---snip 640k" vs "64Kbytes"- why change capitalization on the K? That doesn't mean anything, unless k=1024 and K=1000. But "10mb" vs "1MB" does mean millibits in the first case and megabytes in the second- not likely what you intended. ---snip
-"I just doubled the ram in my Apple II+ by installing 16k of dram" (kilobytes) -"That board is populated with 8Kb chips" -"There are 1024KB in one MB"
My casual usage (may be wrong, corrections welcome):
Back when everything was measured in KiloByte units or less (something measured in megabytes was once as uncommon as measuring in terabytes today), if you just said "k", it was understood that you meant "KiloBytes", with a byte having exactly 2 nibbles:), and one "Kilo" being 1024 (leaving it to the friggin telecom people and drive manufacturers to "redefine" Kilo back to the metric-literal version so they could inflate their numbers).
If you were working with something specifically measured in bits, such as an 8 Kilobit chip, you would add a lower-case "b" to the end; "Kb" or "kb" are both acceptable, everyone knew that "k" meant 1024. For Bytes, you added an upper-case "B"; saying "8 kB" was understandable, but doesn't look right, so typically, if you are going to capitalize the "B", you would capitalise the "K" as well; "8KB".
Later (in the megabyte era) the drive manufacturers had eventually all began using their own versions of "mega" which did not match up with the "real" mega:). So some enterprising person came up with this "mebi/milli/whateverbits" silliness. I don't think this silliness ever was applied to the "Kilo" world.
The simplest way around that ambiguity is to measure everything in bits, rather than bytes.
---snip Aside from that, what you're saying is that you seem to need 10 bits of connection to transmit 8 bits of data. That's a 25% overhead, well in line with the the 10% suggested by joshuac, but completely different from the 600000% increase claimed in the article. ---snip
Ummm, no, the protocol overhead I was talking about was the packet headers for each TCP packet. The overhead he is referring to (I believe) is the overhead for the connection itself. To transmit one 8 bit byte, it can easily take 10 bits (or more).
1 start bit, 8 data bits, 1 stop bit possibly? Or he might be using 8 data bits, 1 stop bit, and some type of parity checking. Or he could have been using something quite obscure for some reason, 8 data bits, 2 stop bits perhaps.
In any case, dividing by 10 is probably the right way to measure bytes per second on his connection. I was talking about the additional overhead that is there for the tcp protocol itself.
In theory; the reality is that the error correction used on most modems during this era (MNP-5 from the multitech world or v.42bis(*) for everyone else) was less than perfect. Enough packets sent, and eventually one would get through with a false positive. For example, if you are using a simple 8 bit checksum, it would be pretty easy for an occaisional bad packet to slip through. The hardware error correction was much better than an 8 bit checksum, but it was still quite possible to lose a long transfer. Hence, the long-lasting popularity of zmodem well into the day of error correcting modems.
Are you counting everything going over your connection (such as the ~5% tcp protocol overhead itself which would not be reduced by "fast tcp") or are you just counting the transfer of the data you see in the end?
Look at the interface statistics on your router when transfering, it will probably show 100% on whatever your slowest interface is (in your case, WAN/DSL side).
And then of course...even if you are using SDSL, your "640k down" will not remain a perfect 640k if there is much noise on the line from outbound traffic. However I suspect this is not the case in your situation.
Actually, the Zmodem that was widely used (real zmodem) maxed out at 1k blocks, but it would steadily scale down to as small as 16 byte blocks (if I recall correctly).
There were variants that did 8k blocks (and often referred to themselves as Zmodem8k), but none of these were true zmodem protocol.
Still, nothing can be quite as fast as ymodem-g:)
A little more on topic; what they are describing does not dynamically scale the packet size, only dynamically adjust the transmission speed up to the point that ack's start slowing down, but (hopefully) before any packets actually get dropped. I suspect disney and such will be quite disappointed if they think they are going to get a 6000x speedup in practical use as hinted at in the articles. Perhaps a 10% speedup for joe blow on a dialup modem, _maybe_. Take a look at your connection some time when downloading a file; you will probably find you can already peg your bandwidth quite nicely.
