The idea of the TPM module is to keep valuable keys in it, right? How do I make a backup of them?
Re:...only if the BIOS chip is replaceable.
on
Phoenix BIOSOS?
·
· Score: 1
One way to flash is to have two "banks". You overwrite the other while running in other.
Another is to have minimal "boot & flash" area which is never overwritten.
This problem has been solved nicely.
I do admit that having huge kernel (like Linux is) in bios is a bit adventurous due to trojans, viruses, etc. But maybe the manufacturer has made the job of keeping up-to-date easy.
Re:The Achilles heel of this...
on
Phoenix BIOSOS?
·
· Score: 5, Insightful
How many FOSS drivers must I mention before you admit Linux does have a problem?
More specifically: how many FOSS drivers *which are not maintained in the kernel tree* must I list?
1. MTP008 temperature sensor was removed from 2.6 (was in 2.4). 2. Peracomm USB ethernet (stopped working while in kernel tree) 3. DIB0700 (and many, many other) based DVB cards - the manufacturer helped making the driver but it still (after over 3 years, in 8.10) is not up-to-date/maintained in the kernel tree. 4. Numerous Wifi cards some of which partially work and some not. 5. Webcams (gspca).
Need I go on?
6. EeePCs... most came with Linux, most drivers still do not work even in 8.10.
Nobody claims this is exclusive to Linux, it is just a lot more pronounced in Linux.
My point is that even when drivers are FOSS and the manufacturer has willingness Linux *users* can and do have problems.
I leave it as an exercise to the reader to find out why and who is to blame.
Sorry, I do not know the "climate" in USA very well.
Here (in Finland) usually the situation is completely opposite, there is a system called "good brother" which means the "if you scratch my back, I'll scratch yours".
So too often the bad guys get off with nothing (+ the money), especially if the scam involves EU (you screw money from EU).
It was created so that CEOs could not claim "I knew nothing about financials".
IMHO same should apply for "I knew nothing about SomeIllegalAct".
If CEO "knows" nothing about some illegal behaviour going on for long time by an employee, he is either an idiot (and should not get huge salary nor golden parachute) or "criminally negligent" or good old fashioned liar.
1. "If you are only worried about loose pointers... drop C/C++." 2. If the shared memory (& semaphores, etc.) between processes are corrupted, I can see only one way out: restart all the processes.
There is one difference more: flushing cache(s)[1]. And yet another one: "threads" in CPU core really require threads, processes won't do.
I have never understood why shared memory between threads is "more fragile" than shared memory between processes, but that's just me.
If you are only worried about loose pointers... drop C/C++.
I'd like the browser to be written in some more robust language than C++. Sure, processes help as long as the code is crap (which Fx is, proved by the amount of security holes. And the plugins, as you said).
[1] This depends on the HW: is there any cache before MMU.
You mean they try to make Cable Ready HD[1], Bly-Ray, HDCP, etc. look almost bearable (compared to the ease of DVD)?
[1] In Finland this means that the TV can decipher HD content. The smart card is "paired" with the TV's serial number so you cannot look the content in another TV set (with the card).
The idea of the virtual machine is that should the OS inside the VM hang or crash or do anything whatsoever the host OS is not affected in any (fatal) way.
This is no longer true (in this case).
Should the VM work as designed, it should not be able to happen.
Same applies for the NAT too, btw. It really does give security as it blocks a lot of unwanted accesses by design. It is not obscurity as the attacker can know how and what is protected. Sure it is not a bit safer than a firewall, but it does give additional security (over direct access).
And mentioning 1. no games (except flash/java on the web) 2. there is 50/50 chance your new HW will not work. 3. there will be intermittent problems in quality (how to copy with scanner & printer so that output is *exactly* same size as input) and functionality (Chinese will printout as garbage). 4. any security fix can render your system un-bootable (latest Ubuntu did that for me:/dev/sda was changed to/dev/hdb).
Sure, there are similar and sometimes worse problems with other OS's, it is just that Linux still has too many of those.
