The first is address space with simple multi-threaded apps.
Under a 32 bit address space, and assuming a 2MB stack, only 1000 threads can be started. It would be reasonable to support more threads, and with a 64 bit address space, it is.
Dealing efficiently and simply with large data (movies) is also facilitated by having a large address space.
If the software can be simplified, the systems can be more reliable. And for applications like a home media center, I believe that this is important.
No, I don't want Oracle. But I do want reliable multimedia, with no "2GB" file size limits. I do want simple, reliable software.
Sure, we only really need a ___ (fill in the blank) bit system! I am sure that we could do with simple 8 bit systems for most tasks, limited to 64KB addressable at any one time. But we have progressed.
The 64 bit consumer processor will give us the ability to run simpler software and build more complex functions on that base. I *do* want voice recognition in my appliances. And stuff like that.
Very nice - integrated sound, video, 10/100 NIC. Low power.
Unfortunately, came with a CPU fan (but that can be removed, and the CPU underclocked).
Can boot from the NIC, so no hard drive or floppy needed. I run one headless as as MP3 player.
Another is a network fax machine.
And etc.
Cheap, slow, good (as in -- didn't crash, or cause great deal of other grief wrt. drivers, etc. at least not with Linux 2.0).
What "sucks" for some is manna for others. No, I *never* intend to install Oracle on these boxes. I *never* intend to benchmark with reference to a PII. Until I can get an underclocked PII mini-mainboard for $10 at least.
And no, the Samuel 3 was not my first choice. That would have been an older Transmeta. The price point wasn't there, though...
You are, of course, refering to the US Constitution in an Icelandic case.
Just to let you know... in my jurisdiction, sharing music files is a legal activity. Movies? maybe (or not). Hasn't been tested yet.
US Constitution? About the same use for me as toilet paper has. Well, it is an interesting read, but it's moot. Of course, we do have idiots spouting "You didn't read me my rights! Miranda!". Too much TV.
Now, I DON'T know Icelandic law, so all I will say is -- we'll see.
Movie: 13 CDN per ticket. 26 for my wife and I. 20 CDN for popcorn/drinks (the theatre doesn't allow food to be brought in). 3 hours baby sitter - 20 - 25 CDN.
Total for a movie session: 70 CDN.
42" projection TV: 1300 CDN surround sound: 250 CDN. Total 1550. DVD rental: 5 dollars. We save 60 - 65 on NOT going to the theatre.
Payback time: ~30 movie trips, which means between 1 and 2 years.
If we BUY the DVDs: saving is 40 for not going. Payback is ~40 movie trips, or 2-3 years.
Given that we intend to hold onto the TV and sound system for 10 years, its a good deal.
If *you* can do a better job supporting the product, go right ahead. Just remember, you are going against the expert.
If no support is needed, or customization, its probably horizontal. Sure, GPL, because that software should be "in the commons". If its vertical, it needs support/customization.
You don't have to supply the software for download... It is permissible to *sell* GPL programs, and only make the source available to people who bought it from you.
You can charge as much as you like for the software. If it is required for someones business, and they buy it, there is pressure for *them* to not redistribute (and lose competitive advantage).
If *you* actually have a contribution, it is reasonable for "people/companies" to pay you. Otherwise, they lose your contribution. Call it a "support model" if you want.
Of course if someone ELSE gets hold of the software, then they can distribute it. Doesn't help the users much -- because "support" isn't available. Still, can be pain in the ass...
It is much better if the *only* interface is the CLI. An interface can be layered with Tcl/Tk (et al) on top to provide the interface. (or python, etc.).
"When Windows 98 was invented, it was reasonably secure."
Best quote of the day! Mod this funny!
Windows 98 wasn't secure. "Press ESC instead of putting in your password". You do remember? This runs in 32 bit mode, but ALWAYS as root. &etc.
The comparision to the car is silly. An alarm can be easily installed in the care, as can a lojack. Problem solved. You don't need to purchase a brand new care in order to benefit from a lojack.
