The reason you can replace something like libc on a running Unix system, is the result of the way the Unix FS works. If you open a file, in this case libc, the kernel sets a reference count on that inode. If you subsequently unlink() (delete) that file, the kernel doesn't actually remove it until the reference count goes to 0. This means already running processes will be unaffected by this change, while new opens would fail.
In the case of a libc upgrade, one unlinks the old file, and replaces it with the new one. New apps start and link against the new libcxx.so. Old apps work as expected.
Windows doesn't work this way, at least not what i've seen -- Mu
My first question is "Do common 32bit binaries Just Work?". Anyone wishing to run 32 bit binary software would best stay clear of Debians port, as it has no real support for "multi-arch". This translates into ugly chroot hacks to get Realplayer, flash, and other binary-only apps working.
I ended up dumping it for Gentoo, which has decent emulation support. Gentoo problem is that many pacakges are marked as unstable for the ADM64, or worse have no support at all. (Not to mention the fact that I think building every package from source is a bit first-year)
Ubuntu has support that seems somewhat inbetween, it's missing libraries needed to run some gtk2 apps, I didn't try and debug it too much, since I was using the live-dvd (which is a great idea).
Often times a bounced message is useuless in dealing with spam, since they often do not have valid return paths. Or worse, they return to sources not the orginator, but a hijacted address.
If you've ever gotten a virus warning for a message you didn't send, you'll know what I mean.
You need to stop them at the IP/SMTP level if you really want to make sure they get the point. It's also a lot more satisfiying to think of a poor spambot getting a reject code.
Canida has a free military, paid for by me. This is what I get instead of healthcare...
You can argue the merrits of this until you pass out... Just realize that nations like Canida and Japan save a lot of money by effectivly realying on our Navy to insure stability.
I'll stop before I loose my calm:) (no I didn't vote for bush) --
Re:And don't call "global warming" accurate scienc
on
A New Ice Age?
·
· Score: 1
Wrong. While it is not possible to run experiments as such, it certainly is possible to make certain predictions based on the underlying physics and look how the predictions turn out based on empirical data. Then the theory is either validated or not - in which case you modify the theory trying to account for the difference. Or, in briefer terms, you apply the basic scientific process.
"Wrong!" Think of your proccess above as unit-testing. It's great because it tests your system in many ways that you expect. Just because you've unit tested though, doesn't mean it will work in the the big picture. There may be many unknowns, misunderstandings, and even outwrite errors (you control all your input data). Now pretend that the system you're writing code for is extreamly large and only partialy documented:)
Bottom line, I'm not adverse to spending some efforts increasing efficency and looking for solutions to percieved problems. I am adverse to taking drastic economically cripelling action when I'm not sure we've a real understanding of the problem.
It's not so much "Harsh" -- as in the blinding light of truth. It's another kind of harsh -- think bad breath.
The important part in this is that you shouldn't decide for other people. You can make many arguments about the better useage of these organs, or sparing the child a (most likely) short painful existance, etc.
Next thing you know, you'll discover you're really arguing for the drowning of baby girls in Mainland China... Or just letting terminal cancer patients die without wasting the money and resources on the out-side-chance that they might recover, or life a few more misriable months.
BTW, I would agree with you in practice (if it were my child) -- Matt
IIRC, the reason for this is that the heads would stick to the platters, back when there was such thing as a "Land Zone". Somehow the spindle knew about this situation and would refuse to start, lest it rip the heads off the seeking arm. Whacking the drive sometimes freed the heads, and the disk would then spin up.
That isn't true. Many "smart" switches are IP aware; and can even do simple filtering bassed on the IP headers.
Which brings me to my next point; there are many hardware devices out there with just-plain-broken IPv6 support. This includes things like IPSE ASICs too.
One point, I'm not sure I agree with it, but It's worth making.
You know that roads like 95 (and most of 128) are funded by the feds, and hence are in part paid for by people who may never even drive on them? This is even worse because they may not even live in the same state, use any services that require that road, or even know anyone who does.
At least this is your state capital, the part of the state that draws the rest together. I don't know what you do for a living, but unless it's raising cattle, you're probobly aware of the fact that Boston is really the only reason this stae is as sucessful as it is. The 495 and 128 belts wouldn't exist without Boston/Cambridge
There are very few tolls. In fact, as a poster said already, most of the tolls are outside the big-dig area of effect. There arn't even any tolls on 93, were 1 of the two tunnels, and the fancy bridge are.
