1) Quickly and silently removing the file, while leaving access hidden inside a link in the/proc filesystem
The Unix Way
2) Failing to remove the file (because you're using it right now) and informing you
The Windows Way, also known as "please reboot for the changes to take effect"
The OS really really should _not_ try to second-guess whats wrong or right, just let the user do it. The running application may still enforce certain rules if it has to.
That's far from true. Linux locks the executable file, i.e. if you attempt to open it for writing, you get an error. You can, however, remove the directory entry, in which case the file is retained as long as the program is still running.
If windows had this we would see a lot less need for reboots after patching the system, and updates would be more robust, too.
But whats really funny is that while this method for updating open and/or locked files has been around a long time, it has only recently found its way into the main memory, where it's called RCU (read copy update) and used for multiprocessor locking. I think there may even be a patent covering it.
It's an perfect example for a trivial patent - it is easily avoided. The problem of clutter in the taskbar has been eliminated by tabs.
I used to have lots of terminals and lots of browser windows open, but now I use tabbed browsing and a terminal program like konsole or multi-gnome-terminal that supports tabs, too.
This is exactly why software patents are bad - most are easily avoided, but still you have to put a lot of work in it to avoid them all, because being trivial, they are often re-invented by programmers confronted with a similar problem.
As recent events have shown trying to secure TCP is not quite so easy, and using TLS/SSL inherits some of TCPs problems. It may also not be the best choice to do it on that layer because every application has to implement (via OpenSSL or else) the whole thing and there are a lot of things to be done wrong (have you tried to use OpenSSL? Its not that easy and there are pitfalls).
The best way to deal with the whole situation is to use IPSec. The operating system deals with this for you, and as user _or_ developer you dont't have to deal with it, at least not much.
Of course IPSec needs a working public key infrastructure (PKI), but there is no way around that anyway - at least not on the long run.
I'm also quite sure a working PKI involves Registrars and the DNS. The Registrars are the one instance that can identify an owner of a domain without doubt, because they registered it for them in the first place. And once the DNS has been fixed (this, too, involves a PKI) there is a good way of fetching the public key for a domain.
The fact alone that TLS is broken in respect to name-based virtual hosts is a good sign that it has to go away. IPv6 is still nowhere near our doorstep and wasting precious IPv4-space for virtual domains because of TLS is plain stupid.
If you get marginal speedup you have never used "real" server hardware. The Bios and other hardware init stuff on our Dell Servers takes (compared to the speed of the machines) ages.
So there is still hope that individual states will have less stringent laws than the directive calls for.
A big chunk of the directive (e.g. the raiding part) is going to be against the constitution in a lot of EU countries.
But the fact that it is in there is worrysome to say the least.
Someone had an agenda and that someone has to do with vivendi.
It's quite simple see. The GPL cannot be tested in court, because if it is found invalid, the defendant can't continue to distribute the copyrighted work. If it is found valid, the defendant either has to abide by it's restrictions or can't continue to distribute the copyrighted work. Do you understand the pattern here ? Anyone who tries to fight it will only lose. That's the beauty of it.
This proves one thing. RMS and his fellows at FSF are true hackers indeed. They even hacked the legal system to make it work for them. The GPL is so clever it brings tears to my eyes:)
cheers,
Nils
So everyone, remember to keep your CVS respository and logs so no one can claim ownership on your code! Well they still can, but you will have some solid evidence.
Now if there only was a way to timestamp a file in a tamperproof manner, something based on cryptography...
If you don't agree, you will when you can't afford to run your air conditioner because oil and coal are scarce and the global temperature is up by 10 degrees...
I was neither agreeing nor disagreeing. Matter of fact i am neither sure nuclear energy is bad, nor am i sure it is good. But i am sure we _will_ see a return of nuclear power in the not-too distant future.
The reason for my post was - i was wondering why you left wast disposal out of your list, and i like the response i got:)
Btw, i dont need air condition because the temperatures in germany are quite moderate:)
Nuclear power is a subject that is near and dear to my heart having spent a part of my life working in the industry for Uncle Sam. There are three real issues with Nuclear power that keep it a hot button issue:
Didn't you forget the forth issue, waste disposal?
> In this case, we have IBM, a DC 10 check. We add a +10 Filing BS patent modifier, and we realize that we'll have to roll a natural 20 to make this check.
> I allready paid my license to SCO, will they be returning my $699?
If you already paid, then you deserve to be ripped off (think auto-LART).
It is all about stupidness. Stupid people click on links in spam, stupid people feed the trolls, stupid people give SCO money, stupid stockmarkets give SCO Execs money, stupid journalists report every fart venting out from a SCO office etc. etc.
If it werent for the stupid ones, all of the above would just go away.
Vtun is still far from being useless. Just turn off vtun encryption and use it via a ssh tunnel. That works very well (i use it for securing wifi) and uses a proven protocol.
I also believe this is good practice and should be a widely accepted policy - re-use of good and proven software is not lame - it is crucial for easy, fun and secure software development. There really is no need for re-inventing the wheel.