And that's the ONLY problem here -- people cracking in. There was/is abolutely no reason for there to be such grevious security issues with Windows 9x. It's not like the 'net was a new idea, or security was a new idea. Windows 9x design may even be the base for a class action suite (to recover the upgrade fees to Windows NT). Especially Windows 98, with IE merged.
Um... no, the C/C++ (and, yes, I am combining them -- I mean "lower level language/intermediate language that translates to machine code") does NOT see what your program is doing. It can only guess.
On the other hand, a re-compiler can watch what is happening, and generate better code.
As an example:
if (condition) {...simple code... }... more code...
When compiled, I cannot tell C (or C++) that "condition" is an error that will occur.. well, should never occur..
So, the optimal sequence would be:...evaluate condition...branch if true to exception...more code
Now, if the "simple code" is small, it may be part of the same cache-line occupied by the condition. Wasting resources.
A re-compiler could tell that this is happening (hey, "condition" hasn't been true for the last 100K iterations! let's optimize for that case!).
This is only a small example; I am sure that you can come up with your own.
SquirrelMail for webmail (serving). Evolution, (or other client of your choice). SquirrelMail uses IMAP.
As to Mail server? I use sendmail. It uses whatever authentication is available. FTP daemon? Same deal.
As to "distro" -- try Redhat 9 (no, not kidding -- its probably the best base for what you are aiming at here, maybe Fedora Core 2).
Server configuration time? Depends on how much you know. An hour, a week -- anyone's guess. There are GUI tools available. On the client? If set up properly, it's a non-issue.
It would be helpful to have a functional spec from you that I can work against (I don't want to raise the bar too high or leave parts unfinished).
But I'll start with a hypothetical. Simple SOHO LAN, 10 machines. Cable/DSL connection to the 'net.
Configure headless server, running DHCP, Web Proxy, BIND, NFS, NIS -- 3 hours to load OS, 1 hour to configure.
Second headless server, FTP, APACHE, SENDMAIL -- 1 hour to grab basic OS image. 1 hour to configure.
Client -- select applications (we assume OpenOffice.org, Mozilla, Evolution -- may be different). Load OS: 4 hours, (including NIS client configuration, proxy configuration). Duplication to 7 other machines: 7 hours.
Total setup time: 17 hours, or 3 working days.
*NOTES* No "ghost" style utility used, or available. Machines start *completely* blank (no OS). Hardware may differ, but each hard-drive has at least 10GB. Each machine has at least 128MB of memory, and ethernet.
So, I noticed someone in the office was playing Solitaire -- and decided to have some fun.
Got something called "Wilson Windoware" (AFAIR). A macro language for Windows. Ran a macro on the computer that polled for SOL window every 10 or so seconds. When SOL was detected, injected window resize commands to it -- slowly at first, increasing in speed, and then a close message.
So, the person saw Solitaire launch, play normally for around 30 seconds, start to shrink, shrink and vanish. Good fun! I wanted to see what Helpdesk support (if any) was logged. None, whatsoever.
The "idea" of an object cannot be owned. If it/is/ owned, it becomes rather useless. As soon as the idea becomes a meme, and crosses into another person, exclusive ownership is lost...
Worse, it can be impossible to say/who/ had the idea first.
The only ownership that can be imposed is government supported monopoly.
And that's the system we have. There is no "natural" or other right.
Where a natural ownership could be shown is in a particular physical manifestation of that idea.
And that's why there is a distinction between theft and (for example) copyright infringement. They are/not/ the same thing at all.
Please stop trying to equate the two issues -- it cannot be done. Consider a teacher who photocopies books and materials for her students. The students are unaware that copyright has been infringed, but are told that copyright has been further granted to them... Are the students guilty of/anything/? No. And why not? In your argument it should be clear (somehow -- it is a/natural/ law according to you) that theft has occurred. But, the students in this case are not liable... A question for you -- why?
With "a book, a song, a chair, a pie", you own the medium. If someone takes your pie, it's theft.
If someone bakes ANOTHER pie, you don't have any say.
If you have a chair, someone else can make a chair, no harm.
