« first day (1348 days earlier)      last day (3607 days later) » 
00:00 - 19:0019:00 - 00:00

12:00 AM
@Kevin Didn't really think that was necessary.. shrug ;)
@terdon Same here! The feedback I got on my first question made me want to stick around.
 
well, I am not satisfied with the current state of the FP, and was fiddling with ideas similar to what Gilles proposed (damn Gilles >:/ )
 
I can see the queue needing some major tweaking on, say, Stack Overflow but removing it everywhere is just a terrible idea.
 
@Braiam Oh yes, an improvement would be great. Just don't remove it!
 
@Braiam Oh improve it by all means, but Shog saying it is useless, bad, terrible, etc is just wrong.
They are, again, looking at the entire network from the eyes of Stack Overflow.
 
That's just SO myopia. Since SO is a few orders of magnitude larger than any other site, the SE team (understandably) focus on it.
 
12:04 AM
@Seth @terdon I'm not satisfied about how users behave in the FP in AU, and you can find proof of it
 
I am just upset they seem to always think "SO needs this so the network must too!"
@Braiam Oh yes, but that isn't the queue's fault!
 
the main problem in that queue is that users misunderstand the use of the queue, and think that the only objective is just to bring the number to 0
 
@Braiam Oh, I don't doubt it. That's why I upvoted Gille's idea (and some similar ones) I just would prefer to have what we have now, over not having anything.
I'd prefer an improved one even more.
 
Exactly.
 
and again Shog is just saying that they might replace the queue with something else, more streamlined @Seth @terdon
is not just removing, but separation of functions according to their expectations
 
12:06 AM
all his ideas are pretty bad, except the third bullet.
 
if the main goal of the FP queue is to provide feedback and this objective is not meet, then I would say that the queue is failing to serve it propose
 
But you see, that is not its main goal. That is one of its main goals.
 
and I'm sure Shog reasoning is similar to that
 
@Braiam I dunno. I like it here, I think it's fine. It also worked great for me as a new user on SU.
As a more experienced user I have often found it a great way of welcoming new users and explaining how the site works to them.
"Separate queues for questions and answers," might make sense though. Newbie questions and newbie answers are two very different beasts.
 
I don't think a whole new queue is a good idea. IMO that would just be too much. Perhaps add an option to filter Qs from As though?
 
12:11 AM
@terdon that's the problem, the queue should function everywhere, you are narrowing your scope to the sites that it might work as intended
@Seth there is....
 
Oh wow, I didn't know that.
 
I really don't know why people do not use these filters if they make the workload more easy to cope with...
 
@Braiam Yes and the main reason it doesn't work everywhere is that it fails on SO as far as I can tell. That, again, is because it has many more users so by extension many more dicks.
 
Anyone has interesting ideas for this question in bash?
0
A: Random script that isn't actually random

RameshBash Script: Initially, I create a new file named people with all the persons name inside it. cat people john jeff emma steve julie Now, I use the below script to select a person name at random. random_character=$(cat ./dev/urandom | tr -cd 'sje' | head -c 10 | tail -c 1) grep "$random_char...

 
@Ramesh I was scared to death when I saw 55 answers and the most upvoted one with only 2 votes
 
12:17 AM
@Braiam, the most upvoted one has 77 votes.
77
A: Random script that isn't actually random

Paul RC It's important to decide who is buying as quickly as possible, so as not to waste precious drinking time - hence C is the obvious choice in order to get maximum performance: #include <stdio.h> #include <stdlib.h> #include <time.h> int main(void) { const char *buyer; int n; srand...

 
another reason to hate php is there...
 
