Don't know about the Kodaks, but HP has lots of photo printers that don't require you to connect to a PC. AFAIK, any of them that include an LCD screen can do this (you need the screen to see what you are doing). They also have some that work in conjunction with a digital camera that use the LCD of the camera as an interface. This requires a certain kind of camera (I believe they are sold as a bundle), so unless you are shopping for a new camera too, you might look elsewhere.
btw, you can connect these to a PC, but you don't have to, and the ones with a bunch of card slots (SD, compact flash, etc.) also function as a general card reader -- not just for photos -- when connected to a PC.
The only issue left is that it did not recognise the wireless card I have installed.. so I can't put it on the network yet.
I installed Ubuntu on an old laptop a few months ago, and also ran into trouble with the wireless card. Mine was a linksys, and they didn't have a driver for Linux. I was able to use the Windows driver on Ubuntu using NdisWrapper. Assuming you are having similar driver trouble, this may work for you too.
IIRC, NdisWrapper was on the install cd (you might have to apt-get or synaptic it to get it on your HD), so you shouldn't have to download it (which is good, since your card doesn't work yet;^).
You'll also need the Windows driver for you card. You can get it off of the CD that came with it, or they are usually available from the manufacturer's website.
Just curious... Would you have the same opinion that the GP was commiting theft (your word, not mine) if s/he was recording the program on VHS? Or is the fact that a digital device is in use somehow make it magically (more) wrong?
In this example for simplicity's sake, I'm going to use a ficticious function to shorten the output. The same logic will apply to any hashing function that gives output of a specific length.
Let's give this function a name, so that people don't assume I'm using SHA-1, and flame me for it. Let's call the function "melandy_hash", and say that for whatever input it receives, it gives a 6 digit hex number for the output (yes, I know that this is ridiculously short, but this is a simplified example).
For the input "This is my input string", you might receive the output 82a78b. For the input "This is another input string", you might receive the output 1721ca. Note that although the values for the output are different, the length is the same, because we specified that the output would be a 6 digit hex number.
To expand this further, and to get directly to the point of your question... say we have a list of strings that we want to hash:
"This is my input string" "This is another input string" "And another one" "And yet another one" "X"
These five strings collectively are the "set" of inputs. When we compute the hashes of these strings using melandy_hash, we get the following output:
82a78b 1721ca 82a78b 1b82ac 97f25b
The above list of hex values is then the "set" of output.
Note that the strings "This is my input string" and "And another one" both have the same hash value. This is known as a collision. As a result, there are really only 4 unique values in out set of output. This means that the set of inputs (with 5 unique elements) is larger than the set of outputs (with 4 unique elements).
Where you may have been confused before is that the output for the string "X" is 97f25b, and for this particular element, the output is larger than the input. The original point that was made was that the number of unique elements in the input set is larger than the number of unique elements in the output set.
A hash is a mapping from a large set of items to a smaller set.
No, it could be the opposite of this, taking a small set of items and mapping them to a larger set. Many hash functions result in a set value of set length and so a hash may be larger that its plaintext.
You're missing the point here. The GP stated that a hash is mapping from a large SET of items to a smaller SET [of items]. You are twisting the words to indicate that s/he said that large ITEMS map to small ITEMS, which is not what was stated. It is true that the hash of the string "a string" is larger than the string itself, but to argue that is to miss the point entirely.
The dog is going to go into a state of extreme alertness and defensiveness and will try to let everyone in the house know about what its discovered irrespective of whether it was trained to be a guard dog or not.
[snip]
Dogs [snip] can tell when there is something "wrong" with their master.
This is absolutely true. Dogs are truly amazing creatures. Dogs in familys with new children are especially intriguing. Most will form an especially close bond with a new baby, and are often more effective at getting a parent's attention than baby monitors. They usually stand guard over the baby whenever someone new is in the house (as often happens with newborns), and if something just doesn't feel right, everyone will know about it.