That sounds an awful lot like a 3 byte instruction to me...when that opcode is hit, the PC is incremented by 3 (or 2 for the zero-page variation on your instruction. BRK, NOP etc. would be what I think of when someone says 1 byte instruction...
But then by your definition, all the instructions on the 6502 (or pretty much any other 8 bit processor) were 1 byte instructions...
---snip ( for example one instruction would look up an address, read the word, add it to the XY register and then look up the resutling number as an address---the definition of a pointer offset included in a single byte machine code instruction ---snip
Ummm, you are just describing a plain ol' DHCP server, there is nothing special about that config.
---snip your computer is assigned a permanent TCP/IP address that stays with your MAC address/Account (Which are linked on a database somewhere) ---snip
That "database somewhere" of IP address to MAC address is the DHCP client lease table on your DHCP server.
Unless you are saying that you IP address follows your _user_ account rather than your NIC's MAC address, in which case your school would truly have something different.
---snip I'm not sure the hardware/software solution that's required to implement this, but on a campus of 5,000+ people, it surely cuts down on tech support calls. ---snip
A 386, a NIC, and the OS of your choice (Hey, even NT 3.5x ran on 386's). If you have got more than 5 machines on your network, it is probably time for a DHCP server.
Just set the lease time to longer than the likely longest period of time a user will have their machine disconnected from the network, and you are fine; the DHCP server will go out of it's way to make certain that MAC address gets the same address re-assigned to it.
---snip but in my experience, trying to manage windows from the command line is a horrendous task. ---snip
For sure, different OS's are more or less admin-friendly. But difficult is not the same thing as impossible.
Trying to manage windows from the CLI _is_ a much more difficult task than with most other OS's. However, with the right tools you can manipulate the registry database, kill or start a process, manipulate files, or just about anything else you would need to do. In any practical real-world situation you would not hassle yourself with trying to work that way, unless it were a task particularly suited to the command line (automation) or you were in a situation where a CLI was all you had available; remote admin of windows machine is almost always done in a GUI (although many of those GUI tools can be ran locally to manipulate a server remotely over a network).
A good admin _can_ accomplish all the same things, but too often people say "it cannot be done" out of ignorance rather than fact.
Woodworking is a hobby I am just starting out on. Ask me to make a particular type of joint with certain tools, and I will say it is impossible. Ask that of someone who has been working with wood his whole life and knows what he is doing, and he can make it happen. I said the tools were not capable of it, but in fact, it was me.
---snip If I have the volume up, not only can I "hear" the screen redraws, but the mouse move, my keyboard pulse, and my network card go to work ---snip
Or your algorithm finish...I used to leave my stereo next to my apple ii; when I had a calculation that was going to take especially long to finish running (plotting fractals at hi-res was a favorite), I'd crank up the stereo on a frequency that picked up the activity well. When the machine was performing a complex task, the sound was very white-noisy. When the machine finished and went into a loop polling the keyboard memory location, the sound was a buzz. I could be in the other room and know when things were done, or in some cases know how _soon_ it would be done.:)
Great, now I'm starting to feel homesick for that old machine, and I'll probably end up unpacking it this weekend (again). Some things an emulator just cannot replace.
---snip Not to mention the remote access aspect of the command line. I'm responsible for maintaining a webserver running Linux; it's so handy to be able to pull up an ssh session and do whatever needs to be done. Let's see the admins of the Windows boxes neighboring it work on their servers from home. ---snip
It's not difficult to setup ssh on windows, and redirect the io of cmd.exe (or most any other CLI shell available for windows, come to think of it) through it if you know what you are doing. The problem isn't so much the OS, it's the type of admins the OS attracts. 90% of windows "admins" do not know how to use the cli and cli apps included with windows, or even use many of the graphical management tools included to their full extent.
With NT 5 native RDP support was added, although you have been able to make any NT machine support RDP for awhile now, (assuming you do not need multi-user support; in that case your stuck with their terminal server distribution, or purchasing something re-written by Citrix and dumping RDP for their protocol), and with NT 5 they made a telnet server an official part of the distribution (not that I would suggest using it). Of course, you could do all this years ago with NT 4 as well...(and with NT 3.5x with a little adaptation, probably).
