=== FFForever-Away is now known as FFForever === asac_ is now known as asac === FFForever is now known as FFForever-Away [01:45] Why do many programs have a "/chrome" sub-directory where they are installed? [01:45] I notice this more in open source applications.. [02:21] New bug: #386172 in postfix (main) "package postfix None [modified: /var/lib/dpkg/info/postfix.list] failed to install/upgrade: subprocess pre-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/386172 === smultron_ is now known as smultron [03:50] i'm getting these errors in dmesg. All was working well, I shut down my server and brought it home from being at school. I couple weeks later, I plug my server in, boot it up and now I'm getting these errors. [03:50] http://pastebin.ubuntu.com/193919/ === s_markow_ is now known as s_markow === FFForever-Away is now known as FFForever === s_markow_ is now known as s_markow [04:36] JDStone: At first look, your 1TB drive /dev/sdc may be going bad?? Or the data cable connecting it isn't firmly connected at both ends? Did you open up the hardware and unplug/replug both ends of the SATA data cable? Might be worth a try. Also, could the drive have sustained a nasty physical shock during its move from school to home?? [04:37] What does smartd say about it? [04:37] Obviously also check the cabling, and if possible simply replace the data and power cables. [04:43] hey guys [04:44] i had a question for 32 bits packages on 64 bits machine [04:44] Go ahead and ask it :) [04:44] woow, so i am not alone [04:44] thanks jmarsden:-) [04:45] i thought nobody is here [04:45] anyway my question is [04:45] soccerGuy: here as well. [04:45] cool [04:45] :-) [04:46] soccerGuy: There are 32bit library compatability layer for 64 bit machines. [04:46] I need to install 32 bit on my amd64, i know chroot can do this or ia32-libs [04:47] however for ia32libs, i need to figure otu each package and their dependency [04:47] I remember that the 'World of Goo' game, which was 32bit only, had some excellent instructions. [04:47] i see ia32-apt-get automatically download necessary libs and dependencies [04:47] Ooo.. I didn't know that. [04:47] however there is not enough documentation how to use that superb package:-( [04:48] * jmarsden will install it and see :) [04:49] I know that my question a little spesific:-) but I was suprised that this magical package is infamous in our Linux community [04:49] http://2dboy.com/forum/index.php/topic,1432.0.html [04:49] jmarsden, let me know how it goes! [04:49] and good luck! [04:50] sudo apt-get install ia32-libs [04:50] you will need that [04:50] pschulz01, yeah i used ia32-libs [04:50] but the problem is I need to locate each package and use dpkg and figure out their dependencies, thanks for suggestion though:-) [04:51] soccerGuy: Did you read /usr/share/doc/ia32-apt-get/README.Debian ? Seems to explain what is going on reasonably well... [04:51] It would be MUCH better to fix the broken package that asks for ia32 libraries. [04:52] twb: True ... I'd guessed the software package concerned could be commercial closed binary-only stuff... [04:52] yes I did jmarsden, but the thing i dont understand how apt-get understand which package(32 or 64) I want him to install [04:52] jmarsden: then all you can do is file a bug. [04:53] jmarsden: but it's important that the user (soccerGuy) at least knows that he should try that approach if it's open to him. [04:54] sorry what you mean twb? [04:55] soccerGuy: line 11 of the README.Debian says: + Library package names are prefixed with ia32- [04:56] So surely you can do sudo apt-get install ia32-libsomeliborother # to force installing the 32bit version. [04:56] But what makes you think you need to figure out all the dependencies by hand... that is what apt does for you... right? [04:56] yes [04:57] it creates 2 sources list [04:57] etc/apt/foreign/sourcelist [04:57] Yes. Make sure they include the repository containing the 32bit package you "need" (please do tell is what it is!) [04:57] and etc/apt/native/sourcelist [04:57] and then use apt-get as normal. [04:57] I know, I read the documentation :) [04:58] haha [04:58] sorry jmarsden [04:58] So ... what exactly is your question or problem? [04:58] so if i run apt-get install ia32-apache2, will it