Scroll bars for these things are retarded. Especially when viewing web pages that require horizontal scrolling. It is no fun at all to have to use a pen to point at a teeny scroll bar at the bottom and drag it back and forth to read a couple of sentences, and then move the pen to a teeny scroll bar on the right to have to move to the next sentence. PDA web browsers/document viewers should take a hint from GhostView and allow a pen click & drag anywhere on the screen drag the page around. Targets should also be large enough such that you can use your index finger to accurately touch them.
I've been thinking about creating a set-based file system. Instead of things being in hierarchal directories, each file is a member of one or more sets. While finding a file, you can specify one or more sets to which that file belongs, rather than trying to drill down a specific directory tree.
For example:
You have an e-mail sent to you from your friend Bob about his dog last month. You could place it in the following sets:
email bob dog august 2003 boring
Say sometime later you remember Bob sent you something boring during the month of August. Simply select the following sets in your file selection box:
bob august boring
And voila! You have all the boring e-mails from Bob in the month of August.
Because of the mathematical nature of sets, many more operations can be performed, such as unions and stuff.
The neato thing is, hierarchal directory structures are a direct subset of sets: To emulate the structure/home/username, you could have sets/home and/home/username, the latter of which would be defined to be a subset of home, so that all files belonging to the set/home/username immediately belong to/home!
Set filesystems can still be access via heirarchal means, too: A set file system driver on something like Linux would have a fixed, single-level directory structure consisting of all the sets (as directories). Accessing 'subfolders' would cause dynamic name binding to those set intersections.
Set filesystems also can be extremely useful for database structures (especially when implemented like ReiserFS!) but I won't get into that here:).
This is nothing new, there's one at ThinkGeek for $17.99 from IOGear. I have it and it works fairly well, except for the fact that no Linux application recognizes horizontal scrollage. Fortunately, I still get use out of the trackball, since the mouse can switch in hardware to 'trackball' mode, where the trackball and the mouse proper switch roles. Moving the mouse up and down to scroll is a lot easier than twiddling a roller or a ball.
If I buy stock in a company (even the one I work for), that doesn't mean that I can freely use any software that they buy from another vendor. Most software comes with per-seat licenses, not per-company. What's to stop music companies from just packaging only a single-user license in a CD? Replace the word 'music' with 'software' in this scheme, and it all falls apart.
Steve Howe of Yes had a digital guitar custom-built by Stepp Ltd. in 1987, but he couldn't quite get the hang of playing it. So now it's on display in the Dangerous Curves exhibit at the Boston Museum of Fine Arts.
It's no wonder these servers have so many problems - there's thirteen of them! They need a lucky #14 - a Bilbo Baggins for their horde of dwarves. That'll stop those DoS attacks and unnecessary requests right away!
After winning $100 in our high-school "battle-of-the-bands", my band of three people decided to make an album. We each chipped in another $66 for a total of $300, and with this money, recorded for 6 hours in a decent, small studio (small room, one employee). We made one hundred copies ourselves using plain-old CD-RW drives. The 100 CDs were $10, and the cases another $30 (I think). The guitarist's mom (who is a graphics designer) did the art, using pictures the guitarist took for his photo class. We printed those on store-bought CD inserts and labels ($10, maybe?), and stuck them together ourselves. Total cost? $350. We sold each CD for $10 ($5 for close friends).
For $350, the quality of the album is pretty darn good, a whole lot better than our self-recorded one. All these recording costs seem simply absolutely ridiculous to me! Of course, professionals don't do 6-hour studio jobs, but then even The Police spent only $2000 for their first signed album.
I go to Worcester Polytechnic Institute, and we just recently had a debate about mandating laptops in college. Trick is, they're not included in the fees. So you could theoretically buy your own, but that's still at least $1500 extra you have to shell out. The debate was very interesting; I'm not going to try to cover it all here, but if you follow the link above, you will find survey results, minutes, and even a video of the debate. Very interesting stuff, good debate. The basic point is, a laptop is nice, but it's definitely not a requirement and would be a waste of money for a lot of people.
The thing about Linux TV ads is that your average person does not want to be shown screen shots of Mozilla, etc., and see that they look exactly like what they're already using. That's boring. They probably won't even think it's Linux. So this is not what we want.
