« first day (1369 days earlier)      last day (3600 days later) » 

slm
12:00 AM
@Ramesh you mean w/ an IP or w/ drivers?
 
Is there a difference?
@slm, I first perform the installation of the boot image (install.img) which is available in the network share. After that process, I carry on the remaining installation from the USB drive. Is my understanding correct?
 
slm
12:27 AM
@Ramesh sounds right. You install the .img to media, boot with it, and then configure it during boot. During install you point it to the HTTP/NFS server w/ the actual CentOS/RHEL image.
 
 
1 hour later…
1:38 AM
The OP is editing my answer to answer my comment. Sigh.
 
@terdon I would have closed that one... long ago
 
@terdon I wanted to notify in chat. But I suddenly remembered these suggested edits will automatically go through from the review queue and so no need to notify :)
 
@Ramesh Yeah, especially when it's your answer being edited :)
@Braiam Understandable. But he's responding to comments and actually trying.
It's a perfectly decent question now that both @Ramesh and I have edited it.
 
0
A: Shell Script - how to scp into remote server and download files and protect password

GillesYou can authorize as many public keys as you like on the server side. Furthermore, you can restrict a key to a specific command on the server side. So generate an SSH key pair on the client, and don't put a password on the private key. Append the public key to the list of authorized keys, and add...

what is rrsync in Gille's answer?
 
can anyone flag this? askubuntu.com/q/493382/169736
 
1:49 AM
The link just opens some code which I do not understand.
 
@Ramesh Flag -> spam ;)
 
Oops. @Briam I meant the rrsync link :)
 
... @Ramesh tab FTW
 
slm
2:09 AM
rrsync: Restricts rsync to subdirectory declared in .ssh/authorized_keys
it's a perl script wrapper
 
@slm, I understood that. I actually wanted to know like how this will work.
So, whenever rsync is being called in ssh this wrapper function will be called?
 
slm
what does rrsync do?
 
Yeah.
It is a wrapper class for rsync, right? So, if rsync command is called in ssh terminal, to restrict the access to particular subdirectories, rrsync will be called?
 
slm
it just wraps rsync and limits only the target directory for writing
# The client uses "rsync -av -e ssh src/ server:dir/", and sshd on the server
# executes this program when .ssh/authorized_keys has 'command="..."'.
# For example:
# command="rrsync logs/client" ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAzGhEeNlPr...
# command="rrsync -ro results" ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAmkHG1WCjC...
#
 
Have I deleted my partition table? I'm trying to install lilo to dual boot slackware & windows8. When I tried to run 'lilo', it complained about my partition table, so I tied running it with '-P ignore'. That didn't do anything, so I tried '-P fix'. Now lilo says, 'Fatal: Partition entry not found.'
 
slm
2:17 AM
so you the remote user would run this: rsync -av -e ssh src/ server:dir/ and assuming you're using the other 1/2 of that key you'd be limited to writing to logs/client
@Ramesh ^^^^^
 
When I run 'parted -l', I get a list of all my partitions, but is this just a cached copy?
 
slm
@WhiteHotLoveTiger - no those are there
not cached
is it mbr or gpt partitioned HDD?
 
Thank god.
gpt
 
slm
you can confirm w/ fdisk -l
 
This is my first time using something with gpt
 
slm
2:19 AM
sudo sfdisk -l (for gpt)
 
sfdisk warns me that it doesn't support GPT, though it goes on to list some odd information
 
@slm thanks. So it works similar to ssh public private keys I suppose.
 
3:04 AM
@terdon, somehow unix.stackexchange.com/questions/141244/… reminds me of,
-3
Q: how to select specific range of line & count the specific occurence of the first charcter w.r.t. each unique second charcter?

yogendra singhHi I have a file like this : #0 0:() 1b:cg*b 1c 0:cg xe #4 0:() 0b:cg*b xc 0:cg 1e #8 0:() 0b:cg*b xc 0:cg xe #12 1b:cg*b xc 0:cg 0e #16 xb:cg*b 1c xe #20 1:() xb:cg*b xc 1:cg 1e #24 x:() xb:cg*b xc xe #28 0:() 1b:cg*b 0c x:cg 0e #29 0:() 0b:cg*b 1c x:cg xe #32 0:() 1b:cg*b Where #0 means at t...

 
@Ramesh Yes. It's beginning to remind me too.
 