@FaheemMitha here is the entirety of my tex related stuff in .emacs
> ;;; AucTex stuff
(require 'tex)
(TeX-global-PDF-mode t)

(setq TeX-auto-save t)
(setq TeX-parse-self t)
(setq-default TeX-master nil)

(add-hook 'LaTeX-mode-hook 'turn-on-auto-fill)
(add-hook 'LaTeX-mode-hook 'flyspell-mode)
(add-hook 'LaTeX-mode-hook 'LaTeX-math-mode)

(add-hook 'LaTeX-mode-hook 'turn-on-reftex)
(setq reftex-plug-into-AUCTeX t)

(setq TeX-PDF-mode t)
 
slm
12:44 AM
@Gilles That's been on my todo list for a LOOONG time.
 
 
3 hours later…
3:47 AM
Can someone try this please?
echo | tee /dev/fd/0 | sed =
Have CTRL+C ready - if it does for you what it was does for me you'll get an endlessly incrementing output like \n\n<nums>\n
Nevermind.
 
4:07 AM
@casey - what?
 
@mikeserv my brain saw fd0 rather than fd/0
i should go to sleep
 
Ahh. Yeah, either that or your brain should...
Did you try it? I just posted a question about it - does it do the same thing for you?
 
> % echo | tee /dev/fd/0 | sed =
tee: /dev/fd/0: Text file busy
1
 
Weird.
 
with an extra newline after that
 
4:09 AM
That's what I'm used to seeing...
 
but no endless incrementing
 
Does this do anything?
printf 'some string\n' | {
    tee /dev/fd/3 | {
        : && sed 's/some/string/'
    }
} 3>&0
I would think it would do the same - text file busy - but I don't get why it doesn't on my end...
 
> % printf 'some string\n' | {
casey@convect src pipe cursh> tee /dev/fd/3 | {
casey@convect src pipe cursh pipe cursh> : && sed 's/some/string/'
casey@convect src pipe cursh pipe cursh> }
casey@convect src pipe cursh> } 3>&0
tee: /dev/fd/3: Text file busy
string string
ignore the messy zsh promping :)
@mikeserv for the record I'm doing this inside rxvt-unicode
though that shouldn't have an effect
 
@derobert, I have tried to cover the concepts of ext3 file system that we were discussing today.
0
Q: debug mkfs.ext3 command output

RameshI was creating a new file system in my external HDD. While formatting, I had to format this partition to the remaining available partition which is somewhere around 850GB. Now, I created an ext3 file system in this partition. This is the output of my mkfs.ext3 command. mkfs.ext3 /dev/sdb3 mke2f...

 
4:29 AM
@casey Thanks
 
@casey - I don't think think so either - I tried it in the console, too. It has either something to do with my kernel version and the it way it handles pipes or with the version of tee that I have. It's weird.
But I used to see that text file busy error a lot which is what surprised me just now.
 
@mikeserv I'm using coreutils 8.21 and kernel 3.12.13 with nothing too exotic enabled (mostly just things turned off or compiled in vs being a mod)
 
I just installed a fresh arch os the other day. I'll get those numbers too...
 
@FaheemMitha hope it helps
 
uname -r
3.14.6-1-ARCH
pacman -Q coreutils
coreutils 8.22-4
 
4:40 AM
@casey I'll try running with a minimal .emacs today.
 
4:56 AM
@casey at least with regular latex (default) it works. I'll try with pdflatex.
 
 
3 hours later…
8:12 AM
I have been using vim for over two decades and just now discovered :norm.
I'm not sure whether to hang my head and cry in a corner or dance a jig, but I do know :g/pattern/norm @m and I are going to be good friends.
 
 
3 hours later…
11:33 AM
@Caleb you can also just have a macro call itself, and have the beginning of the macro be /pattern
 
@Patrick I'm afraid I don't quite follow.
 
@Caleb well your goal is to run the m macro across all instances of pattern right? When recording the m macro, have the first action of the macro be /pattern, then the last action is @m
 
@Patrick How will it know when to stop recursing?
 
@Caleb - thanks very much.
 