Take Microsoft commercials for example. They either involve multicolored flying butterflies or people discussing their crazy business solutions. These are obviously the kinds of ads that are succesful, and the kind that we should duplicate.
IBM has already duplicated the latter type in their Linux-based server ads: remember the one where the servers have gone missing, but they've simply all been replaced with one server running Linux? So we're pretty much all set in that department.
That leaves the first type. Pretty butterfly. Dancing happy people. Hm. For some reason that doesn't seem to fit our image all too well. Okay then, take Apple's ad campaign. Remember the "Think Different" ads? I've found the people who most want to switch to Linux are those who just don't want to be fettered by Microsoft. Rather than showing facts, show the symbolic advantage of Linux over Windows:
Show hordes of distraught people in chains and handcuffs trying to follow a fluttering butterfly climbing up a mountain or through pricker bushes or something, and a few others who have cast off their chains and are skipping along happily after a large penguin running through a field, a pied-piper sort of thing. The fact that we have a fuzzy mascot is actually probably our best weapon in convincing the "average" person to switch.
Parallel Illiad's idea, and show a Star Destroyer or something chasing after a small lonely X-Wing (perhaps piloted by Tux). The Star Destroyer gets a BSOD, GPF, or some other easily recognizable bane of Windows users (but not Windows-specific, to avoid lawsuits). It needs to reboot Windows. The X-Wing comes around, and ten seconds later (emphasize that Windows needs lots of time to reboot, which it does) KABOOM! no more X-Wing.
The main thing to remember is, people don't care about facts. Maybe businesses do, but IBM's got that covered already (if they stop, just follow their example). But people don't. They like colors. They like humor. They like special effects. They like fuzzy. They like Star Wars. Oh, and they really like free internet. A RedHat- or Mandrake-sponsored version of MSN would be wonderful.
But face it. People want entertainment. Facts aren't entertainment. Facts are when you get another sandwich. Entertainment keeps you glued to the screen those extra thirty seconds: thirty seconds of Linux.
Wow, first PetsWarehouse suing people for badmouthing their company, now this. It almost makes me ashamed to be a US citizen, that our country lets things like this happen. This is what destroys our economy, large conglomerates suing the pants off of little businesses. This helps nothing. If only patents still numbered in the thousands, not the millions (or is it billions?) that they do now.
I agree for the most part with Pinball Wizard, except that there is a difference between One-Time Pad and a similar system that uses incremental chunks from a huge key.
With one-time pad, a new key needs to be generated and delivered each time a new message is to be sent. Assuming that the key and the message, being of equal size, are both sent over the same insecure medium, there is still a chance that they can both be intercepted and the message decoded.
The advantage of using small parts of a huge key one after another is that the huge key needs to be transferred once. Example:
You generate a 32MB key, copy it onto a memory stick, and physically give it to your friend. Now you can encrypt over 32,000 1KB messages using that key, without ever having to transfer the key again. This means you can send your messages over a thouroughly insecure medium, but they cannot be cracked because the key is never transmitted over this medium.
"Why not just transmit smaller keys over a secure medium?" you might ask. In that case, you might as well just transfer the message itself over that same medium, and save yourself some bandwidth.
This is all based on an assumption of how "Multi-Time Pad" encryption works, of course.
Most modern printers are terrible - they don't support PostScript, they have no internal memory, they hold a miniscule amount of paper, and they get jammed often. My family's Lexmark inkjet is case in point - it holds about 30 sheets, has no memory, and only uses Lexmark's "jnl" format. Laser printers are somewhat better, but I've no expreience with them.
Me? I use an Apple Imagewriter II. Sure, it doesn't support PS, but that's what ghostscript is for (does a nice job, too). Never jams, has unlimited paper supply (the paper is stored externally), almost never gets jammed, and even has 2KB memory in it, upgradable to 32KB! Most printers die after a few years, but this one's twelve years old and running strong!
I'm not a lawyer, but my assumption is that patents only become infringed if those ideas are used in commercial software, which Linux is not. Using patented algorithms in Linux is just like using patented algorithms in some code you wrote up at home. If anything, companies such as RedHat are infringing on the patent, but even that's iffy.
