UPDATE:
So I edited this a little:
cat /tmp/prompt
OUTPUT:
ps1() {
IFS=/$IFS ; set -- \
${PWD%"${last="${PWD##/*/}"}"}
IFS=${IFS#?}
printf "${1+%.1s/}" "$@"
printf "$last > "
}
PS1='$(ps1)${PS2c##*[$((PS2c=0))-9]}'
PS2='$((PS2c=PS2c+1)) > '
So now I can:
ENV=/...
I understand what makes your answer tick, and I have a hard time figuring out what you're doing, so your answer is likely to be incomprehensible to someone who hasn't seen that trick before
@Braiam what, you want me to harass the mods again?
I have a large number of identical fanless PCs running debian 6 (ARM). Most of these are connected via comcast and work ok. There are some that are connected to 'WiMax' modems and are having communication issues.
Specifically: if I ssh to one of these and try a command like 'ps -ax' I'll get abo...
UPDATE:
So I edited this a little. I stuck it in /tmp for now but I think I'm going to keep it for myself, too. It's here:
cat /tmp/prompt
PROMPT SCRIPT:
ps1() {
IFS=/$IFS ; set -- \
${PWD%"${last="${PWD##/*/}"}"}
IFS=${IFS#?}
printf "${1+%.1s/}" "$@"
printf "$last >...
And when I finally find a comment that mentions OCSP, its still wrong (like all comments on news sites, it seems). Latest version of Firefox, at least, enables OCSP by default. And has for a while, I believe.
I'm pretty sure the web of trust gives you one of a few problems—you have no trust path to a lot of stuff (you only trust people you're very confident in); you just a get a trust path to an organization you trust, and everyone gets that organization to verify identities (I think they call that a CA); or you cast such a wide web of trust that anyone with some money can get verified as gmail via bribery.
And there is talk of doing things like that—e.g., of warning when a site's certificate is suddenly signed by a different CA
But there are a lot of things we don't really know how to solve. E.g., can NSA get a court order for Verisign (or any CA in the US) to give them a certificate for a web site of their choice? (And similarly for other governments.)
And you can't not accept Verisign certs because, well, you won't have connectivity anymore.
@derobert it's not, but it's the way the system is implemented today.
the thing with PKI is that it solves two completely separate things that people often put together. it solves the need to have encryption, and it solves the need to identify the website you're talking to.
PKI solves the second. it is generally used to solve the first (but not always - self-signed certificates can solve this too). there's an important distinction.
@strugee Encryption is pretty worthless if you don't know who you're sending the encrypted message to.
Now, there is something browser vendors could fix immediately: there is absolutely no reason that an unverified, encrypted connection should be scarier than an unverified, unencrypted connection. But all browsers have huge warnings for the former, but not the later.
That would allow a lot more people to turn on SSL.
And it'd at least make the NSA have to mount active attacks instead of passive.
(or we could deploy dnssec, which solves the #(!(@#* issue)
DNS-based Authentication of Named Entities (DANE) is a proposed protocol to allow X.509 certificates, commonly used for Transport Layer Security (TLS), to be bound to DNS names using Domain Name System Security Extensions (DNSSEC).
It is proposed in RFC 6698 as a way to authenticate TLS client and server entities without a certificate authority (CA).
Background of the problem
TLS/SSL encryption is currently based on certificates issued by certificate authorities (CAs). Within the last few years, a number of CA providers suffered serious security breaches, allowing the issuance of certi...
... That largely makes CAs obsolete, if we actually deployed it.
I actually have a DNSSEC verifier addon installed but it doesn't work on any of my machines because there isn't support in the underlying OS (either because it didn't ship with Windows or because I haven't set it up yet)
I read somewhere that all DNS root zones now supported DNSSEC...
Anyway, fixing (i.e., removing) the self-signed certificate warnings is some pretty low-hanging fruit to some notable security improvements. You could even make some pretty large improvements, if you included a warning when the cert changes (with some way to disable it, e.g., by signing the new cert with the old one).
If you use Firefox you can also opt-in for Shumway. It is available in Nightly build (and by extension), with some quirks proper of something that isn't quite cooked, but would work as full replacement of Flash given enough time.
There are other full alternatives like GNU Gnash but lack of suppo...
@strugee The Debian packages build almost everything as a module, so each build takes a while. And they repeat the full build multiple times to make different variants—e.g., i486, i686, amd64, etc.
There used to be even more...
Obviously, if you customize your own kernel, you can turn a lot of those ms to ns, and also not build the variants you don't need.