BUT...do many NT "admins" know how to do this, or even have the slightest clue about where they would go to get started? Nope. And from what I have observed, in more than a few places having a _little_ more knowledge about the OS they "support" would have made their lives a lot easier...but they don't bother, they just keep slogging on with what little knowledge someone force fed them at a certification class.
Unix admins on the other hand tend to actually take an interest in knowing what makes their OS tick. Partly that the OS is much more open (for some Unix OS's, you can even get the source code easily). Partly that in the Windows world, the interface to a server appears identical to the interface to a workstation, which gives confidence to users who want to become an admin where they deserve no confidence at all:).
A good admin can make up for a weakness on any platform, whether it be Netware, Linux, Windows, or Lantastic, without needing to blame his/her tools. A lousy "admin" can also more than compensate for all the strengths of a platform. What matters most is the competency of the admins that tend to work with those systems, not really the systems themselves.
That all said, given a choice, I know which platform is takes me much less effort to secure than the others...heck, you might almost say it is "secure by default":)
Re:Who cares about Salon? Will anyone save The WEL
on
Salon Asks for Help
·
· Score: 1
Did you even read my post?:)
Who cares about Salon? Will anyone save The WELL?
on
Salon Asks for Help
·
· Score: 2, Insightful
In the grand scheme of things, Salon is just another online newsjournally type site, if they go away, we are not losing anything unique. The Whole Earth 'Lectronic Exchange on the otherhand, has been around since the 80's...I remember having the The WELL on my dial list as I went through my nightly BBS romp.
Unfortunately, through various twists and turns, The WELL ended up under Salon's ownership. If Salon goes under, does this mean The WELL will also? That would be horrible and ironic end for something as insignificant as Salon to take down a piece of internet history with it...in 10 years (assuming Salon shuts down, which it likely will) few people will remember or miss Salon...but 20 years down the road, The WELL is still remembered for it's place in the early internet, and I know there are plenty of people who will miss it.
Otoh, maybe more people like Salon than I realize; but somehow I doubt it is the same as The WELL.
I hope if Salon goes under, that The WELL will somehow be preserved. As for Salon.com, I could care less.
---snip... but back then, I was learning DOS 3.3 (for Apple II) and AppleWorks, becuase "everyone in the future will be using this stuff!"...
Anyway, I took that seriously, and made damn sure that I *knew* to enter the proper date when Appleworks was starting up, and that I *had* to make sure I had the right disks in the drives. ---snip
---snip Anyway, to get to my point, I wonder how relavent the things that they learn now will be a few years after they graduate - and I hope it is *concepts* that they learn, instead of cookie cutter "type CATALOG to see a what's on your disk, insert your disk and type PR#6 to start AppleWorks" stuff... ---snip
I'm totally with you, the way things are "taught" is a big pet peeve of mine.
My school had a similar class to the one you describe, but for various reasons I never found my way into it. I was fortunate enough to have an Apple IIe of my own at home, with a few random reference books for various things. Instead of the rote class learning, I was teaching myself 6502 assembly (because I had a reference to 6502 opcodes and had found a way to get myself into the miniassember that came with Integer Basic), and learning the various subroutines that could be called in DOS 3.3. When I got my hands on 3.3E by begging and stealing from our nazi computer "teacher" at my elementary school, I remember the joy of decompiling 20 instructions at a time to get glimmer of what the minor differences were in the code. As time went on, I taught myself how to automate writing much of my assembly by using WPL, a very under-appreciated scripting language internal to AppleWriter II (well, Don Lancaster knew how good it is). Later, (once I had gotten my hands on the reference books) I taught myself the in's and outs of high level languages like Integer Basic and Applesoft; when I managed to sneak off with a copy of Apple Logo, I learned everything I could about that, because "it's fun to learn what makes things tick!"....meanwhile, the class learned the syntax of various DOS 3.3 commands.
Anyway, I took nothing very seriously, at that formative age (10 years old) I had found a toy that had limitless possibilities, that could be reprogrammed to perform any task you could conceive. The class was being taught how to operate a tool within narrow confines of specific pre-decided tasks.
