This is what Coda does. Coda supports disconnected operation, resyncronization, and distributed servers.
With coda, when you plug in your laptop (or whatever), you work in online mode, with a local cache. When you pull the (network) plug, you keep the local cache but make changes offline. When you next come online, you resync.
Now, if only it were stable enough... Coda (and all of Odyssey) could really use some nice testing and hacking. And probably porting.
It's nearly entirely consumer's faults that most "consumer" products are absolute crap. Most consumers don't bother figuring out what it will cost to RUN the printer, only what the printer costs. Then they go buy the cheapest one.
If consumers actually compared the total cost of ownership of the printer, as they should, they would realize they are getting screwed over on cartridges. They would realize that the $40 printer isn't really cheap.
But most consumers don't. That's why printer manufactorers have to resort to things like damn-expensive non-standard ink tanks to make any money. There is no way you can sell a $300 black-and-white printer to the average consumer.
Even if it gets 16 pages a minute, has a duplexer, and costs $0.02/page (+paper). Like mine.
Even though at $.15/page, that $40 inkjet costs $790 by the first bale. A lot more if you include the fact that laser paper is $20/bale, while inkject paper is absurd.
Hell, most consumers probably can't do the math to figure the above. But they did it, they'd realize that the only excuse to ever buy a $40 printer is if you print once in a blue moon. Even then, a new cartridge ($25-40) is needed every six months due to drying.
Sure, so basicly, write the software under, e.g., the BSD licence, new version. Then distribute it to a third party (and only that party). They then release under the GPL to the rest of the world.
Why do admins not update? Probably because most distros do all of the following with updates:
Make it hard.
Break things.
Have updates that don't work.
All three of those describe my experiences with RedHat. Until 7.0, it wasn't easy to see if your software was up-to-date. Even with 7.0, you still have to worry about them breaking things. And half the time the dependencies don't work out, especially if I've isntalled something myself. Which is needed quite often, because RedHat doesn't have to great a variety of software in their distro. With older versions, you had to go to updates.redhat.com and chase them down yourself. Half the time, I still have to due to dependency problems, etc. No surprise, anyone ever noticed how hard it is to report a bug in RedHat?
There is a very easy way to do this: The proxy knows (or should know) four pieces of data:
1) The destination IP address
2) The destination hostname
3) The destination path
4) The request type
From there, it's pretty easy to get it right, even in the face of malice. First, the proxy checks how it handles that request type: Cache GET, pass-through POST, pass-through everything else (probably). Next, it uses the set:
{dest. IP addr, dest hostname, dest path,
request type} to determine if the object is cached. If so, it returns the cached copy. Otherwise, it grabs a copy from the destination IP address and follows normal caching procedure (check Expires, Cache-control, ETag, etc.)
The _only_ changes here are that the IP address is part of the object identifier and that no name lookups are done. Not doing name lookups even saves CPU power! OTOH, you lose some cache efficiency to sites with multiple IP addresses. But not much.
No. ISP 'transparent' proxying works by a router notcing requests on a given port and silently redirecting to a proxy server.
The router will see port 80. It will redirect to the proxy. The proxy will do who-knows-what when confronted with an invalid URL. Hopefully, passthrough. Maybe. Maybe not.
[ BTW: You can do all this with iptables under Linux. Read the manpage. Try it out. Then you'll understand it. ]
No doubt it was little fun for the other 10K people who got a mailbombing.
It ain't fun when everyone on a big To: or Cc: line gets into a reply war either. Reply to all, "unsubscribe"...
It really just wastes more bandwidth, more disk space, and most importantly, more time.
I realize spam is damn annoying. Especially a lot of it. There are some spammers who, were they to, ummm, suddenly cease to be amongst the living, I would not be at all saddened; neh, I'd be thrilled. That actually describes a lot of them;-)
But don't turn one spam into a mailbombing. For every extra email the spammer gets (is he even subscribed to the list?), innocent people get 10,000. One hundred mails and you just wasted something like 80 gigabits of bandwidth, disk space, and time.
