Our (Progeny's) implementation of APT is reasonably secure.
We have SSL support, derived from a
patch by Tomas Pospisek.
We have added authentication and authorization.
Debian's APT currently supports the signing of repositories.
Regarding package signatures: that's more relevant to the package
installation tools than to APT itself, but long ago, Progeny wrote
and contributed to Debian debsigs, a tool for
allowing GPG signatures to be embedded in packages.
I've been using Emacs for years, but I tried Eclipse for a while. It was too buggy (or unpredictable, which to me was buggy) for my tastes, but there's one advantage to a Java IDE that significantly decreases development time over a tool like Emacs: continuous compilation.
Eclipse will show you whenever you've left out an import, or mistyped a variable name, or pick your favorite goof. Using Emacs, I would type for a while, compile, get errors, track them down, compile again, etc. Seeing errors on the fly and fixing them, without having to wait for a Java compiler to start and execute, is addictive and tremendously more efficient.
Yes, yes, yes. I bought the DVD for this movie years before I had a DVD player.
Too many funny scenes to pick a favorite, but when the giant and Dean meet, the giant smiles and says "Deeeaaannnn"...that just kicks my ass every time.
The phone that wouldn't hang up; Dean in the middle of the road; "That hurt"; "Check please".
PGI does support ia64 as well as i386, and
developers outside of Progeny are working on
powerpc. The design is modular, to minimize
the work required to make it functional on
other architectures (although "minimize" should
not imply that it's easy).
We hope to have ia64 CDs available shortly,
but given the relative market shares of the
two platforms, we wanted to make the i386 images
available without waiting for ia64.
Other recent developments at Progeny include the release of
Discover 2.0, a cross-platform extensible hardware identification
library and tool; Progeny Graphical Installer (PGI) 1.0, which
contrary to its name is properly an installer creation system;
and the announcement of Platform Services, a subscription service
that makes it easier for companies to develop and maintain
Linux-powered products and services.
> I'm not unhappy because XFree86 4.2 is not in Debian stable. I'm not even unhappy because it is not in testing. > I'm unhappy because XFree86 4.2, after 6 months, is not even in unstable. That's just bogus.
Well, it has been in experimental for over a month.
Also check out the Debian changelog for X
(/usr/share/doc/xserver-xfree86/changelog.Debian.g z).
It may give you some idea of how much effort it took
to fine-tune 4.1.
On the other hand, why am I wasting time on an
asshole who derides others' work as "bogus"?
(Disclaimer: I work with Branden Robinson, the
maintainer for X in Debian, so I have some
appreciation for the amount of work he puts into
X.)
"No OS, other than an enterprise-level UNIX, has ever cost $300."
From bestbuy.com:
Microsoft Windows XP Professional: $299.99
Microsoft Windows XP Home Edition: $199.99
Every now and again, I contemplate buying
Windows for occasional home use, but
sanity always sets in when I see the up-front
cost, not to mention how much I'd pay for
Office or just about any other major commercial software.
It's not an options menu, but you can
configure the available flags when starting
the server. Check the man page.
On casual inspection, the following appears to
remove phantom zone and guided missiles:
bzfs -f PZ -f GM
Ancient ASCII games - Hunt?
on
Nethack 3.4.0
·
· Score: 1
On the subject of ASCII games, does anyone
still play the great multi-player shoot-'em-up
game, hunt?
Stab 'em or nuke 'em; all it takes is a bit
more ammo.
Unfortunately, even google can't get around the
fact that "hunt" is a rather generic word, so
I've been unable to track down any active
servers. I have found lots of references to
hunt the wumpus.
The Atlantic Monthly has a fantastic article
this month about the Native American world
pre-Columbus.
It's not available online to non-subscribers, but
I would highly recommend picking up the magazine.
It has always been one of the best subscription
deals around; it wasn't all that long ago that
it was $20 for 2 years.
Serious waves of deja vu just came crashing
through my monitor.
I should go collect all those old posts and
assemble them as a collection of essays. Perhaps
they'd be valuable for college courses in critical
thinking.
Been there, done that, lived out of a car because
I was addicted to MUDs - 10 years ago.
Wrecked my education, but taught me how to program in C.
Overall, I'd rather have learned in school and
have gotten a degree back when I could attend
school full-time.
Note, however, that the article was remarkably
perceptive about the potential for business
efficiency improvements, although it sounds
as if there were sufficient existing examples
to make such predictions straightforward.
The Economist contains some of the finest
reporting and analysis anywhere. I've wanted
to subscribe to it for years; with luck, I'll
bite the bullet soon.
Well, we didn't use Outlook, and all SMTP traffic
was filtered through two different platforms with
virus scanning on at least one.
The company also did periodic checks on phone
lines to make certain only authorized modems
would respond.
As someone else has pointed out, the UID question
is a non-starter. It requires root privileges to
bind a service to a port lower than 1024, but
not to run the service past that point.
Outgoing connections were very limited as well,
and as the other response pointed out, firewalls
handle those relatively securely.