The book and song are different. And that's copyright. You can't copy them (you CAN copy the chair and the pie). Someone else, never having read the book, or having heard the song can produce them, but the gov. doesn't permit them to copy either.
I run my own mail server (vanity addresses). Up until very recently my ISP mail server allowed me to use their SMTP server to relay -- and forced me to authenticate. Which worked well.
Recently, my ISP switched to "yahoo" for mail services. yahoo uses qmail, and is configured to not allow any relaying at all. Which means I am back to directly sending email.
Oh well, back to the drawing board (I don't want my email to ORIGINATE from yahoo.com, I just want them to relay my email properly).
1 - Downloading music is not/stealing/, it/may/ be copyright infringement.
2 - depending on where you live, it may not even be copyright infringement, but may be perfectly legal (and moral).
3 - It/is/ your property -- a "shrink-wrap license" is not, and can not, be a valid contract. However, be very careful about backups -- a copy is allowed to install, and a copy/may/ be allowed to execute (not always, strangely). Another copy for backup/may/ be allowed, depending on where you live.
"To generate accurate statistics, Windows Update evaluates a Globally Unique Identifier (GUID) that is stored on your computer to uniquely identify it. The GUID does not contain any information that can be used to identify you."
Um... the GUID *is* an indentifier for that particular computer. It will not identify an individual *only* in the in the case of a shared computer. Other than that, "they" know who you are. Even if your IP address changes (eg. you move, or use DHCP), the GUID WILL NOT CHANGE. So the GUID is better suited as an identifier.
I call "bs". Which application are you referring to? I assume that it's not java, but a windows dll style plugin (activex). In which case, IE and Windows/is/ the platform. IE/Solaris and IE/Mac wouldn't work. Not smart, tying to a single vender/that/ tightly. But those boxes should not be "internet" enabled.
And, most of that crud should be gone. Even MS is pointing you at.net
Yup, its an attack vector. Real Story: A major vendor of network bandwidth wanted to remotely bring up router boxes... using dhcp etc. And remotely boot those boxes. Over the internet.
So, here is the attack vector:
Hacker intercepts communication, capturing digitally signed OS that boots the router box. Now, this is useless, because all the hacker can do is wait for the box to reloaded, and then feed it the same OS image.
Imagine, though, that a little time has gone by. Now, some vulns are known in that particular core. Now, when the router box is rebooted, the hacker can put BACK the vulnerable OS, and rehack.
So, each download image must be secured against the box, to avoid replay attacks. Yes, this was brought up up the review of the "load OS over the internet" project.
Of course, all names have been filed off, to protect.... well, whoever.
Of course no mass market data producer is going to re-crypt data for each individual. That would mean that the OS/CONTENT would have to be re-encrypted after registration, *or* every retail copy would be different, *or* there would be no retail copies (only pre-installed).
No need for Oracle, no need for 64 bit...
Interesting idea. I'd mod this idea up.
But... there are other uses for 64 bit-ness.
The first is address space with simple multi-threaded apps.
Under a 32 bit address space, and assuming a 2MB stack, only 1000 threads can be started. It would be reasonable to support more threads, and with a 64 bit address space, it is.
Dealing efficiently and simply with large data (movies) is also facilitated by having a large address space.
If the software can be simplified, the systems can be more reliable. And for applications like a home media center, I believe that this is important.
No, I don't want Oracle. But I do want reliable multimedia, with no "2GB" file size limits. I do want simple, reliable software.
Sure, we only really need a ___ (fill in the blank) bit system! I am sure that we could do with simple 8 bit systems for most tasks, limited to 64KB addressable at any one time. But we have progressed.
The 64 bit consumer processor will give us the ability to run simpler software and build more complex functions on that base. I *do* want voice recognition in my appliances. And stuff like that.
Ratboy.
I bought several VIA Samuel 3 main boards.
Very nice - integrated sound, video, 10/100 NIC. Low power.
Unfortunately, came with a CPU fan (but that can be removed, and the CPU underclocked).
Can boot from the NIC, so no hard drive or floppy needed. I run one headless as as MP3 player.
Another is a network fax machine.
And etc.