I just graduated high school, yet never had a powerful graphing calculator (Casio's aren't terribly programmable). But everyone I knew who had a TI had no clue what more than half the functions on it did; they merely used them to play games (as the few who owned PDAs did). Unfortunately, their power is dulled by the fact that they are so slow; an equivalently-priced PDA can do the same types of calculations in 1/10th the time. (I can't wait to stick a Scheme interp. on my Zaurus!) PDAs are currently banned because they are "programmable". But so are all graphing calculators. On SATs, the only things that are banned are devices housing QWERTY keyboards, which most PDAs don't. Also, TIs can be programmed (and come with) more functionality than your average Palm. Even my Zaurus comes with only a 4-function calculator app! Back on the topic of the CASIO, I left it at home nearly every other day of school, if even that infrequently. Yet I survived through every math and physics class often without it. Because of graphing calculators, most kids don't even know what a parabola looks like, let alone how to draw one. Most people even forget fractions and long division, and rather write the answer the calculator gives them, like "3.999999999" rather than "4".
Both calculators and PDAs are tools, and should/not/ be used as learning tools. Kids learn to use them to do math, rather than the actual underlying concepts. Don't allow 4-function calculators until algebra; don't allow graphing calculators until calculus; don't allow scheme-based RPN symbolic integration magic twiddles until set theory!
My favorite alternative is Arachne. It's a fullscreen graphical browser available for DOS and for Linux using SVGAlib/GGI. It supports tables like nothing I've ever seen, is extremely fast, very reliable, and supports ftp/smtp/pop3/etc. I used to even run the DOS version on a 386SX/20 with 4MB ram - a little bit slow, but it ran fine! It does not yet support Java(Script) yet, but JavaScript is in the works. It is among my favorites for Linux, and is the BEST for DOS.
IMHO, the best way to implement a "universal" translator is using an intermediate language capable of expressing most, if not all, concepts. This could be either a pseudo-language understandable only by computers, or a real language, such as Esperanto. Then, all non-directly-translatable phrases, such as "there are", "thank you", and other such abstract concepts, would only have to have one "mapping" per language, linking that language to the pseudo-language. The real time / two-way stuff is nothing; just work off of previous "talk & type" things, and find a way to prevent feedback loops.
I believe the Sharp Zaurus comes with Java - if not, you could load it with Kaffe, then get a Java-based NES emulator. I don't know how slow this would be (shouldn't matter with NES, and Kaffe is fast), but a similar approach would work on many other PDAs, too.
I certainly agree! I've never typed on a ergo keyboard for more than five minutes at a time, but I've immediately run into the same "which-side-of-the-break?" problem, especially since I tend to use both hands to type on either side of the keyboard, rather than keeping each hand confined to its area. As to the neccessity of ergo keyboards, I find that in order the prevent various disorders/pain, just keep your wrists in line with your arms and your fingers and you'll be fine - the keyboard I'm on now has a little plastic wristwrest that does a good job of that. On other computers, I'll use a piece of wood. Angling your hands as if you were using an ergo keyboard helps, too. So does Dvorak:). Cogito, ergo ne ergo!
How could such a game be terribly different from another other first person shooter, especially those that were specifically designed to simulate an army experience (minus the bathroom duty, etc.). I'm guessing the sole difference is that you'd see army advertisments everywhere, and the music would be bad. And I'm sure the won't have boot camp instructors yelling their heads off, etc., but rather the classic "shaking-jaw-scrolling-text-so-slow-can't-speed-it -up-but-can-skip-it-with-a-keypress" mission briefs... "Hey, did you know that these violent games that you love to shoot people in are based on real life? In the next level, there are no extra lives - join the Army!"
Slashdot Mirror
The object /cm/cs/who/dmr/otherunix.html does not exist on this server.
/cm/cs/who/dmr/ does not exist on this server.
/cm/cs/who/ does not exist on this server.
/cm/cs/ does not exist on this server.
/cm/ does not exist on this server.
The object
The object
The object
The object
Talk about brute force.
