If you have some public SSH keys, you should really upload them to Launchpad.
This is one of the lesser-known but extremely useful features of
Launchpad. When I'm adding accounts for people to one of my many,
various machines, I always have to figure out a secure channel to
upload/download keys or deal with the temporary exposure of a plaintext
password. Since Launchpad takes great care to encrypt both upload and
download of sensitive data (including keys) it makes an ideal place to
exchange them.
So, if you think you are the sort of person who I might need to give an
account to for some reason one day, save us both some time and go upload
your keys right now.
One of the first inklings the world had of the upcoming release of the game
"Halo" was a cryptic website showing various obscure quotations which
appeared to be from some ancient religious text. Pre-acquisition, I
always loved their sense of style; I am a huge fan of the Marathon
series. My favorite quote from the collection, which stuck with me for
years, was:
Microsoft's real problem is the much simpler business problem that their flagship product is clearly inferior its lower cost alternatives. It doesn't help them much that the "lower" cost is, in many cases, zero, but that's beside the point. I have made, and will make again, arguments about freedom and scaling cost and a number of other social issues surrounding code in general and operating systems especially, but all that is going to end up being peripheral. I'm pretty sure that I'd continue to use the Ubuntu-derived operating systems for my own computers even if they somehow cost more than the Microsoft equivalents.
Of course, we've all heard this kind of thing before from the archetypical linux snobs on slashdot and other forums, people who seem incapable of recognizing that any value can be created by people who disagree with them. Personally, I have purchased a Windows license - not just a computer which happened to have it included - as recently as last month. (Although I must admit that my sense of practical compromise goes only so far.)
I can certainly understand some skepticism that Microsoft will topple and Linux will be ubiquitous in the home and office any time soon. Certainly that is not going to happen soon (for sufficiently software-industry-ish definitions of "soon" — which means they could be out of business but for cash in the bank within a year). With Windows marketshare hovering somewhere between 80 and 95 percent, depending on who you believe, Canonical has a long way to go to fix bug 1.
On the other hand, one could make the case that it's happened already. In the office? Dozens of large companies, Red Hat foremost among them, make a comfortable living on corporate Linux deployments. Sure, in most cases it's not on anyone's desktop, but it is making inroads.
In the home? There are already 1 million PS3s in the US, and the current generation console market is staying rather cool - if it heats up to the levels of the previous one, the numbers could be more like 50 million. Even when it's running games, the PS3 is a linux box, and to some people, it is in a much more literal sense.
Mark Shuttleworth himself says that there are about 8 million Ubuntu users alone. That's not just a measurement error.
As with the Twisted site, however, the interesting thing here is not the number, but the trend. Of course pundits and the mainstream press are going to look at the numbers they can see and not even consider the possibility that Windows could be dislodged. By the time the idea seems reasonable and possible, though, the adoption curves will be in the process of going exponential as the new network effects combine with the advantages of the platform that already, now, have people migrating away from the current network effects.
If you're watching the arc of an arrow in flight it's quite fast. Hard to predict, and often effectively invisible thanks to its motion. By the time it's stopped moving and is easy for everyone to see, it's already arrived.
Our conviction is like an arrow already in flight.Today, I recalled this quotation as I noted a few things going on in my daily trip around the tubes:
Your life will only last until it reaches you.
-
Dell will be offering an option to preinstall Ubuntu on its PCs.
"Linux-based PCs was an "overwhelming" request" ... "We heard loud and clear from customers that they wanted this," Cook said. And of those who wanted Linux, "80 percent came back and said Ubuntu," Cook said.
-
While Microsoft is putting a substantial marketing push behind
Silverlight, even mainstream press are asking: what about
Linux? Of course, developer after developer are immediately
asking the same thing.
"Linux lets us say to the customer, 'This thing is ubiquitous,'" ... "Otherwise, we have to build a version of the Web site for Linux, which is expensive."
-
The
Twisted website's traffic from Linux has broken 20%. (non-Windows
traffic, including MacOS, is at 36%). The interesting thing about
this is not the number - I realize that we are on the cutting edge of
technology and so on - but the trend. People are migrating away
from Windows. Nobody is going back.
Microsoft's real problem is the much simpler business problem that their flagship product is clearly inferior its lower cost alternatives. It doesn't help them much that the "lower" cost is, in many cases, zero, but that's beside the point. I have made, and will make again, arguments about freedom and scaling cost and a number of other social issues surrounding code in general and operating systems especially, but all that is going to end up being peripheral. I'm pretty sure that I'd continue to use the Ubuntu-derived operating systems for my own computers even if they somehow cost more than the Microsoft equivalents.
Of course, we've all heard this kind of thing before from the archetypical linux snobs on slashdot and other forums, people who seem incapable of recognizing that any value can be created by people who disagree with them. Personally, I have purchased a Windows license - not just a computer which happened to have it included - as recently as last month. (Although I must admit that my sense of practical compromise goes only so far.)
I can certainly understand some skepticism that Microsoft will topple and Linux will be ubiquitous in the home and office any time soon. Certainly that is not going to happen soon (for sufficiently software-industry-ish definitions of "soon" — which means they could be out of business but for cash in the bank within a year). With Windows marketshare hovering somewhere between 80 and 95 percent, depending on who you believe, Canonical has a long way to go to fix bug 1.