0
Q: dmesg message repeat

peterretiefI keep getting this message repeated in dmesg [215259.269481] [UFW BLOCK] IN=eth0 OUT= MAC=01:00:5e:00:00:01:00:26:f2:46:24:8e:08:00 SRC=192.168.1.1 DST=224.0.0.1 LEN=36 TOS=0x00 PREC=0x00 TTL=1 ID=1711 DF PROTO=2 [215270.008461] CPU2: Core temperature above threshold, cpu clock throttled (...

Interesting. The question has managed to attract only 11 views.
The pastebin content has attracted 208 views.
 
 
6 hours later…
8:54 AM
@terdon So, have you come across the Worm story? I'm guessing no.
 
 
3 hours later…
11:41 AM
@slm The editing efforts from polym on closed items give a nice boost to the reviewed Reopen Votes, but I am not sure how "Leave Closed" affects them WRT to a real reopen attempt by the OP. Do you know how easy it is to reopen something after a first reopen attempt has failed?
 
slm
12:14 PM
@Anthon see this meta
4
Q: Six reopen votes cast resulted in tie with question remaining closed

Jason SturgesWhen voting, doesn't a minimum number of votes required for action have to reached? For example, five votes must be cast to reopen closed questions. Wouldn't this mean a maximum of nine votes could be cast before reaching a consensus? I'm confused by this reopen vote in which three reopen and ...

3 leave it closes, closes it, period. Always needs 5 to reopen. Multiple attempts that fail do not impact it.
A mod can immediately VtO too 8-)
 
12:42 PM
@slm, ok, but that doesn't really answer my question. Once it is voted to be left closed, Can the OP or someone else reinstate another round of reopen reviews (based on more substantial modifcations). If the latter is not the case, we should discourage editing of close votes 'for the looks'.
 
slm
12:53 PM
@Anthon - once you've burned your VtO that's it. You do not get another one, so I guess the A would be no. WRT the editing, I do not know what happens in that scenario.
I often edit closed for looks as you say merely b/c I hate having crap on the site that looks...well...crappy.
My perspective is that all the Q's and A's are boulders that we're trying to roll up a hill, if I can push one a little bit closer to the top then I do so.
 
@slm. I have done such edits as well, often opening a window to do so while reviewing a "Close Votes' item, but if a closed question has only once chance of being reopened, then that should not be waisted on some nice but insubstantial edits. We could of course try it out on an old item.
 
slm
@Anthon I'll research our general Q of what happens to a user's past VtO wrt the editing of a closed Q
I'm heading to work now, when I have a chance around lunch I'll look.
 
1:25 PM
@FaheemMitha Sorry, was away, no I haven't. You made me curious, as you knew you would, though.
 
 
1 hour later…
2:31 PM
The text in the chat history is bouncing...
 
@Patrick Are you drunk perhaps?
 
No, not drunk. Though sanity is in question.
hrm. it went away once I dismissed the 'new question' notification bubble in the top left
 
slm
2:48 PM
@Patrick mine is too
I'm using Chrome
 
How long before comments in chat get dropped in the mailbox?
 
slm
mine continues even after dismissing the notification bubble.
 
Let's find out: Wed Jul 9 16:50:56 CEST 2014
terdon@oregano ~ $
 
@slm Does that mean we're both a tad bit insane?
yup, mine is continuing to do it now. another bubble came up, dismissed, and it's still bouncing
Resizing the window stopped it this time
that might have been it the first time too. The first time I also closed the chrome debug pane at the same time I dismissed the bubble. So it's likely resizing the window that stops it
 
@Patrick What does bouncing mean here?
@terdon Ah, the experimental method. But whose mailbox is this supposed to get dropped into?
 
slm
2:56 PM
resizing works for me as well.
but as soon as a new msg shows up in chat it reverts to bouncing
it's jittering a few pixels up/down
I highly suspect it has to do w/ whatever trick is used to force the browser to reload this page
 
still wondering what bouncing means.
something is moving up and down?
 
slm
the entire contents of the page for chat
 
@slm Wow. Does that look as weird as it sounds?
 
slm
it looks like the page is being aggressively reloaded and you can pick up a hint of this visually. It's always done this to a degree for though
 
@slm Oh
 
2:59 PM
@FaheemMitha Yours. It won't if you type anything after my ping though.
 