My dad is blind (no, he doesn't have a dog, but I wish he would get one) and just from experience of me leading him around, it's not possible to do it synthetically. There's simply too much information to process... too many stimuli to evaluate. While a robot can navigate hallways, maybe even curbs, it can't judge if someone is holding a door open for you, or is going to let it slam in your face. I have seen guide dogs do this. It's uncanny how well they can judge a stranger's intentions.
Oh, and for those not in the know, guide dogs will only eliminate on command, so they won't be dropping little brown presents on the floor of the A&P. That's one of the first things they are taught.
I haven't looked into the prices on these specific chips, but this is what the project is looking at:
Xilinx 3S4000 or Xilinx 3S1500
The major benefit of the 3S4000 is more real estate... more room to work with multipliers, etc. The benefit of the 3S1500 is that Xilinx has released a Webpack development tool for it. It's probably cheaper too, but as I said, I haven't checked prices.
I for one don't even care if they don't ever produce a single card themselves, as long as they get the evaluation FPGA board and all the source designs/code complete.
For what it's worth, the evaluation board is what they intend to produce in November. At that point they're looking for funding for a production run.
This is probably way redundant by now, but here goes anyway...
The interval between the counter changing is waaaaay too regular.
That's because it is regular. Check out the source. It periodically checks the current real number through XML, resets the counter to that number, then adjusts the estimated interval that the counter should use until the next update.
You didn't really expect them to have a gazillion clients constantly polling some poor server somewhere to catch each and every single download, did you?
I used to work in that industry too. Typically, bound material would be cut into loose sheets... you basically sacrifice a book to get the images electronically. Also, any decent high volume scanner can scan both sides of a sheet at once, so there's no flipping.
As an unrelated asidem some even scan in color, but your storage requirements go way up if you do anything other than bitonal (even greyscale eats up the bytes pretty quick).
I was moderated down heavily for stating that MD5 was broken for any and all purposes before. Now I think I feel at least somewhat vindicated.
Your statement that MD5 was broken for "any and all" purposes is pretty broad. The bottom line in security is that circumventing a security measure should be more expensive (in terms of money, time, etc.) than the value of what it is trying to protect. If you are trying to protect something that is particularly valuable, then yes, you should go to lengths to ensure that it is adequately protected. On the other hand, if you are trying to protect something that holds no real value, why go to the time and effort?
So you were modded down for it. On this point, I will agree with you. You are entitled to your opinion, and I don't think that you should be squelched for it if your opinion differs from those that have mod points. Others, however, are entitled to post a dissenting opinion, much in the way that I am doing now.
Ok, maybe not "make customers use...", (you're going to get a lot of resistance to that here) but at least let them use the *one port* on the modem to plug in a switch/router/whatever that has a firewall, plus room for more computers to plug in.
The drill is*... 1. Get everything set up according to Comcast's view on how the world should work (e.g. one computer per connection, unless you want to pay extra -- I don't).
2. As soon as they leave (or you get off the phone, depening on how you chose to install), break out the switch/firewall, and plug the computer into that.
3. Spoof the MAC address on the NIC in the computer as the external MAC on the switch (works on Linksys products -- others too, I assume).
4. Plug the switch into the modem. Externally, this just looks like you un-plugged your computer for a few minutes. Note that this is not so much a matter of the MAC must be exactly the same as when they set up your account (people do periodically buy new computers, after all), but rather the coding in the MAC not announce to Comcast "I'm a Linksys Cable/DSL router wih 4 port switch! Guess how many computers are plugged into me! Why don't you bill me extra?"
5. Configure the firewall built into the switch/router/whatever.
6. Make whatever repairs you need to on the computer that was online with no external firewall (installing your favorite Linux distro is a good start).
7. Plug in the other computer(s) that Comcast wanted to charge you an additional "fee" for.
8. Profit!!! -- wait, no, step 8 is just "Don't get screwed!!! (relates both to Comcast and malware)"
* at least according to Comcast rules 2 years ago -- they may have slackened up the rules since.