Anyway, this is a previous employer, but I still
don't want to get too detailed about their
security setup. Suffice it to say that I don't
think a paranoid firewall policy is unreasonable,
particularly for a large company.
Am I the only one who liked the movie, but hated
the climactic scene? I felt the physical conflict
that finished out the scene was much more dramatic
in the book; the movie version was just wretched.
I've been the guy blocking open port
requests for months on end.
I'm afraid it's a necessary evil. Without the
bureaucracy to prevent quick firewall changes,
it would be more difficult to conduct a security
review before the port has been opened. Of course,
once a port is open and services are running,
turning it off becomes a near impossibility.
You'd be amazed how little some software vendors
seem to understand firewalls and the Internet.
I've seen requests to open up a half-dozen
ports through the firewall, including incoming
ports to provide data that could be retrieved
from the inside, were the software to allow it.
If your firewall is in the right hands, all
those bureaucratic hurdles are a force for
good, not evil. Unfortunately, that's a fairly
large caveat.
I feel cool; I've worked for BBN and Eli Lilly
and Company, so I've been involved with 4
class A networks.
Eli Lilly and Company is a large pharmaceutical
firm who has had an Internet connection since the
late 80s, long before most non-technology companies of similar size.
As I understand the history from someone who
should have known,
Eli Lilly originally applied for a class B
address space back in the late 80s/early 90s,
but Jon Postel himself suggested that they ask for
a class A instead.
Postel later criticized Lilly (among others)
for not returning
the extra addresses.
It should be pointed out that renumbering
40,000+ computers is a non-trivial task,
and handing back portions of the address space
would likely cause other headaches. To be
honest, I'm not certain anyone has actually
formally asked Lilly to turn the space over.
Just as proprietary software will continue
to exist, closed hardware will almost certainly
continue to exist as well.
Why not just apply the appropriate business
model to a company's needs? Open source
hardware won't put anyone out of business,
unless the open products turn out to
be better than the closed ones...and isn't
that what capitalism is all about? May the
best product win?
Slashdot Mirror
Our (Progeny's) implementation of APT is reasonably secure.
Regarding package signatures: that's more relevant to the package installation tools than to APT itself, but long ago, Progeny wrote and contributed to Debian debsigs, a tool for allowing GPG signatures to be embedded in packages.
I've been using Emacs for years, but I tried Eclipse for a while. It was too buggy (or unpredictable, which to me was buggy) for my tastes, but there's one advantage to a Java IDE that significantly decreases development time over a tool like Emacs: continuous compilation.
Eclipse will show you whenever you've left out an import, or mistyped a variable name, or pick your favorite goof. Using Emacs, I would type for a while, compile, get errors, track them down, compile again, etc. Seeing errors on the fly and fixing them, without having to wait for a Java compiler to start and execute, is addictive and tremendously more efficient.
Yes, yes, yes. I bought the DVD for this movie years before I had a DVD player.
Too many funny scenes to pick a favorite, but when the giant and Dean meet, the giant smiles and says "Deeeaaannnn"...that just kicks my ass every time.
The phone that wouldn't hang up; Dean in the middle of the road; "That hurt"; "Check please".
And, of course, "No following" and "Superman".
Must go watch for 20th time.
> It was the armies that came later that wiped out/corraled up the Indians.
Well, technically the Indians had already been wiped out by the explorers. 95% or so...what a staggering number. Stalin was right.
Since Debian typically only needs to be installed once and dist-upgraded thereafter to keep current, why not use the PGI-based installer?
Even if you find the upgrading capabilities inadequate, the next version of Debian will have a better installer than the current one.
PGI does support ia64 as well as i386, and developers outside of Progeny are working on powerpc. The design is modular, to minimize the work required to make it functional on other architectures (although "minimize" should not imply that it's easy).
We hope to have ia64 CDs available shortly, but given the relative market shares of the two platforms, we wanted to make the i386 images available without waiting for ia64.
Other recent developments at Progeny include the release of Discover 2.0, a cross-platform extensible hardware identification library and tool; Progeny Graphical Installer (PGI) 1.0, which contrary to its name is properly an installer creation system; and the announcement of Platform Services, a subscription service that makes it easier for companies to develop and maintain Linux-powered products and services.
Progeny has just released a new ISO for a woody install using PGI. You can find it at http://archive.progeny.com/progeny/pgi/.
We heard your plea.
> I'm not unhappy because XFree86 4.2 is not in Debian stable. I'm not even unhappy because it is not in testing.
> I'm unhappy because XFree86 4.2, after 6 months, is not even in unstable. That's just bogus.
Well, it has been in experimental for over a month.
See http://lists.debian.org/debian-devel/2002/debian-d evel-200204/msg01343.html
for some idea of the challenges of maintaining
X for Debian.
Also check out the Debian changelog for X (/usr/share/doc/xserver-xfree86/changelog.Debian.g z).
It may give you some idea of how much effort it took
to fine-tune 4.1.