@Caleb when the pattern hits end of document, it'll abort
 
11:37 AM
@mikeserv np
 
@Kevin - do you think it's possible that this is related to that tee thing?
 
just another way of skinning the cat. It just feels more intuitive to me, thats all
 
@Patrick Huh. I see. I still think that's a little brittle to setup an you can't do in a piece at a time, and :norm has other advantages such as operating on visual select ranges.
 
@mikeserv how in the world you got Caleb first ping?
 
Hmmm...
 
11:40 AM
@Braiam On your meta post?
 
oh, I think I know how... Caleb edited the comment
 
@Braiam - I dunno. What did I do? I'm not sure I understand the question.
 
@Caleb yeah
 
@Braiam Yup. I pinged one user, then edited to add the second one and ping the second user and took off the first @. It's a dirty hack but it works.
 
I never got a ping on that one.
I just happened to stop by...
 
11:42 AM
@mikeserv Doh. I did it too fast then.
 
@Caleb - all's well that ends well, I guess.
Thanks for clarifying that by the way.
 
12:07 PM
@casey that's an odd one. You can't call open on a pipe, but I would have expected a different error code, probably EINVAL
weird, susv3 says “pipes cannot be open()-ed” but doesn't say which error code open should return
 
 
2 hours later…
2:04 PM
Does the OP need NFS configured here?
0
Q: How to use my login password to do two things

LenzI have a cifs drive that I mount regularly, and it requires a password for it. Since it's annoying to type my password every time, I wrote a script that sets the PASSWD environment variable, mounts the drive, and then I clear it (cifs checks the PASSWD environment variable before asking). It wo...

 
2:55 PM
Is asking why I'm getting sshd entries from a blocked (by iptables) subnet on topic for U&L?
 
@CanadianLuke is windows somehow involved?
 
Nope. IPs from China, even though I (thought, at least) blocked the entire range
So, safe to ask on the site. Thanks
 
slm
@Ramesh Provided a couple of avenues
@CanadianLuke I would say that's on topic
 
@slm, was just going through your answer only.
 
@slm As I write the question title, it suddenly appears! :)
 
slm
3:08 PM
The search for previously asked Q's is pretty good 8-)
 
@slm, even if these packages are already installed, if the OP doesn't have root access, he won't be able to edit the /etc/nsswitch.conf and /etc/auto.master files right?
 
slm
@Ramesh Yes, that is a "major" dilemma
 
OK, very quick question... I found out I needed -I instead of -A on the rules to block them right away. I tried putting them further up the iptables.rules file. Where can I start the -I INPUT -s ip.add.re.ss/8 -j DROP entries?
 
@CanadianLuke I suggest you try denyhost. It should do all this for you:
   DenyHosts  is  a python program that automatically blocks ssh attacks by adding entries to /etc/hosts.deny.  DenyHosts will
   also inform Linux administrators about offending hosts, attacked users and suspicious logins.
 
Hi Friends
I need help in troubleshoot DNS error
 
3:12 PM
Hi, you can post your question in the site :) By this way, your question will be reached to more people and you can get your issue resolved quicker.
 
actually when we are opening carsizzler.com it take to much time but the same time when we open 103.225.77.2 it open very quickly without any delay.
 
@terdon, by blocking ssh attacks mean blocking invalid login attempts to the server?
 
slm
@CanadianLuke I would suggest using fail2ban
 
Well, I know fail2ban failed to ban many "attacks"
 
@CanadianLuke It shouln't really matter how high up the file they are. As long as you don't override them with a later rule, they should go into effect.
 
3:15 PM
I set it to ban on 2 incorrect SSH attempts, but it never seemed to matter
@Caleb The "nat" table is not intended for filtering, the use of DROP is therefore inhibited.