Which is why they come down on only a few companies. As long as they don't come down on many people, the bean counters look at the situation and ask which is the cheapest way:
1) Switch completely to free software
2) Implement internal software auditing,
configuration control, etc.
3) Ignore the issue, hope it doesn't happen
to us.
Remember, your expected costs are how much it'd cost time the chance of it happening. If three out of 700,000 people who get the nastygram actually get audited, well, that's a pretty low expected cost. Still scary. So, in order to avoice big liabilities, they think of doing something. (1) gets thrown out as being expensive. That leaves (2) and (3) to be considered. The BSA just has to pull a balancing act --- and a rather easy one at that --- to try and make it (2). That gets their members the most money.
The only chance of (1) happening is for a very small (a few people) company, or for a new company. Not much chance of it, as long as the BSA plays its hand right.
The Kursk is many times larger than any spacecraft. The Kursk goes through water, which has a lot more drag than space. So, no, without some hard facts I don't see why the amount of uranium needed for space travel would be a problem.
How many mrem per person per year per kilogram of uranium released in the stratosphere are we even talking about? Does it compare to the 10mrem you can easily get from taking a trip on a jet?
With PURE random data, this won't work. Why anyone would want to transmit gigabytes worth of pure random data is beyond me. A signal worth compressing isn't going to be purely random.
There is use to transmitting large amounts of random data; it's called encryption. Any good ciper should make encrypted data (minus any plaintext headers, etc.) be impossible to tell from random data. Especially if it is a stream cipher! Anything else is a weakness. If you can come up with lossless compressor that works against a cipher, that cipher should not be used.
Data compression, btw, also has this effect, though it's not a weakness but an opertunity to compress missed; any lossless compressor that produces significantly compressable output is pretty pitiful.
Use the correct tool for the job. Perl, for example, would not be my tool of choice for a compiler. INTERCAL would not be my tool of choice for anything except torture, and job security.
The important thing is to use the right tool for the job. If you use the wrong tool, the code will suck more than it otherwise would. And that, of course, is the programmers fault for choosing the wrong tool.
Why does it have to be unplayable in a CD-ROM drive? That's not the only way to stop someone from ripping a disc. It's also not the most effective. Consider, the steps to encoding a packet of MP3 audio are:
Read digital audio data
Convert to MP3, by running through
psycho-acoustics model, etc.
Write out MP3 data
The digital audio data can be obtained from a number of sources, and one of them is certainly the analog data. So you can not stop this step, unless you stop releasing music. Stopping (3) isn't doable either, because you have no control over the hardware or software used.
The most productive would be to stop (2). You just need to introduce, like Macrovision, distortions that are not audible but still screw up MP3 encoding. This would be a copy protection system under the DMCA, and creating an MP3 encoder to get around it (as could probably easily be done) would be illegal. I am very surprised this has not been tried, and hope I haven't given them any ideas!
Nah, it tells you that you must dial the area code before this number, and blah, blah, blah, because, blah, blah, blah, even when you didn't used to have to, blah, blah, blah. Repeat.
Translation: We're the phone company. It's not like you have a choice. PS: Enjoy carpel tunnel;-)
I have never visted those sites --- and both seem to be down now (for obvious reasons) --- so I don't know how those sites are, but a lot of the other sites I visit make you click an `I agree' button.
Those agreements usually reference the site's privacy policy. They expect that those agreements are binding on you, and for that to be the case, wouldn't the same agreement be binding on them to?
Further, I've seen sites that display their privacy policy or other promiss to never give away and/or sell the data on the order page, too. Are they not by making the promiss when I order --- and send them money --- forming a contract with me not to sell the data?
Do the consumers of these sites have any recourse? If they don't, then how should a privacy policy be constructed such that it is legally enforceable?
It's hard for me to immagine that the silly little links at the bottom of a page saying that "by using this site you agree to..." could possibly be valid if their privacy policies aren't.
