I have a previous version of the Fon router, which allows SSH by default. Unfortunately, as another has mentioned, it also allows/requires Fon to have root access to your router by default, so as far as I'm concerned you can't trust the device.
Also, the only way to access your wired network from the wireless is to allow ALL wireless users to have that access. Well, okay, you could do things like SSH out to a machine on the Internet, SSH back in, and set up port forwarding that way, but nobody would ever do that:). And your own wireless access is treated the same as everybody else's-- you have to log in every time. Annoying in combination with Firefox2's ability to resume sessions-- it loads the Fon login redirection page for every tab you had open.
They've been promising a firmware fix which would allow two SSIDs with different configurations for a long time, but last I checked it still isn't out.
The upshot of this is that I thought I would be getting a nifty solution which would let me share my access while covering my own needs. Instead I really have to run two routers, one for me, and one for everybody else. And despite the fact that I live in a pretty densely populated area, in about six months the number of people who have signed on to the Fon router, besides me, is zero. Oh, correction: the buddy who told me about Fon came by and tried to sign in with his account, which he is supposed to be able to do as a "Linus" user. That didn't work either.
In summary... it's more work and their system is not transparent or secure (oh yeah, there's no encryption on the wifi connections). It's a nifty idea, but I can't really recommend it.
It's odd, I realized the other day as SuSE, for no reason (I moved a server to another room and plugged it into an external RAID; no other changes), decided to interrupt its boot process to detect ~12 "new" pieces of hardware-- most of which it had needed to boot)... SuSE is becoming more and more like Windows. The steaming pile of bugs that is YAST, the obtuse config files, the inconsistent behaviours (exactly the same actions have varying results). It's about like Redhat's Linuxconf from around 5.2. Under Debian, Gentoo, Redhat, Fedora, Ubuntu, if you have a problem, documentation is easy to find, but usually the config files and system utilities are plenty. Under SuSE... there's often only one way to do it, and they won't tell you what it is. If you manually edit config files, don't ever run YAST again.
That's a great idea! I hate that all my DVDs play whenever I tell them to. If only some random cokehead executive, or hacker, or bug, or network problem could occasionally stop my DVDs from playing. And if it adds cost to the system, so much the better! Now if you'll excuse me, I'm off to give myself a nice hard cockpunch.
I get panics on my MBP, in the same few memory locations, when loading the wifi heavily. For instance, using scp or nfs will trigger a panic within a few minutes. This only happens when the wifi is on; ethernet works fine. I've been waiting to take it in for repair until such a time as Apple A) acknowledges all the other problems (heat, whine) I want fixed on this POS, and B) I can stand to do without my beloved POS for a week or so.
Right... I think we pretty much agree, you just went into more detail. Though another threat would be that in a standard configuration somebody could log in from the tainted PC to your stick, then grab or use your keys, or take root through sudo or a local exploit-- anyway, I'm sure you know this and my original message was not really intended for people who know this stuff already. The worst risks in this case require either foreknowledge of your setup, live monitoring, or that the passwords and other information you type during this session are sensitive elsewhere.
First, this looks like a really cool product, so I feel like kind of a jerk for picking this nit, but I will anyway because social skills are not my forte:
"For example, users who need secure remote access from insecure PCs could plug in the ThumbStix, ssh into it, and connect to remote VPNs from there, Hughes suggests."
I guess it could provide a layer of obfuscation. If you know what specific attack is being used, and you're sure that the insecure PC is not running a keylogger or sniffer, it could help. But if the first link in the chain is broken, don't trust the chain.
Ha, no. The footnote specifies the length and the bitrate of the song. Changing the bitrate will affect the number of songs one can store in exactly the same way as changing the length. Since I don't think there market segment which insists on being able to store at least 2000 songs, all exactly four minutes long, there is no reason they can't say "Based on 4 minutes per song and 128kbps average bitrate."
Um. Or they could specify the capacity in bytes, which will come in handy as these devices are used for other types of media and programs. I think there's a precedent for that.
Almost the only time advertisements influence me are when they convince me NOT to buy from a company. If an ad is offensively loud, shrill, intrusive, or stupid, I make a note of that company. For example, when Quizno's decided that their product was best represented by a retarded mutant singing rodent, which carried the implication that you're eating... retarded mutant mice... I stopped eating there until about half a year after the last time I saw one.
