The story is about running 10Mbps over Cat1 cable (non-twisted pair) or random wires (non-coaxial).
To clarify: the twist is only for preventing differences in the cable and local environment from affecting one wire of a balanced pair more than the other. The pair won't magically radiate more if you untwist it (and you were specifically talking about line losses due to radiation, whereas the previous poster was talking about RFI). Your 3500' run of cat1 may only do 5mbit, but a 3500' run of cat3 (which a lot of phone line is these days) may well do 15.
Straight bare metal is pretty much the definition of an antenna.
If that's all there was to it, we'd all have to live closer to powerplants. Read up on transmission line theory. The losses (radiation) for travelling waves are a few orders of magnitude lower than if a standing wave were created on a resonant line.
If this is really VDSL-based, there will be several modulated sine waves in use.
Depending on the power you'd need to push your signal down a mile of barbed wire
Easily determined by the required bitrate, available bandwidth, and noise floor. Millivolts, although they'll probably use a couple volts (like standard 10/100/1000baseT) to make the parts cheaper.
with a transmitting antenna a mile long
Properly-designed transmission line does not radiate (much). This is primarily done by either running a balanced signal down two twisted conductors (twisted pair) or running an unbalanced signal inside a grounded shield (coax).
the MS Windows print model was designed to work well with inkjet printer, you cant just send commands ad hoc to the driver to anypary of a page, you must send a complete bands of information to the driver in order from top to bottom. It's kinda of a kludge.
This is completely false.
The commands are GDI commands. Your word processor (for example) sends the same commands to render fonts on a word processor screen as it does to render them on a printed page. Only after your word processor has sent all the commands for a page can the print driver begin sending that page to the printer.
The windows print "driver" (filter really is more accurate) interprets these commands into the printer's native language (PCL, postscript, page-description-language-of-the-month, etc). Then it sends that to the windows parallel port driver (with certain exceptions - some drivers for cheapo printers talk to the printer on their own) in whatever-sized chunks it likes. Typically for laser printers, a driver would send an entire page to the printer, for dot-matrix, daisy-wheel, and inkjet printers, it would send a line or two at once.
Having a common interface to do both your display rendering and your print rendering makes sense, and Microsoft didn't come up with the idea. I'm not sure who did, but NeXT was touting "display postscript" a very long time ago.
I'd hate for my corporate desktop to be recording everything I'd say for the posterity of HR
Chances are pretty good that when you need to swear profusely, your computer would be rebooting after a crash (and therefore would not be running lipreading software).
On the other hand, recording for posterity everything HR says can be very entertaining.
HTTP can't really offer all that FTP does in terms of file transport.
HTTP really is all that.
HTTP/1.1 supports, among other things, file resuming via a standardized header (Range:) and pipelining (whereas FTP's control port+data port means n+1 TCP connections). HTTP can give you a file compressed the way you want it - and in the language you asked for - without filename hacks. HTTP's If-Modified-Since: header makes it more cacheable. In addition, most HTTP server implementations are more flexible - they can authenticate against things other than the local account database, and there is a widely implemented standard for HTTP over SSL - HTTPS. CGI is also more pervasive and useful than SITE EXEC.
So basically Your beef is with braindead managers and only secondarily with the crappy appliances?
No, it just seems braindead managers manage to put appliances on my desk more often. And yes, I did quit my job when I had enough of this. And yes, at my next job I inherited the support role for dozens of things I didn't spec, including a couple appliances. That's life.
have You ever seen an appliance perform as promised?
Yes, but I like to do things with servers that people don't promise. Examples:
Watchguard Firebox II
Performs as advertised. Is actually a Linux 2.0.33 system sitting on 7MB of flash. Must be configured through its own management app. Last time I moved it to a different IP block, it took several hundred mouse clicks to change all the incoming NAT rules to different external IP's. On the plus side, it let me stage all these changes and then apply them to the firebox all at once. Management app is the only way to get bandwidth stats as Watchguard does not believe in SNMP. Does not support any dynamic routing protocol. Must be rebooted for any firewall rule change (however reboots only take 26 seconds).