Now, almost 15 years later, what I learned then on that Apple IIe was invaluable; what I learned that was truly valuable was not how to interface with a disk ii controller and count clock cycles for timing in my ML loops, it was that I learned something about learning. The class had learned how to be told what to do.
The most valuable thing that IIe taught me is that you are fooling yourself if you tell yourself you "know" everything about a subject. When people say they "knew" DOS 3.3 because typing "catalog,d2" showed them the contents of the second drive on the current controller, I realized they were selling themselves short; there was always more to learn; they didn't "know DOS". But since they assumed they knew everything, there was no room for them to learn more.
The next most valuable thing I discovered was how to pull something apart and learn how it works, without a master plan in front of you. Too many people have been taught to "learn" by being shown an example, and then emulating. It's faster, it get's the grade school concert band able to push out a few notes in time for their parents to be proud during the winter concert, but rote knowledge is a poor subsititue for actual understanding. Type "pr#6" to boot off the floppy in drive 1, slot 6...does that actually teach you anything about what is going on, or are you just mechanically following directions? When all you learn is to follow directions, inovating when given an unexpected problem is very difficult. When you understand what is happening, you give yourself many more choices, and much more control.
Anyhow, I learned many of the same subjects that the computer class at my grade school set out to learn. But I suspect that over time I got much more out of my learning experience than those students did, simply because of the way they were forced to learn.
To make this slightly on-topic, does anyone know how the Redhat classes are taught? Do the teach you think unix, or do they teach you the syntax of commands?
Yes, if they thought something was seriously wrong right after liftoff (which they checked, and concluded nothing serious had happened), it is possible to get a second shuttle up in the time of the mission if everything went on schedule with that second launch.
If they had the ability to inspect the shuttle from outside, and they thought it necessary (remember, the insulation torn off the main tank was ruled unimportant, and nothing else went visibly wrong with the launch), then maybe (assuming the problem was damaged tiles and was visible from outside) they might have found the problem before attempting rentry, and would have been able to figure out some miraculous Apollo 13-style save.
But then just lugging the already existing robotic arm (which wasn't taken on this trip) along would have accomplished the same self-inspection tasks easier and better than this microsattelite could. The microsattelite is pretty cool in itself tho, but on the orbiter, just take the arm with a camera.
---snip So you already know, without knowing the actual extent of any theoretical damage that a minimal approach rather than the normal one would have made no difference? What would have stopped them from, e.g. getting another shuttle up, stopping by the space station, etc.? ---snip
I doubt there is a much more "minimal" approach than what is already used by the shuttle. As for getting another shuttle up, the soonest an emergency launch of atlantis could be performed would be a week, assuming everything went well. Considering how often launch dates have to be pushed back even under ideal conditions, actually pulling off an emergency launch in a week is a long shot.
And no, stopping by the space station was not an option, Columbia was in a much lower orbit than the iss and orbiting at a different angle (I know there is better term for that, anyone?).
There might have been _something_ they could have done, but those three options can pretty much be ruled out.
and what is _really_ scary is looking at the this list, it looks like plenty of admins have been accessing this system from home; the log dates back to 1-jan-2002. If you are a lazy cracker, grep for all the lines with "DSL" in them, and probably 80-90% of those hosts are home workstations of military sysadmins of one type or another. If they are dumb enough to leave logfiles of users accessing a server used for military network administration open to the public, imagine what their home computers are like...
What's even more depressing is that it looks like some of these guys use AOL...
---snip Given that this thing is intended to glide like airplane, except in water, I wonder what it would take to make it able to fly in air? Probably a lot of engine power that it doesn't have, and a lot less weight.:( ---snip
...and for the hydrofoils that are designed to create downward force when moving through the water to start making upward force when being used as airfoils, amongst other problems...but a flying submersible would rock. The police vtol jet plane/wheeled land vehicle (I think)/submersible shown briefly in the movie "AI" was pretty neat...
more interesting than other parts of the movie, come to think of it.
Slashdot Mirror
---snip
:), and one "Kilo" being 1024 (leaving it to the friggin telecom people and drive manufacturers to "redefine" Kilo back to the metric-literal version so they could inflate their numbers).