Now if only ssl were so integrated into the operating system that i could use select() on a ssl-socket created with socket(), and thus making writing of ssl-enabled apps as easy as non-ssl-enabled ones, that would be great!
You might want to try NuppelVideo. It produces rather big files but works like a charm. I use cheap bttv-based cards too and with NuppelVideo i have no frame-drops and video and audio are in perfect sync.
The.nuv-Files as output can then be converted to whatever you want using Mplayer.
The scheduler in 2.6.xx is hyperthreading-aware. It knows that switching a process from one hyperthread to another on the same cpu is less expensive than switching to another physical cpu (becaus both first- and second-level cache reside on-die), but it also tries to balance load on physical cpus.
While >=2.4.19 supported hypterthreading up to a certain point it happend that two processes were running on the same cpu while the other (physical) cpu was running idle. This does not happen with the new ht-aware scheduler.
Look here for a (compressed) version of the initial discussion.
This is surely the case, but is it really "precious" if nobody cared enough to make sure it was archieved properly?
You can look at this from the other side - data gets filtered. The good and really important stuff (e.g. Shakespeare) will be saved because everyone knows about it and cares, the mediocre and bad stuff gets sorted out if nobody cares about it. That is a kind of democratic process, compareable to modding slashdot post up or down based on content.
Of course there will always be some stuff that was worth saving that gets lost, but processes of these magnitude tend to be "messy". It is kind of self-regulatory, whatever is important enough will be saved because it is, and nearly everything important lost will be thought of by somebody else if you give it enough time.
There is only so much information that can effectively be used by humankind, so there needs to be a process that sorts out the stuff that should be save. And as humankinds capability to manage and use information increase, so will its capability of storing it long term.
But of course it is still true that you need to put work and resources into the process so that it works efficient enough for science and art and literature to "improve" at a reasonable rate.
And here is on for the microsoft zealots to think about: the data that will be hardest to keep for a long time is the one stored in undocumented, proprietary closed-source formats like.doc etc. Then again, who cares about _your_ documents;-P
Granted, single-sign-on is convenient, but you can achieve nearly the same convenience using mozilla (and probably any other browser).
Just create a random password for every service you use, log in once and let mozilla store the username/password pair in the password manager and make sure access to the password manager is password protected. That way you only have to remember one password, but you still have different passwords for different sites which are reasonable secure (for they are random generated).
Now if Mozilla supported tracking the age of passwords and telling you to generate a new password for a site once the password reaches a certain age, that would be great!
Slashdot Mirror
ok mr. smartypants, if you want to be pedantic do it right: GMT has been replaced by UTC
The Unix Way
The Windows Way, also known as "please reboot for the changes to take effect"
The OS really really should _not_ try to second-guess whats wrong or right, just let the user do it. The running application may still enforce certain rules if it has to.
If windows had this we would see a lot less need for reboots after patching the system, and updates would be more robust, too.
But whats really funny is that while this method for updating open and/or locked files has been around a long time, it has only recently found its way into the main memory, where it's called RCU (read copy update) and used for multiprocessor locking. I think there may even be a patent covering it.
It's an perfect example for a trivial patent - it is easily avoided. The problem of clutter in the taskbar has been eliminated by tabs.
I used to have lots of terminals and lots of browser windows open, but now I use tabbed browsing and a terminal program like konsole or multi-gnome-terminal that supports tabs, too.
This is exactly why software patents are bad - most are easily avoided, but still you have to put a lot of work in it to avoid them all, because being trivial, they are often re-invented by programmers confronted with a similar problem.
As recent events have shown trying to secure TCP is not quite so easy, and using TLS/SSL inherits some of TCPs problems. It may also not be the best choice to do it on that layer because every application has to implement (via OpenSSL or else) the whole thing and there are a lot of things to be done wrong (have you tried to use OpenSSL? Its not that easy and there are pitfalls).
The best way to deal with the whole situation is to use IPSec. The operating system deals with this for you, and as user _or_ developer you dont't have to deal with it, at least not much.
Of course IPSec needs a working public key infrastructure (PKI), but there is no way around that anyway - at least not on the long run.
I'm also quite sure a working PKI involves Registrars and the DNS. The Registrars are the one instance that can identify an owner of a domain without doubt, because they registered it for them in the first place. And once the DNS has been fixed (this, too, involves a PKI) there is a good way of fetching the public key for a domain.
The fact alone that TLS is broken in respect to name-based virtual hosts is a good sign that it has to go away. IPv6 is still nowhere near our doorstep and wasting precious IPv4-space for virtual domains because of TLS is plain stupid.
If you get marginal speedup you have never used "real" server hardware.
The Bios and other hardware init stuff on our Dell Servers takes (compared to the speed of the machines) ages.
But the fact that it is in there is worrysome to say the least.
Someone had an agenda and that someone has to do with vivendi.