Procom CD Tower (the hardware I played with was ~18 months old)
Mostly performs as advertised. Has two 20GB HDD's and a 5-slot CD changer in a baby AT case with a LCD on the front panel. Requires a patch to be installed on your NT4 PDC (!) to communicate NT groups to its hacked version of Samba. Occasionally it would lose its WINS registration. This was fixed in newer versions of Samba, and I would have loved to take advantage of Samba's new winbind functionality on that CD tower, but the only way to get a shell on that beast would be to put the hard disks in a different system and mount the filesystem that way.
Smith, Gardner, and Associates Mail-order And Cataloguing System (MACS) Shipping Confirmation Station (oh baby yeah!:)
Performed as advertised, which wasn't what we needed. Unfortunately it was all SGA had and we were rather locked in. It was an enlight desktop case with a clone Pentium MMX board, 8MB RAM, a brand new (!) 200MB hard drive, and a custom circuit board that functioned as a keyboard wedge for the scan gun and retransmitted serial port data (from the weigh scale) as keyboard data. For this hunk of junk they charged us $5000+. Couldn't be used with the Symbol LS4000 scanguns without a bunch of AT to PS/2 adapters. Wound up replacing the thing with a more recent ATX system, a software keyboard wedge, and a custom serial cable (please take this moment to remember the dozens of serial mice whose tails were chopped in two that day:), saving the company $16,000+ (they needed five of them).
Why are You assuming that any appliance would provide just crappy tools that run or crash in an environment You don't use? And that the hardware is crappy and crashes and is unmaintainable and the service contract won't be fullfilled?
Experience. I can go into detail if you like.
Aren't these all just the reasons You should write a req spec and not just order some crappy piece of shit from Joes Garage
If I only maintained things I personally made the decision to purchase at my job, I would be a lot happier. Unfortunately the management ranks are filled with people who think like you (well, about the depending on external companies thing, not the specing stuff out thing), and these type of boxes just tend to show up on my desk.
With any luck, a positive effect of this little recession will be to put a stop to the "brain drain" effect that led managers in certain areas to believe that rented brains were superior to hired brains in the first place. Such a viewpoint has always had negative effects on department morale.
However, why are You specing ssh? You don't need ssh, You need an interface for configuring and maintaining the system from remote workstation using secure protocol.
An interface that doesn't piss me off. Why should I have to test out the functionality of all the pieces of a new black box for every single thing I have to implement when I know of a system that will work? Why should I have to load yet another crappy management tool onto a Windows desktop or put up with yet another clunky web management interface that doesn't do what I want? Why should I have to tell management that this one machine can't alert me when something non-fatal happens to it, like a fan or disk dying?
Almost all appliance manufacturers attempt to reinvent the manageability wheel. None of them put the magnitude of effort into it that the OS vendors do. They put a friendly-looking interface on top of existing software and stop there. They're much less likely to fix a bug in that interface or upgrade their firmware to include a fix for a known bug in one of those pieces of existing software. Chances are very slim that you'll see new firmware released to support new standards as they become more popular, as the mentality those vendors like to encourage is that of appliance replacement.
Your assertion that it takes a well-manned IT department to set up a server using normal hardware and install proxy serving software (or whatever) on it is amusing.
If You need a http proxy in enterprise environment, You don't set up a computer to do the task. You get an appliance with service contract.
If you actually have experience with HTTP proxy software, this will only piss you off. Same for firewall appliances, CD tower appliances, file server appliances, web server appliances, you name it.
The right way to do it is to go with what you're familiar with and what you know works. Not what comes in a black box that you suspect might be half-assed. Not what might work in six months.
Spec'ing out the equivalent appliances for two fully-redundant 100GB+ squid proxy servers that you can ssh into, with hot-swap RAID, redundant power supplies, network connectivity of your choice (gig ethernet, for example), etc. with a hardware vendor service agreement that specifies 4-hour parts replacement turnaround time may be an exercise in futility.
If it was really "innocent till proven guilty" then why is the US bombing Afghanistan?
Ok, lemme spell it out for the slow people...
The "common law" western court system as we know it is national in scope. That means that sovereign nations each independantly employ justice systems as they see fit.