How often does the page get reloaded? I thought they must be using some fancy ajax thing.
 
@terdon - please consider replacing the phrase "formatting abuse" with "excessive formatting." Maybe it's just me, but the word "abuse" is not only "excessively" dramatic, but carries with it some serious, unnecessary connotations. Thanks. :)
 
@Chap Oh, sorry, that's a classic SE meme. That's what that kind of thing is called. No offense meant, I can see why it would be surprising if you weren't aware of it.
 
@slm I've never seen it do it before today
 
slm
I've seen it come and go
 
3:07 PM
@Chap have a look here for examples of how this phrase is used on the SE sites:
Anyway, I really did not mean to be rude and I'm sorry if I bothered you.
 
@terdon Yes, I can't wait 1/2 hr or whatever it is to type something here, sorry.
 
:)
Takes a few seconds, maybe a minute anyway.
 
Let's just ask someone. @Gilles, how long does it take before a chat message is dropped in a mailbox?
@terdon Rather more than that, I'd say.
I assume Gilles would know. He seems to know everything about this stuff.
 
@FaheemMitha A few minutes then. Not too long.
 
3:35 PM
@terdon Well, this one arrived in 25 min or less. Wasn't watching my inbox.
 
Wow!
 
Getting an upper bound is easy. Getting the exact time via this method is less easy.
 
Yeah
 
@terdon Well, I was just wondering if you had read it, and if so, what you thought of it. As I said already, I don't recommend looking at it unless you want to lose a week of your life. I closed it after two days of not getting enough sleep. I think I was around 1/3 of the way through. Perhaps on your next Bahamas vacation. You could catch up on your reading listening to the sound of the waves...
 
4:21 PM
Anyone know how SELinux works?
 
@Ramesh using unicorn powder... with some red....
 
@Braiam, thanks. I changed the configuration file of mysql and when I restarted it did not work.
 
"it did not work." exactly how?
 
After I disabled the SELinux and restarted the machine, it worked. So I was wondering, SELinux prevents someone from changing configuration files, is it?
The mysqld daemon failed to start.
 
with what kind of error?
 
4:41 PM
@Ramesh It's a mystery.
Personally, I would not use it unless I had nuclear secrets, or was forced to.
 
SElinux is a giant freaking pain in the ass. Ask any sys admin who's used it how they handle it, and I'd bet you the majority of their answers are "I disable it"
 
@Patrick Presumably it is useful for some things. But it was developed by the NSA, right?
So probably its real purpose is to spy on you.
 
@FaheemMitha the concept is nice, but it's extremely cumbersome and difficult to maintain
 
@Patrick he he. I always do this. vim /etc/Selinux/config change enforcing to disabled. :p
 
if your system never ever ever ever changes, and it does the same thing for years on end, using SELinux might be worth it
 
4:47 PM
So how it works?
 
that's not a simple answer
 
Does anyone use Microsoft Outlook web application here?
I just scheduled a meeting for 3.30 PM - 4.30 PM CT time and it sent out the meeting invite as morning 5 AM - 6 AM.
 
5:20 PM
@FaheemMitha 15min, IIRC
why didn't I get an inbox notification from that message?
 
@Gilles Do you have a reference?
 
@FaheemMitha source: Gilles's memory, private communication
@mirabilos You may want to comment about mksh on unix.stackexchange.com/questions/29554/…
 
@FaheemMitha that sounds right to me too
 
yeah 15 sounds right.
 
@mirabilos And about mksh's support for process substitution on unix.stackexchange.com/questions/141367/…
 
5:25 PM
@Gilles Ok.I thought it might be documented somewhere. Thanks.
 
@FaheemMitha it probably is. Have you searched MSE?
 
@Gilles no, let me look.
@Gilles a quick search doesn't bring anything up. but i suck at searching.
According to
76
Q: Should we be allowed to edit comments?

wafflesI am a horrendous speller, and my mild dyslexia does not help. Sometimes I typo in my comments, I look at it a few days later and see something that screams to be fixed. But to fix it I would have to get rid of all my up votes and lose continuity. Should we be allowed to fix minor typos in co...

the amount of time one is allowed to edit comments is 5 min, but in practice is seems to be 3.
 