Cheap, slow, good (as in -- didn't crash, or cause great deal of other grief wrt. drivers, etc. at least not with Linux 2.0).
What "sucks" for some is manna for others. No, I *never* intend to install Oracle on these boxes. I *never* intend to benchmark with reference to a PII. Until I can get an underclocked PII mini-mainboard for $10 at least.
And no, the Samuel 3 was not my first choice. That would have been an older Transmeta. The price point wasn't there, though...
Ratboy.
Interesting.
You are, of course, refering to the US Constitution in an Icelandic case.
Just to let you know... in my jurisdiction, sharing music files is a legal activity. Movies? maybe (or not). Hasn't been tested yet.
US Constitution? About the same use for me as toilet paper has. Well, it is an interesting read, but it's moot. Of course, we do have idiots spouting "You didn't read me my rights! Miranda!". Too much TV.
Now, I DON'T know Icelandic law, so all I will say is -- we'll see.
Ratboy666
What is "illegal media"? In my jursidiction, that would include child pornography.
The GOVERNMENT has the responsibility to take down people trading in illegal media.
As to trading in copyrighted material -- not against any law (at least not US, Canadian, etc. It may be in other jurisdictions).
So, what ARE you saying?
Ratboy.
Cost breakdown:
Movie: 13 CDN per ticket.
26 for my wife and I.
20 CDN for popcorn/drinks (the theatre doesn't allow food to be brought in).
3 hours baby sitter - 20 - 25 CDN.
Total for a movie session: 70 CDN.
42" projection TV: 1300 CDN
surround sound: 250 CDN.
Total 1550. DVD rental: 5 dollars. We save 60 - 65 on NOT going to the theatre.
Payback time: ~30 movie trips, which means between 1 and 2 years.
If we BUY the DVDs: saving is 40 for not going. Payback is ~40 movie trips, or 2-3 years.
Given that we intend to hold onto the TV and sound system for 10 years, its a good deal.
Ratboy.
Um...
Ok, I have a wireless connection on a computer (wlan0) via USB. The computer also has a PCI ethernet (eth0).
I want to "share" the wireless connection to my server (and then acess the 'net).
I run "parprouted wlan0 eth0", and it "just works". And if its two ethernet cards, standard bridging can be used "man brctl".
Big deal. But now I *am* curious. Can Windows share a wireless connection? Or just wired?
Ratboy.
first, you can run programs with local libpath. second, multiple library versions can co-exist. third (under solaris) apis are versioned.
Sure, sell it and undercut the original vendor...
If *you* can do a better job supporting the product, go right ahead. Just remember, you are going against the expert.
If no support is needed, or customization, its probably horizontal. Sure, GPL, because that software should be "in the commons". If its vertical, it needs support/customization.
Ratboy
You don't have to supply the software for download... It is permissible to *sell* GPL programs, and only make the source available to people who bought it from you.
You can charge as much as you like for the software. If it is required for someones business, and they buy it, there is pressure for *them* to not redistribute (and lose competitive advantage).
If *you* actually have a contribution, it is reasonable for "people/companies" to pay you. Otherwise, they lose your contribution. Call it a "support model" if you want.
Of course if someone ELSE gets hold of the software, then they can distribute it. Doesn't help the users much -- because "support" isn't available. Still, can be pain in the ass...
Ratboy.
It is much better if the *only* interface is the CLI. An interface can be layered with Tcl/Tk (et al) on top to provide the interface. (or python, etc.).
Look at the way Redhat does configuration.
(we do want usability, right?)
Ratboy.
You mean StarOffice didn't edit HTML? You mean OpenOffice doesn't use XML? You mean that Excel actually does stats correctly?
Ratboy
"When Windows 98 was invented, it was reasonably secure."
Best quote of the day! Mod this funny!
Windows 98 wasn't secure. "Press ESC instead of putting in your password". You do remember? This runs in 32 bit mode, but ALWAYS as root. &etc.
The comparision to the car is silly. An alarm can be easily installed in the care, as can a lojack. Problem solved. You don't need to purchase a brand new care in order to benefit from a lojack.