The airport tunnel which is in many ways a bigger deal, has a fairly expesnive toll, but it's not an often made trip for most people.
The real question is, will they REALLY have the heart to tear down the old 93? Seems like they just moved the problem underground, and that they need more bandwidth:) -- Mu
Also, legit users who have dynamic IPs are often wise enough to take corrective measures sould they encounter a problem. That makes Dynamic IP DBLS rather a rather safe and effective counter-measure. -- Matt
SPAM on my 6 year old email address exceeded 200 messages a day, a few of which regularly made it past Spam-Ass. The moment I changed my MX to use blacklists (both Dynamic IP and known-open relay), SPAM throughput dropped by at least 40%. And as aothes above have pointed out, without tweaking, SPAM-Ass uses RBLs.
I would love for there to be a clean solution to this, but there presently isn't one. I'd rather see a few rejects a minute, than waste CPU and bandwidth tagging a message for the user...
As long as the coast of SPAM is born by the recipient, or recipents ISP, things will continue to get worse.
DJB had a suggestion here: http://cr.yp.to/im2000.html
As a former employee, I think they are doomed too. Linux is only a minor problem though, the real issue is the maturity of the market and thier inability to add value to their current products. Their products will simply become a commodity.
Of course, they way MS is doomed is most companies dream: they can coast along with their current products (and sizable cash and investements) for about 20 years before they see the train at the end of the tunnel.
True enough, it's been said that they are their biggest competitor. How many shops are going to upgrade to WinXP from Win2k, or Office 2k to Office XP?
emerging markets. Sorry to burst your bubble, but it has nothing to do with linux being a "major" competitor...
I have no bubble, I'm pretty much un-affected by whatever MS does these days. From time to time I need to spend effort getting my video card working so I can play games, but other than that, I'm a Unix admin. -- Matt
I'd like to purpose one possible interpretation of this data: They are doomed, and they know it.
Why are they doomed? If a majority of there profits come from the OS and the Office suite, what happens when there is a major competitor, like Linux / Open Office?
The fact that they are trying to hard to break into new markets, indicates to me that they feel the cow has a limited life span.
No matter how much cash a company has, they cannot continue to spend money on failing venturers, share-holders don't like that...
Anyway, spin it how you like. I've felt this way for a while. I don't see them as having a real monopoly, there are alternatives, people just find the MS assurance worth the price, for now. It's not as if they have or infrastructure, real capital that no one else can match (like railroads, or telephone poles). They make software, and they make a lot of money making some of it, if suddnely the world doesn't want it anymore, they have nothing.
-- Matt
Re:Disk buffers & memory subsystem updated??
on
New Linux 2.5 Benchmarks
·
· Score: 3, Informative
Linux does not honor the sticky bit.
man chmod:...and the Linux kernel ignores the sticky bit on files. Other ker-
nels may use the sticky bit on files for system-defined purposes. On
some systems, only the superuser can set the sticky bit on files. -- Matt
The thing I dislike the most about Norton Ghost, is hat it's DOS based. Getting networking working, for SMB image transfer is not always easy...
Cloning PC-Unix boxes (Linux, etc), doesn't really require any special software though... When I need a new node for our EDA cluster, I boot tomsrbt, and run fdisk, and then kick off a script that pulls down an.tar.gz, and takes edits various/etc files to change hostname, IP, etc. Chroot, run lilo, and your done. -- Matt
Remeber this, taxes are rarely really cut once they are instated.
Also, has anyone noticed that the Good People who run the Govment, always threaten to cut popluar programs first, so they can justify thier continued theft.
My number one expense is income tax! It used to be housing, but I moved to a smaller place...
Slashdot Mirror
Perhaps everyone knows this, but...
The reason you can replace something like libc on a running Unix system, is the result of the way the Unix FS works. If you open a file, in this case libc, the kernel sets a reference count on that inode. If you subsequently unlink() (delete) that file, the kernel doesn't actually remove it until the reference count goes to 0. This means already running processes will be unaffected by this change, while new opens would fail.
In the case of a libc upgrade, one unlinks the old file, and replaces it with the new one. New apps start and link against the new libcxx.so. Old apps work as expected.