Error occurred at line: 4
# more iptables.rules
# Generated by iptables-save v1.4.14 on Wed Jun 18 08:03:11 2014
*nat
:PREROUTING ACCEPT [60143:14159825]
-I INPUT -s 166.0.0.0/8 -j DROP
 
@terdon Blocking at the SSH daemon level is WAY different than blocking at the iptables filtering level.
@CanadianLuke Uuuu, ya that shouldn't be in the NAT table section.
If you are dealing with a structured rules file like that then the section matters. I usually just script it myself so that's not one I run into.
 
slm
@CanadianLuke That's entirely possible but I find it suspect, thinking it might be more of a configuration issue
 
Damnit. Well, I'm gonna give denyhosts a try anyways
 
slm
The damnit isn't directed to us, right?
 
No, of course not. Canadians would never say that to a person/people! Only to crappy programs :)
 
3:20 PM
Is it also necessary to specify the iptables rules in the correct order? i mean, if we have allowed some packets in the beginning of rule that rule would be processed before denying right?
 
Correct...
K, I'm going to scrap those rules from the iptables section, and try denyhosts. That takes effect immediately, correct?
 
slm
@CanadianLuke would seem so
http://denyhosts.sourceforge.net/
 
That seems to be a much better solution than the regular iptables!
Just watching the logs. I see the failed attemps on auth.log, and then 30 seconds later, they're stopped on the denyhosts log :)
Thanks guys
 
I personally much prefer iptables. If a host is attacking my SSH, i want to ban them from the entire server, not just ssh
 
slm
Yeah I would agree w/ Patrick, it's better to deal with attacks via the firewall. I would try and dig in as to why fail2ban isn't working for you. Add some logging to confirm that the rules are working the way you think.
 
3:32 PM
anyone has any idea why the perl syntax highlight doesn't work here? unix.stackexchange.com/a/137855/41104
 
@Caleb Wouldn't it be the way to go though? If you want to block ssh login attempts, denyhosts and hosts.deny seem the right approach. I say this from the depths of my ignorance though, am I wrong?
@Braiam Syntax error here: $config_line[0] =~ s[^(\w+)://][];
 
Well, denyhosts seems to work. But I guess the real question is, where does it block first? iptables or hosts.deny?
 
You seem to be using [ as delimiters but are not closing the s///. It should be s[^(\w+)://][[ or something, depending on what you're trying to do.
 
@terdon weird, I'm copying from the apt-mirror script directly
 
@CanadianLuke denyhosts/hosts.deny uses the tcpwrappers library. it blocks at the application level
 
3:39 PM
@CanadianLuke According to its man page it's hosts.deny but if anyone has anything else to say, believe them. I am NOT an expert on this kind of thing.
 
it's supposed to detect *://* where the first * is the protocol and the second is the domain
 
@Braiam Really? It looks like a syntax error to me unless this is some kind of weird idiomatic thing.
WHat script is this?
 
/usr/bin/apt-mirror
 
@CanadianLuke iptables! It's WAY up the chain from when the SSH daemon gets it.
 
@Patrick, so iptables work on network layer level, is it?
 
3:42 PM
@Ramesh yes, iptables is the earliest point of entry into the system
 
@Braiam The one I have does not use that syntax. Can you show me the relevant line?
 
@terdon Usually if you have the option you want to block stuff like that as far UP the software stack and networking chain as you can without breaking anything else.
 
@Caleb So get fail2ban working, and hope it works?
 
@Caleb sounds reasonable.
 
3:42 PM
@terdon ^
 
@CanadianLuke, not sure if this helps. forum.slicehost.com/index.php?p=/discussion/4676/…
 
@CanadianLuke It would do automatically what you were just trying to configure manually, which has some advantages. And you don't have to "hope" it works, either it does or it doesn't. If you configure it right it will.
 
@Braiam Yes it does. Ok, well that's what's confusing the color highlighting but I have no idea why that works. Some kind of weird esoteric perl syntax apparently.
AAA! It's giving two character classes!
Huh, cool. So, apparently, this is correct perl syntax:
 
Would it be OK for me to post my fail2ban configuration files to Pastebin and link here, to ask for help, if I can't see anything blatantly obvious?
 
s[characterclass1][characterclass2]
 
3:47 PM
@CanadianLuke Sure. For the sake of semi-private network or config details you can make it non-indexed with a short expiry; but it shouldn't be a real issue either way.
 
ugh, I have problems writing a tag excerpt for apt-mirror:
> questions about the tool to mirror/clone Debian repositories, apt-mirror
or
> tool for mirroring Debian-based archives locally.
 