No, stupid, of course it is going to crush tokio once it has grown to its full size.
cheers, Nils
So everyone, remember to keep your CVS respository and logs so no one can claim ownership on your code!
Well they still can, but you will have some solid evidence.
Now if there only was a way to timestamp a file in a tamperproof manner, something based on cryptography...
I wanted to mention it because it is easier to remember.
The reason for my post was - i was wondering why you left wast disposal out of your list, and i like the response i got
Btw, i dont need air condition because the temperatures in germany are quite moderate
This is such a good thing, as somehow nobody has ever been able to invent a door that opens on its own. But now Johnny 5 comes to the rescue!
> In this case, we have IBM, a DC 10 check. We add a +10 Filing BS patent modifier, and we realize that we'll have to roll a natural 20 to make this check.
:)
Just do a "take 20" if you have the time
> I allready paid my license to SCO, will they be returning my $699?
If you already paid, then you deserve to be ripped off (think auto-LART).
It is all about stupidness. Stupid people click on links in spam, stupid people feed the trolls, stupid people give SCO money, stupid stockmarkets give SCO Execs money, stupid journalists report every fart venting out from a SCO office etc. etc.
If it werent for the stupid ones, all of the above would just go away.
Vtun is still far from being useless.
Just turn off vtun encryption and use it via a ssh tunnel. That works very well (i use it for securing wifi) and uses a proven protocol.
I also believe this is good practice and should be a widely accepted policy - re-use of good and proven software is not lame - it is crucial for easy, fun and secure software development. There really is no need for re-inventing the wheel.
Now if only ssl were so integrated into the operating system that i could use select() on a ssl-socket created with socket(), and thus making writing of ssl-enabled apps as easy as non-ssl-enabled ones, that would be great!
Dear Slashdot Editor,
i know these computer thingies are complicated and, at times, confusing.
However, you might want to notice the distinction between the operating system (OS), which is called GNU/Linux, and the kernel called "Linux".
With that in mind, re-read your headline please.
kindest regards,
Nils Juergens
P.S: How about "How to upgrade your operating system to linux 2.6"?
> Maybe thay could take a few dollars that they were
> going to spend on new spacesuits and spend it on
> saving the enviroment we have.
Five minutes after the environment is saved,
a big asteroid hits earth, leaving nothing but
ruins for aliens to discover.
You might want to try NuppelVideo. It produces rather big files but works like a charm. I use cheap bttv-based cards too and with NuppelVideo i have no frame-drops and video and audio are in perfect sync.
.nuv-Files as output can then be converted to whatever you want using Mplayer .
The
Replying to point (2):
The scheduler in 2.6.xx is hyperthreading-aware.
It knows that switching a process from one hyperthread to another on the same cpu is less expensive than switching to another physical cpu (becaus both first- and second-level cache reside on-die), but it also tries to balance load on physical cpus.
While >=2.4.19 supported hypterthreading up to a certain point it happend that two processes were running on the same cpu while the other (physical) cpu was running idle. This does not happen with the new ht-aware scheduler.
Look here for a (compressed) version of the initial discussion.
Funny, thats exactly what i thought on opening the article.
But i think he does this so nobody would notice him snooping around on a linux conference.
This is surely the case, but is it really "precious" if nobody cared enough to make sure it was archieved properly?
.doc etc. ;-P
You can look at this from the other side - data gets filtered. The good and really important stuff (e.g. Shakespeare) will be saved because everyone knows about it and cares, the mediocre and bad stuff gets sorted out if nobody cares about it. That is a kind of democratic process, compareable to modding slashdot post up or down based on content.
Of course there will always be some stuff that was worth saving that gets lost, but processes of these magnitude tend to be "messy". It is kind of self-regulatory, whatever is important enough will be saved because it is, and nearly everything important lost will be thought of by somebody else if you give it enough time.
There is only so much information that can effectively be used by humankind, so there needs to be a process that sorts out the stuff that should be save. And as humankinds capability to manage and use information increase, so will its capability of storing it long term.
But of course it is still true that you need to put work and resources into the process so that it works efficient enough for science and art and literature to "improve" at a reasonable rate.
And here is on for the microsoft zealots to think about: the data that will be hardest to keep for a long time is the one stored in undocumented, proprietary closed-source formats like
Then again, who cares about _your_ documents
thanks for reading this,
Nils
Why, of course! But with the os i use, mozilla stores its stuff in my home dir, which i have to backup now and then anyway.
:)
And i also use el-cheapo software raid5 with ide drives, so it does not happen too often i actually need my backup.
Backup of important files is always implied, but of course you cant mention it often enough
Granted, single-sign-on is convenient, but you can achieve nearly the same convenience using mozilla (and probably any other browser).
Just create a random password for every service you use, log in once and let mozilla store the username/password pair in the password manager and make sure access to the password manager is password protected. That way you only have to remember one password, but you still have different passwords for different sites which are reasonable secure (for they are random generated).
Now if Mozilla supported tracking the age of passwords and telling you to generate a new password for a site once the password reaches a certain age, that would be great!