have to prove my innocence. It's "innocent before proven guilty", right?
Obviously, some hackers will find a way to crack the file format pretty much the day it is announced, and the BitTorrents will continue. That's OK; a little piracy never hurt anybody.
I'm glad we agree. So why do we need it?
The idea is to protect regular people -- folks who just want to buy or rent a song or movie and play it without a big hassle and without giving control of their computers over to some other company
How the fuck does this "protect" them more than "no DRM at all"?
and to help the big publishers feel comfortable about moving towards digital distribution.
If that's its entire purpose, then it can rot in hell.
Do you think the average college helpdesk is prepared to answer random Linux questions?
At my community college, they were prepared for the specific Linux question I had. Unfortunately, the only answer they had was "We don't know." They were glad to have a solution once I solved it, though.
a longer explanation.
Due to a few bad decisions in college and the economy imploding at work, I am now going to Indian Hills Community College, trying to at least keep it together for a semester or two before I invest in a four-year college again.
I have no idea whether it's required, but there is a Unix/Linux class. But then, there's also a Visual Basic class.
Aside from the fact that a few things (VB?) will require Windows -- though they at least have the decency to have an MSDNAA license, meaning free copies of Windows and nearly everything needed -- there's also the fact that Macs sort of get a passing reference saying "We hope it works, but we can't support you," and Linux gets no mention.
The wireless fails out of the box with Ubuntu, yet works with Windows. Talking to the help desk, they basically said "We don't support that, we don't know much, but our vendor assures us that the problem is not with our equipment, but with Linux."
To get it working again, I had to switch DHCP clients. Neither dhclient nor dhcpcd worked, but udhcpc did. It's worth mentioning, this is not a common problem -- I used NetworkManager's point and click interface pretty much everywhere, and it worked pretty much everywhere, from hotel rooms to hospitals to crappy little Linksys routers -- it even worked if I plugged into ethernet in school -- everywhere except the school wireless.
The conclusion to this story? I mailed the helpdesk again with my findings, and with the little script I wrote to disable NetworkManager, bring up wlan0 manually, and run udhcpc. They seemed very glad to have a solution.
So, I'm not really sure what to make of it. On the one hand, it was obviously a priority, and I was pretty much left to fend for myself. On the other hand, no one actually has a problem with me using Linux, most of the time.
I realize that doesn't answer your questions about printing or VPNs -- I haven't had to do either yet. Printing, I've only done from lab computers (all Windows, naturally), and they don't require a VPN, though my personal VPN works fine from the school wireless. Their website is an abomination, but it mostly works fine in Chrome, with only one place so far which requires Firefox, and I haven't had to use IE yet, except on lab computers.
Just for fun, another anecdote: Iowa State University, when I was there, had a lab full of top-of-the-line Linux workstations. In the classes I was taking, they were used mainly to run rdesktop, which seems profoundly retarded, but I never had a problem due to running Linux or OS X. This was around 2005-2006.
Erm... seems like you might've responded to the wrong comment, or blatantly misread mine? I'm the atheist, responding to "God-boy", who never mentioned that particular argument (nor did I).
I admire your goals, but being a dick about it isn't going to change minds.
Because sometimes it works better? Even in Unix world, there's Emacs as a testament
...to the fact that some people would much rather use vim.
Moreover, Emacs isn't monolithic. Aside from the fact that it's just a set of Lisp scripts for a much simpler editor (TECO), there's the fact that most of the more interesting things you might want to do with it are also going to call some external program at some point.
One obvious example: Version control. I very much doubt anyone's actually written bindings for Emacs for all the various version control systems it supports. Specifically, I know it supports Git, and I know Git has no simple library or API beyond its commandline interface -- in other words, if you use Git with Emacs, that means Emacs is calling git commands.
Small programs, doing one thing well. One unified interface acting as a frontend for these small programs. Sounds exactly like the Unix philosophy to me -- indeed, the best of both worlds.
There are some that are NOT under my control.
Potential security hole right there.
I cn not install software on those systems. I can not add anything on those systems. Further not all logins are weblogins and some that are only work on very locked down IE machines
Look at shvytejima's reply -- that is, PasswordHasher is available both as an extension and as a standalone HTML page -- thus, easy to host yourself, if you don't trust someone else to. Visit with your locked down IE, type your password and the base domain, copy/paste the hash into the page (or program) you needed.
So by increasing the security on their side by doing all the things that are possible, they actually have decreased it in the end.
Which is exactly why the important thing here is to allow tighter security, not require it. The hardware key is a great example -- I'd stick it on my keychain and be happy to have it. Most people would probably physically attach it to the machine it's used from. Make it optional and we can both be happy.
The right way to solve the problem users have with their passwords is educating those users, not writing off a service as "insecure" because the people using it are. After all, social engineering is always possible.
So if after say 20 years of intensive computer usage by non-geeks what we do now does not work, I would suggest we should start looking for something else.