@FaheemMitha no, it's 5min
you may be averaging that with chat messages, for which it's 2min
 
@Gilles Hmm, could have sworn it was 3. Ok.
@Gilles Ah, yes, I meant chat, sorry. I thought chat was 3? It is 2? Bummer.
 
5:44 PM
@FaheemMitha I'm trying not to look into it but you're making it hard :)
 
@FaheemMitha we all suck at searching when it's on Stack Exchange
read: SE search sucks
0
Q: load sudo privileges on unix box

user1643087We used to have sudo all in unix servers, n now for security concerns, it will be be revoked and can be provided only to specific file system the command were earlier loaded in the format/usr/bin* /clocal/iw-backup/, this restricted us to use the commands with any argument later we loaded in th...

voting to close as: unbelievably unclear what you're asking
 
@terdon Don't worry about it. I think Gilles' number is probably right.
 
I meant the worm thing.
 
@terdon Sorry, I thought you meant the SE thing.
@terdon Ah, right. Forgot about that.
Possible feature request: extend the reply thing to allow people to link without pinging the linkee. What do people think?
If that is too obscure, I can expand.
This one would be handy.
13
Q: Show original message on click, inline, for chat replies

PëkkaWhen somebody replies to an older message in chat, you have to click the "in reply to" arrow and jump to the message to get the context. That makes you leave the current position in the transcript, and old content possibly even needs to be loaded first. That's often a bit disrupting, especially w...

 
6:06 PM
@FaheemMitha I've often wished for that too. I doubt they'd do it though, it would probably complicate things too much. Well worth asking for though.
 
@terdon Ok. Currently searching through meta to see if it has come up before.
 
diff --git a/src/webkit/webkit_resources.gyp b/src/webkit/webkit_resources.gyp
index 6e46651..2ec6fdb 100644
--- a/src/webkit/webkit_resources.gyp
+++ b/src/webkit/webkit_resources.gyp
@@ -28,25 +28,18 @@
           },
           'includes': [ '../build/grit_action.gypi' ],
         },
-        ]
-    },
-
-    {
-      'target_name': 'webkit_strings',
-      'type': 'none',
-      'variables': {
-        'grit_out_dir': '<(SHARED_INTERMEDIATE_DIR)/webkit',
-      },
-      'actions': [
         {
what a messy diff :(
 
Interesting. I didn't know one could self-reply manually.
9
A: Block the ability to manually ping oneself in chat

GillesReplying to one's own post is useful. Some chat messages are part of a conversation, that's what replies are for. If you post more than one message as part of a reply on a busy group, it's nice to be able to link them. Self-replies are also useful when you want to add a tidbit of information to a...

So, how do I find the magic :123456?
@terdon doesn't look like it. posting now.
Not sure I understand this.
0
Q: Exim4 Configuration for Webserver

user47227I successfully configured exim4 to handle mail only for the localhost (just for forms on the website). I tested it by sending messages from the CLI to myself on another domain. My problem is that the server has the FQDN of www.example.com and I need to be able to send mail to person@example.com....

Is he asking how to use an email server?
What he is asking for just looks like standard functionality, if I am not mistaken.
 
6:27 PM
@FaheemMitha I use this:
12
Q: Chat Reply Helper for Stack Exchange sites

Oliver Salzburg Press : to start replying to a previous message. Press ↑ as many times as you need to mark the desired message. What does it do? First of all, the Chat Reply Helper for Stack Exchange sites provides a simple key combination to select a message to reply to. This removes the need to grab the...

2
 
@terdon That is very handy!
 
Yes it is!
 
The only problem: I had to restart Firefox, resulting in my 10+ open browsers from 5 different Workspaces being all on one and the same desktop after the restart Right click Move to Workspace, right click Move to Workspace....
 
@terdon
0
Q: Using the chat reply function without pinging

Faheem MithaIf you use the reply function in chat, it currently does two things. It links your message to an earlier message. It pings the person who wrote the message you are linking to. In some cases, it can be useful to decouple these two. Specifically, one might want to link to someone elses message ...

Is this clear?
 
6:36 PM
@FaheemMitha looks good
 
@terdon Thanks, that looks handy. Trying it now.
 
@terdon Great.
 
