What I envision is that when you move the mouse quickly across the screen, instead of seeing 6 simultaneous pointer images in a neat line, the input handler (Windows mouse driver) would blur the image from the last displayed position to the current one. This would be good, IMHO, because something I see at college once in a while is the mouse run up to full speed... when you expect it to go 1/4 of the way across the screen and it slams into the opposite wall, a blur would give a clue as to where the pointer went.
The blur itself need be nothing fancy... just take the average color of the trailing two mouse pixels and smack down a couple polygons of that color from current to last position.
Actually, I was rather surprised to hear about/see those ballots. In my state, we use automatic voting machines... walk in, pull red handle, pull down one lever per column, push red handle, leave. No holes, no alignment problems.
There are two problems, though. When you pull a lever, a red-on-white "X" is visible, implying "No." And any propositions are up above the candidate lists, visually lumped with the voting directions. Only 1/3 of the voters that turned out actually voted on the proposition this year (to run $3.8 billion into debt to fix roads.)
By "clean up the damn system" you mean campaign finance reform, right? Nobody cares about low-probability negative events like the Earth getting hit by an asteroid or the "wrong" candidate winning the election anyway. (They do care about low-probability positive events, like winning the lottery.)
If you can make it, someone will break it. I have absolute faith that something can be encrypted such that {DNA|quantum|brute force} methods are all infeasible. What if your cipher doesn't have discrete states?
Software patents enable Compuserve, for example, to patent a compression algorithm or a program that reads or writes a specific file format.
[href:Why there are no GIF files on GNU web pages] Both Unisys and IBM have the LZW patent. CS didn't do their homework, and Unisys didn't bother anyone until 1995 (8 years after the creation of the GIF format) when it was thoroughly entrenched. This is Patenting Done Wrong: the _end result_ was patented.
Then there's PKZIP: the actual implementation was patented. Other software can't use PKWare source code, but they (WinZip, WinRar, others) can create the ZIP file format without infringing on the PK patent. Patenting Done Right: the _implementation_ was patented.
There's always another way to do something (ask Rube Goldberg)... end-result patents are the only bad ones. Copyrights, well... I'm commenting on patents.
We're being forced to use Network Essentials in class. The problem is, it's inconsistent---after describing the job of each layer of the OSI model, they then say that TCP goes _only_ with the transport layer. Hello? TCP's jobs are ensuring reliable delivery (transport layer) AND establishing a session (session layer)! What do they think SYN/ACK/FIN are for? It's also Pro-MS: they only mention tools in the interoperability section that allow NT to work with Netware---NOTHING of the reverse. In the Survey of NOSes chapter, WinNT and Novell each get 3 pages of decription, Apple gets 1 page, Banyan Vines gets almost a page, and *nix, Win9x, and OS/2 each get a paragraph. Under creating accounts, NT gets 4 pages with illustrations, and *nix just gets 2 paragraphs (which never mention (user|group)(add|mod|del).
</rant>
More objectively, it's just old. Who needs to know everything (terminator resistance) about Thinnet and Thicknet? ArcNet? This from the third edition (2000)... shouldn't the bugs be out by now?
As supporting evidence for your distribution model, I would like to point out the piles of shareware on my hard drive that I paid for... oh, wait...... I didn't. Why? Convenience.
If something is convenient in Form X, Marketing can always sell it. If two competing products are convenient, then the one with the biggest edge will win (Betamax vs. VHS). If the edge isn't big enough, both proliferate (Phillips vs. slotted screws).
Ogg Vorbis is actually two layers: layer "Ogg" controls the timing during playback, and "Vorbis" is the actual compressed audio. When Vorbis is done and they move on to video, it'll be Ogg X, using the Ogg timing layer with the X video compression. (X=placeholder. They'll come up with something far more creative for a name.)
No URL, but it was on the old site (before they moved to xiph.org)
You are assuming that DigitalColon^H^H^H^H^HConvergence cannot change its business model. The market is supposed to change in response to pressures... not invoke the legal system and destroy the pressure.
...LACK of RESPECT for OTHER people's LIFE, and LIBERTY.