Yes it would be nice if all broadband connected computers (ok, at least all of them that are in homes) were behind a hardware firewall -- the kind that isn't "off by default" when you have to reinstall your OS. But that's just not realistically going to happen. There are the "you can't tell ME what to do with MY computer" bunch, then there's the group that don't want to be forced to use a specific product, but actually know what they're doing, and finally the clueless that don't want to pay extra for it (e.g. my parents). They want security, but they want it for free.
Target does (at least in Indiana - USA, they do). One time in particular, I signed a little sloppy, and they asked for photo id because the sigs didn't match (normal on the back of the card, sloppy on the terminal).
But your observation is very astute. Target is the only retailer that I recall even *looking* at the back of the card.
Re:Not a problem (yet)
on
SHA-1 Broken
·
· Score: 2, Interesting
Sounds prudent to me, as long as you have the CPU time to spare.
CPU time on legit transactions is not really an issue. Hashes are designed to be computationally expensive to crack, and not computationally expensive to use normally.
So for legit use you have the following (assuming use of MD5 and SHA1):
Time to compute ONE hash via MD5 + Time to compute ONE hash via SHA1 = Ain't much
Compared to what a brute force would have to do:
Time to compute MD5 hashes until you find a match + Time to compute SHA1 hashes based on MD5 matches until you find a match = A whole lot
... or (depending on which is less expensive)...
Time to compute SHA1 hashes until you find a match + Time to compute MD5 hashes based on SHA1 matches until you find a match = A whole lot
Either way, it's a lot more difficult brute force plaintext that will match hashes for both MD5 and SHA1 than for either alone, but the additional CPU time on the system for legitimate use is negligible.
I was right there with you until you regurgitated the *AA marketspeak...
we are talking about thousands of [3-dollar] illegal copies sold, instead of thousands of [15-dollar] legitimate ones
You're not trying to say that every person that purchases a copy for 3 USD would have purchased it for 15 USD if the cheap one were not availiable, are you?
That stinks of *AA logic to me.
There is this novel new economic concept called demand. You might want to give it a look-see.
Slashdot Mirror
Don't know about the Kodaks, but HP has lots of photo printers that don't require you to connect to a PC. AFAIK, any of them that include an LCD screen can do this (you need the screen to see what you are doing). They also have some that work in conjunction with a digital camera that use the LCD of the camera as an interface. This requires a certain kind of camera (I believe they are sold as a bundle), so unless you are shopping for a new camera too, you might look elsewhere.
btw, you can connect these to a PC, but you don't have to, and the ones with a bunch of card slots (SD, compact flash, etc.) also function as a general card reader -- not just for photos -- when connected to a PC.
You forgot...
9. Profit!!!
The only issue left is that it did not recognise the wireless card I have installed.. so I can't put it on the network yet.
;^).
x .php/Installation
I installed Ubuntu on an old laptop a few months ago, and also ran into trouble with the wireless card. Mine was a linksys, and they didn't have a driver for Linux. I was able to use the Windows driver on Ubuntu using NdisWrapper. Assuming you are having similar driver trouble, this may work for you too.
IIRC, NdisWrapper was on the install cd (you might have to apt-get or synaptic it to get it on your HD), so you shouldn't have to download it (which is good, since your card doesn't work yet
Here's a link to the project...
http://ndiswrapper.sourceforge.net/
Here's the install instructions on the wiki...
http://ndiswrapper.sourceforge.net/mediawiki/inde
You'll also need the Windows driver for you card. You can get it off of the CD that came with it, or they are usually available from the manufacturer's website.
Good luck,
-m
Just curious... Would you have the same opinion that the GP was commiting theft (your word, not mine) if s/he was recording the program on VHS? Or is the fact that a digital device is in use somehow make it magically (more) wrong?
That would be great, except updating the text can be a pain.
It's still stuck on "CHA"
Ok, bear with me for a moment...