And that's the ONLY problem here -- people cracking in. There was/is abolutely no reason for there to be such grevious security issues with Windows 9x. It's not like the 'net was a new idea, or security was a new idea. Windows 9x design may even be the base for a class action suite (to recover the upgrade fees to Windows NT). Especially Windows 98, with IE merged.
Ratboy.
"Omega Hacker"
Of course ONLY you could conceive of the functionality locked within hardware.
Of course ONLY the manufacturer could supply good firwware for an embeded system.
Of course we can't imagine the timing requirements.
What arrogance!
Let you in on a little secret -- there are better programmers out there.
Why not loosen up, and try supporting the hacker ethos?
Ratboy
Um... no, the C/C++ (and, yes, I am combining them -- I mean "lower level language/intermediate language that translates to machine code") does NOT see what your program is doing. It can only guess.
...simple code... ... more code...
...evaluate condition ...branch if true to exception ...more code
On the other hand, a re-compiler can watch what is happening, and generate better code.
As an example:
if (condition) {
}
When compiled, I cannot tell C (or C++) that "condition" is an error that will occur.. well, should never occur..
So, the optimal sequence would be:
Now, if the "simple code" is small, it may be part of the same cache-line occupied by the condition. Wasting resources.
A re-compiler could tell that this is happening (hey, "condition" hasn't been true for the last 100K iterations! let's optimize for that case!).
This is only a small example; I am sure that you can come up with your own.
Ratboy.
Windows and "generic drivers"... In a word, NO. [There are standard drivers, but Linux has more].
However, everything does come with a driver disk. Quality of the driver may be very iffy, but its there.
As to blaming Linux... not a trust issue at all. The manufacturers simply don't produce linux drivers [and there are notable exceptions].
So, if Linux doesn't have the driver, it doesn't work -- and the user has nothing except "Linux" to blame.
Ratboy
SquirrelMail for webmail (serving). Evolution, (or other client of your choice). SquirrelMail uses IMAP.
As to Mail server? I use sendmail. It uses whatever authentication is available. FTP daemon? Same deal.
As to "distro" -- try Redhat 9 (no, not kidding -- its probably the best base for what you are aiming at here, maybe Fedora Core 2).
Server configuration time? Depends on how much you know. An hour, a week -- anyone's guess. There are GUI tools available. On the client? If set up properly, it's a non-issue.
It would be helpful to have a functional spec from you that I can work against (I don't want to raise the bar too high or leave parts unfinished).
But I'll start with a hypothetical. Simple SOHO LAN, 10 machines. Cable/DSL connection to the 'net.
Configure headless server, running DHCP, Web Proxy, BIND, NFS, NIS -- 3 hours to load OS, 1 hour to configure.
Second headless server, FTP, APACHE, SENDMAIL -- 1 hour to grab basic OS image. 1 hour to configure.
Client -- select applications (we assume OpenOffice.org, Mozilla, Evolution -- may be different). Load OS: 4 hours, (including NIS client configuration, proxy configuration). Duplication to 7 other machines: 7 hours.
Total setup time: 17 hours, or 3 working days.
*NOTES* No "ghost" style utility used, or available. Machines start *completely* blank (no OS). Hardware may differ, but each hard-drive has at least 10GB. Each machine has at least 128MB of memory, and ethernet.
Does this help?
Ratboy
So paint me a shade of confused. What is "single sign-on"?
How does it differ from NIS? Or RADIUS?
Just curious.
Ratboy.
So, I noticed someone in the office was playing Solitaire -- and decided to have some fun.
Got something called "Wilson Windoware" (AFAIR). A macro language for Windows. Ran a macro on the computer that polled for SOL window every 10 or so seconds. When SOL was detected, injected window resize commands to it -- slowly at first, increasing in speed, and then a close message.
So, the person saw Solitaire launch, play normally for around 30 seconds, start to shrink, shrink and vanish. Good fun! I wanted to see what Helpdesk support (if any) was logged. None, whatsoever.
Still, good for a laugh.
Ratboy.
The "idea" of an object cannot be owned. If it /is/ owned, it becomes rather useless. As soon as the idea becomes a meme, and crosses into another person, exclusive ownership is lost...