I'd love to be able to use tools such as XConfigurator. It'd be great, if only they supported my monitor. They sort of support it -- RedHat's tool, for example, let's me select my monitor (an Apple Performa Plus Display), but then there are no valid modeliens in the config file for it... what a lot of help. Maybe most of the world doesn't use 640x480 66.7Hz vsync monitors.
The tools need some serious improvement. Especially since XFree86's FBDev should be able to figure out some acceptable parameters to at least get something on screen; the kernel, open firmware, and the MacOS can.
That's not all. For some reason, even once I write a proper modeline by hand, it won't run in anything more than 8bit color. The X server just crashes whenever I try to start gnome or gimp or who knows what else. Wonderfull. I'm sure a beginning user will be real happy with this. Even 8-bit is a PITA to get the gamma correction right. Maybe 4.0 will be better.
The Xpmac X server, which can autodetect all the settings, and even get the gamma correction reasonable, dose not like anything but 8-bit color. Sure, it can be told to use 16 or 24-bit, but it does not pass that info along to the hardware. And if one DARES to use vmode on the console to set it to 16 or 24-bit color, Xpmac decides that a 66.7Hz refresh rate is far to low, and tries to smoke my hardware. How nice. Only way to get sanity back is to hit the reboot key.
Speaking of vmode, mine starts up at ``0 8''. I'm not even going to try and guess what that 0 means; 6 is the correct value. Attempting to set the vmode to 0 8 give an error (of course). I bet the beginner will be real happy with this.
RPM today decided to install (as a upgrade) a half-downloaded package. You'd think it'd sanity check before installing, no? Of course, you can't even use rpm --verify on an uninstalled package, AFAICT.
gmc croaks every upgrade unless you delete the previous version's config files. And today after an upgrade I have not been able to get it to keep the positions of icons on my desktop (and yes, I checked the preferences. Auto-positioning is off). I even deleted every configuration file remotely related to GNOME. No help.
</RANT>
And I've used Linux for a while. A computer newbie would puke. Linux has a long way to go.
I've never understood the wish to share a scanner like this. What good does it do if you have to go run to the scanner to put the document in, then run back to another computer to hit 'scan', then run to the scanner to put the next page in, etc?
Perhaps I've missed something... but I don't know what. I could see if it was one scanner in the middle of a bunch of computers -- perhaps in a computer lab -- but over a whole house?
Looks promissing. Maybe now we can get a probe to pluton (let's not make it one of those cheaper missions, though).
However, how do they plan to scale this to larger objects? It seems like the sail would break. Also, how much energy would a large laser across the solar system take?
Let's just hope Congree doesn't kill -9 this project.
...but why is it that so-called "lawmakers" fail to realize the great potential set forth by the existence of the internet? After all, at what point in human history has so much information been instantly attainable by anyone on the planet?>
They do realize the great potential of the Internet. They do realize that never before has so much information has been availible. And above all they realize it is a threat to their power.
The Internet is run by pure voluntary association. If we agree to exchange packets, we will; if I don't want to touch your packets, I won't. If I want to recieve email from a certain host, fine. I'll do so. If not, I'll deny the connection. Power on the internet it over yourself and your property. This is quite the opposite of how government -- their power -- works, which is by power over other people and other people's property. The Internet is, plain and simple, a threat to their power. And they don't like that.
When you mix tyrrany and information, what happens? The tyrrany dies out. The Internet is mixing in a serious free-flow of information. So much for their power.
They have no power over the Internet. They can manage to stop a porn store from opening on Main Street, for example. But they can't stop someone from openeing www.porn.com. Major loss of power -- and they don't like it.
If you want to discuss things without the Internet with a large group of people, how must you do it? You need to all get together. Quite expensive, especially if you're spread out geographicly. But with the Internet, there are mailing lists, newsgroups, chats, etc., which can be held with anyone in the world. No need to set up meeting places. No need to travel. No need for passports (a favorite tool of control, no doubt). You just type. Or, with increasing broadband, talk. And it's quite cheap -- FAR cheaper than say an international teleconference! Do you think that they like thta?