On the other hand, why am I wasting time on an asshole who derides others' work as "bogus"?
(Disclaimer: I work with Branden Robinson, the maintainer for X in Debian, so I have some appreciation for the amount of work he puts into X.)
"No OS, other than an enterprise-level UNIX, has ever cost $300."
From bestbuy.com:
Microsoft Windows XP Professional: $299.99
Microsoft Windows XP Home Edition: $199.99
Every now and again, I contemplate buying Windows for occasional home use, but sanity always sets in when I see the up-front cost, not to mention how much I'd pay for Office or just about any other major commercial software.
It's not an options menu, but you can configure the available flags when starting the server. Check the man page.
On casual inspection, the following appears to remove phantom zone and guided missiles:
bzfs -f PZ -f GM
On the subject of ASCII games, does anyone still play the great multi-player shoot-'em-up game, hunt?
Stab 'em or nuke 'em; all it takes is a bit more ammo.
Unfortunately, even google can't get around the fact that "hunt" is a rather generic word, so I've been unable to track down any active servers. I have found lots of references to hunt the wumpus.
FWIW, that's CNN's universal "Asteroid threat" image, so no scaling factors really apply here.
The Atlantic Monthly has a fantastic article this month about the Native American world pre-Columbus.
It's not available online to non-subscribers, but I would highly recommend picking up the magazine. It has always been one of the best subscription deals around; it wasn't all that long ago that it was $20 for 2 years.
Serious waves of deja vu just came crashing through my monitor.
I should go collect all those old posts and assemble them as a collection of essays. Perhaps they'd be valuable for college courses in critical thinking.
Don't forget when it was named (nicknamed?) "gay cancer".
And the band played on...
Been there, done that, lived out of a car because I was addicted to MUDs - 10 years ago.
Wrecked my education, but taught me how to program in C. Overall, I'd rather have learned in school and have gotten a degree back when I could attend school full-time.
Note, however, that the article was remarkably perceptive about the potential for business efficiency improvements, although it sounds as if there were sufficient existing examples to make such predictions straightforward.
The Economist contains some of the finest reporting and analysis anywhere. I've wanted to subscribe to it for years; with luck, I'll bite the bullet soon.
Well, we didn't use Outlook, and all SMTP traffic was filtered through two different platforms with virus scanning on at least one.
The company also did periodic checks on phone lines to make certain only authorized modems would respond.
As someone else has pointed out, the UID question is a non-starter. It requires root privileges to bind a service to a port lower than 1024, but not to run the service past that point.
Outgoing connections were very limited as well, and as the other response pointed out, firewalls handle those relatively securely.
Anyway, this is a previous employer, but I still don't want to get too detailed about their security setup. Suffice it to say that I don't think a paranoid firewall policy is unreasonable, particularly for a large company.
Am I the only one who liked the movie, but hated the climactic scene? I felt the physical conflict that finished out the scene was much more dramatic in the book; the movie version was just wretched.
I've been the guy blocking open port requests for months on end.
I'm afraid it's a necessary evil. Without the bureaucracy to prevent quick firewall changes, it would be more difficult to conduct a security review before the port has been opened. Of course, once a port is open and services are running, turning it off becomes a near impossibility.
You'd be amazed how little some software vendors seem to understand firewalls and the Internet. I've seen requests to open up a half-dozen ports through the firewall, including incoming ports to provide data that could be retrieved from the inside, were the software to allow it.
If your firewall is in the right hands, all those bureaucratic hurdles are a force for good, not evil. Unfortunately, that's a fairly large caveat.
From a purely selfish standpoint, defending Dmitry now makes it easier to prevent the same injustice from happening to you.
Next time China arrests and convicts a visiting scholar for espionage, any U.S. protests could easily be rebutted with a comparison to Sklyarov.
Why free Dmitry? If for no other reason, to regain some semblance of self-respect as a nation.
I feel cool; I've worked for BBN and Eli Lilly and Company, so I've been involved with 4 class A networks.
Eli Lilly and Company is a large pharmaceutical firm who has had an Internet connection since the late 80s, long before most non-technology companies of similar size.
As I understand the history from someone who should have known, Eli Lilly originally applied for a class B address space back in the late 80s/early 90s, but Jon Postel himself suggested that they ask for a class A instead.
Postel later criticized Lilly (among others) for not returning the extra addresses.
It should be pointed out that renumbering 40,000+ computers is a non-trivial task, and handing back portions of the address space would likely cause other headaches. To be honest, I'm not certain anyone has actually formally asked Lilly to turn the space over.
Actually, to play the devil's advocate, you stated "At least those of us who are US citizens".
The phrase "At least" doesn't carry the same weight as "only"; traditionally, "at [the] least" has implied a superset, not a subset.
Just as proprietary software will continue to exist, closed hardware will almost certainly continue to exist as well.
Why not just apply the appropriate business model to a company's needs? Open source hardware won't put anyone out of business, unless the open products turn out to be better than the closed ones...and isn't that what capitalism is all about? May the best product win?