:). So some enterprising person came up with this "mebi/milli/whateverbits" silliness. I don't think this silliness ever was applied to the "Kilo" world.
640k" vs "64Kbytes"- why change capitalization on the K? That doesn't mean anything, unless k=1024 and K=1000. But "10mb" vs "1MB" does mean millibits in the first case and megabytes in the second- not likely what you intended.
---snip
-"I just doubled the ram in my Apple II+ by installing 16k of dram" (kilobytes)
-"That board is populated with 8Kb chips"
-"There are 1024KB in one MB"
My casual usage (may be wrong, corrections welcome):
Back when everything was measured in KiloByte units or less (something measured in megabytes was once as uncommon as measuring in terabytes today), if you just said "k", it was understood that you meant "KiloBytes", with a byte having exactly 2 nibbles
If you were working with something specifically measured in bits, such as an 8 Kilobit chip, you would add a lower-case "b" to the end; "Kb" or "kb" are both acceptable, everyone knew that "k" meant 1024. For Bytes, you added an upper-case "B"; saying "8 kB" was understandable, but doesn't look right, so typically, if you are going to capitalize the "B", you would capitalise the "K" as well; "8KB".
Later (in the megabyte era) the drive manufacturers had eventually all began using their own versions of "mega" which did not match up with the "real" mega
The simplest way around that ambiguity is to measure everything in bits, rather than bytes.
---snip
Aside from that, what you're saying is that you seem to need 10 bits of connection to transmit 8 bits of data. That's a 25% overhead, well in line with the the 10% suggested by joshuac, but completely different from the 600000% increase claimed in the article.
---snip
Ummm, no, the protocol overhead I was talking about was the packet headers for each TCP packet. The overhead he is referring to (I believe) is the overhead for the connection itself. To transmit one 8 bit byte, it can easily take 10 bits (or more).
1 start bit, 8 data bits, 1 stop bit possibly? Or he might be using 8 data bits, 1 stop bit, and some type of parity checking. Or he could have been using something quite obscure for some reason, 8 data bits, 2 stop bits perhaps.
In any case, dividing by 10 is probably the right way to measure bytes per second on his connection. I was talking about the additional overhead that is there for the tcp protocol itself.
In theory; the reality is that the error correction used on most modems during this era (MNP-5 from the multitech world or v.42bis(*) for everyone else) was less than perfect. Enough packets sent, and eventually one would get through with a false positive. For example, if you are using a simple 8 bit checksum, it would be pretty easy for an occaisional bad packet to slip through. The hardware error correction was much better than an 8 bit checksum, but it was still quite possible to lose a long transfer. Hence, the long-lasting popularity of zmodem well into the day of error correcting modems.
Are you counting everything going over your connection (such as the ~5% tcp protocol overhead itself which would not be reduced by "fast tcp") or are you just counting the transfer of the data you see in the end?
Look at the interface statistics on your router when transfering, it will probably show 100% on whatever your slowest interface is (in your case, WAN/DSL side).
And then of course...even if you are using SDSL, your "640k down" will not remain a perfect 640k if there is much noise on the line from outbound traffic. However I suspect this is not the case in your situation.
Actually, the Zmodem that was widely used (real zmodem) maxed out at 1k blocks, but it would steadily scale down to as small as 16 byte blocks (if I recall correctly).
:)
There were variants that did 8k blocks (and often referred to themselves as Zmodem8k), but none of these were true zmodem protocol.
Still, nothing can be quite as fast as ymodem-g
A little more on topic; what they are describing does not dynamically scale the packet size, only dynamically adjust the transmission speed up to the point that ack's start slowing down, but (hopefully) before any packets actually get dropped. I suspect disney and such will be quite disappointed if they think they are going to get a 6000x speedup in practical use as hinted at in the articles. Perhaps a 10% speedup for joe blow on a dialup modem, _maybe_. Take a look at your connection some time when downloading a file; you will probably find you can already peg your bandwidth quite nicely.
Ummm, hello, can we say latency? A swarm of LEO sattelites is really the only way to go.