More complaints about PDA GUIs:
Scroll bars for these things are retarded. Especially when viewing web pages that require horizontal scrolling. It is no fun at all to have to use a pen to point at a teeny scroll bar at the bottom and drag it back and forth to read a couple of sentences, and then move the pen to a teeny scroll bar on the right to have to move to the next sentence. PDA web browsers/document viewers should take a hint from GhostView and allow a pen click & drag anywhere on the screen drag the page around. Targets should also be large enough such that you can use your index finger to accurately touch them.
As to "GPLed QT alternatives", there's also Opie.
You can do this in DOS, too, (just set the filenames to the same FAT chain), but scandisk certainly doesn't like it....
I've been thinking about creating a set-based file system. Instead of things being in hierarchal directories, each file is a member of one or more sets. While finding a file, you can specify one or more sets to which that file belongs, rather than trying to drill down a specific directory tree.
/home/username, you could have sets /home and /home/username, the latter of which would be defined to be a subset of home, so that all files belonging to the set /home/username immediately belong to /home!
:).
For example:
You have an e-mail sent to you from your friend Bob about his dog last month. You could place it in the following sets:
email
bob
dog
august
2003
boring
Say sometime later you remember Bob sent you something boring during the month of August. Simply select the following sets in your file selection box:
bob
august
boring
And voila! You have all the boring e-mails from Bob in the month of August.
Because of the mathematical nature of sets, many more operations can be performed, such as unions and stuff.
The neato thing is, hierarchal directory structures are a direct subset of sets: To emulate the structure
Set filesystems can still be access via heirarchal means, too: A set file system driver on something like Linux would have a fixed, single-level directory structure consisting of all the sets (as directories). Accessing 'subfolders' would cause dynamic name binding to those set intersections.
Set filesystems also can be extremely useful for database structures (especially when implemented like ReiserFS!) but I won't get into that here
Good thing they didn't skimp and only take one year to come up with it; it might have sucked.
(w/ apologies to Scott Adams)
This is nothing new, there's one at ThinkGeek for $17.99 from IOGear. I have it and it works fairly well, except for the fact that no Linux application recognizes horizontal scrollage. Fortunately, I still get use out of the trackball, since the mouse can switch in hardware to 'trackball' mode, where the trackball and the mouse proper switch roles. Moving the mouse up and down to scroll is a lot easier than twiddling a roller or a ball.
There's a major (future) flaw in this:
If I buy stock in a company (even the one I work for), that doesn't mean that I can freely use any software that they buy from another vendor. Most software comes with per-seat licenses, not per-company. What's to stop music companies from just packaging only a single-user license in a CD? Replace the word 'music' with 'software' in this scheme, and it all falls apart.
"The roughly 339lb car (*with 517lb driver*) with 1500 watts of power won by nearly 5 hours"
:)
Was I the only one who read it like this?
Anyways maybe some of the technologies used in this car could be applied to `standard' electric cars.
Steve Howe of Yes had a digital guitar custom-built by Stepp Ltd. in 1987, but he couldn't quite get the hang of playing it. So now it's on display in the Dangerous Curves exhibit at the Boston Museum of Fine Arts.
It's no wonder these servers have so many problems - there's thirteen of them! They need a lucky #14 - a Bilbo Baggins for their horde of dwarves. That'll stop those DoS attacks and unnecessary requests right away!
After winning $100 in our high-school "battle-of-the-bands", my band of three people decided to make an album. We each chipped in another $66 for a total of $300, and with this money, recorded for 6 hours in a decent, small studio (small room, one employee). We made one hundred copies ourselves using plain-old CD-RW drives. The 100 CDs were $10, and the cases another $30 (I think). The guitarist's mom (who is a graphics designer) did the art, using pictures the guitarist took for his photo class. We printed those on store-bought CD inserts and labels ($10, maybe?), and stuck them together ourselves. Total cost? $350. We sold each CD for $10 ($5 for close friends).
For $350, the quality of the album is pretty darn good, a whole lot better than our self-recorded one. All these recording costs seem simply absolutely ridiculous to me! Of course, professionals don't do 6-hour studio jobs, but then even The Police spent only $2000 for their first signed album.