Windows doesn't work this way, at least not what i've seen
--
Mu
My first question is "Do common 32bit binaries Just Work?". Anyone wishing to run 32 bit binary software would best stay clear of Debians port, as it has no real support for "multi-arch". This translates into ugly chroot hacks to get Realplayer, flash, and other binary-only apps working.
I ended up dumping it for Gentoo, which has decent emulation support. Gentoo problem is that many pacakges are marked as unstable for the ADM64, or worse have no support at all. (Not to mention the fact that I think building every package from source is a bit first-year)
Ubuntu has support that seems somewhat inbetween, it's missing libraries needed to run some gtk2 apps, I didn't try and debug it too much, since I was using the live-dvd (which is a great idea).
Were getting there.
--
Mu
The 4th of July is called "Independence Day" for a reason. I think you've over-estimated the hight of your point of view.
--
Mu
Often times a bounced message is useuless in dealing with spam, since they often do not have valid return paths. Or worse, they return to sources not the orginator, but a hijacted address.
If you've ever gotten a virus warning for a message you didn't send, you'll know what I mean.
You need to stop them at the IP/SMTP level if you really want to make sure they get the point. It's also a lot more satisfiying to think of a poor spambot getting a reject code.
Canida has a free military, paid for by me. This is what I get instead of healthcare...
:)
You can argue the merrits of this until you pass out... Just realize that nations like Canida and Japan save a lot of money by effectivly realying on our Navy to insure stability.
I'll stop before I loose my calm
(no I didn't vote for bush)
--
Wrong. While it is not possible to run experiments as such, it certainly is possible to make certain predictions based on the underlying physics and look how the predictions turn out based on empirical data. Then the theory is either validated or not - in which case you modify the theory trying to account for the difference. Or, in briefer terms, you apply the basic scientific process.
:)
"Wrong!" Think of your proccess above as unit-testing. It's great because it tests your system in many ways that you expect. Just because you've unit tested though, doesn't mean it will work in the the big picture. There may be many unknowns, misunderstandings, and even outwrite errors (you control all your input data). Now pretend that the system you're writing code for is extreamly large and only partialy documented
Bottom line, I'm not adverse to spending some efforts increasing efficency and looking for solutions to percieved problems. I am adverse to taking drastic economically cripelling action when I'm not sure we've a real understanding of the problem.
--
It's not so much "Harsh" -- as in the blinding light of truth. It's another kind of harsh -- think bad breath.
The important part in this is that you shouldn't decide for other people. You can make many arguments about the better useage of these organs, or sparing the child a (most likely) short painful existance, etc.
Next thing you know, you'll discover you're really arguing for the drowning of baby girls in Mainland China... Or just letting terminal cancer patients die without wasting the money and resources on the out-side-chance that they might recover, or life a few more misriable months.
BTW, I would agree with you in practice (if it were my child)
--
Matt
IIRC, the reason for this is that the heads would stick to the platters, back when there was such thing as a "Land Zone". Somehow the spindle knew about this situation and would refuse to start, lest it rip the heads off the seeking arm. Whacking the drive sometimes freed the heads, and the disk would then spin up.
--
Matt
DLT has remarkable resilience to stray magnetic fields, or in the case of my testing, even stray bulk-tape erasers.
They have an excellent archive length, over 30 years.
Here is what you'll want if you're looking to erase them -- that or an MRI machine:
http://www.athana.com/ddequip/v92.htm
Your ISP doesn't want you to run a server; and they arn't going to change thier policy even if they have the address space to do so.
My ISP (RCN) filters ports 80 and 25, for example. Even though I have a real public IP address.
--
Mu
That isn't true.
:)
Many "smart" switches are IP aware; and can even do simple filtering bassed on the IP headers.
Which brings me to my next point; there are many hardware devices out there with just-plain-broken IPv6 support. This includes things like IPSE ASICs too.
The transition will be painful
--
One point, I'm not sure I agree with it, but It's worth making.
You know that roads like 95 (and most of 128) are funded by the feds, and hence are in part paid for by people who may never even drive on them? This is even worse because they may not even live in the same state, use any services that require that road, or even know anyone who does.
At least this is your state capital, the part of the state that draws the rest together. I don't know what you do for a living, but unless it's raising cattle, you're probobly aware of the fact that Boston is really the only reason this stae is as sucessful as it is. The 495 and 128 belts wouldn't exist without Boston/Cambridge
--
Mu
There are very few tolls. In fact, as a poster said already, most of the tolls are outside the big-dig area of effect. There arn't even any tolls on 93, were 1 of the two tunnels, and the fancy bridge are.