@terdon oh my god I love that extension
 
Hello! :D
 
6:39 PM
@strugee Yes, I have some Ubuntu packages. I don't really see that as a problem. As you must know by now, you're far more security conscious than I.
$ dpkg-query -S ubuntu | awk '{print $1}' | sru
aptdaemon:
apturl-common:
bsdmainutils:
cairo-dock-core:
gnome-doc-utils:
gnome-pkg-tools:
gparted:
lintian,
lintian:
live-usb-install:
mintinstall-icons:
mint-x-icons:
multisystem:
murrine-themes:
pastebinit:
shutter:
software-properties-common:
synaptic:
ttf-ubuntu-font-family:
Also, I don't quite get the point of that post. It seems to be saying that Canonical evil so anything remotely related to them is also evil. I don't buy that.
They make no specific claims apart from the (clearly outdated) points about the now defunct Ubuntu One and then just say that LMDE includes a tiny number of Ubuntu packages and, therefore, it must be unsafe. Without actually having demonstrated that any of those packages are unsafe and in fact only saying that one of them is for sure an exact copy of the Ubuntu one.
Their points about Testing sound more reasonable but, again, not that bothered really.
In any case, I would never recommend LMDE to a newbie, I was talking about Mint.
 
@terdon is lmde and mint very different?
from a usability standpoint?
 
@FaheemMitha The former is based on Debian, the latter on Ubuntu.
 
@terdon this is true
 
@terdon And that makes a difference?
 
Roughly, LMDE is a Debian and Mint is an Ubuntu flavor.
 
6:45 PM
@FaheemMitha not very much for usability, no
 
@FaheemMitha Yeah! Of course. Mint is much more polished and user friendly. LMDE is a stock Debian with some green added.
@strugee Dunno, a lot of tools are not around that the Ubuntu peeps love.
I had to fight it into submission considerably more than any Ubuntu or Mint install.
 
I still think Mint sounds yummy, whatever anyone else says.
I must have had a Positive Mint Experience as a child, or something.
 
PME?
Better than PMS to be sure.
 
@terdon Trying that extension. Didn't work.
@terdon Trying again.
 
@FaheemMitha You need to reload the chat window, then add a : as the first character of your message, then hit the Up button to choose which message to reply to.
 
6:49 PM
Ok, I think it works for me. Thanks for the tip, @terdon
 
:)
 
@terdon Right.
 
ok
@terdon hmm
 
Yes, this is a definite improvement.
@FaheemMitha Gosh, I feel like a power user now! I'm replying to my own message!
So, I could have a conversation with myself, if I wanted to. Web chat solipsism.
 
6:53 PM
@terdon I think it is just security people in paranoia mode. It is all about circles of trust.
 
@strugee None of them actually know. These are random people off the internet each spiting their own crap. They may be right, or not but I don't see why you think they are. They themselves are not sure:
> Mint might be an exception and need further investigation, but afaik it does not have an independent repository. (ofc LMDE is fine, because it uses debian testing repository and debian is not known to do weird things and has a strong ethical commitment... so maybe change Linux Mint to "Linux Mint Debian Edition")
 
@terdon /me shrugs
 
The points about Unity and sending data are completely irrelevant to Mint. That's Ubuntu bull.
 
I don't have a super strong opinion on it.
just playing devil's advocate a little, I suppose
 
I can see where they are coming from. This is all about the US govt spying on people. Like I said, paranoia mode.
 
6:58 PM
yep
 
Not such a bad thing on this unhappy little planet. As long as you don't let it go overboard. I have a mad old uncle who lives alone and is convinced everyone is out to get him / trying to kill him. That's going overboard.
Amusing. I didn't know coreutils had a factor program.
 
7:30 PM
What about this one?
45
Q: SE Chat Modifications -- Keyboard navigation and commands for chat

Tim Stone Screenshot Use /command shortcuts to perform common chat tasks: See message history inline: Easily preview replied-to messages: And much, much more... About Legends tell of a prolific Meta Stack Overflow chatter who despised using their mouse above all things. In an effort to keep t...

 
7:54 PM
Hey guys, do you know where I post meta questions for all SO sites?
 
ah ok thanks @Patrick
Oh wow I just saw that my meta question appears to affect only my settings
well
 
@Gnouc Your argument still makes no sense. You're saying it fails on the OPs machine because he's on a 32-bit CPU, and the number is bigger than 32-bit. But it's also bigger than 64-bit, so it shouldn't work on 64-bit either. But it does, meaning the code can handle numbers of any size. If the code can handle numbers of any size, than it doesn't matter if he's on 32-bit or 64-bit.
 