Depends what you mean by "what we do now".
Everything you've just mentioned is fairly stupid design, where the systems are entirely outside your control as a user, and you are "forced" to have a level of security they will accept, leading to you doing things like writing passwords down. About all I could suggest from your side is, that text file should be encrypted with a password you can remember.
But we weren't talking about that. We were talking about webmail -- and most webmail is either smart enough to use OpenID, or not smart enough to do more than a straight username/password. In either case, it's possible to be both reasonably secure (seemingly-random password) and convenient (hashing, save password, etc).
It's certainly possible to be at least secure enough that no one's going to pull it directly off a hard drive or a sticky note, nor guess it based on your pet's name, favorite movie, dictionary attack, etc.
So I think I'm justified in not feeling a lot of sympathy for someone who had their webmail account cracked by guessing a password. I can sort of see someone not knowing to force https, at least when using wireless (hint: https://mail.google.com/, and Gmail will remember it once you login), but it's not that difficult to deal with a secure password, especially in that context.
I prefer a browser I can easily extend myself to one that has "everything" built in, where "everything" might not include something Opera hasn't thought of.
If you can't tell the difference between the server and the client, how the hell are you a competent admin?
Then you have users who don't even know "Open in new tab exists", so they keep clicking on the left mouse button, thereby using only one-two tabs.
That's why performance of a single tab was also measured.
Other users (like me for example) got used to clicking the middle mouse button so all new pages are opened in new tab.
So opening eight tabs is still a useful measure of how quickly it can open a new tab.
Scientifically, it would be relatively easy to write a software that would
So write it.
I think your definition of "relatively" is a bit odd, considering we're talking about the difference between that and just telling the browser to open eight tabs, and we're talking about a tech reporter, not necessarily a programmer, doing the measurement.
Yeah, I'm on Jaunty 64-bit. And I'm using the Chromium nightly PPA -- Google for it -- means I have to download maybe 10-20 megs per day, but it also means I don't have to compile anything.
I'm aware of that project.
However, Chrome exists now, and the extensions are usable in the nightlies.
Is Electrolysis in a similar state? That is, could I actually use a Firefox nightly with Electrolysis enabled as a reasonably reliable daily browser?
I am not allowed to save them. I must memorize them.
Nonsense. While Chrome doesn't seem to have this yet, Firefox and Konqueror come with encrypted password stores out of the box.
That is, you enter one master password, and it then remembers all your passwords for you.
I also have a friend who wrote a Firefox extension, which I'm seriously considering replicating (or finding, if he ever published it), which would take one master password that he'd remember, combine it with the domain, and computer a hash. Thus, nothing is ever stored, but there's still only one password to remember.
This scheme prevents a breach at one website from compromising others, so long as your master password and/or local password store is safe. And it's a lot easier to try to keep that safe than to try to create and memorize tons of random passwords.
Finally, there is the option to use client-side certificates and/or OpenID, with services that support them. This would allow you to choose whatever means of authentication you like, passwords or otherwise.
The point is, you're not allowed to save them somewhere obvious in plain text, or especially, taped to your monitor.
It is a known fact that people are stupid. If you make something that proves that fact, then the problem is not the moron users, but the designers.
But trying to idiot-proof it is the wrong approach, or at least, should not be a priority. As the saying goes, they'll always build a better idiot.
No, the right approach is to increase the ease with which someone could use the system properly, and how far "properly" extends. After you've done that -- in this case, after OpenID is ubiquitous -- then you can worry about how to dumb it down to where an idiot can use it.
But if you design the system for an idiot in the first place, you're both creating more idiots, and in this case (using passwords and "pet's name" security questions), making the system less secure and/or less convenient for experienced users.
That's bitten me once -- and only once -- since I started doing it.
My bank requires security questions.
It then picks a random security question when I login, as part of their wish-it-was two-factor authentication scheme.
Indeed it is, but it works well enough, except for this one issue.
I suppose it's possible that this is one thing which hasn't been ported to 64-bit, and I should try a 32-bit Flash 10. But it is kind of frustrating that this "alpha" works flawlessly for everything I throw at it, including an "alpha" version of 64-bit Chromium, except for fullscreen.
In other words, that XKCD is still relevant.
Especially because it's not even Linux's fault. It's whatever asshat they have working for Adobe, who somehow claims using XVideo (what mplayer uses to play fullscreen 1080p flawlessly while Flash struggles with 480) would be impossible because of some bullshit like color spaces. Which might've been a valid argument, if Gnash hadn't already implemented XVideo support.
Citation needed.
Also, no it's not. Even if I ran the slowest and most unstable distro ever made, it doesn't explain why Flash 9 would work, and Flash 10 would turn into a slideshow.
Yeah... then honor those settings, and create ~/.config/uzbl or whatever it is.
Having it just fail out of the box is not acceptable, no matter what you're trying to comply with. It is quite literally a non-starter.
You must have a very strange definition of "bulletproof"...