Even a funny, clever ad will not make me buy something I don't want or need.
About the best an advertiser can hope for from me is to not offend me too much. It's very rare that an ad informs me of something I'm not already aware of. The only exceptions are Google's text ads, which I only see when I'm specifically looking to buy, and occasionally trailers for movies (and yet, I rarely go to see movies anymore, because of the 20 minutes of ads before the film, the shrieking babies, etc.) As far as TV goes, movie trailers make sense, but I don't see the purpose of most others. Is there anybody out there who hasn't made up their mind on Coke vs. Pepsi? Am I that unique in only reacting negatively to ads? Is the average consumer really that stupid? Oh... back to that again. Never mind.
...would like fraud investigations to cease? I wonder why. I wonder why this CEO doesn't want this type of crime investigated, and why he doesn't want people to understand that there are victims in this type of crime.
Everything you see, like this new evidence, comes via light. Some light sources are closer and more recent than others. But do you discard all evidence that comes in the form of photons?
Okay, I know I'm simplifying by leaving out hallucinations, faulty organs, the false light you detect if you close your eyes and press your eyeballs with a finger, etc. Jeez, so many nitpickers around here...
Often sleazy little tibits like these are inserted at the last minute into a bill hundreds or thousands of pages long. When the votes are taken, nobody is even aware of the changes. Congress critters complain about having voted for something they didnt know about, but none of them seems to want to address the problem.
So, why doesn't Congress use a revision control system? When the day comes to vote on a bill, you check for changes since the last time you read it. If there are changes, you know who made them and when. Your basic audit trail.
I suspect that one of the reasons something like this hasn't been implemented yet is that most politicians are habitual defectors rather than cooperators; they may not want their enemies to be able to use dirty tricks, but they'd like to be able to do it themselves.
Ah, besides. Can you imagine Congressional debate on whether to use CVS, svn, or... what am I thinking? Free software wouldn't even be on the table.
Where do we want the semantic web most? Search, for now (later, AI:) Which means that it would take a long, long time for it to become useful if we wait for a large proportion of sites in the world to do the rework themselves.
But, what if we use the same technologies that allow for p2p annotation of sites (like Greasemonkey, de.licio.us), to, with few clicks, vote on the relation of the important ideas in a site? You'd have to have a credibility/karma/trust system, as some of the most important relations you'd want to search on are commercial/noncommercial, advertisement/review, question/answer. And these would definitely be targets for spammers, advertisers, pranksters, and other habitual defectors.
Most people would not bother definining the relationships, but you'd only need a fraction of a percent to do the initial reviews. People who are not averse to cooperation, for example the people who work on OSS or wikis, will see that if they have sunk 15 minutes of searching into finding a page, doing a minute or two of markup/commentary can save thousands of others that much time. A larger group will be willing to vote on the accuracy of the links in two situations: they find the links very useful, or very misleading. The more misleading karma you've built up, the less your vote is worth, which should take care of spammers (but still leaves political/religious debates wide open for mismoderation).
And then, of course, Google ties it together by using these tags to enable semantic searches:
"I want English or German language ANSWERS to the question, how do I get the Marvell SATA driver working under RHEL 3.0, written between 2004 and 2005".
The semantic search should know that RHEL 3.0 is aka "Redhat Enterprise Linux 3.0", and "Marvell SATA" could be mv_sata or aar81xx.
If everything is selfish, the word "selfish" has no meaning. Words are supposed to communicate, define, delineate. If you've got a definition for a word that renders it completely useless, that's probably not the definition people are using.
Good to see Sun still makes its business decisions based on what might hurt other companies rather than what's good for Sun. I just hope nobody tells them about the GPL before they make the one of the most laughable mistakes of the century. Oh, and how does "owning" SuSE Linux equate to a boost for Solaris on PPC? The source is available under the GPL if you're curious... unless they plan to encourage adoption of Solaris by knifing a Linux distro.
Anyway, Sun, why bother? Why not wait another couple of years, and then let Novell buy you?
Bash can now store timestamps in the history and save them to the history file. This alone is worth the upgrade for me. The option to erase duplicates is pretty nice too.