8:10 PM
@Patrick it might matter, uintmax_t is a 128-bit type on some 64-bit platforms
 
@Gilles That number is bigger than even 128-bit
It's slightly less than 256-bit
 
@Patrick see my edit
 
@Patrick: I read the doc info factor and read the factor source. I think the problem is that factor must convert string to int first. In 32-bit machine, this check always fails and return LONGINT_OVERFLOW.
 
@Gnouc I think that Patrick point is: 32-bit machine = 32-bit cpu? or 32-bit kernel? or 32-bit build?
 
@Braiam for the sake of the whole discussion here, 32-bit machine = 32-bit build of factor, the others are irrelevant
@Gnouc what's relevant here is not that the machine is 32-bit, it's the size of uintmax_t. Which is 64-bit on most common platforms, 32-bit and 64-bit alike.
 
8:23 PM
@Gilles: so why factor always fails in 32-bit machine when number greater than 2**64-1?
 
@Gnouc because 2**64-1 is the size limit on (most) 32-bit machines
but 2**64-1 is also the size limit on (most) 64-bit machines
so the 32-bitness is not relevant
 
@Gilles: Why it does not fail in 64-bit machine?
 
@Gnouc it does
 
@Giles: No, check it!
 
@Gnouc I did, it does
 
8:26 PM
"Yes." "No." "Yes!!" "NO!!"
 
@Gilles: Do you check if factor is built with GNU MP?
 
At least as of coreutils 8.13, the argument has to fit in uintmax_t. Which, on my Debian wheezy amd64, is a 64-bit type. This is the case even if factor is compiled with GMP support.
 
@Gilles: From info factor
If `factor' is built without using GNU MP, only single-precision
arithmetic is available, and so large numbers (typically 2^64 and
above) will not be supported. The single-precision code uses an
algorithm which is designed for factoring smaller numbers.
 
that sentence is misleading since even if factor is built with GMP, it doesn't support bignums
 
@Gilles: So if factor support bignum and not compile with GMP, can it handle number larger than 2**64-1?
 
8:44 PM
@Gnouc no, unless uintmax_t is larger than that on your platform
at least as of coreutils 8.13, I haven't checked other versions
 
Can you give portion of code show this problem?
 
@Gnouc show what problem?
 
I read the manual and the bug report but nothing point this out?
"that sentence is misleading since even if factor is built with GMP, it doesn't support bignums"
I have used factor in variant of 32-bit machine
and it always fails when number greater than 2**64-1
@Gilles: And another thing
From GMP home page, It says "GMP is a free library for arbitrary precision arithmetic, operating on signed integers, rational numbers, and floating-point numbers. There is no practical limit to the precision except the ones implied by the available memory in the machine GMP runs on. GMP has a rich set of functions, and the functions have a regular interface."
So GMP itself support bignum?
 
@Gnouc yes (that's what it's for!)
 
probably factor uses GMP only for factorization
 
8:56 PM
er, no, scratch that
factor does call a GMP parsing function, if it uses GMP
So if compiled with GMP support, it supports effectively arbitrary-sized numbers
 
...even if it's 32-bit
 
factor 5326497351283633029120 works for me
((2**62)*5*3*7*11)
on an unrelated note, does anyone else think I did something wrong here?
 
@Gilles: Can you explain more about xstrtoumax function use in factor source?
 
@Gnouc do you know strtoumax? strtoul?
 
Yes, but I think some thing has done in factor source, for function xstrtoumax
Because I have read and tested that the limit of factor in 32-bit machine always 2**64-1
even if it was built with GMP
 
9:10 PM
xstrtoumax is coreutils's own version of strtoumax
An x prefix is a common GNU convention for thin wrappers around standard library functions, for which a full implementation is sometimes also provided in case the function isn't available on some platform. Here, it isn't actually a wrapper: the implementation from the coreutils source tree is always used.
 
both Python and Common Lisp for example can support numbers of arbitrary length.
On 32 bit too, afaik.
 