There is no such system that is international in scope. Instead, it works kind of like the playground in grade school. You make friends (allies), make agreements with them (treaties), if you're rich you can give them some lunch money (foreign aid), and if you're big/brave/foolish enough you can also bully people around (sanctions, war, etc). All these playground-like social conventions are called "international law".
It'll be this way until sovereignity breaks down and all nations succumb to a global empire that can enforce its own laws wherever it wants. Then you won't have war, you'll just have rebels and revolutionaries. Doesn't that sound like fun?
The Hardware Issue- Do you have surround sound? That would be a major question.
No, it's not, not if you're playing music originally from CD. CD's are stereo. Not 4-channel, not 5.1. Do you expect your surround system to magically figure out what speaker to send a signal to?
With that said, four-speaker stereo can significantly increase the size of your room's "sweet spot" and reduce the stereo distortion effect you hear when turning your head. Add a subwoofer for deep bass response, and that's about the most you'll need for accurate playback of any two-channel source.
Here's a link to the Feist Publications vs. Rural Telephone Service Co 1991 US Supreme Court ruling on phone book copyrightability. Note they mention originality as a constitutional requirement for copyright protection - an outline of the US is only copyrightable if it has an original element to it (otherwise it doesn't promote the arts and sciences).
If a million monkeys type out the source code to MS Office, Microsoft can't sue. Likewise, if you happen to create a series of dual-tone meta frequency notes using a touch-tone phone using non-copyrighted material (a phone book, your memory, etc), then that's an independant creation. Now if a telemarketer overheard you dialing, and recorded it (made a copy), then you might have something.
IANAL (and I know the whole point was to be funny anyway).
ISA is a bus - all the pins on all the slots are wired to each other, straight across. Go to an electronics shop, get an ISA perfboard card, put the regulator on that, and have it supply power to the other slots.
That way, if it doesn't work, and there is no smoke discoloration, you can hide your tracks if you need to send it back for warranty repair:)
Slashdot Mirror
"Our right to shop for housewares in a safe environment outweighs his right to anal sovereignty. This is America, dammit."
Or dropping it from one meter onto a surface such as concrete that causes it to accelerate to zero m/s in .0001 seconds.
That's why rubber/plastic is so useful for things like this - it can dramatically lengthen acceleration times, reducing G's to sane levels.
To clarify: the twist is only for preventing differences in the cable and local environment from affecting one wire of a balanced pair more than the other. The pair won't magically radiate more if you untwist it (and you were specifically talking about line losses due to radiation, whereas the previous poster was talking about RFI). Your 3500' run of cat1 may only do 5mbit, but a 3500' run of cat3 (which a lot of phone line is these days) may well do 15.
Straight bare metal is pretty much the definition of an antenna.
If that's all there was to it, we'd all have to live closer to powerplants. Read up on transmission line theory. The losses (radiation) for travelling waves are a few orders of magnitude lower than if a standing wave were created on a resonant line.
If this is really VDSL-based, there will be several modulated sine waves in use.
Depending on the power you'd need to push your signal down a mile of barbed wire
Easily determined by the required bitrate, available bandwidth, and noise floor. Millivolts, although they'll probably use a couple volts (like standard 10/100/1000baseT) to make the parts cheaper.
with a transmitting antenna a mile long
Properly-designed transmission line does not radiate (much). This is primarily done by either running a balanced signal down two twisted conductors (twisted pair) or running an unbalanced signal inside a grounded shield (coax).
This is completely false.
The commands are GDI commands. Your word processor (for example) sends the same commands to render fonts on a word processor screen as it does to render them on a printed page. Only after your word processor has sent all the commands for a page can the print driver begin sending that page to the printer.
The windows print "driver" (filter really is more accurate) interprets these commands into the printer's native language (PCL, postscript, page-description-language-of-the-month, etc). Then it sends that to the windows parallel port driver (with certain exceptions - some drivers for cheapo printers talk to the printer on their own) in whatever-sized chunks it likes. Typically for laser printers, a driver would send an entire page to the printer, for dot-matrix, daisy-wheel, and inkjet printers, it would send a line or two at once.
Having a common interface to do both your display rendering and your print rendering makes sense, and Microsoft didn't come up with the idea. I'm not sure who did, but NeXT was touting "display postscript" a very long time ago.