That sounds an awful lot like a 3 byte instruction to me...when that opcode is hit, the PC is incremented by 3 (or 2 for the zero-page variation on your instruction. BRK, NOP etc. would be what I think of when someone says 1 byte instruction...
But then by your definition, all the instructions on the 6502 (or pretty much any other 8 bit processor) were 1 byte instructions...
---snip
( for example one instruction would look up an address, read the word, add it to the XY register and then look up the resutling number as an address---the definition of a pointer offset included in a single byte machine code instruction
---snip
There was a one byte opcode that did this?
nah, CDMA is way better.
Perhaps you are thinking of mousetext? Special characters shaped appropriately to make drawing borders and such much easier.
Ummm, you are just describing a plain ol' DHCP server, there is nothing special about that config.
---snip
your computer is assigned a permanent TCP/IP address that stays with your MAC address/Account (Which are linked on a database somewhere)
---snip
That "database somewhere" of IP address to MAC address is the DHCP client lease table on your DHCP server.
Unless you are saying that you IP address follows your _user_ account rather than your NIC's MAC address, in which case your school would truly have something different.
---snip
I'm not sure the hardware/software solution that's required to implement this, but on a campus of 5,000+ people, it surely cuts down on tech support calls.
---snip
A 386, a NIC, and the OS of your choice (Hey, even NT 3.5x ran on 386's). If you have got more than 5 machines on your network, it is probably time for a DHCP server.
Just set the lease time to longer than the likely longest period of time a user will have their machine disconnected from the network, and you are fine; the DHCP server will go out of it's way to make certain that MAC address gets the same address re-assigned to it.
Hell, Netware 3 did it, nothing new. I _believe_ most good RAID controllers do this for their individual drives, as well.
Or just change your preferences...
---snip
but in my experience, trying to manage windows from the command line is a horrendous task.
---snip
For sure, different OS's are more or less admin-friendly. But difficult is not the same thing as impossible.
Trying to manage windows from the CLI _is_ a much more difficult task than with most other OS's. However, with the right tools you can manipulate the registry database, kill or start a process, manipulate files, or just about anything else you would need to do. In any practical real-world situation you would not hassle yourself with trying to work that way, unless it were a task particularly suited to the command line (automation) or you were in a situation where a CLI was all you had available; remote admin of windows machine is almost always done in a GUI (although many of those GUI tools can be ran locally to manipulate a server remotely over a network).
A good admin _can_ accomplish all the same things, but too often people say "it cannot be done" out of ignorance rather than fact.
Woodworking is a hobby I am just starting out on. Ask me to make a particular type of joint with certain tools, and I will say it is impossible. Ask that of someone who has been working with wood his whole life and knows what he is doing, and he can make it happen. I said the tools were not capable of it, but in fact, it was me.
---snip
:)
If I have the volume up, not only can I "hear" the screen redraws, but the mouse move, my keyboard pulse, and my network card go to work
---snip
Or your algorithm finish...I used to leave my stereo next to my apple ii; when I had a calculation that was going to take especially long to finish running (plotting fractals at hi-res was a favorite), I'd crank up the stereo on a frequency that picked up the activity well. When the machine was performing a complex task, the sound was very white-noisy. When the machine finished and went into a loop polling the keyboard memory location, the sound was a buzz. I could be in the other room and know when things were done, or in some cases know how _soon_ it would be done.
Great, now I'm starting to feel homesick for that old machine, and I'll probably end up unpacking it this weekend (again). Some things an emulator just cannot replace.
---snip
:).
:)
Not to mention the remote access aspect of the command line. I'm responsible for maintaining a webserver running Linux; it's so handy to be able to pull up an ssh session and do whatever needs to be done. Let's see the admins of the Windows boxes neighboring it work on their servers from home.
---snip
It's not difficult to setup ssh on windows, and redirect the io of cmd.exe (or most any other CLI shell available for windows, come to think of it) through it if you know what you are doing. The problem isn't so much the OS, it's the type of admins the OS attracts. 90% of windows "admins" do not know how to use the cli and cli apps included with windows, or even use many of the graphical management tools included to their full extent.