@terdon s[regexp][replacement] rather. In Perl, if the opening delimiter of s, m, q, etc. is an opening parenthesis/bracket/brace then the end of the quoted unit is the matching closing delimiter
You can write s[[[:alpha:]]]}uc$&}ge to uppercase a string. If you hate the next guy who'll read your script.
here's a slightly more practical example:
m{@ARGV}?$c+=1:2for<STDIN>;eval(_PEOBFQZzCJUKMYLHXW^q[/",!6du9&-hn.508i|4])
For, um, some value of practical. But this is an actual Perl program that does something useful.
 
4:02 PM
Hi guys.
 
Anyone got an idea about the fail2ban config file?
 
4:18 PM
where should I post a Linux security hackerish question?
 
@Ramesh depends on the question... could be here, Server Fault, or infosec
... and I guess a few more, possibly, depending on the question (say, crypto)
 
I will post here in that case. It's just too much info to paste in the chat window actually :)
 
well, you could paste the title, or a short summary
oh noes, I have lost 2 precious Ubuntu rep from "user was removed". Wonder how I got 2 rep from a user anyway...
suggested edit, maybe?
 
@derobert, I have posted the question.
 
@Braiam Is that really useful?
I mean, it'd be useful if it told you what type of questions to tag with that tag. Like, should I tag a question about using an apt-mirror repository, or only ones about configuring apt-mirror?
 
4:34 PM
@derobert using an "apt-mirror" repository is the same as using any other apt mirror ;). it's about the tool that mirrors repository, apt-mirror
also, there are experts in the use of the tool
 
Yep. My point was your proposed tag wiki excerpt doesn't answer that question, it should.
 
@derobert btw, how many debian bug reports would you say you have filed?
 
@FaheemMitha No idea... I haven't counted.
 
I switched email addresses around 2011, and the BTS doesn't make it easy, but I'd estimate around 200 or so.
@derobert I archive all my bug reports, so I could count them, I suppose.
 
bugs.debian.org/cgi-bin/… ... well, there is my current email address
(no, I don't really care if its in the chat log, its not like I could get any more spam than I already do)
 
4:48 PM
Hey, that might be a good job for a perl script. Go through an mbox file, and count the number of unique email addresses of the form number@bugs.debian.org. If I posted it, I bet it would get upvotes. People are mad about scripts here. Though I might get ticked off for not trying first.
 
bugs.debian.org/cgi-bin/… ... older email address (now dead)
 
@derobert that looks like a fair number, more than mine, probably. How long have you been using Debian?
 
ways to get amusing amount of spam:
- give your email to whatever says it will give you something "free" when asked
- publish your email in plain text online
- send bugs to any bug tracker
 
@FaheemMitha Since Potato, I think
 
Hmm, maybe 500 in total. Any standouts?
 
4:50 PM
Nah, that's over 700 in total... Standouts, not sure.
 
@derobert ok
What about ones with patches? Those are usually more interesting.
 
Here is a fun one, still AFAIK unresolved: bugs.debian.org/cgi-bin/bugreport.cgi?bug=717190
 
You could put these in your election statement. Hey, look at me, I'm a good citizen!
 
(from yet another email address...)
 
Hey, I'm now off-topic for the chat room! Bad Faheem.
 
4:52 PM
@FaheemMitha It'd go with the "hey, I'm a BSD committer!" guy
 
@derobert I've got one against aptitude from 2002. Still unresolved.
Others too, but that may be the most prominent.
 
In actually find bug reports a good way of determining people's technological literacy. And of course, whether they give a damn.
 
anyone an expert with GRUB ?
 