Consider this: I invent a perpetual motion machine. I sell it in a black box. Then somebody "reverse engineers" it. The information is now free, and demand drops considerably (as people build their own.) Now... is that disrespecting my life, as my income and quality of life drops? Is it disrespecting my liberty to sell however I please?
Consider this quote from the page you linked to: "(Rand presumably got royalties from the sale, and I'm sure it wasn't sold without her permission!)" But if someone duplicates Atlas Shrugged and sells copies, it _is_ sold without permission or payment! It is also not freedom of speech/press---Congress is making no law about who can print it. That's all the 1st Amendment covers... Congress making laws.
The whole anti-any-IP argument is flawed: it takes much less effort to copy than create. I favor _limited_ IP. Anyone can use it to preserve life; the monopoly disappears in 7 years; the granters must understand the device or be shown a working implementation (the inventor of the transformer was laughed out of the office for trying to patent a short circuit, because that's what the patent examiner saw it as. Soon after, the laws were reformed so the PE need not understand the device.)
Does anyone know of methods for distinguishing between various inks/toners placed on a page?
I would start with looking at the reflected spectrum of various wavelengths (infrared..ultraviolet) from the back of the paper. Any wavelength that makes the two toners reflect differently will do for reading. The problem is resolution... just telling the toners apart is easier than getting an exact picture. Maybe a laser could scan it, and reflectivity would be sent to a computer for image reconstruction.
that's where the problem lies; in using mac addresses to form IP addresses.
This idea is screaming, "LAME!" MAC addresses are only 48 bits; IPv6 allows 128 bit addresses. If the IPv6 designers thought 48 bits would be enough, they should go back and listen to Bill Gates say, "640K of memory is all anyone will ever need." They should look at our current 32 bit addressing scheme. They should look at me, connecting over a 14.4 modem.
wrt DickBreath's reply, I don't know what'll replace Ethernet... but I envision going back to coax and using broadband on it (instead of baseband.) If you can cram video information for 50 cable channels on it, it should be reasonably high bandwidth.
Preferably one where you could run any untrusted app in a chroot jail. Is that even possible under Win32?
Under Windows 9x, the closest you could come is a separate partition. (Got Partition Magic?) But if it's "sprinkling bits", it would have to know where to look for them (Registry entries), and they'd probably show up as files anyway (otherwise Defrag would kill your Napster.) The best way to hide a file would be as a.dll in \windows\system...
If the newbie is clueless enough to put up a server with everything loaded on it despite the numerous FAQs and warnings, I don't have a lot of sympathy and DON'T blame RedHat.
OK, look at it from a newbie perspective. You bought Linux without ever checking hardware (hell, you don't even know what it is, so if X wants to know refresh timings, you're SOL.) You have a 500 page manual and a CD that seems to tell you what to do on-screen. Then you get to the point where you have to make a choice: do you decide to try and figure out where in the manual to look, or just bonk something at random? </simulation> Unfortunately, many newbies _will_ just bonk something. I suggest a second, smaller booklet titled "Foo Linux Installation Guide" which briefly runs through the installation and makes security notes on the way. It'd look something like this:
Next, you will meet a screen listing the possible daemons (services) that can be started when booting. These services MUST be configured to be safe! They are:
inetd - inetd will start other daemons whenever anyone tries to connect to them over the Internet. See page 311 in the manual for configuring inetd.
ftpd - Serves files to anyone over the Internet that connects with an FTP client. For security, the default directory for ftpd to get files from is/ftproot, which is empty. See page 314 in the manual for configuring ftpd.
And so on. Then newbies would have security info and a brief description of what it does on hand and usable during installation, plus a pointer to more information.
If I see another retarded consumer user say that compaqs suck, I will scream.
Then scream. I will resent any company who never bothered to make a 56k modem for their computer.
I did spend an entire Saturday wrangling with a 3rd party 56k, trying to install it on COM1:. Nothing worked.
People please understand that HOME PCs that you get in one wrapped up plastic shell from BigStoreCo are going to be shit, regardless of brand.
That's why I'm going custom-built next time I can afford a computer. It's just that I have a good reason to be pissed at Compaq.
Compaq makes what I consider to currently be the world's best Intel-based servers, in terms of functionality, cost, and price/performance.