In this example for simplicity's sake, I'm going to use a ficticious function to shorten the output. The same logic will apply to any hashing function that gives output of a specific length.
Let's give this function a name, so that people don't assume I'm using SHA-1, and flame me for it. Let's call the function "melandy_hash", and say that for whatever input it receives, it gives a 6 digit hex number for the output (yes, I know that this is ridiculously short, but this is a simplified example).
For the input "This is my input string", you might receive the output 82a78b. For the input "This is another input string", you might receive the output 1721ca. Note that although the values for the output are different, the length is the same, because we specified that the output would be a 6 digit hex number.
To expand this further, and to get directly to the point of your question... say we have a list of strings that we want to hash:
"This is my input string"
"This is another input string"
"And another one"
"And yet another one"
"X"
These five strings collectively are the "set" of inputs. When we compute the hashes of these strings using melandy_hash, we get the following output:
82a78b
1721ca
82a78b
1b82ac
97f25b
The above list of hex values is then the "set" of output.
Note that the strings "This is my input string" and "And another one" both have the same hash value. This is known as a collision. As a result, there are really only 4 unique values in out set of output. This means that the set of inputs (with 5 unique elements) is larger than the set of outputs (with 4 unique elements).
Where you may have been confused before is that the output for the string "X" is 97f25b, and for this particular element, the output is larger than the input. The original point that was made was that the number of unique elements in the input set is larger than the number of unique elements in the output set.
Make sense now?
You're missing the point here. The GP stated that a hash is mapping from a large SET of items to a smaller SET [of items]. You are twisting the words to indicate that s/he said that large ITEMS map to small ITEMS, which is not what was stated. It is true that the hash of the string "a string" is larger than the string itself, but to argue that is to miss the point entirely.
Here's the wikipedia article...
http://en.wikipedia.org/wiki/Eye_floater
but that's clearly a typo and not a spelling error
Umm... ever think that icepick was referring to "spelling errors" in the general sense, of which typos are a sub-class?
Way to nit-pick.
The dog is going to go into a state of extreme alertness and defensiveness and will try to let everyone in the house know about what its discovered irrespective of whether it was trained to be a guard dog or not.
[snip]
Dogs [snip] can tell when there is something "wrong" with their master.
This is absolutely true. Dogs are truly amazing creatures. Dogs in familys with new children are especially intriguing. Most will form an especially close bond with a new baby, and are often more effective at getting a parent's attention than baby monitors. They usually stand guard over the baby whenever someone new is in the house (as often happens with newborns), and if something just doesn't feel right, everyone will know about it.
My dad is blind (no, he doesn't have a dog, but I wish he would get one) and just from experience of me leading him around, it's not possible to do it synthetically. There's simply too much information to process... too many stimuli to evaluate. While a robot can navigate hallways, maybe even curbs, it can't judge if someone is holding a door open for you, or is going to let it slam in your face. I have seen guide dogs do this. It's uncanny how well they can judge a stranger's intentions.
Oh, and for those not in the know, guide dogs will only eliminate on command, so they won't be dropping little brown presents on the floor of the A&P. That's one of the first things they are taught.
I haven't looked into the prices on these specific chips, but this is what the project is looking at:
Xilinx 3S4000 or
Xilinx 3S1500
The major benefit of the 3S4000 is more real estate... more room to work with multipliers, etc. The benefit of the 3S1500 is that Xilinx has released a Webpack development tool for it. It's probably cheaper too, but as I said, I haven't checked prices.
I for one don't even care if they don't ever produce a single card themselves, as long as they get the evaluation FPGA board and all the source designs/code complete.
For what it's worth, the evaluation board is what they intend to produce in November. At that point they're looking for funding for a production run.
This is probably way redundant by now, but here goes anyway...
The interval between the counter changing is waaaaay too regular.
That's because it is regular. Check out the source. It periodically checks the current real number through XML, resets the counter to that number, then adjusts the estimated interval that the counter should use until the next update.