Apple's Appletalk drivers are up to version 60.something. Not marketing inflation, though -- no one ever sees those version numbers without looking for them quite dillegently (i.e., without some programming work).
There went my plans send MicroSoft to the other side of the universe quickly by going through the middle... now we'll have to wait a little longer. Ak. Awefull! Or maybe -- just maybe -- they'll fall off the edge:)
Also takes out some good sci-fi plots, unfortunately.
Hmmm... I wonder if the paper companies felt threatened way back when there was the idea of a paperless world from computers? I'm too young to remember.
But anyway, will e-comerce really replace the mall? Um, no. If it would, why didn't the telephone and catalogues? Because there is more at a mall than sometimes-pretty pictures, a blurb written by the marketing team, and an 'order' button.
Give me an e-commerce store where you can chat with friends -- and I don't mean type, I mean look at [in three dimensions, damn it!] and talk to them. -- get something to eat, and shop all in the same bloated building. Maybe with a T1 and some neat hardware you can manage the chat and the shopping, but it still won't be real.
Give me an e-commerce site where you can actually touch and sometimes even try out the stuff before buying it. Maybe with that nice T1 I could try out software by using an X server, but how will I know how it works on my computer? And what if it's not software?
I'm sure there's more, but I don't hang out at malls. Oh well.
Malls need to recognize what they do best -- and do it. I'm sure they'd love to be the only source for merchandice, but there's already a Walmart down the street. And a Fatbrain.com on the 'Net. Both are cheaper. Somehow, I think people go to the mall for more than the products they sell. Maybe they should get their pollsters out and research what it is if they don't know:)
And besides, who can't figure out www.StoreName.com by now?
Slashdot Mirror
Mozilla does have popunder tabs. It's called "Load links in the background", and it's under Preferences->Navigator->Tabbed Browsing.
HTH.
Why are you removing IPs with 255 as the second or third octet? Those are perfectly valid. Indeed, the company I work for has some assigned to it.
Also, 192.0.0.0/8 isn't invalid, only 192.168.0.0/16.
Also, you shouldn't scan 224.0.0.0/7 (multicast)
This is what Coda does. Coda supports disconnected operation, resyncronization, and distributed servers.
With coda, when you plug in your laptop (or whatever), you work in online mode, with a local cache. When you pull the (network) plug, you keep the local cache but make changes offline. When you next come online, you resync.
Now, if only it were stable enough... Coda (and all of Odyssey) could really use some nice testing and hacking. And probably porting.
It's nearly entirely consumer's faults that most "consumer" products are absolute crap. Most consumers don't bother figuring out what it will cost to RUN the printer, only what the printer costs. Then they go buy the cheapest one.
If consumers actually compared the total cost of ownership of the printer, as they should, they would realize they are getting screwed over on cartridges. They would realize that the $40 printer isn't really cheap.
But most consumers don't. That's why printer manufactorers have to resort to things like damn-expensive non-standard ink tanks to make any money. There is no way you can sell a $300 black-and-white printer to the average consumer.
Even if it gets 16 pages a minute, has a duplexer, and costs $0.02/page (+paper). Like mine.
Even though at $.15/page, that $40 inkjet costs $790 by the first bale. A lot more if you include the fact that laser paper is $20/bale, while inkject paper is absurd.
Hell, most consumers probably can't do the math to figure the above. But they did it, they'd realize that the only excuse to ever buy a $40 printer is if you print once in a blue moon. Even then, a new cartridge ($25-40) is needed every six months due to drying.
Sure, so basicly, write the software under, e.g., the BSD licence, new version. Then distribute it to a third party (and only that party). They then release under the GPL to the rest of the world.