Translating GDI to postscript code under wine is as trivial as using an existing Windows driver for a postscript printer.
UNIX apps don't send GDI commands - they usually send postscript commands.
So unless someone wants to write a postscript to GDI filter, that approach won't work.
Oh, and things that need to communicate directly with your hardware (like this printer driver) may not be able to run in wine anyway.
Chances are pretty good that when you need to swear profusely, your computer would be rebooting after a crash (and therefore would not be running lipreading software).
On the other hand, recording for posterity everything HR says can be very entertaining.
HTTP really is all that.
HTTP/1.1 supports, among other things, file resuming via a standardized header (Range:) and pipelining (whereas FTP's control port+data port means n+1 TCP connections). HTTP can give you a file compressed the way you want it - and in the language you asked for - without filename hacks. HTTP's If-Modified-Since: header makes it more cacheable. In addition, most HTTP server implementations are more flexible - they can authenticate against things other than the local account database, and there is a widely implemented standard for HTTP over SSL - HTTPS. CGI is also more pervasive and useful than SITE EXEC.
Let FTP die the death it has so long deserved.
Actually, gasoline has an energy density ten times that of dynamite. Your gallon of gas fumes (almost 3kg) is closer to 30kg of dynamite.
And yes, it's a very big boom.
They say they're aiming for 10% efficiency.
How is something that is 10% efficient one thousand times less efficient than anything?
I'll assume you meant sampling rates, not bitrates, as you were talking about CD's vs. vinyl.
A 48khz sampling rate effectively implements a 24khz lowpass filter. Bass isn't affected. Ultrasonic treble is.
No, it just seems braindead managers manage to put appliances on my desk more often. And yes, I did quit my job when I had enough of this. And yes, at my next job I inherited the support role for dozens of things I didn't spec, including a couple appliances. That's life.
have You ever seen an appliance perform as promised?
Yes, but I like to do things with servers that people don't promise. Examples:
Watchguard Firebox II
Performs as advertised. Is actually a Linux 2.0.33 system sitting on 7MB of flash. Must be configured through its own management app. Last time I moved it to a different IP block, it took several hundred mouse clicks to change all the incoming NAT rules to different external IP's. On the plus side, it let me stage all these changes and then apply them to the firebox all at once. Management app is the only way to get bandwidth stats as Watchguard does not believe in SNMP. Does not support any dynamic routing protocol. Must be rebooted for any firewall rule change (however reboots only take 26 seconds).
Procom CD Tower (the hardware I played with was ~18 months old)
Mostly performs as advertised. Has two 20GB HDD's and a 5-slot CD changer in a baby AT case with a LCD on the front panel. Requires a patch to be installed on your NT4 PDC (!) to communicate NT groups to its hacked version of Samba. Occasionally it would lose its WINS registration. This was fixed in newer versions of Samba, and I would have loved to take advantage of Samba's new winbind functionality on that CD tower, but the only way to get a shell on that beast would be to put the hard disks in a different system and mount the filesystem that way.
Smith, Gardner, and Associates Mail-order And Cataloguing System (MACS) Shipping Confirmation Station (oh baby yeah! :)
Performed as advertised, which wasn't what we needed. Unfortunately it was all SGA had and we were rather locked in. It was an enlight desktop case with a clone Pentium MMX board, 8MB RAM, a brand new (!) 200MB hard drive, and a custom circuit board that functioned as a keyboard wedge for the scan gun and retransmitted serial port data (from the weigh scale) as keyboard data. For this hunk of junk they charged us $5000+. Couldn't be used with the Symbol LS4000 scanguns without a bunch of AT to PS/2 adapters. Wound up replacing the thing with a more recent ATX system, a software keyboard wedge, and a custom serial cable (please take this moment to remember the dozens of serial mice whose tails were chopped in two that day :), saving the company $16,000+ (they needed five of them).
I could go on, but I'm out of time for now.
Experience. I can go into detail if you like.