I go to Worcester Polytechnic Institute, and we just recently had a debate about mandating laptops in college. Trick is, they're not included in the fees. So you could theoretically buy your own, but that's still at least $1500 extra you have to shell out. The debate was very interesting; I'm not going to try to cover it all here, but if you follow the link above, you will find survey results, minutes, and even a video of the debate. Very interesting stuff, good debate. The basic point is, a laptop is nice, but it's definitely not a requirement and would be a waste of money for a lot of people.
The thing about Linux TV ads is that your average person does not want to be shown screen shots of Mozilla, etc., and see that they look exactly like what they're already using. That's boring. They probably won't even think it's Linux. So this is not what we want.
Take Microsoft commercials for example. They either involve multicolored flying butterflies or people discussing their crazy business solutions. These are obviously the kinds of ads that are succesful, and the kind that we should duplicate.
IBM has already duplicated the latter type in their Linux-based server ads: remember the one where the servers have gone missing, but they've simply all been replaced with one server running Linux? So we're pretty much all set in that department.
That leaves the first type. Pretty butterfly. Dancing happy people. Hm. For some reason that doesn't seem to fit our image all too well. Okay then, take Apple's ad campaign. Remember the "Think Different" ads? I've found the people who most want to switch to Linux are those who just don't want to be fettered by Microsoft. Rather than showing facts, show the symbolic advantage of Linux over Windows:
Show hordes of distraught people in chains and handcuffs trying to follow a fluttering butterfly climbing up a mountain or through pricker bushes or something, and a few others who have cast off their chains and are skipping along happily after a large penguin running through a field, a pied-piper sort of thing. The fact that we have a fuzzy mascot is actually probably our best weapon in convincing the "average" person to switch.
Parallel Illiad's idea, and show a Star Destroyer or something chasing after a small lonely X-Wing (perhaps piloted by Tux). The Star Destroyer gets a BSOD, GPF, or some other easily recognizable bane of Windows users (but not Windows-specific, to avoid lawsuits). It needs to reboot Windows. The X-Wing comes around, and ten seconds later (emphasize that Windows needs lots of time to reboot, which it does) KABOOM! no more X-Wing.
The main thing to remember is, people don't care about facts. Maybe businesses do, but IBM's got that covered already (if they stop, just follow their example). But people don't. They like colors. They like humor. They like special effects. They like fuzzy. They like Star Wars. Oh, and they really like free internet. A RedHat- or Mandrake-sponsored version of MSN would be wonderful.
But face it. People want entertainment. Facts aren't entertainment. Facts are when you get another sandwich. Entertainment keeps you glued to the screen those extra thirty seconds: thirty seconds of Linux.
Wow, first PetsWarehouse suing people for badmouthing their company, now this. It almost makes me ashamed to be a US citizen, that our country lets things like this happen. This is what destroys our economy, large conglomerates suing the pants off of little businesses. This helps nothing. If only patents still numbered in the thousands, not the millions (or is it billions?) that they do now.
I agree for the most part with Pinball Wizard, except that there is a difference between One-Time Pad and a similar system that uses incremental chunks from a huge key.
With one-time pad, a new key needs to be generated and delivered each time a new message is to be sent. Assuming that the key and the message, being of equal size, are both sent over the same insecure medium, there is still a chance that they can both be intercepted and the message decoded.
The advantage of using small parts of a huge key one after another is that the huge key needs to be transferred once. Example:
You generate a 32MB key, copy it onto a memory stick, and physically give it to your friend. Now you can encrypt over 32,000 1KB messages using that key, without ever having to transfer the key again. This means you can send your messages over a thouroughly insecure medium, but they cannot be cracked because the key is never transmitted over this medium.
"Why not just transmit smaller keys over a secure medium?" you might ask. In that case, you might as well just transfer the message itself over that same medium, and save yourself some bandwidth.
This is all based on an assumption of how "Multi-Time Pad" encryption works, of course.
Most modern printers are terrible - they don't
support PostScript, they have no internal memory,
they hold a miniscule amount of paper, and they
get jammed often. My family's Lexmark inkjet is
case in point - it holds about 30 sheets, has no
memory, and only uses Lexmark's "jnl" format.
Laser printers are somewhat better, but I've no
expreience with them.