@marti no.
 
4:56 PM
@marti somewhat, though do see the room description before asking here in chat...
 
well its related to GRUB right?
I am having huge trouble
0
Q: Unable to access BIOS on Lenovo

martiI have a Lenovo G580 on which I installed Windows 8 and several linux distros which have messed with my booting.The last I used was boot-repair tool. I have an EFI partition on sda2.The Grub load and I can boot to windows or Linux Mint or Mageia. It seems now I cannot access BIOS by my regular F...

 
Your link in that question is 404-compliant
you might also be able to get into the BIOS by booting it from grub, which you can do with EFI.
 
@derobert Here is mine, old address -> bugs.debian.org/cgi-bin/…
 
How can I get into BIOS once grub-efi has already booted? is it that fwsetup cmd? doesnt seem to work
 
5:00 PM
 
@marti yeah, fwsetup from the grub prompt should get you into the firmware interface
or "BIOS" as you're calling it
 
But the first time I call the command nothing happens...after that it just say cmd not found
command not found
 
Hmmm. It works on my machines... All two of them that have UEFI enabled.
 
so you mean after that command it reboots and gets into UEFI settings?
 
5:08 PM
@marti Well, it doesn't reboot, it goes directly to the settings. But I confess I've only ever used the menu item (which just runs that command). I haven't actually tested typing fwsetup at the grub command prompt.
But it should work the same. I can test that on one of my UEFI machines tonight, if you'd like
 
It should be the same.But why wouldnt it work for me? So have I really messed up my NVRAM variables?
 
@marti I'd guess there is a bug in the firmware. Well, all firmware has bugs, tons of them... So, rather, you've hit a bug.
 
@Gilles Oh. Nice, I didn't know that, thanks.
 
@derobert The proprietary firmware for routers seems particularly terrible.
 
@FaheemMitha Not by firmware standards!
 
5:15 PM
so any workarounds? should I dare flash my BIOS ?
 
@derobert Dunno. I don't have that much experience with firmware.
 
@marti I have no idea if that efibootmgr -b 0 -B in the thread you linked is going to get you back into setup, or just get nothing booting.
Flashing the BIOS could be interesting w/o having access to the interface...
@marti if you can find out what the first entry should be, you could put that in the boot order, after Linux, and maybe that'd work.
 
So, fail2ban just isn't starting, even though it says it is. Now to troubleshoot...
 
I could flash it within windows.However if something goes wrong with the flash or with that efibootmgr command , I wont be able to boot anything and will get bricked
@derobert How can I find that entry?
 
Yep. I'd guess if you could reset the NVRAM it'd start with sane defaults...
@marti Well, if that guy is still around on the forum, he could grab it from his machine
 
5:19 PM
I tried to trace that guy.lol.I came across this sysmagazine.com/posts/169985
Its a translated version.The same guy Adiost had this issue...lucky for him he had his warranty and replaced it.
@der
@derobert How do I reset NVRAM?
 
@marti On desktops its easy, you open it up and pull the battery. Not sure on laptops. Some are easy enough (pull the battery [the main one] and unplug from AC adapter, wait 10 minutes)
others, you'd have to disassemble
 
Disassemble and take out the CMOS battery?
 
I think it's pretty likely that even if that efibootmgr line makes nothing work, that the NVRAM reset would fix that.
No guarantee on that, of course.
I suggest you write down your current efi boot vars, though
so you can restore them, but at higher positions, from a recovery disc
 
How do I get my efi boot vars? also I cant get into a recovery disk as I cant change my boot medium right?
 
@marti Well, you can try F12 (or 10, or 11) to get a boot popup
you could also try some other keys to get into the setup menu (del, for example)
but in theory, once you remove all the boot entries, you're supposed to be able to get into the firmware setup menu again.... that's when you'd use the recovery disc
 
5:32 PM
I did try...none of them work...thats the reason for my panic...F2 for bios and F11 for changing media both dont work.
 
you get your current settings by running efibootmgr without any arguments
 
running efibootmgr gives me a blank output
 
you're running it as root?
 
yes
 
that's weird, as that guide is telling you to delete the first entry... but there aren't any...
 