Unfortunatley we'll be modded to hell...
Re:Java is safe, mysql is safe...
on
Oracle Buys Sun
·
· Score: 1
I'm more worried about NFS than anything else.
There seems to be very little pNFS enthusiasm outside Sun... I hope I am wrong.
This far I've been told "in extremely cheap machine Linux would make more sense" and "buy more expensive peripherals". Wow.
OEM's cannot compile for the distributions as every minor-minor (security patch) version *requires* different binary. The binary on the CD which comes along the peripheral would certainly not work on any up-to-date distribution. Good? No. Acceptable? IMNSHO no. Preferable? Kernel developers think so.
You did not address my issue, quite contrary. The issue is that there is no "guaranteed to work with Linux HW". The issue is that practically nothing works perfectly out-of-the-box - unless you search and pay more and are lucky. Printers are just nice example.
BTW, have you ever used non-Postscript printer with Linux? I recommend you try to print e.g. Arabic or Chinese. If this is what you mean by "works"...
Don't you get it? It is far too expensive to compile a driver for every distribution & minor version there is. This is the reason why e.g. DVB makers do not give Linux support. And if Asus is not OEM, then who is?
Your advice is hilarious, I can imagine what Asus would think about that... "sure average joe can use Linux - the new drivers do not need to be compiled on the Netbook, they can use distcc".
Slashdot Mirror
I gave you one particular product line on which the WiFi does not "just work" (EeePC).
I am happy to hear you have had no problems. I would be much, much more happier if I did not have any problems either.
Wrong.
1. You need to compile the module.
2. You need to recompile the module on every kernel (security) patch.
3. How do I update the OS? I hate security holes.
5. Explain that idea in the kernel developers list, you will be hammered to death.
Other than that, interesting idea. I think OpenBoot might be better ... or maybe not, don't know enough.
FUD ... hmmm.
The idea of the TPM module is to keep valuable keys in it, right?
How do I make a backup of them?
One way to flash is to have two "banks". You overwrite the other while running in other.
Another is to have minimal "boot & flash" area which is never overwritten.
This problem has been solved nicely.
I do admit that having huge kernel (like Linux is) in bios is a bit adventurous due to trojans, viruses, etc. But maybe the manufacturer has made the job of keeping up-to-date easy.
How many FOSS drivers must I mention before you admit Linux does have a problem?
More specifically: how many FOSS drivers *which are not maintained in the kernel tree* must I list?
1. MTP008 temperature sensor was removed from 2.6 (was in 2.4).
2. Peracomm USB ethernet (stopped working while in kernel tree)
3. DIB0700 (and many, many other) based DVB cards - the manufacturer helped making the driver but it still (after over 3 years, in 8.10) is not up-to-date/maintained in the kernel tree.
4. Numerous Wifi cards some of which partially work and some not.
5. Webcams (gspca).
Need I go on?
6. EeePCs ... most came with Linux, most drivers still do not work even in 8.10.
Nobody claims this is exclusive to Linux, it is just a lot more pronounced in Linux.
My point is that even when drivers are FOSS and the manufacturer has willingness Linux *users* can and do have problems.
I leave it as an exercise to the reader to find out why and who is to blame.
Sorry, I do not know the "climate" in USA very well.
Here (in Finland) usually the situation is completely opposite, there is a system called "good brother" which means the "if you scratch my back, I'll scratch yours".
So too often the bad guys get off with nothing (+ the money), especially if the scam involves EU (you screw money from EU).
You do know why SOX was created?
It was created so that CEOs could not claim "I knew nothing about financials".
IMHO same should apply for "I knew nothing about SomeIllegalAct".
If CEO "knows" nothing about some illegal behaviour going on for long time by an employee, he is either an idiot (and should not get huge salary nor golden parachute) or "criminally negligent" or good old fashioned liar.
1. "If you are only worried about loose pointers ... drop C/C++."
2. If the shared memory (& semaphores, etc.) between processes are corrupted, I can see only one way out: restart all the processes.