Don't confuse Proliant with Presario.
I'm not confusing them. I'm only working within my range of experience (Presario 7222.) If they don't want me browsing any faster than 28.8, I'm not supporting them. If they wanted (more of) my money, they shouldn't have built this PoS.
There's a reason I use FTP for site management: if someone wants my site taken down, I can move it easily.
If you keep server logs, you could locate whoever uploaded the offending piece and redirect fire unto them. After all, if you didn't post it, you shouldn't need to be responsible for it.
The rant applies to any company that pulls this same type of crap. the fact that its a free product is irrelevant.
No it isn't. If it's a loss leader, they lose money every time you buy one! Every time you buy one, it means they have to produce another and hope they sell to someone who will pay for the other stuff. So they take 2x loss for 1x profit. Then they will revise their business model or die.
There will be a dnet client. Then all is not lost.:)
Seriously... if AMD is teaming up with Compaq, I will NOT be supporting Sledgehammer. I know what it's like to have a Compaq.
-- LoonXTall
Re:When will the world accept that x86 is a deaden
on
Pentium IV Problems?
·
· Score: 1
Let x86 die.
I have been thinking about killing the x86 in a backwards-compatible way. It goes like this:
Build a new system, with a 64bit RISC CPU (maybe with a handful of 128b registers).
Add a cheap x86^H^H^H^H^H^H^H^H^HDuron.
Make the BIOS have an mmu64 call, which makes a region of RAM visible to the 64b chip with the specified protections. Then you can have shared memory between 32b and 64b processes for communication, and the 64b chip can execute the thread.
As more and more stuff is ported to the 64b chip, it picks up more and more of the processing; Moore's Law holds.
Linux (and anything else opensource) can run on the 64b chip because its source is available to cross-compile. There is no need for the x86.
Linux OEM systems become available for less than Windows systems, because they have no x86 any more.
The fate of Windows is left to the reader as an exercise.
I predict that IA-64 will survive only because of sizeof(Intel). Although predicated execution is a cool concept, it's been a holy grail of compiler construction (making sequential code parallel) for many years.
VSN (volume serial number) is based on the date/time of formatting. It's there so that WinDOS can read the VSN and still determine if the same disk is in the drive. This is because they use inferior drives that can't trigger a DISKCHANGE event.
Slashdot Mirror
What I envision is that when you move the mouse quickly across the screen, instead of seeing 6 simultaneous pointer images in a neat line, the input handler (Windows mouse driver) would blur the image from the last displayed position to the current one. This would be good, IMHO, because something I see at college once in a while is the mouse run up to full speed... when you expect it to go 1/4 of the way across the screen and it slams into the opposite wall, a blur would give a clue as to where the pointer went.
The blur itself need be nothing fancy... just take the average color of the trailing two mouse pixels and smack down a couple polygons of that color from current to last position.
-- LoonXTall
Clarification: 1/3 of voters that went into a booth voted on whether or not to go into debt. A mojority of those voted No.
-- LoonXTall
Actually, I was rather surprised to hear about/see those ballots. In my state, we use automatic voting machines... walk in, pull red handle, pull down one lever per column, push red handle, leave. No holes, no alignment problems.
There are two problems, though. When you pull a lever, a red-on-white "X" is visible, implying "No." And any propositions are up above the candidate lists, visually lumped with the voting directions. Only 1/3 of the voters that turned out actually voted on the proposition this year (to run $3.8 billion into debt to fix roads.)
-- LoonXTall
By "clean up the damn system" you mean campaign finance reform, right? Nobody cares about low-probability negative events like the Earth getting hit by an asteroid or the "wrong" candidate winning the election anyway. (They do care about low-probability positive events, like winning the lottery.)
-- LoonXTall
If you can make it, someone will break it. I have absolute faith that something can be encrypted such that {DNA|quantum|brute force} methods are all infeasible. What if your cipher doesn't have discrete states?
-- LoonXTall
Software patents enable Compuserve, for example, to patent a compression algorithm or a program that reads or writes a specific file format.
[href:Why there are no GIF files on GNU web pages] Both Unisys and IBM have the LZW patent. CS didn't do their homework, and Unisys didn't bother anyone until 1995 (8 years after the creation of the GIF format) when it was thoroughly entrenched. This is Patenting Done Wrong: the _end result_ was patented.