install 32bits apache? [04:58] i didnt know how to use the package [04:59] Getting biarch in dpkg would be nice... [04:59] Well, apache2 is not a library, so that's not quite what the docs say... [04:59] i thought it can download binary packages [04:59] soccerGuy: sudo apt-get install whatever-weird-package-you-need.deb # should pull in whaetver other 32 big dependencies it needs... [05:00] soccerGuy: So did you actually try it? Just apt-get the package you really need?? [05:00] i didnt try what you said, i will give a try [05:01] so i still need to locate the package [05:01] soccerGuy: No, you just need to know its name -- but if you don't know that, how do you know that you "need" a 32 bit version of it??! [05:01] i mean if i type apt-get install apache2, i can get apache2 for 64 bits and i dont need to figure out dependencies [05:01] What is the package you are really trying to use that is 32bit only? [05:02] I need to install apache2 webserver [05:02] That is not 32bit only! [05:02] no, it is not 32 bit only [05:02] Trust me the 64 bit version works fine. [05:02] i need to use 32 bit version for memory concern [05:02] So... what is the real issue here? [05:02] real issue is ram [05:03] Are you 100% sure the 32bit version will use significantly less RAM than the 64bit version? Where did you get that information from? [05:03] I got from various articles [05:04] I have only 256MB VPS [05:04] Have you tried using the 64bit version and just removed all the modules you do not need, configuring apache carefully, etc? [05:04] yes [05:04] I optimize the server [05:05] according to their tuning documentation [05:05] By the time you load all the relevant 32bit libs etc you might find you are not really saving much RAM... can you point me to an article suggesting the use of 32bit apache on a 64bit server to save RAM? [05:05] so now i am rereading that line 11 [05:05] sure [05:05] one sec [05:07] soccerGuy: Whoever's telling you to use 32-bit versions of apache to save RAM is silly. [05:07] infinity: That is my sense too, but let's read his sources to be sure :) [05:07] soccerGuy: While it's technically true that the data structures are smaller, and it will save a bit of memory usage, you'll also lost all the 64-bit registers and lose a whack of performance in the process. [05:07] http://stackoverflow.com/questions/935443/64bit-vs-32bit-ubuntu-on-low-memory-systems-why-does-ubuntu-64bit-eat-my-ram and http://www.scribd.com/doc/363677/Benchmarks-AMD64-in-32bit-mode-vs-64bit-mode-Ubuntu [05:08] phew [05:08] you can find lots of benchmark on internet [05:09] anywayI still dont get how to use this packe, so you are sayin apt-get install package name will be enough [05:09] The first one is about using a 32bit system as a whole, vs using a 64bit system as a whole. It specifically includes someone saying using a 32bit app on a 64bit system consumes extar RAM because of all the 32bit libs you load just for it... [05:09] soccerGuy: The benchmarks you just showed us showed amd64 LAMP systems outperforming i386... [05:10] hey guys, I will show my results after i figure out how to use this package [05:10] soccerGuy: I see nothing in those articles recommending running 32bit apache on 64bit Linux. [05:10] OK, have fun :) [05:11] soccerGuy: Anyhow. If you want to test 32-bit and 64-bit, I'd recommend testing a pure i386 install and a pure amd64 install. Doing anything bi-arch just to run a 32-bit apache will definitely never be a win. [05:11] infinity: Agreed, but soccerGuy apparently doesn't have the RAM for doing both at once on his VPS :) [05:12] :-) [05:12] jmarsden, did you install any 32bits application by using that package, if you do please just give me an example, thats all i need [05:13] apache is big package, so i want to install dnsutils [05:13] in 32 bits [05:13] soccerGuy: No -- you never provided me with the name of a 32bit-only application I could try out for you :) [05:13] can u give me command which i should run [05:14] dnsutils? [05:14] dnsutils is available as a 64bit package. [05:14] apt-get install dnsutils install 64 bits [05:14] they have 32 bits too [05:14] RIght. [05:14] Yes. [05:14] how can i install 32 bits one [05:15] If you are realy desperate, tweak your apt sources files to exclude the 64 bit one, I