/who/ had the idea first.
/not/ the same thing at all.
/anything/? No. And why not? In your argument it should be clear (somehow -- it is a /natural/ law according to you) that theft has occurred. But, the students in this case are not liable... A question for you -- why?
Worse, it can be impossible to say
The only ownership that can be imposed is government supported monopoly.
And that's the system we have. There is no "natural" or other right.
Where a natural ownership could be shown is in a particular physical manifestation of that idea.
And that's why there is a distinction between theft and (for example) copyright infringement. They are
Please stop trying to equate the two issues -- it cannot be done. Consider a teacher who photocopies books and materials for her students. The students are unaware that copyright has been infringed, but are told that copyright has been further granted to them... Are the students guilty of
Ratboy.
"Medium" vs. "Message".
With "a book, a song, a chair, a pie", you own the medium. If someone takes your pie, it's theft.
If someone bakes ANOTHER pie, you don't have any say.
If you have a chair, someone else can make a chair, no harm.
The book and song are different. And that's copyright. You can't copy them (you CAN copy the chair and the pie). Someone else, never having read the book, or having heard the song can produce them, but the gov. doesn't permit them to copy either.
Marshall McLuhan:
"The Medium is the Message"
From the man who coined the term "media".
Ratboy
Not true
I run my own mail server (vanity addresses). Up until very recently my ISP mail server allowed me to use their SMTP server to relay -- and forced me to authenticate. Which worked well.
Recently, my ISP switched to "yahoo" for mail services. yahoo uses qmail, and is configured to not allow any relaying at all. Which means I am back to directly sending email.
Oh well, back to the drawing board (I don't want my email to ORIGINATE from yahoo.com, I just want them to relay my email properly).
ratboy
1 - Downloading music is not /stealing/, it /may/ be copyright infringement.
/is/ your property -- a "shrink-wrap license" is not, and can not, be a valid contract. However, be very careful about backups -- a copy is allowed to install, and a copy /may/ be allowed to execute (not always, strangely). Another copy for backup /may/ be allowed, depending on where you live.
2 - depending on where you live, it may not even be copyright infringement, but may be perfectly legal (and moral).
3 - It
ratboy.
"To generate accurate statistics, Windows Update evaluates a Globally Unique Identifier (GUID) that is stored on your computer to uniquely identify it. The GUID does not contain any information that can be used to identify you."
Um... the GUID *is* an indentifier for that particular computer. It will not identify an individual *only* in the in the case of a shared computer. Other than that, "they" know who you are. Even if your IP address changes (eg. you move, or use DHCP), the GUID WILL NOT CHANGE. So the GUID is better suited as an identifier.
Ratboy.
I call "bs". Which application are you referring to? /is/ the platform. IE/Solaris and IE/Mac wouldn't work. Not smart, tying to a single vender /that/ tightly. But those boxes should not be "internet" enabled.
.net
I assume that it's not java, but a windows dll style plugin (activex). In which case, IE and Windows
And, most of that crud should be gone. Even MS is pointing you at
ratboy
Boot time vulns...
Yup, its an attack vector. Real Story: A major vendor of network bandwidth wanted to remotely bring up router boxes... using dhcp etc. And remotely boot those boxes. Over the internet.
So, here is the attack vector:
Hacker intercepts communication, capturing digitally signed OS that boots the router box. Now, this is useless, because all the hacker can do is wait for the box to reloaded, and then feed it the same OS image.
Imagine, though, that a little time has gone by. Now, some vulns are known in that particular core. Now, when the router box is rebooted, the hacker can put BACK the vulnerable OS, and rehack.
So, each download image must be secured against the box, to avoid replay attacks. Yes, this was brought up up the review of the "load OS over the internet" project.
Of course, all names have been filed off, to protect.... well, whoever.
Of course no mass market data producer is going to re-crypt data for each individual. That would mean that the OS/CONTENT would have to be re-encrypted after registration, *or* every retail copy would be different, *or* there would be no retail copies (only pre-installed).
Ratboy.