Then there's PKZIP: the actual implementation was patented. Other software can't use PKWare source code, but they (WinZip, WinRar, others) can create the ZIP file format without infringing on the PK patent. Patenting Done Right: the _implementation_ was patented.
There's always another way to do something (ask Rube Goldberg)... end-result patents are the only bad ones. Copyrights, well... I'm commenting on patents.
-- LoonXTall
We're being forced to use Network Essentials in class. The problem is, it's inconsistent---after describing the job of each layer of the OSI model, they then say that TCP goes _only_ with the transport layer. Hello? TCP's jobs are ensuring reliable delivery (transport layer) AND establishing a session (session layer)! What do they think SYN/ACK/FIN are for? It's also Pro-MS: they only mention tools in the interoperability section that allow NT to work with Netware---NOTHING of the reverse. In the Survey of NOSes chapter, WinNT and Novell each get 3 pages of decription, Apple gets 1 page, Banyan Vines gets almost a page, and *nix, Win9x, and OS/2 each get a paragraph. Under creating accounts, NT gets 4 pages with illustrations, and *nix just gets 2 paragraphs (which never mention (user|group)(add|mod|del).
</rant>
More objectively, it's just old. Who needs to know everything (terminator resistance) about Thinnet and Thicknet? ArcNet? This from the third edition (2000)... shouldn't the bugs be out by now?
-- LoonXTall
Then it would be your own stupid fault for not patenting it.
It _can't_ be patented if there are no patents.
-- LoonXTall
As supporting evidence for your distribution model, I would like to point out the piles of shareware on my hard drive that I paid for... oh, wait...... I didn't. Why? Convenience.
If something is convenient in Form X, Marketing can always sell it. If two competing products are convenient, then the one with the biggest edge will win (Betamax vs. VHS). If the edge isn't big enough, both proliferate (Phillips vs. slotted screws).
-- LoonXTall
Let's just declare the RIAA a monopoly, and make its members compete. Or declare it a public utility and regulate them.
-- LoonXTall
Ogg Vorbis is actually two layers: layer "Ogg" controls the timing during playback, and "Vorbis" is the actual compressed audio. When Vorbis is done and they move on to video, it'll be Ogg X, using the Ogg timing layer with the X video compression. (X=placeholder. They'll come up with something far more creative for a name.)
No URL, but it was on the old site (before they moved to xiph.org)
-- LoonXTall
You are assuming that DigitalColon^H^H^H^H^HConvergence cannot change its business model. The market is supposed to change in response to pressures... not invoke the legal system and destroy the pressure.
-- LoonXTall
Consider this: I invent a perpetual motion machine. I sell it in a black box. Then somebody "reverse engineers" it. The information is now free, and demand drops considerably (as people build their own.) Now... is that disrespecting my life, as my income and quality of life drops? Is it disrespecting my liberty to sell however I please?
Consider this quote from the page you linked to: "(Rand presumably got royalties from the sale, and I'm sure it wasn't sold without her permission!)" But if someone duplicates Atlas Shrugged and sells copies, it _is_ sold without permission or payment! It is also not freedom of speech/press---Congress is making no law about who can print it. That's all the 1st Amendment covers... Congress making laws.
The whole anti-any-IP argument is flawed: it takes much less effort to copy than create. I favor _limited_ IP. Anyone can use it to preserve life; the monopoly disappears in 7 years; the granters must understand the device or be shown a working implementation (the inventor of the transformer was laughed out of the office for trying to patent a short circuit, because that's what the patent examiner saw it as. Soon after, the laws were reformed so the PE need not understand the device.)
-- LoonXTall
Does anyone know of methods for distinguishing between various inks/toners placed on a page?
I would start with looking at the reflected spectrum of various wavelengths (infrared..ultraviolet) from the back of the paper. Any wavelength that makes the two toners reflect differently will do for reading. The problem is resolution... just telling the toners apart is easier than getting an exact picture. Maybe a laser could scan it, and reflectivity would be sent to a computer for image reconstruction.
-- LoonXTall
This protocol can support up to 15 (not 16, which seems strange to me.)