With NT 5 native RDP support was added, although you have been able to make any NT machine support RDP for awhile now, (assuming you do not need multi-user support; in that case your stuck with their terminal server distribution, or purchasing something re-written by Citrix and dumping RDP for their protocol), and with NT 5 they made a telnet server an official part of the distribution (not that I would suggest using it). Of course, you could do all this years ago with NT 4 as well...(and with NT 3.5x with a little adaptation, probably).
BUT...do many NT "admins" know how to do this, or even have the slightest clue about where they would go to get started? Nope. And from what I have observed, in more than a few places having a _little_ more knowledge about the OS they "support" would have made their lives a lot easier...but they don't bother, they just keep slogging on with what little knowledge someone force fed them at a certification class.
Unix admins on the other hand tend to actually take an interest in knowing what makes their OS tick. Partly that the OS is much more open (for some Unix OS's, you can even get the source code easily). Partly that in the Windows world, the interface to a server appears identical to the interface to a workstation, which gives confidence to users who want to become an admin where they deserve no confidence at all
A good admin can make up for a weakness on any platform, whether it be Netware, Linux, Windows, or Lantastic, without needing to blame his/her tools. A lousy "admin" can also more than compensate for all the strengths of a platform. What matters most is the competency of the admins that tend to work with those systems, not really the systems themselves.
That all said, given a choice, I know which platform is takes me much less effort to secure than the others...heck, you might almost say it is "secure by default"
Did you even read my post? :)
In the grand scheme of things, Salon is just another online newsjournally type site, if they go away, we are not losing anything unique. The Whole Earth 'Lectronic Exchange on the otherhand, has been around since the 80's...I remember having the The WELL on my dial list as I went through my nightly BBS romp.
Unfortunately, through various twists and turns, The WELL ended up under Salon's ownership. If Salon goes under, does this mean The WELL will also? That would be horrible and ironic end for something as insignificant as Salon to take down a piece of internet history with it...in 10 years (assuming Salon shuts down, which it likely will) few people will remember or miss Salon...but 20 years down the road, The WELL is still remembered for it's place in the early internet, and I know there are plenty of people who will miss it.
Otoh, maybe more people like Salon than I realize; but somehow I doubt it is the same as The WELL.
I hope if Salon goes under, that The WELL will somehow be preserved. As for Salon.com, I could care less.
Ok, your wrong. It's a bunch of helium :)
we're giving up... :)
---snip ... but back then, I was learning DOS 3.3 (for Apple II) and AppleWorks, becuase "everyone in the future will be using this stuff!"...
,d2" showed them the contents of the second drive on the current controller, I realized they were selling themselves short; there was always more to learn; they didn't "know DOS". But since they assumed they knew everything, there was no room for them to learn more.
Anyway, I took that seriously, and made damn sure that I *knew* to enter the proper date when Appleworks was starting up, and that I *had* to make sure I had the right disks in the drives.
---snip
---snip
Anyway, to get to my point, I wonder how relavent the things that they learn now will be a few years after they graduate - and I hope it is *concepts* that they learn, instead of cookie cutter "type CATALOG to see a what's on your disk, insert your disk and type PR#6 to start AppleWorks" stuff...
---snip
I'm totally with you, the way things are "taught" is a big pet peeve of mine.
My school had a similar class to the one you describe, but for various reasons I never found my way into it. I was fortunate enough to have an Apple IIe of my own at home, with a few random reference books for various things. Instead of the rote class learning, I was teaching myself 6502 assembly (because I had a reference to 6502 opcodes and had found a way to get myself into the miniassember that came with Integer Basic), and learning the various subroutines that could be called in DOS 3.3. When I got my hands on 3.3E by begging and stealing from our nazi computer "teacher" at my elementary school, I remember the joy of decompiling 20 instructions at a time to get glimmer of what the minor differences were in the code. As time went on, I taught myself how to automate writing much of my assembly by using WPL, a very under-appreciated scripting language internal to AppleWriter II (well, Don Lancaster knew how good it is). Later, (once I had gotten my hands on the reference books) I taught myself the in's and outs of high level languages like Integer Basic and Applesoft; when I managed to sneak off with a copy of Apple Logo, I learned everything I could about that, because "it's fun to learn what makes things tick!"....meanwhile, the class learned the syntax of various DOS 3.3 commands.