Why do admins not update? Probably because most distros do all of the following with updates:
All three of those describe my experiences with RedHat. Until 7.0, it wasn't easy to see if your software was up-to-date. Even with 7.0, you still have to worry about them breaking things. And half the time the dependencies don't work out, especially if I've isntalled something myself. Which is needed quite often, because RedHat doesn't have to great a variety of software in their distro. With older versions, you had to go to updates.redhat.com and chase them down yourself. Half the time, I still have to due to dependency problems, etc. No surprise, anyone ever noticed how hard it is to report a bug in RedHat?
Fortunately, a solution has existed for a while. I have been using it exclusivly for a while.
There is a very easy way to do this: The proxy knows (or should know) four pieces of data:
1) The destination IP address
2) The destination hostname
3) The destination path
4) The request type
From there, it's pretty easy to get it right, even in the face of malice. First, the proxy checks how it handles that request type: Cache GET, pass-through POST, pass-through everything else (probably). Next, it uses the set:
{dest. IP addr, dest hostname, dest path,
request type}
to determine if the object is cached. If so, it returns the cached copy. Otherwise, it grabs a copy from the destination IP address and follows normal caching procedure (check Expires, Cache-control, ETag, etc.)
The _only_ changes here are that the IP address is part of the object identifier and that no name lookups are done. Not doing name lookups even saves CPU power! OTOH, you lose some cache efficiency to sites with multiple IP addresses. But not much.
No. ISP 'transparent' proxying works by a router notcing requests on a given port and silently redirecting to a proxy server.
The router will see port 80. It will redirect to the proxy. The proxy will do who-knows-what when confronted with an invalid URL. Hopefully, passthrough. Maybe. Maybe not.
[ BTW: You can do all this with iptables under Linux. Read the manpage. Try it out. Then you'll understand it. ]
No doubt it was little fun for the other 10K people who got a mailbombing.
;-)
It ain't fun when everyone on a big To: or Cc: line gets into a reply war either. Reply to all, "unsubscribe"...
It really just wastes more bandwidth, more disk space, and most importantly, more time.
I realize spam is damn annoying. Especially a lot of it. There are some spammers who, were they to, ummm, suddenly cease to be amongst the living, I would not be at all saddened; neh, I'd be thrilled. That actually describes a lot of them
But don't turn one spam into a mailbombing. For every extra email the spammer gets (is he even subscribed to the list?), innocent people get 10,000. One hundred mails and you just wasted something like 80 gigabits of bandwidth, disk space, and time.
Which is why they come down on only a few companies. As long as they don't come down on many people, the bean counters look at the situation and ask which is the cheapest way:
1) Switch completely to free software
2) Implement internal software auditing,
configuration control, etc.
3) Ignore the issue, hope it doesn't happen
to us.
Remember, your expected costs are how much it'd cost time the chance of it happening. If three out of 700,000 people who get the nastygram actually get audited, well, that's a pretty low expected cost. Still scary. So, in order to avoice big liabilities, they think of doing something. (1) gets thrown out as being expensive. That leaves (2) and (3) to be considered. The BSA just has to pull a balancing act --- and a rather easy one at that --- to try and make it (2). That gets their members the most money.
The only chance of (1) happening is for a very small (a few people) company, or for a new company. Not much chance of it, as long as the BSA plays its hand right.
Perhaps the cvs(1) manpage would help?
The Kursk is many times larger than any spacecraft. The Kursk goes through water, which has a lot more drag than space. So, no, without some hard facts I don't see why the amount of uranium needed for space travel would be a problem.
How many mrem per person per year per kilogram of uranium released in the stratosphere are we even talking about? Does it compare to the 10mrem you can easily get from taking a trip on a jet?
There is use to transmitting large amounts of random data; it's called encryption. Any good ciper should make encrypted data (minus any plaintext headers, etc.) be impossible to tell from random data. Especially if it is a stream cipher! Anything else is a weakness. If you can come up with lossless compressor that works against a cipher, that cipher should not be used.