Slashdot Mirror
Yeah, it took several months for a change of address to show up on their map. From what I read on their forums, that's standard. Needs work :)
Also, the only way to access your wired network from the wireless is to allow ALL wireless users to have that access. Well, okay, you could do things like SSH out to a machine on the Internet, SSH back in, and set up port forwarding that way, but nobody would ever do that :). And your own wireless access is treated the same as everybody else's-- you have to log in every time. Annoying in combination with Firefox2's ability to resume sessions-- it loads the Fon login redirection page for every tab you had open.
They've been promising a firmware fix which would allow two SSIDs with different configurations for a long time, but last I checked it still isn't out.
The upshot of this is that I thought I would be getting a nifty solution which would let me share my access while covering my own needs. Instead I really have to run two routers, one for me, and one for everybody else. And despite the fact that I live in a pretty densely populated area, in about six months the number of people who have signed on to the Fon router, besides me, is zero. Oh, correction: the buddy who told me about Fon came by and tried to sign in with his account, which he is supposed to be able to do as a "Linus" user. That didn't work either.
In summary... it's more work and their system is not transparent or secure (oh yeah, there's no encryption on the wifi connections). It's a nifty idea, but I can't really recommend it.
It's odd, I realized the other day as SuSE, for no reason (I moved a server to another room and plugged it into an external RAID; no other changes), decided to interrupt its boot process to detect ~12 "new" pieces of hardware-- most of which it had needed to boot)... SuSE is becoming more and more like Windows. The steaming pile of bugs that is YAST, the obtuse config files, the inconsistent behaviours (exactly the same actions have varying results). It's about like Redhat's Linuxconf from around 5.2. Under Debian, Gentoo, Redhat, Fedora, Ubuntu, if you have a problem, documentation is easy to find, but usually the config files and system utilities are plenty. Under SuSE... there's often only one way to do it, and they won't tell you what it is. If you manually edit config files, don't ever run YAST again.
So... good match for MS.
That's a great idea! I hate that all my DVDs play whenever I tell them to. If only some random cokehead executive, or hacker, or bug, or network problem could occasionally stop my DVDs from playing. And if it adds cost to the system, so much the better! Now if you'll excuse me, I'm off to give myself a nice hard cockpunch.
Okay, so I'm not a good person.
Nope, didn't happen under Ubuntu and memory and the rest of the hardware tested fine.
Oh, wait, you're a trolldouche.
I get panics on my MBP, in the same few memory locations, when loading the wifi heavily. For instance, using scp or nfs will trigger a panic within a few minutes. This only happens when the wifi is on; ethernet works fine. I've been waiting to take it in for repair until such a time as Apple A) acknowledges all the other problems (heat, whine) I want fixed on this POS, and B) I can stand to do without my beloved POS for a week or so.
Right... I think we pretty much agree, you just went into more detail. Though another threat would be that in a standard configuration somebody could log in from the tainted PC to your stick, then grab or use your keys, or take root through sudo or a local exploit-- anyway, I'm sure you know this and my original message was not really intended for people who know this stuff already. The worst risks in this case require either foreknowledge of your setup, live monitoring, or that the passwords and other information you type during this session are sensitive elsewhere.
Meh. I still want one.
First, this looks like a really cool product, so I feel like kind of a jerk for picking this nit, but I will anyway because social skills are not my forte:
"For example, users who need secure remote access from insecure PCs could plug in the ThumbStix, ssh into it, and connect to remote VPNs from there, Hughes suggests."
I guess it could provide a layer of obfuscation. If you know what specific attack is being used, and you're sure that the insecure PC is not running a keylogger or sniffer, it could help. But if the first link in the chain is broken, don't trust the chain.
Ha, no. The footnote specifies the length and the bitrate of the song. Changing the bitrate will affect the number of songs one can store in exactly the same way as changing the length. Since I don't think there market segment which insists on being able to store at least 2000 songs, all exactly four minutes long, there is no reason they can't say "Based on 4 minutes per song and 128kbps average bitrate."
Um. Or they could specify the capacity in bytes, which will come in handy as these devices are used for other types of media and programs. I think there's a precedent for that.
for people not familiar with degrees Celsius
The word you're searching for is "Americans"...