Anyway, I took nothing very seriously, at that formative age (10 years old) I had found a toy that had limitless possibilities, that could be reprogrammed to perform any task you could conceive. The class was being taught how to operate a tool within narrow confines of specific pre-decided tasks.
Now, almost 15 years later, what I learned then on that Apple IIe was invaluable; what I learned that was truly valuable was not how to interface with a disk ii controller and count clock cycles for timing in my ML loops, it was that I learned something about learning. The class had learned how to be told what to do.
The most valuable thing that IIe taught me is that you are fooling yourself if you tell yourself you "know" everything about a subject. When people say they "knew" DOS 3.3 because typing "catalog
The next most valuable thing I discovered was how to pull something apart and learn how it works, without a master plan in front of you. Too many people have been taught to "learn" by being shown an example, and then emulating. It's faster, it get's the grade school concert band able to push out a few notes in time for their parents to be proud during the winter concert, but rote knowledge is a poor subsititue for actual understanding. Type "pr#6" to boot off the floppy in drive 1, slot 6...does that actually teach you anything about what is going on, or are you just mechanically following directions? When all you learn is to follow directions, inovating when given an unexpected problem is very difficult. When you understand what is happening, you give yourself many more choices, and much more control.
Anyhow, I learned many of the same subjects that the computer class at my grade school set out to learn. But I suspect that over time I got much more out of my learning experience than those students did, simply because of the way they were forced to learn.
To make this slightly on-topic, does anyone know how the Redhat classes are taught? Do the teach you think unix, or do they teach you the syntax of commands?
Yes, if they thought something was seriously wrong right after liftoff (which they checked, and concluded nothing serious had happened), it is possible to get a second shuttle up in the time of the mission if everything went on schedule with that second launch.
In that case, I totally agree with you :)
If they had the ability to inspect the shuttle from outside, and they thought it necessary (remember, the insulation torn off the main tank was ruled unimportant, and nothing else went visibly wrong with the launch), then maybe (assuming the problem was damaged tiles and was visible from outside) they might have found the problem before attempting rentry, and would have been able to figure out some miraculous Apollo 13-style save.
But then just lugging the already existing robotic arm (which wasn't taken on this trip) along would have accomplished the same self-inspection tasks easier and better than this microsattelite could. The microsattelite is pretty cool in itself tho, but on the orbiter, just take the arm with a camera.
---snip
So you already know, without knowing the actual extent of any theoretical damage that a minimal approach rather than the normal one would have made no difference? What would have stopped them from, e.g. getting another shuttle up, stopping by the space station, etc.?
---snip
I doubt there is a much more "minimal" approach than what is already used by the shuttle. As for getting another shuttle up, the soonest an emergency launch of atlantis could be performed would be a week, assuming everything went well. Considering how often launch dates have to be pushed back even under ideal conditions, actually pulling off an emergency launch in a week is a long shot.
And no, stopping by the space station was not an option, Columbia was in a much lower orbit than the iss and orbiting at a different angle (I know there is better term for that, anyone?).
There might have been _something_ they could have done, but those three options can pretty much be ruled out.
and what is _really_ scary is looking at the this list, it looks like plenty of admins have been accessing this system from home; the log dates back to 1-jan-2002. If you are a lazy cracker, grep for all the lines with "DSL" in them, and probably 80-90% of those hosts are home workstations of military sysadmins of one type or another. If they are dumb enough to leave logfiles of users accessing a server used for military network administration open to the public, imagine what their home computers are like...
What's even more depressing is that it looks like some of these guys use AOL...
---snip :(
Given that this thing is intended to glide like airplane, except in water, I wonder what it would take to make it able to fly in air? Probably a lot of engine power that it doesn't have, and a lot less weight.
---snip
...and for the hydrofoils that are designed to create downward force when moving through the water to start making upward force when being used as airfoils, amongst other problems...but a flying submersible would rock. The police vtol jet plane/wheeled land vehicle (I think)/submersible shown briefly in the movie "AI" was pretty neat...
more interesting than other parts of the movie, come to think of it.