Me? I use an Apple Imagewriter II. Sure, it
doesn't support PS, but that's what ghostscript
is for (does a nice job, too). Never jams, has
unlimited paper supply (the paper is stored
externally), almost never gets jammed, and even
has 2KB memory in it, upgradable to 32KB! Most
printers die after a few years, but this one's
twelve years old and running strong!
I'm not a lawyer, but my assumption is that
patents only become infringed if those ideas are
used in commercial software, which Linux is not.
Using patented algorithms in Linux is just like
using patented algorithms in some code you wrote
up at home. If anything, companies such as RedHat
are infringing on the patent, but even that's iffy.
I just graduated high school, yet never had a powerful graphing calculator (Casio's aren't terribly programmable). But everyone I knew who had a TI had no clue what more than half the functions on it did; they merely used them to play games (as the few who owned PDAs did). Unfortunately, their power is dulled by the fact that they are so slow; an equivalently-priced PDA can do the same types of calculations in 1/10th the time. (I can't wait to stick a Scheme interp. on my Zaurus!)
/not/ be used as learning tools. Kids learn to use them to do math, rather than the actual underlying concepts. Don't allow 4-function calculators until algebra; don't allow graphing calculators until calculus; don't allow scheme-based RPN symbolic integration magic twiddles until set theory!
PDAs are currently banned because they are "programmable". But so are all graphing calculators. On SATs, the only things that are banned are devices housing QWERTY keyboards, which most PDAs don't. Also, TIs can be programmed (and come with) more functionality than your average Palm. Even my Zaurus comes with only a 4-function calculator app!
Back on the topic of the CASIO, I left it at home nearly every other day of school, if even that infrequently. Yet I survived through every math and physics class often without it. Because of graphing calculators, most kids don't even know what a parabola looks like, let alone how to draw one. Most people even forget fractions and long division, and rather write the answer the calculator gives them, like "3.999999999" rather than "4".
Both calculators and PDAs are tools, and should
My favorite alternative is Arachne. It's a fullscreen graphical browser available for DOS and for Linux using SVGAlib/GGI. It supports tables like nothing I've ever seen, is extremely fast, very reliable, and supports ftp/smtp/pop3/etc. I used to even run the DOS version on a 386SX/20 with 4MB ram - a little bit slow, but it ran fine! It does not yet support Java(Script) yet, but JavaScript is in the works. It is among my favorites for Linux, and is the BEST for DOS.
IMHO, the best way to implement a "universal" translator is using an intermediate language capable of expressing most, if not all, concepts. This could be either a pseudo-language understandable only by computers, or a real language, such as Esperanto. Then, all non-directly-translatable phrases, such as "there are", "thank you", and other such abstract concepts, would only have to have one "mapping" per language, linking that language to the pseudo-language. The real time / two-way stuff is nothing; just work off of previous "talk & type" things, and find a way to prevent feedback loops.
I believe the Sharp Zaurus comes with Java - if not, you could load it with Kaffe, then get a Java-based NES emulator. I don't know how slow this would be (shouldn't matter with NES, and Kaffe is fast), but a similar approach would work on many other PDAs, too.
I certainly agree! I've never typed on a ergo keyboard for more than five minutes at a time, but I've immediately run into the same "which-side-of-the-break?" problem, especially since I tend to use both hands to type on either side of the keyboard, rather than keeping each hand confined to its area. As to the neccessity of ergo keyboards, I find that in order the prevent various disorders/pain, just keep your wrists in line with your arms and your fingers and you'll be fine - the keyboard I'm on now has a little plastic wristwrest that does a good job of that. On other computers, I'll use a piece of wood. Angling your hands as if you were using an ergo keyboard helps, too. So does Dvorak :).
Cogito, ergo ne ergo!
How could such a game be terribly different from another other first person shooter, especially those that were specifically designed to simulate an army experience (minus the bathroom duty, etc.). I'm guessing the sole difference is that you'd see army advertisments everywhere, and the music would be bad. And I'm sure the won't have boot camp instructors yelling their heads off, etc., but rather the classic "shaking-jaw-scrolling-text-so-slow-can't-speed-it -up-but-can-skip-it-with-a-keypress" mission briefs...
"Hey, did you know that these violent games that you love to shoot people in are based on real life? In the next level, there are no extra lives - join the Army!"