:)
The airport tunnel which is in many ways a bigger deal, has a fairly expesnive toll, but it's not an often made trip for most people.
The real question is, will they REALLY have the heart to tear down the old 93? Seems like they just moved the problem underground, and that they need more bandwidth
--
Mu
That's CA not MTA :(
--
Read before you post ^_^
Who is the CA in this case? :(
If you accept self-signed certs, then we are simply back were we started
If not, then we have yet another cash-cow for Verigisn and friends.
Now, perhaps a Free MTA with working revocation would work. This way one could cancle a cert of a known spammer.
--
Matt
Also, legit users who have dynamic IPs are often wise enough to take corrective measures sould they encounter a problem. That makes Dynamic IP DBLS rather a rather safe and effective counter-measure.
--
Matt
To me, your argument sounds like trolling.
SPAM on my 6 year old email address exceeded 200 messages a day, a few of which regularly made it past Spam-Ass. The moment I changed my MX to use blacklists (both Dynamic IP and known-open relay), SPAM throughput dropped by at least 40%. And as aothes above have pointed out, without tweaking, SPAM-Ass uses RBLs.
I would love for there to be a clean solution to this, but there presently isn't one. I'd rather see a few rejects a minute, than waste CPU and bandwidth tagging a message for the user...
As long as the coast of SPAM is born by the recipient, or recipents ISP, things will continue to get worse.
DJB had a suggestion here:
http://cr.yp.to/im2000.html
Just download the install floppies (3 of them), and boot. It will be painless if your hardware is like mine.
I still run Linux on almost all the PC hardware I deal with, FreeBSD has a real nice feel about it, but just doesn't have everything I need.
--
Matt
As a former employee, I think they are doomed too. Linux is only a minor problem though, the real issue is the maturity of the market and thier inability to add value to their current products. Their products will simply become a commodity.
Of course, they way MS is doomed is most companies dream: they can coast along with their current products (and sizable cash and investements) for about 20 years before they see the train at the end of the tunnel.
True enough, it's been said that they are their biggest competitor. How many shops are going to upgrade to WinXP from Win2k, or Office 2k to Office XP?
--
Matt
emerging markets. Sorry to burst your bubble, but it has nothing to do with linux being a "major" competitor...
I have no bubble, I'm pretty much un-affected by whatever MS does these days. From time to time I need to spend effort getting my video card working so I can play games, but other than that, I'm a Unix admin.
--
Matt
I'd like to purpose one possible interpretation of this data: They are doomed, and they know it.
Why are they doomed? If a majority of there profits come from the OS and the Office suite, what happens when there is a major competitor, like Linux / Open Office?
The fact that they are trying to hard to break into new markets, indicates to me that they feel the cow has a limited life span.
No matter how much cash a company has, they cannot continue to spend money on failing venturers, share-holders don't like that...
Anyway, spin it how you like. I've felt this way for a while. I don't see them as having a real monopoly, there are alternatives, people just find the MS assurance worth the price, for now. It's not as if they have or infrastructure, real capital that no one else can match (like railroads, or telephone poles). They make software, and they make a lot of money making some of it, if suddnely the world doesn't want it anymore, they have nothing.
--
Matt
Linux does not honor the sticky bit.
...and the Linux kernel ignores the sticky bit on files. Other ker-
man chmod:
nels may use the sticky bit on files for system-defined purposes. On
some systems, only the superuser can set the sticky bit on files.
--
Matt
When they come into practical use, we will all be saved.
--
Matt
The thing I dislike the most about Norton Ghost, is hat it's DOS based. Getting networking working, for SMB image transfer is not always easy...
.tar.gz, and takes edits various /etc files to change hostname, IP, etc. Chroot, run lilo, and your done.
Cloning PC-Unix boxes (Linux, etc), doesn't really require any special software though... When I need a new node for our EDA cluster, I boot tomsrbt, and run fdisk, and then kick off a script that pulls down an
--
Matt
Remeber this, taxes are rarely really cut once they are instated.
Also, has anyone noticed that the Good People who run the Govment, always threaten to cut popluar programs first, so they can justify thier continued theft.
My number one expense is income tax! It used to be housing, but I moved to a smaller place...
--
Matt