Here's the code to parse the argument:
static bool
print_factors (char const *s)
{
  uintmax_t n;
  strtol_error err = xstrtoumax (s, NULL, 10, &n, "");

#if HAVE_GMP
  enum { GMP_TURNOVER_POINT = 100000 };

  if (err == LONGINT_OVERFLOW
      || (err == LONGINT_OK && GMP_TURNOVER_POINT <= n))
    {
      mpz_t t;
      mpz_init (t);
      if (gmp_sscanf (s, "%Zd", t) == 1)
        {
          debug ("[%s]", _("using arbitrary-precision arithmetic"));
          print_factors_multi (t);
          return true;
        }
      err = LONGINT_INVALID;
it calls xstrtoumax, and if GMP support is used and xstrtoumax complained that the number is too big, then the code calls a GMP function to do the parsing
@FaheemMitha So does C. The difference is that in C, you need to use a nonstandard library, whereas Common Lisp and Python have it built into the core language.
 
@Gilles Right. And factor needs help with that?
 
@Gilles: So why it always fails in 32-bit machine?
 
@Gnouc because that machine has a version of factor compiled without GMP support
 
9:17 PM
@Gilles: It's strange, I check that it is compiled with GMP in my tested 32-bit machine
 
9:29 PM
@slm, could you please have a look at these comments?
 
9:51 PM
@Gnouc What distribution?
 
@CristianCiupitu No, you didn't. Your edit on the question was fine IMO. You might have been a little overzealous with the answer. Most of your edits there were unneeded and did not really improve the post substantially but you were not out of line as such.
I would say the poster certainly overreacted.
 
what parts do you think were unneeded?
 
@CristianCiupitu Well, you basically changed the style rather than actually fixing any problems. Again, nothing wrong with that, just not strictly speaking necessary.
 
by the way, I would have linked to the home pages, but I couldn't find the home pages of 2 programs, so...
right
 
Seriously, you did nothing wrong. Don't worry about it.
 
9:54 PM
@Gnouc i have the same problem, but the debian factor is not compiled with bignum support because they don't want to depend on GMP. personally i think it is not a big deal.
 
ok, thank you
 
@CristianCiupitu In fact, looking at it again, I saw you added those extra links which are quite useful so yeah, you did improve more than just the style.
People get worked up about their posts being edited sometimes but that's their problem, not yours. Editing is very important on the SE network.
 
I agree, I got one of my answers edited once.
 
:16506413 I think some sites have a limit for users under a certain reputation. I don't think we do but I;m not sure. I'd have to check.
 
and it's funny because it was style edit basically
 
9:58 PM
@CristianCiupitu Heh, most of my answers have been edited. The first time it happened I was indignant and then had a closer look and saw that the edit improved my answer immensely and then learned to welcome edits :)
 
@Tim yes, the limit of 6 a day is site specific. I don't know which sites have it:
7
A: Is the question/day limit per site or across the network?

Nick CraverThe limit mentioned in the blog post is a per-site one.

 
Same here with my answer and when I look back at my original content it makes me wonder what was I thinking then.
 
@Tim you've been on quite the question-asking spree today, if I am not mistaken.
 
10:26 PM
@terdon @terdon you might want to look at this meta.stackexchange.com/q/164899/213575
 
Another penalty shootout......
 
10:47 PM
argentina vs germany!!!!!!
 
@Braiam Perfect, thanks! And of course, I already had it in my favorites. I should learn to check them every now and again.
sigh
 
Tim
Where (under which dir) does "dpkg -l <packagename>" find the info of the package?
 
@Tim dpkg database
 
Tim
is it under /usr?
 
@Tim no, is not a directory at all
 
10:52 PM
/var/lib/dpkg/status probably
 
Tim
when installing the package, there will be some files copied to some directories under /usr.
is one of the dir what helps dpkg to find the info of the package?
 
@Tim the dpkg system reads its own database, an entry is created each time you install a package using apt/dpgk. When you install from source, dpkg has no knowledge of anything you install and no amount of fiddling about short of modifying the database will change that.
@Braiam is there any way to add a file to the database short of manually editing it?
 
@terdon What database?
 
@Tim exactly what are you trying to do
 
Tim
"man packagename" does rely on the package files under /usr/local/man?
 
10:56 PM
actually is man binary just that most of the time the binary and the package are called the same
 
Tim
where "man packagename" looks up the manpage of the package?
 
@Tim /usr/share/man
 
Tim
Is that place same for packages installed from aptget and from make?
 
@Tim I'm smelling that you are trying to do something and instead of asking us how to do it, you are asking us how to do what you think needs to be done, so let me put it clearly: what's your goal?
 