Data compression, btw, also has this effect, though it's not a weakness but an opertunity to compress missed; any lossless compressor that produces significantly compressable output is pretty pitiful.
Use the correct tool for the job. Perl, for example, would not be my tool of choice for a compiler. INTERCAL would not be my tool of choice for anything except torture, and job security.
The important thing is to use the right tool for the job. If you use the wrong tool, the code will suck more than it otherwise would. And that, of course, is the programmers fault for choosing the wrong tool.
Why does it have to be unplayable in a CD-ROM drive? That's not the only way to stop someone from ripping a disc. It's also not the most effective. Consider, the steps to encoding a packet of MP3 audio are:
The digital audio data can be obtained from a number of sources, and one of them is certainly the analog data. So you can not stop this step, unless you stop releasing music. Stopping (3) isn't doable either, because you have no control over the hardware or software used.
The most productive would be to stop (2). You just need to introduce, like Macrovision, distortions that are not audible but still screw up MP3 encoding. This would be a copy protection system under the DMCA, and creating an MP3 encoder to get around it (as could probably easily be done) would be illegal. I am very surprised this has not been tried, and hope I haven't given them any ideas!
Nah, it tells you that you must dial the area code before this number, and blah, blah, blah, because, blah, blah, blah, even when you didn't used to have to, blah, blah, blah. Repeat.
;-)
Translation: We're the phone company. It's not like you have a choice. PS: Enjoy carpel tunnel
Those agreements usually reference the site's privacy policy. They expect that those agreements are binding on you, and for that to be the case, wouldn't the same agreement be binding on them to?
Further, I've seen sites that display their privacy policy or other promiss to never give away and/or sell the data on the order page, too. Are they not by making the promiss when I order --- and send them money --- forming a contract with me not to sell the data?
Do the consumers of these sites have any recourse? If they don't, then how should a privacy policy be constructed such that it is legally enforceable?
It's hard for me to immagine that the silly little links at the bottom of a page saying that "by using this site you agree to..." could possibly be valid if their privacy policies aren't.
--
And I've used Linux for a while. A computer newbie would puke. Linux has a long way to go.
--
Perhaps I've missed something... but I don't know what. I could see if it was one scanner in the middle of a bunch of computers -- perhaps in a computer lab -- but over a whole house?
Please enlighten me.
--
However, how do they plan to scale this to larger objects? It seems like the sail would break. Also, how much energy would a large laser across the solar system take?
Let's just hope Congree doesn't kill -9 this project.
--
They do realize the great potential of the Internet. They do realize that never before has so much information has been availible. And above all they realize it is a threat to their power.
--
inflation, though -- no one ever sees those version numbers without looking
for them quite dillegently (i.e., without some programming work).
Hmm... How old is emacs?
--
Try it with "slash dot":
There went my karma...
--
Also takes out some good sci-fi plots, unfortunately.
--
But anyway, will e-comerce really replace the mall? Um, no. If it would, why didn't the telephone and catalogues? Because there is more at a mall than sometimes-pretty pictures, a blurb written by the marketing team, and an 'order' button.
Give me an e-commerce store where you can chat with friends -- and I don't mean type, I mean look at [in three dimensions, damn it!] and talk to them. -- get something to eat, and shop all in the same bloated building. Maybe with a T1 and some neat hardware you can manage the chat and the shopping, but it still won't be real.
Give me an e-commerce site where you can actually touch and sometimes even try out the stuff before buying it. Maybe with that nice T1 I could try out software by using an X server, but how will I know how it works on my computer? And what if it's not software?
I'm sure there's more, but I don't hang out at malls. Oh well.
Malls need to recognize what they do best -- and do it. I'm sure they'd love to be the only source for merchandice, but there's already a Walmart down the street. And a Fatbrain.com on the 'Net. Both are cheaper. Somehow, I think people go to the mall for more than the products they sell. Maybe they should get their pollsters out and research what it is if they don't know :)
And besides, who can't figure out www.StoreName.com by now?
--