suppose. But this is a 100% pointless thing to be doing. [05:15] You can't. [05:15] We don't support multi-arch package installs. [05:15] If you want random things compiled for i386 on an amd64 system, you get to build them yourself at this point. [05:15] according to ia32-apt-get, it can do this job, isnt it jmarsden? [05:16] For packages which are in the 32bit world but NOT in the 64bit one, yes. [05:16] You have not gfiven us the name of such a package yet. [05:16] (You can dpkg --force-architecture, but it will drive apt insane later) [05:16] dnsutils isnt a package? [05:16] It has a 64bit version. [05:16] ok i think i almost have my mailserver setup correctly using dovecot/postfix/ssl [05:16] soccerGuy: You need to supply the name of something that is 32bit *only* for this to work. [05:17] i'm tailing my /var/log/mail.log and can see when i start thunderbird imap-login and it logs in my user...but i can't recieve mail [05:17] umm, jmarsden, I thought this ia32-apt-get can install 32 bits package eventough that package has 64 bits versoion, maybe i am wrong [05:17] soccerGuy: given the ridiculous number of filthy dependencies you seem to have, would it be easier for you to just run your proprietary ia32 app within a virtual environment? [05:18] i used to recieve regular mail on port 143, but now i'm trying to setup ssl so i forwarded 993 to my mailserver instead of 143, do i need to forward both ports? or just the ssl 993 port? [05:18] twb: In 256MB total RAM in a VPS? Are you serious :) [05:18] I am already in virtual enviroment:-) [05:18] jmarsden: sorry, I didn't see those stats. [05:18] soccerGuy: then ask for a second VPS ;-P [05:18] haha twb [05:18] biarch is still a massive pain on apt/dpkg. [05:19] My usual policy is that if a stupid proprietary app wants, say, RHEL 4.2, then just give it that and leave it the hell alone. [05:19] yes twb, I am agree with you, but this package claim that it can figure out dependencies, thats why i am interested [05:20] soccerGuy: ia32-apt-get is intended for (I think) cases where there is ONLY a 32bit version of something (say a flash plugin) ... if figures out the deps for that and grabs them for you. [05:20] soccerGuy: ia32-apt-get will be a hack; it certainly won't be perfect. This MUST be the case, because the underlying dpkg infrastructure doesn't support biarch properly. [05:21] soccerGuy: out of interest, what is this proprietary thing you're trying to install? [05:21] twb: Scroll back... he wants to install 32bit apache because he thinks it will use less RAM... [05:22] jmarsden: you are making fun with me [05:22] :-) [05:22] No, I'm serious.... you are asking us to help you do something that makes no sense at all. [05:22] I will prove you guys that is true once i install apache 32 bits on my server:-) [05:23] As I said earlier, have fun :) [05:23] thanks guys [05:23] I will be digging [05:23] jmarsden: really? ahaha [05:24] jmarsden: I thought he wanted 32-bit apache because he was installing something else, like vmware-server, that wanted to run (say) a 32-bit mod_idiotic. [05:24] Me too, at first... took a while for me to understand the reason behind the request. [05:25] soccerGuy: If you're really that concerned about RAM usage, why aren't you just reinstalling the whole systrem as 32-bit? [05:25] soccerGuy: (Still, the performance loss from dropping the amd64 extensions will cripple you far more than the extra RAM will help you, and I stand by that) [05:26] infinity: they only provide 64 bits [05:26] soccerGuy: Yes, so? [05:26] I think I need to use chroot [05:27] so how can i install application 32 bits? [05:27] soccerGuy: My colo host only provides CentOS. This doesn't stop me from debootstrapping Ubuntu, and doing a pivot_root install before I whack CentOS completely. :P [05:27] thanks for suggestion infinity [05:27] :-) [05:28] *shrug* [05:28] My contention is that if you don't understand bi-arch on Linux enough that you're asking the questions you are, what you're trying to do will only end in tears. [05:28] But, I wash my hands of it, and intend to enjoy a long nap. :P [05:29] we will see infinity [05:29] we will see [05:43] im trying to setup SSL in Dovecot, now does the ssl_cert_file have to be a .pem file? while creating my keys it spit