Surely the Arch kernels must have most everything built, else your first step to installing Arch would be install a different distro first and build a kernel :-P
I only see 3 on Debian now: 486, 686, and amd64. There used to be a lot more. The kernel has been getting better at not needing different configs for performance to not suck.
@strugee well, i686 performs much better than i486, but won't run everywhere.
Its gotten much better as those variants have either become obsolete (who uses PAE anymore? You use a 64-bit chip if you want more than 3G) or there isn't a huge overhead anymore (smp)
$ head -n21 /etc/kernel/postinst.d/apt-auto-removal
#!/bin/sh
set -e
# Author: Steve Langasek <steve.langasek@canonical.com>
#
# Mark as not-for-autoremoval those kernel packages that are:
# - the currently booted version
# - the kernel version we've been called for
# - the latest kernel version (determined using rules copied from the grub
# package for deciding which kernel to boot)
# - the second-latest kernel version, if the booted kernel version is
# already the latest and this script is called for that same version,
... so apparently its actually only some kernels that aren't auto-removed. Or at least, that's what testing is doing. Possibly wheezy was different.
I am trying to display only the data section of an udp packet using tcpdump.
In some other words, is it any way to filter the header section of the udp package?
The below command
sudo tcpdump -Aq -i lo udp port 1234
returns:
E..".J@.@.U~.........v.....!HELLO
How can I discard the E..".J@...
I know that there are already a lot of questions about this issue, but I didn't find the right answer to my problem.
I have a Debian 7.3 Server with WebMin installed.
I configured a cron job to run every minute as root user, so i run the command crontab -l when i'm logged in as root it shows:
...
@Gilles - thank you. First of all, can you answer my question about using cat? If what you suggested is better I'll try it, but I need to know for sure if it will work. Second, it's cool that you know so much more about Linux than I do, but I just asked a question as to why the installer chose a different partition. I did say what I did using the installer if you read the question. I asked out of curiosity and I was afraid that /dev/sda2 on the original disk and /dev/sda3 on the target disk might cause problems. A simple "that won't cause problems" would have been enough. :) — Tensigh14 mins ago
not sure if it should be reopened. Y'all who've used the CentOS installer, do you understand the question?
I like destruction. Not in the smash-it-up kinda way, though that can have its finer points, but more as the reverse of construction. I like things broken into pieces - not just broken - because I like to know what goes on inside a thing that makes it tick in the first place. If I'm lucky - and I...
You should try to out-do him. Tell us all about your childhood being bought up by wolves.
I've been having a strange issue recently, where I would delete files from a location (i.e. Desktop), and then empty the Trash. However, a day or so later (I'm not sure exactly what triggers it), the files reappear with new permissions (644).
How do I either figure out what's causing it, or stop...
in an expression, which represents better my wording: if (crap = true) then close; if (crap = true and closed = true) then delete or if (crap = true) then close; if (closed = true) then delete
That sounds like a filesystem badly malfunctioning. But actually, I wonder if this person is running from a CD or something, and an overlay filesystem is malfunctioning...
I'm not familiar with Mint's setup, so I'm going to avoid being drug into that comment thread, but grep ~ /proc/mounts would be interesting. Or grep /home /proc/mounts. Or df -T ~.
Also interesting would be to confirm using ls etc. that the deletes are actually happening.
Swearing is actually pretty impersonal. Everyone (English speakers, anyway) uses the same words. It would be a lot more interesting if people swore in exotic foreign languages, or better, invented their own swear words.
No direct answer, sorry, but it looks like http://stackoverflow.com/questions/8475694/how-to-specify-in-crontab-by-what-user-to-run-script might be useful to you.
who upvoted this??
and it's not even the first posts queue reviewer, though that guy needs a telling off too
I am trying to display only the data section of an udp packet using tcpdump.
In some other words, is it any way to filter the header section of the udp package?
The below command
sudo tcpdump -Aq -i lo udp port 1234
returns:
E..".J@.@.U~.........v.....!HELLO
How can I discard the E..".J@...
I know that there are already a lot of questions about this issue, but I didn't find the right answer to my problem.
I have a Debian 7.3 Server with WebMin installed.
I configured a cron job to run every minute as root user, so i run the command crontab -l when i'm logged in as root it shows:
...
No direct answer, sorry, but it looks like http://stackoverflow.com/questions/8475694/how-to-specify-in-crontab-by-what-user-to-run-script might be useful to you.