5:35 PM
@casey The pdf version works too. Looks like something in my .emacs breaks preview for these journal class files. Gosh, isn't software fun?
 
I think cat /sys/firmware/efi/efivars should give you all the EFI vars... is that empty too?
(I unfortunately do not have any EUFI machines in front of me...)
Actually, I do have one that I can wake from network and ssh into. That'll work...
 
I remember grub2-efi install also gave an ouput saying Magei in 0000 something.However its empty now.
the command u gave: cat: /sys/firmware/efi/efivars/: Is a directory
 
Ok, cat on that won't work. It's a directory. And is empty on my fully-working UEFI machine.
root@Forest:~# efibootmgr -v
BootCurrent: 0000
Timeout: 0 seconds
BootOrder: 0000
Boot0000* debian        HD(1,800,3c800,7f2234ed-541d-4950-8f01-e9e2890ce71a)File(\EFI\debian\grubx64.efi)
that's what that should look like
 
so should I try adding an entry then?
 
@marti no, honestly, at the moment, since its not broken, I think the wisest course of action is to not touch it...
provided of course that you haven't just removed that entry, it has been booted recently, correct?
 
5:44 PM
yes.
 
BTW: Turns out the directory is /sys/firmware/efi/vars/, and it should have a bunch of subdirectories (one per EFI variable)
 
@derobert not sure what the issue was here -> bugs.debian.org/cgi-bin/bugreport.cgi?bug=409383
 
@derobert I can only see del_var and new_var file
 
@FaheemMitha I think it was that some of the servers in the rotation were slow and/or down.
Normally, you connect to the first address DNS returns... that lets tricks like DNS round robin load balancing work
 
@derobert Oh, I see. Actually, I'm surprised apt always picked the same one.
Wouldn't one expect a random selection?
@derobert Yes, that is what I'd expect too. Does it still do that? You could always reopen it.
Well, create a new bug I guess.
 
5:48 PM
It appears there is a reason for that behavior.
 
@derobert Which is? I didn't follow the discussion.
 
The servers aren't actually in sync for at least 4h a day.
No idea if that's still the case.
I know I had fun with that when I had squid in the middle...
 
@Ramesh I guess so. I really know next to nothing about networking, sorry.
 
@derobert um, not following. what servers?
 
Squid would actually do the round robin, I think. So I'd get all sorts of hash sum mismatches from, e.g., a newer release file on one server
@FaheemMitha The Debian mirrors
 
5:50 PM
@derobert the only two files are del_var and new_var
 
@marti Wow, that's broken
@marti check through dmesg see if there is anything about failing to read the EFI vars
... and, just to confirm, you are booting with UEFI, not the BIOS compatibility mode?
 
@derobert So, the idea is not to have people trying different servers in a single run, which might not be in sync? How long does apt remember the individual address, and how does it manage it? Caching?
 
@FaheemMitha Oh, I suspect it just sorts the DNS result it gets.
 
@derobert Hmm, and picks the lowest, or something?
 
yeah its uefi boot
 
5:52 PM
@FaheemMitha Yeah. Or something deterministic, at least. Hopefully not lowest, or that mirror would collapse!
 
@derobert Well, if it was deterministic with no input, it would always pick the same mirror. Does apt still do that, btw? Would make a good question here. Probably wouldn't be as popular as the cow one, though.
 
@derobert Here is my dmesg shrib.com/uefi
 
@FaheemMitha Well, it needs to always pick the same mirror on a particular machine, at least for a while. So you could always hash the machine name, or IP, or something like that to pick one. You want it to be different on different machines, of course, to prevent overloading one mirror.
No idea if apt still does that.
 