out a .crt and .key not a .pem [05:56] billybigrigger: https://help.ubuntu.com/9.04/serverguide/C/dovecot-server.html [05:56] when i try to telnet localhost imaps i don't get * OK DOVECOT [05:56] jmarsden::: thanks bro, i'm reading the server guide [05:56] telnet does not decrypt the SSL for you :) [05:56] Boo, telnet. Yay for netcat/socat [05:56] i can see in my logs, im logging in ok [05:57] Jun 11 22:54:01 alixandra dovecot: imap-login: Login: user=, method=PLAIN, rip=192.168.1.101, lip=192.168.1.103, TLS [05:57] jmarsden: he might be using it without SSL [05:57] Apparently not... [05:57] i created a self signed cert, pointed dovecot.conf to them in my /etc/ssl dir, enabled ssl, and enabled plaintext_auth [05:58] just like the guide says [05:58] forward my router from port 143 to 993 [05:58] billybigrigger: Try openssl s_client -connect localhost:993 # instead of telnet [05:58] but i can't recieve mail, i can send mail, but im using my isp's smtp server for that, so never a problem there [05:58] booya [05:59] jmarsden::: that works good [05:59] That's why I suggested it :) [05:59] hehe [05:59] Now you can type IMAP commands at it to your hearts content to debug your IMAP server setup. [06:00] * billybigrigger googles imap command [06:00] BTW if that works then you have working SSL certs already (they may be self signed, and not for your domain, etc etc but they are working for SSL at some very basic level of "Working". [06:05] Can I use python scripting to read and write files on Ubuntu Server 9.04? [06:05] well [06:05] working with apache of course [06:05] like using python in the cgi bin? [06:06] hmm [06:06] i can't seem to pass any imap commands [06:06] select BAD Error in IMAP command received by server. [06:06] list BAD Error in IMAP command received by server. [06:14] If you don't understand the basics of IMAP why are you telnetting to an IMAP port in the first place? [06:14] Just exit out and use whatever way of debugging your setup you are most comfortable with. [06:16] You could use mutt or some other IMAP-capable email client if you prefer. [06:17] your losing me [06:17] im not an imap-guru :P [06:18] Then don't try to telnet to the imap port and type IMAP commands in the first place... I didn't suggest that, you did it... [06:18] Since the issue is not what you said it was (SSL cert files)... what *is* the actual problem? [06:18] i can't recieve mail [06:19] What happens when someone sends you email -- how far does it get? Does it end up in your mailbox on the server? [06:19] no [06:19] im assuming its being bounced back [06:20] assuming is bad. Where can I send a test email to that you think should reach you? [06:20] billybigrigger@thefrozencanuck.ca [06:20] i can send email from this pc to my hotmail account fine [06:20] but sending from hotmail, the mail hasn't shown up yet, it's been a good hour atleast [06:21] i shouldn't have to screw with my mx record should i? [06:21] You are not troubleshooting all that well, it would seem. I just sent you a test and my server logs show: Jun 11 22:20:44 eclipse5 postfix/smtp[16633]: 0B1895685EA: to=, relay=thefrozencanuck.ca[68.146.139.247]:25, delay=4.5, delays=0.01/0.04/2.9/1.6, dsn=2.0.0, status=sent (250 2.0.0 Ok: queued as DD185916E9) [06:22] So you received it fine, it is somewhere on the server at IP address 68.146.139.247 . [06:22] well then it must be somewhere else, i have been in the process of switching the mailserver from one pc to another [06:22] yeah [06:22] i have 2 mailservers behind that router [06:22] i've pointed the ports from cabo:143 to alixandra:993 [06:22] since the new server is using ssl [06:22] and the old one had no auth, just plaintext [06:22] which worked fine... [06:23] WHich server is listening on port 25 on 68.146.139.247 [06:23] bah, my bad [06:24] i forward the wrong port, i forwarded apache to alixandra not smtp [06:24] smtp was forwarded to the old server [06:25] When you fix that tell me and I'll send you another test email. [06:26] k [06:26] done a few minutes ago.... [06:26] now when i send from hotmail i get a Relay access denied [06:26] same with your test email [06:27] may i paste your email address here? [06:27] Right, now my logs show: Jun 11 22:26:45 eclipse5 postfix/smtp[15577]: 106E15692DD: to=, relay=thefrozencanuck.ca[68.146.139.247]:25, delay=1.7, delays=0.01/0/1.3/0.45, dsn=5.7.1, status=bounced (host thefrozencanuck.ca[68.146.139.247] said: 554 5.7.1 : Relay access denied (in reply to RCPT TO command)) [06:27] hmmm [06:27] well i have smtp relay setup to my isp [06:28] So your postfix config is in need of adjustment regarding smtpd_recipient_restrictions, I would think. [06:28] but i thought i had that configured for outgoing email.... [06:28] smtp relay is when *you* send... [06:28] ya thats what i thought [06:28] Your MX points to your server IP at 68.146.139.247 [06:28] So fix postfix on that server to allow me to send you email :) [06:30] well i see smtpd_recipient_restrictions = reject_unknown_sender_domain, reject_unknown_recipient_domain, [06:30] and some more [06:30] why would a default config reject unknown sender domain? [06:31] That's unlikely to be the problem... read the docs to understand what each restriction really does. [06:32] You may want to check your myorigin and mydestination lines in /etc/postfix/main.cf ? [06:33] myorigin = /etc/mailname [06:33] mydestination = alixandra, localhost.localdomain, localhost, $mydomain [06:33] seems ok [06:34] billybigrigger: To answer your reject unknown sender domain question: http://www.postfix.org/postconf.5.html#reject_unknown_sender_domain [06:35] billybigrigger: Is $mydomain your actual domain or is that literally what is. [06:35] this doesn't make sense, i've copied the postfix/dovecot configs over from a working server, and changed the forwarded ports to the correct pc... [06:35] no $mydomain was in there already [06:35] billybigrigger: What does hostname -d say ? [06:35] i just added alixandra [06:36] So far we are missing your actual domain name :) [06:36] cg.shawcable.net [06:36] ahh [06:36] That's not the same thing as thefrozencanuck.ca ... so fix your poxyfix config so I can send email to you ... [06:36] s/poxyfix/postfix/ [06:36] try now [06:37] Nope., still relay denied. What did you change and did you reload afterwards? [06:39] added alixandra.cg.shawcable.net to mydomains and restarted postfix [06:40] You asked me to send to billybigrigger@thefrozencanuck.ca. So thefrozencanuck.ca needs to be one of the domains you accept email for... and it still isn't. [06:40] try now [06:41] OK, you got it. Can you see it? [06:42] sent you a reply [06:42] that's really odd.... [06:42] im going to check the config for the old mailserver [06:43] myorigin = /etc/mailname [06:43] mydestination = alixandra, localhost.localdomain, localhost, $mydomain [06:43] thats from my oldserver [06:43] and incoming mail worked perfectly fine [06:43] On the old server, hostname -d # says what? [06:45] nothing [06:45] haha [06:46] Strange. Somewhere on that server, either in /etc/mailname or in the hostname you must have thefrozencanuck.ca for it to have worked. [06:46] hmmm [06:46] i don't recall setting that [06:46] but yes your right [06:47] But you received email over SMTP into that server for user@thefrozencanuck.ca on it? Well, anwyay... now you have a working email server. [06:47] What about hostname -f? [06:51] cabo [07:01] is there any ghosting software preinstalled on ubuntu server? [07:04] The dd command could be considered ghosting software, I suppose? [07:05] dd if=/dev/sda of=/dev/sdc # will do an image copy of disk sda to disk sdc, for example [07:31] New bug: #375149 in tomcat6 "tomcat6 needs debug start mode with jpda (dup-of: 375493)" [Undecided,New] https://launchpad.net/bugs/375149 === pschulz01 is now known as pschulz01_away [10:54] postfix logs - what rotates them? doesn't seem to be logrotate -(8.04) [11:09] hmm, found the info - in Debian/Ubuntu its syslog, not logrotate that does it for mail.log (see: http://www.ducea.com/2006/06/06/rotating-linux-log-files-part-1-syslog/) === pschulz01_away is now known as pschulz01 [12:03] Hi, iam having 300sec delay on *some* session_start() calls in php5 using ubuntu server (virtual esx setup). Any clues? :/ I am lost on this one [12:04] never had this problem on our non-virtual debian box [12:04] could someone take a look at this and hit me with ideas of how this could happen ? thanks http://paste.ubuntu.com/194251/ [12:12] anyone [12:12] ? [12:18] I've got a problem with pam_mkhomedir , it