On the other hand, one could make the case that it's happened already. In the office? Dozens of large companies, Red Hat foremost among them, make a comfortable living on corporate Linux deployments. Sure, in most cases it's not on anyone's desktop, but it is making inroads.
In the home? There are already 1 million PS3s in the US, and the current generation console market is staying rather cool - if it heats up to the levels of the previous one, the numbers could be more like 50 million. Even when it's running games, the PS3 is a linux box, and to some people, it is in a much more literal sense.
Mark Shuttleworth himself says that there are about 8 million Ubuntu users alone. That's not just a measurement error.
As with the Twisted site, however, the interesting thing here is not the number, but the trend. Of course pundits and the mainstream press are going to look at the numbers they can see and not even consider the possibility that Windows could be dislodged. By the time the idea seems reasonable and possible, though, the adoption curves will be in the process of going exponential as the new network effects combine with the advantages of the platform that already, now, have people migrating away from the current network effects.
If you're watching the arc of an arrow in flight it's quite fast. Hard to predict, and often effectively invisible thanks to its motion. By the time it's stopped moving and is easy for everyone to see, it's already arrived.
This week, some people are saying some nice things
about Twisted.
"A big component to that has been our use of Twisted Python. We're pretty reliant on the Twisted framework, and we use it for our base-line management software that we use to run the great majority of production services that we have, our monitoring infrastructure and the next-generation thing that we have coming, which is a suite of programs that will automate the upgrade process for us," Kelley said.
Python Slithers into Systems, eWeek.com
"Twisted is one of our favorite pieces of technology. We’ve used it in Sleevenotez, and we had a major project last year, that was unfortunately canned for budget reasons after we’d done about 2 months modelling and spec work, that we designed in Twisted. I’ve been using it for years and I really think it’s one of the finest software systems I’ve ever come across."
Twisted "ready for the big time", isotoma blog
class A(object):
def m(self, alpha):
print 'A.m', alpha
class B(A):
def m(self, alpha, beta):
print 'B.m', alpha, beta
super(B, self).m(alpha)
class C(B):
def m(self, alpha, beta, gamma):
print 'C.m', alpha, beta,
gamma
super(C, self).m(alpha, beta)
class D(B):
def m(self, alpha, beta, gamma, delta):
print 'D.m', alpha, beta, gamma,
delta
super(D, self).m(alpha, beta)
class E(C, D):
def m(self, alpha, beta, gamma, delta, epsilon):
print 'E.m', alpha, beta, gamma,
delta, epsilon
x = ???
super(E, self).m(alpha, *x)
E().m(1, 2, 3, 4, 5)
The UNIX security model sucks.
Unfortunately, we are likely to be stuck with it for the next hundred quintillion years, it will outlive the sun and possibly humanity as we flee to other stars and trade our technology to species across the galaxy. I understand this fact. I can live with it.
Still, we must be able to do better than the current tools, like sudo. I have had a variety of Twisted-based ideas for this kicking around in the back of my head for a while.
Imagine a Twisted daemon that ran at boot,
Such a daemon could be used for more than just 'sudo'. Most of the tasks currently reserved for 'init', such as run-parts, could be run as "nobody" instead, with
Since security rules could be implemented in Python, it would be easy to have flexible policy declarations, like, "/usr/bin/foobar can always run /usr/sbin/bazqux processes as the 'foobar' user when run by people in the 'xyz' group". This avoids giving unrestricted system access to either members of the 'xyz' group, or anyone who can exploit the 'foobar' executable. Ideally programs could be distributed with their own security rules rather than, as sudo does, making separating privileges the administrator's responsibility.
Of course I have no time to implement this, nor to advocate it to the dozens of very high-profile projects which would need to adopt it in order for it to be useful. I wish that I could, though, every time sudo lets me run two commands as root in a row because it would be too inconvenient to type my password a second time.
Unfortunately, we are likely to be stuck with it for the next hundred quintillion years, it will outlive the sun and possibly humanity as we flee to other stars and trade our technology to species across the galaxy. I understand this fact. I can live with it.
Still, we must be able to do better than the current tools, like sudo. I have had a variety of Twisted-based ideas for this kicking around in the back of my head for a while.
Imagine a Twisted daemon that ran at boot,
seteuid and
setegid to "nobody", but retaining root privileges.
spawnProcess already supports switching UIDs for your
subprocess. Instead of running subprocesses directly, you could run a
Twisted client program which would connect to the root daemon and ask it to
do something for you.Such a daemon could be used for more than just 'sudo'. Most of the tasks currently reserved for 'init', such as run-parts, could be run as "nobody" instead, with
start-stop-daemon asking to run specific
commands as root. You could eliminate just about every "suid" binary
by having all the binaries themselves be non-SUID, but distributed with
security rules that allow their execution in specific restricted
contexts.Since security rules could be implemented in Python, it would be easy to have flexible policy declarations, like, "/usr/bin/foobar can always run /usr/sbin/bazqux processes as the 'foobar' user when run by people in the 'xyz' group". This avoids giving unrestricted system access to either members of the 'xyz' group, or anyone who can exploit the 'foobar' executable. Ideally programs could be distributed with their own security rules rather than, as sudo does, making separating privileges the administrator's responsibility.
Of course I have no time to implement this, nor to advocate it to the dozens of very high-profile projects which would need to adopt it in order for it to be useful. I wish that I could, though, every time sudo lets me run two commands as root in a row because it would be too inconvenient to type my password a second time.