@derobert Right.
this would only apply if one was to be using a pool like https.us.debian.org i guess
E.g.
host http.us.debian.org
http.us.debian.org is an alias for ftp.us.debian.org.
ftp.us.debian.org has address 64.50.236.52
ftp.us.debian.org has address 64.50.233.100
ftp.us.debian.org has address 128.61.240.89
ftp.us.debian.org has address 128.30.2.36
ftp.us.debian.org has IPv6 address 2610:148:1f10:3::89
 
@FaheemMitha yep, that's the pool I was using
@marti [ 0.353172] efivars: get_next_variable: status=8000000000000002 ... that's why there are no variables, kernel got an error trying to read them
 
6:02 PM
so my UEFI is buggy?
 
@marti yes, definitely. Just like everyone else's. Yours just happened to have a bug you hit...
@marti Anyway, my suggestion is to leave it as-is for now. If it does stop booting, then I'd suggest a NVRAM reset by pulling all the batteries, followed by a firmware update if available.
 
Okay thanks a lot derobert.Do let me know if something does ever strike.Although the fwsetup.why wouldnt that work?
 
@marti I would assume because the firmware is currently very confused and not working
BTW: Here is documentation on how the EFI vars interfaces are supposed to work. uefidk.intel.com/blog/accessing-uefi-variables-linux ... it might come in handy if you wind up having to do an NVRAM reset.
 
Thanks derobert.Here is a bug report similar to my issue.bugs.launchpad.net/ubuntu/+source/linux/+bug/1082418
 
6:18 PM
WOOT!
100% [21 Translation-en bzip2 0 B]    2,979 PB/s 0s^C%
btw, testing mirrors are broken... or at least temporary down
 
@derobert I f i ever manage to get into my UEFI should I reset it to Default settings?
 
@marti No idea. There is a lot to be said for "it's working, don't fiddle with it"
@Braiam Debian? Working here.
There are a bunch of Debian mirrors, and they're run pretty independently. So it's unlikely for them all to be down...
 
W: Failed to fetch xz:/var/lib/apt/lists/partial/http.debian.net_debian_dists_testing_main_binary-a‌​md64_Packages Hash Sum mismatch
@derobert is the index files itself
 
@Braiam Working here. I've found sometimes clearing out the downloaded lists (rm /var/lib/apt/lists/*) helps with that error.
Or, since you have less bandwidth, maybe just the one list with the error
 
@derobert why deleting them all if it's just one of the files that it's wrong? also, that would not help since mine is in the .../partial directory
there, just deleted the one that was giving problems
 
6:29 PM
@Braiam Well, because I have enough bandwidth that deleting and redownloading all of them is quicker than figuring out the right one to delete :-P
 
Google gives me targeted searches, just looked for "cocoa" and the first result was Apple API
 
@Braiam what are you searching for Cocoa for?
 
@derobert make chocolate...?
 
Ah, I thought your two messages had some relation. And was confused.
@Braiam Besides, if you want to make chocolate, you need:

 The Frying Pan

Sometimes hot, always heavy. (cooking.stackexchange.com)
:-D
Any way, lunch time. @FaheemMitha maybe this time after lunch, I'll actually get to copy editing!
 
slm
6:52 PM
@Braiam can I see your recommendations on bandcamp?
@Braiam - I found my feed, what's your name
I only have the D Fast - Sunday recommendation from a few weeks ago
oh boy it's evan
1
Q: How do I find out if my wireless card supports 5ghz?

Evan CarrollI have the 03:00.0 Network controller: Intel Corporation Centrino Wireless-N 2200 (rev c4) Hwo do I find out if that card/driver support 5 GHz?

 
@slm This one time at band camp...
My X server crashed for no reason. A bit worrying. Why would this happen? Maybe my graphics card is dying? It is quite old. The logs just said server terminated unexpectedly.
 
00:00 - 19:0019:00 - 00:00

« first day (1348 days earlier)      last day (3607 days later) »