You didn't really expect them to have a gazillion clients constantly polling some poor server somewhere to catch each and every single download, did you?
It's Marine One, actually.
http://en.wikipedia.org/wiki/Marine_1
I used to work in that industry too. Typically, bound material would be cut into loose sheets... you basically sacrifice a book to get the images electronically. Also, any decent high volume scanner can scan both sides of a sheet at once, so there's no flipping.
As an unrelated asidem some even scan in color, but your storage requirements go way up if you do anything other than bitonal (even greyscale eats up the bytes pretty quick).
So how many people clicked "No" to the question at the bottom of the linked page (Was This Information Useful) ?
By the way, MSFT seems to like an encoding scheme where Every Word Must Be Capitalized.
If not for that pesky law of diminishing returns, that might be useful.
Your statement that MD5 was broken for "any and all" purposes is pretty broad. The bottom line in security is that circumventing a security measure should be more expensive (in terms of money, time, etc.) than the value of what it is trying to protect. If you are trying to protect something that is particularly valuable, then yes, you should go to lengths to ensure that it is adequately protected. On the other hand, if you are trying to protect something that holds no real value, why go to the time and effort?
So you were modded down for it. On this point, I will agree with you. You are entitled to your opinion, and I don't think that you should be squelched for it if your opinion differs from those that have mod points. Others, however, are entitled to post a dissenting opinion, much in the way that I am doing now.
AMEN!
Ok, maybe not "make customers use...", (you're going to get a lot of resistance to that here) but at least let them use the *one port* on the modem to plug in a switch/router/whatever that has a firewall, plus room for more computers to plug in.
The drill is*...
1. Get everything set up according to Comcast's view on how the world should work (e.g. one computer per connection, unless you want to pay extra -- I don't).
2. As soon as they leave (or you get off the phone, depening on how you chose to install), break out the switch/firewall, and plug the computer into that.
3. Spoof the MAC address on the NIC in the computer as the external MAC on the switch (works on Linksys products -- others too, I assume).
4. Plug the switch into the modem. Externally, this just looks like you un-plugged your computer for a few minutes. Note that this is not so much a matter of the MAC must be exactly the same as when they set up your account (people do periodically buy new computers, after all), but rather the coding in the MAC not announce to Comcast "I'm a Linksys Cable/DSL router wih 4 port switch! Guess how many computers are plugged into me! Why don't you bill me extra?"
5. Configure the firewall built into the switch/router/whatever.
6. Make whatever repairs you need to on the computer that was online with no external firewall (installing your favorite Linux distro is a good start).
7. Plug in the other computer(s) that Comcast wanted to charge you an additional "fee" for.
8. Profit!!! -- wait, no, step 8 is just "Don't get screwed!!! (relates both to Comcast and malware)"
* at least according to Comcast rules 2 years ago -- they may have slackened up the rules since.
Yes it would be nice if all broadband connected computers (ok, at least all of them that are in homes) were behind a hardware firewall -- the kind that isn't "off by default" when you have to reinstall your OS. But that's just not realistically going to happen. There are the "you can't tell ME what to do with MY computer" bunch, then there's the group that don't want to be forced to use a specific product, but actually know what they're doing, and finally the clueless that don't want to pay extra for it (e.g. my parents). They want security, but they want it for free.
But your observation is very astute. Target is the only retailer that I recall even *looking* at the back of the card.
Geez, that's almost 1.85 petafurlongs!
(according to the Google Calculator)
So for legit use you have the following (assuming use of MD5 and SHA1):Compared to what a brute force would have to do:... or (depending on which is less expensive)
You're not trying to say that every person that purchases a copy for 3 USD would have purchased it for 15 USD if the cheap one were not availiable, are you?
That stinks of *AA logic to me.
There is this novel new economic concept called demand. You might want to give it a look-see.
Or if you prefer, that's 81,944.6962 fortnights according to the Google Calculator.