Tim
no, you are smelling too much. I am organizing my knowledge.
 
11:00 PM
@Tim Don't follow. You mean local vs Debian binary package installs?
In that case, of course not.
the FHS generally recommends /usr/local for local installs
 
@FaheemMitha /var/lib/dpkg OK, not a database as in SQL but that's what it's called, the dpkg database.
 
Tim
configure, make and make install install packages under /usr/local, while apt/dpkg install packages under /usr, doesn't it
 
@Tim No. It's nowhere near as standard as that.
 
@terdon What was the question again?
@Tim no.
It depends.
 
configure is a shell script and any developer can choose to have it work in any way they please basically. Convention has it that user-installed programs go to /usr/local but even that is often ignored.
 
Tim
11:02 PM
Are those two paths by default?
 
There is a fair amount of fragmentation. Look at the FHS, which most linux distributions try to follow. Debian tries hard.
 
In any case, nothing to do with whether it was installed manually or by apt/dpkg
9 mins ago, by terdon
@Braiam is there any way to add a file to the database short of manually editing it?
 
@Tim what terdon said. sometimes /opt is used.
 
@FaheemMitha I was wondering how one could trick dpkg into thinking something had been installed as a package.
 
@terdon The usual way is equivs.
 
11:04 PM
@terdon IMO dpkg checksums the files
 
I personally have never used it, because I just do it the hard way, with binary packages.
@terdon I assume you mean from the pov of telling dpkg and friends - this package has been installed, get on with it?
 
@Tim, I've seen your last few questions and it looks like you're a little confused about how package management works. As @Braiam said before, what are you actually trying to do?
@Braiam What do you mean?
 
@terdon Maybe he is just trying to understand what is going on?
 
@terdon the /var/lib/dpkg/info/*.list have checksums, ergo you can't edit them
 
@FaheemMitha That won't make dpkg -l print it though, will it?
@Braiam Ah, yes, that makes sense.
 
11:06 PM
@terdon Print out what?
If you mean internal package information, then no.
There is no sensible way of faking that, afaik. And little point trying.
 
@FaheemMitha One of Tim's questions was about getting a manually compiled and installed library to be listed with dpkg -l
 
@terdon Well, you can add an entry with equivs, and then, yes, dpkg -l will list it.
 
Exactly, little point trying. Which is why @Braiam and I have been asking @Tim what his actual objective is.
 
but that is about as far as you can go. you can't get a listing of the files in the package for example.
 
@FaheemMitha Will it? So I could simply set any arbitrary executable as an alternative and then it'd be listed by dpkg? I really doubt that.
 
Tim
11:09 PM
Hi, I think I got it wrong in my old note:
files of a self built program are copied into multiple directories:

executable binaries: $prefix/bin
libraries (.so, .a): $prefix/lib

headers (.h): $prefix/include
source (.c, .cpp), if specify debug during compilation or for later uninstall/reinstall/repair: $prefix/src

configuration files: $prefix/etc

resources (pics,sounds,icons): $prefix/share
mans: /usr/share/man, not $prefix/man
package installation infromation for being tracked by OS package system (.pc): $prefix/lib/pkgconfig
 
@terdon no, you manually create a dummy package. like I said, I've never done it.
It is just a placeholder in the system, so that other stuff can depend on it, essentially
 
@FaheemMitha Ah, yes, with a dummy package I guess it would work.
 
Tim
Can you point out the correct ones?
 
@Tim .pc files only makes sense for libraries/headers, a normal system can go without pkgconfig
 
a use case, is texlive, which the tex.sx people prefer to use from upstream rather than the debian version.
 
11:12 PM
@Tim Pretty much all of them. But most of those are conventions, not rules.
Also, what @Braiam said. For example compare the number of .pc files I have to the number of installed packages:
terdon@oregano ~ $ ls /usr/lib/pkgconfig | wc
     46      46     884
terdon@oregano ~ $ dpkg -l | grep ^ii | wc
   3774   38094  584722
terdon@oregano ~ $
The point is that if you install a program from source, you can place everything wherever you like. apt/dpkg won't know about it, they only know about packages installed through them.
It might be possible to somehow make dpkg aware of such packages but it really isn't going to be worth the effort and I have never seen a configure script that attempted that.
 

« first day (1369 days earlier)      last day (3600 days later) »