I'd guess it supports 16 nodes, with 1 node being the chipset.
-- LoonXTall
that's where the problem lies; in using mac addresses to form IP addresses.
This idea is screaming, "LAME!" MAC addresses are only 48 bits; IPv6 allows 128 bit addresses. If the IPv6 designers thought 48 bits would be enough, they should go back and listen to Bill Gates say, "640K of memory is all anyone will ever need." They should look at our current 32 bit addressing scheme. They should look at me, connecting over a 14.4 modem.
wrt DickBreath's reply, I don't know what'll replace Ethernet... but I envision going back to coax and using broadband on it (instead of baseband.) If you can cram video information for 50 cable channels on it, it should be reasonably high bandwidth.
-- LoonXTall
Preferably one where you could run any untrusted app in a chroot jail. Is that even possible under Win32?
Under Windows 9x, the closest you could come is a separate partition. (Got Partition Magic?) But if it's "sprinkling bits", it would have to know where to look for them (Registry entries), and they'd probably show up as files anyway (otherwise Defrag would kill your Napster.) The best way to hide a file would be as a .dll in \windows\system...
-- LoonXTall
If the newbie is clueless enough to put up a server with everything loaded on it despite the numerous FAQs and warnings, I don't have a lot of sympathy and DON'T blame RedHat.
OK, look at it from a newbie perspective. You bought Linux without ever checking hardware (hell, you don't even know what it is, so if X wants to know refresh timings, you're SOL.) You have a 500 page manual and a CD that seems to tell you what to do on-screen. Then you get to the point where you have to make a choice: do you decide to try and figure out where in the manual to look, or just bonk something at random? </simulation> Unfortunately, many newbies _will_ just bonk something. I suggest a second, smaller booklet titled "Foo Linux Installation Guide" which briefly runs through the installation and makes security notes on the way. It'd look something like this:
And so on. Then newbies would have security info and a brief description of what it does on hand and usable during installation, plus a pointer to more information.
-- LoonXTall
If I see another retarded consumer user say that compaqs suck, I will scream.
Then scream. I will resent any company who never bothered to make a 56k modem for their computer.
I did spend an entire Saturday wrangling with a 3rd party 56k, trying to install it on COM1:. Nothing worked.
People please understand that HOME PCs that you get in one wrapped up plastic shell from BigStoreCo are going to be shit, regardless of brand.
That's why I'm going custom-built next time I can afford a computer. It's just that I have a good reason to be pissed at Compaq.
Compaq makes what I consider to currently be the world's best Intel-based servers, in terms of functionality, cost, and price/performance.
Don't confuse Proliant with Presario.
I'm not confusing them. I'm only working within my range of experience (Presario 7222.) If they don't want me browsing any faster than 28.8, I'm not supporting them. If they wanted (more of) my money, they shouldn't have built this PoS.
-- LoonXTall
There's a reason I use FTP for site management: if someone wants my site taken down, I can move it easily.
If you keep server logs, you could locate whoever uploaded the offending piece and redirect fire unto them. After all, if you didn't post it, you shouldn't need to be responsible for it.
-- LoonXTall
The rant applies to any company that pulls this same type of crap. the fact that its a free product is irrelevant.
No it isn't. If it's a loss leader, they lose money every time you buy one! Every time you buy one, it means they have to produce another and hope they sell to someone who will pay for the other stuff. So they take 2x loss for 1x profit. Then they will revise their business model or die.
-- LoonXTall
There will be a dnet client. Then all is not lost. :)
Seriously... if AMD is teaming up with Compaq, I will NOT be supporting Sledgehammer. I know what it's like to have a Compaq.
-- LoonXTall
Let x86 die.
I have been thinking about killing the x86 in a backwards-compatible way. It goes like this:
The fate of Windows is left to the reader as an exercise.
I predict that IA-64 will survive only because of sizeof(Intel). Although predicated execution is a cool concept, it's been a holy grail of compiler construction (making sequential code parallel) for many years.
-- LoonXTall
----------------------
Price * Watts
Multiply by 1.3 for DDR SDRAM, divide by 50 for RDRAM.
-- LoonXTall
-- LoonXTall