Almost the only time advertisements influence me are when they convince me NOT to buy from a company. If an ad is offensively loud, shrill, intrusive, or stupid, I make a note of that company. For example, when Quizno's decided that their product was best represented by a retarded mutant singing rodent, which carried the implication that you're eating... retarded mutant mice... I stopped eating there until about half a year after the last time I saw one.
Even a funny, clever ad will not make me buy something I don't want or need.
About the best an advertiser can hope for from me is to not offend me too much. It's very rare that an ad informs me of something I'm not already aware of. The only exceptions are Google's text ads, which I only see when I'm specifically looking to buy, and occasionally trailers for movies (and yet, I rarely go to see movies anymore, because of the 20 minutes of ads before the film, the shrieking babies, etc.) As far as TV goes, movie trailers make sense, but I don't see the purpose of most others. Is there anybody out there who hasn't made up their mind on Coke vs. Pepsi? Am I that unique in only reacting negatively to ads? Is the average consumer really that stupid? Oh... back to that again. Never mind.
Glad I don't own Netapp stock right about now.
Okay, I know I'm simplifying by leaving out hallucinations, faulty organs, the false light you detect if you close your eyes and press your eyeballs with a finger, etc. Jeez, so many nitpickers around here...
So, why doesn't Congress use a revision control system? When the day comes to vote on a bill, you check for changes since the last time you read it. If there are changes, you know who made them and when. Your basic audit trail.
I suspect that one of the reasons something like this hasn't been implemented yet is that most politicians are habitual defectors rather than cooperators; they may not want their enemies to be able to use dirty tricks, but they'd like to be able to do it themselves.
Ah, besides. Can you imagine Congressional debate on whether to use CVS, svn, or... what am I thinking? Free software wouldn't even be on the table.
But, what if we use the same technologies that allow for p2p annotation of sites (like Greasemonkey, de.licio.us), to, with few clicks, vote on the relation of the important ideas in a site? You'd have to have a credibility/karma/trust system, as some of the most important relations you'd want to search on are commercial/noncommercial, advertisement/review, question/answer. And these would definitely be targets for spammers, advertisers, pranksters, and other habitual defectors.
Most people would not bother definining the relationships, but you'd only need a fraction of a percent to do the initial reviews. People who are not averse to cooperation, for example the people who work on OSS or wikis, will see that if they have sunk 15 minutes of searching into finding a page, doing a minute or two of markup/commentary can save thousands of others that much time. A larger group will be willing to vote on the accuracy of the links in two situations: they find the links very useful, or very misleading. The more misleading karma you've built up, the less your vote is worth, which should take care of spammers (but still leaves political/religious debates wide open for mismoderation).
And then, of course, Google ties it together by using these tags to enable semantic searches:
"I want English or German language ANSWERS to the question, how do I get the Marvell SATA driver working under RHEL 3.0, written between 2004 and 2005".
The semantic search should know that RHEL 3.0 is aka "Redhat Enterprise Linux 3.0", and "Marvell SATA" could be mv_sata or aar81xx.
If somebody's going to do something evil, we should help them do it effectively!
...at least be nice about it?
That's the most idiotic idea I've ever seen in my life. And I read slashdot!
Cool, maybe it can enable the turn signals on those Lexuses. Lexus: So insulated from the road, you can forget you're driving!
If everything is selfish, the word "selfish" has no meaning. Words are supposed to communicate, define, delineate. If you've got a definition for a word that renders it completely useless, that's probably not the definition people are using.
In firefox, you can use CTRL-J or CTRL-k to take you to the search bar. So you do lose a little real estate, but save a keystroke.
Good to see Sun still makes its business decisions based on what might hurt other companies rather than what's good for Sun. I just hope nobody tells them about the GPL before they make the one of the most laughable mistakes of the century. Oh, and how does "owning" SuSE Linux equate to a boost for Solaris on PPC? The source is available under the GPL if you're curious... unless they plan to encourage adoption of Solaris by knifing a Linux distro.
Anyway, Sun, why bother? Why not wait another couple of years, and then let Novell buy you?
In conclusion: Assclowns.
Bash can now store timestamps in the history and save them to the history file. This alone is worth the upgrade for me. The option to erase duplicates is pretty nice too.