Aren't these all just the reasons You should write a req spec and not just order some crappy piece of shit from Joes Garage
If I only maintained things I personally made the decision to purchase at my job, I would be a lot happier. Unfortunately the management ranks are filled with people who think like you (well, about the depending on external companies thing, not the specing stuff out thing), and these type of boxes just tend to show up on my desk.
With any luck, a positive effect of this little recession will be to put a stop to the "brain drain" effect that led managers in certain areas to believe that rented brains were superior to hired brains in the first place. Such a viewpoint has always had negative effects on department morale.
An interface that doesn't piss me off. Why should I have to test out the functionality of all the pieces of a new black box for every single thing I have to implement when I know of a system that will work? Why should I have to load yet another crappy management tool onto a Windows desktop or put up with yet another clunky web management interface that doesn't do what I want? Why should I have to tell management that this one machine can't alert me when something non-fatal happens to it, like a fan or disk dying?
Almost all appliance manufacturers attempt to reinvent the manageability wheel. None of them put the magnitude of effort into it that the OS vendors do. They put a friendly-looking interface on top of existing software and stop there. They're much less likely to fix a bug in that interface or upgrade their firmware to include a fix for a known bug in one of those pieces of existing software. Chances are very slim that you'll see new firmware released to support new standards as they become more popular, as the mentality those vendors like to encourage is that of appliance replacement.
Your assertion that it takes a well-manned IT department to set up a server using normal hardware and install proxy serving software (or whatever) on it is amusing.
If you actually have experience with HTTP proxy software, this will only piss you off. Same for firewall appliances, CD tower appliances, file server appliances, web server appliances, you name it.
The right way to do it is to go with what you're familiar with and what you know works. Not what comes in a black box that you suspect might be half-assed. Not what might work in six months.
Spec'ing out the equivalent appliances for two fully-redundant 100GB+ squid proxy servers that you can ssh into, with hot-swap RAID, redundant power supplies, network connectivity of your choice (gig ethernet, for example), etc. with a hardware vendor service agreement that specifies 4-hour parts replacement turnaround time may be an exercise in futility.
(And yes, proxy failover can work much like DNS failover - the client is smart enough to figure it out.)
Ok, lemme spell it out for the slow people...
The "common law" western court system as we know it is national in scope. That means that sovereign nations each independantly employ justice systems as they see fit.
There is no such system that is international in scope. Instead, it works kind of like the playground in grade school. You make friends (allies), make agreements with them (treaties), if you're rich you can give them some lunch money (foreign aid), and if you're big/brave/foolish enough you can also bully people around (sanctions, war, etc). All these playground-like social conventions are called "international law".
It'll be this way until sovereignity breaks down and all nations succumb to a global empire that can enforce its own laws wherever it wants. Then you won't have war, you'll just have rebels and revolutionaries. Doesn't that sound like fun?
"...removing Microsoft's IPX/SPX compatible protocol and Novell's IPX 32-bit protocol from the dialup adapter bindings resolved the error."
Perhaps that documentation is out of date. Support is a lot more pervasive than that now.
Cisco claims to have added support for ssh for the 3600's as of IOS 12.1.
You're not very aware. Cisco Foundry Juniper [fill in the blank here]
No, it's not, not if you're playing music originally from CD. CD's are stereo. Not 4-channel, not 5.1. Do you expect your surround system to magically figure out what speaker to send a signal to?
With that said, four-speaker stereo can significantly increase the size of your room's "sweet spot" and reduce the stereo distortion effect you hear when turning your head. Add a subwoofer for deep bass response, and that's about the most you'll need for accurate playback of any two-channel source.
Not necessarily. The alternative is interprocess communication with a privileged process.
Here's a link to the Feist Publications vs. Rural Telephone Service Co 1991 US Supreme Court ruling on phone book copyrightability. Note they mention originality as a constitutional requirement for copyright protection - an outline of the US is only copyrightable if it has an original element to it (otherwise it doesn't promote the arts and sciences).
IANAL (and I know the whole point was to be funny anyway).
ISA is a bus - all the pins on all the slots are wired to each other, straight across. Go to an electronics shop, get an ISA perfboard card, put the regulator on that, and have it supply power to the other slots.
That way, if it doesn't work, and there is no smoke discoloration, you can hide your tracks if you need to send it back for warranty repair :)