I do NOT want 30 different processes, all firefox tabs, using up all my cores just to run spam animations.
I do. I really, really DO.
Rather than have just one core being slowed down to a crawl making *ME* wait.
Bloody, hell that's why I bought dual core!
There is one difference more: flushing cache(s)[1].
And yet another one: "threads" in CPU core really require threads, processes won't do.
I have never understood why shared memory between threads is "more fragile" than shared memory between processes, but that's just me.
If you are only worried about loose pointers ... drop C/C++.
I'd like the browser to be written in some more robust language than C++. Sure, processes help as long as the code is crap (which Fx is, proved by the amount of security holes. And the plugins, as you said).
[1] This depends on the HW: is there any cache before MMU.
I am not willing to say "A" is sane. Perhaps "A in a sandbox" would be sane, but not "with full (user/admin) privileges".
Users just cannot be trusted not to run "britneynude".
Sorry, but *lack of correlation* does prove there is no causation.
Or to be more accurate: if there is, it is so small that it is lost in noise. And this is all the GP said ...
Amen!
You mean they try to make Cable Ready HD[1], Bly-Ray, HDCP, etc. look almost bearable (compared to the ease of DVD)?
[1] In Finland this means that the TV can decipher HD content. The smart card is "paired" with the TV's serial number so you cannot look the content in another TV set (with the card).
The idea of the virtual machine is that should the OS inside the VM hang or crash or do anything whatsoever the host OS is not affected in any (fatal) way.
This is no longer true (in this case).
Should the VM work as designed, it should not be able to happen.
Same applies for the NAT too, btw. It really does give security as it blocks a lot of unwanted accesses by design. It is not obscurity as the attacker can know how and what is protected. Sure it is not a bit safer than a firewall, but it does give additional security (over direct access).
And mentioning /dev/sda was changed to /dev/hdb).
1. no games (except flash/java on the web)
2. there is 50/50 chance your new HW will not work.
3. there will be intermittent problems in quality (how to copy with scanner & printer so that output is *exactly* same size as input) and functionality (Chinese will printout as garbage).
4. any security fix can render your system un-bootable (latest Ubuntu did that for me:
Sure, there are similar and sometimes worse problems with other OS's, it is just that Linux still has too many of those.
Unfortunatley we'll be modded to hell ...
I'm more worried about NFS than anything else.
There seems to be very little pNFS enthusiasm outside Sun ... I hope I am wrong.
Education is the only way to resolve this, really.
No. Sandboxing is another way (and therefore much, much better).
You got me! You are good!
"Linux has better fonts", "TeX" and "OEM must make own distro" ... wow, I had no clue you were pulling my leg before those.
Thanks, my week had been a bit down, I needed a good laugh.
You did not try to print Chinese, did you.
This far I've been told "in extremely cheap machine Linux would make more sense" and "buy more expensive peripherals". Wow.
OEM's cannot compile for the distributions as every minor-minor (security patch) version *requires* different binary.
The binary on the CD which comes along the peripheral would certainly not work on any up-to-date distribution.
Good? No. Acceptable? IMNSHO no. Preferable? Kernel developers think so.
Ten points and a parrot stamp, your answer was 100% correct!
Now you will, maybe, see the problem with Linux.
You did not address my issue, quite contrary. The issue is that there is no "guaranteed to work with Linux HW". The issue is that practically nothing works perfectly out-of-the-box - unless you search and pay more and are lucky. Printers are just nice example.
BTW, have you ever used non-Postscript printer with Linux? I recommend you try to print e.g. Arabic or Chinese. If this is what you mean by "works" ...
Don't you get it? It is far too expensive to compile a driver for every distribution & minor version there is. This is the reason why e.g. DVB makers do not give Linux support. And if Asus is not OEM, then who is?
Utter bull.
The mathematics do not address situations like results of temporary calculations being stored in hard disk and never overwritten.
Your advice is hilarious, I can imagine what Asus would think about that ... "sure average joe can use Linux - the new drivers do not need to be compiled on the Netbook, they can use distcc".