does not seam to create a homedir with pam_mount using NFS [12:18] I have no idea on how to debug either , i entered debug in pam-session but cant see anything about pam_mkhomedir [12:37] Appiah_: pam_mkhomedir and NFS sounds tricky. I'd use autofs to mount, but I'd also use something other than pam_mkhomedir to make the actual directories [12:39] well in pam_mount.conf.xml it does say [12:39] I think it's using pam_mkhomedir? either way none of them seams to work [12:41] drurew: apparently hwaddress line in interfaces is not working. And udev seems confused. Do you see any anomalies in kernel log? [12:46] cankoy:http://paste.ubuntu.com/194285/ [12:46] cankoy:Please complain to your hardware vendor. Switching to a random MAC. [12:46] Please complain to your hardware vendor. Switching to a random MAC. [12:52] peter_: Oh, hi :) [12:52] peter_: Anders just mentioned your name a minute ago and here you are :) [12:52] peter_: 300s timeout sounds suspiciously like a DNS thing. [12:53] soren: hi, thanks. and happy birthday :) [12:54] drurew: I guess a new eth interface is created everytime MAC is changed (check with 'ifconfig -a'). You need to find a way to stop that, specifically forcedeth driver seems to be the problem. [12:54] peter_: HAhah :) Thanks :) [12:56] http://xfliu.blogspot.com/2007/08/linux.html found that [12:56] im gonna source a driver [12:56] wish me luck [12:56] :) [13:01] drurew: that looks quite old, I suggest you search ubuntuforums, launchpad bug reports, etc before patching a kernel module. [13:02] you right [13:07] cankoy: undoing a mac spoofing .... this is gonna be hard [13:36] kirkland: sure, I'll add something to the virt section [14:04] Hi. Does anyone know how to split a internet connection let's say ip's from 192.168.1.2-192.168.1.20 to have 30% from bandwidth and rest 70% [14:04] ? [14:09] al_paun: using traffic shaping? [15:27] would we expect kvm to work on 8.04? [15:27] I'm starting a guest directly and kvm just seems to chomp 100% CPU and not really do anything [15:32] huh, having said that, it appears to do very much the same thing on karmic [16:36] Hi [16:44] I need to have a PHP file exectuted every 10 seconds. Is the best way to do that to write a script that runs 5 times and sleeps 10 seconds at every iteration and then have a cron job run that script every minute? The script downloads a bunch of xml files and stores them locally [16:44] i'd do it with cron [16:45] ivoks: does cron support 10 seconds run? [16:45] ivoks: hi btw [16:45] mathiaz: nope, 1 minute [16:45] ivoks: 1 minute is the lowest denominator :( [16:45] ivoks: IIRC cron only support the minute granularity [16:45] hi [16:46] oh, 10 seconds... right.. sorry :) [16:49] is it worth looking into writing a daemon? [16:49] well, technicaly, it's doable with cron :) [16:50] but that would be a wicked hack, not a solution [16:51] man cron: cron then wakes up every minute, examining all stored crontabs, checking each command to see if it should be run in the current minute. [16:51] right [16:51] so: [16:52] * * * * * sleep 1; do_somethinw [16:52] * * * * * sleep 2; do_something [16:52] * * * * * sleep 3; do_something [16:52] ... [16:52] would be a wicked hack :) [16:52] ivoks: are you using "wicked" as a positive is? :P [16:53] i think forking a script into background that uses an infinite loop and a sleep(10) which thus executes your command every ten seconds, would kill much less performance. [16:56] Hecate: thanks I'll look into it! :) [16:56] ivoks: if I can't get that script working I'll use that, thank you :) [16:56] :) [16:57] you're welcome [17:58] i did the "encrypt home directory" on install of Server 9.04. Is there a way to back out of that with out re-installing? [17:58] I found this: http://ubuntuforums.org/showthread.php?t=1135796 [17:59] but want to confirm before I really break something [18:04] jdstrand: hey - who sponsored ufw to debian? [18:05] mathiaz: kees [18:05] kees is my man ;) [18:05] * jdstrand waves to kees [18:06] jdstrand: when was the first release of ufw available in Ubuntu? hardy? [18:06] mathiaz: yes. See https://wiki.ubuntu.com/UbuntuFirewall for versions in Ubuntu [18:07] (and other stuff) [18:08] jdstrand: great - thank ya [18:08] np [18:10] troglobyte: *FIRST* you need to copy all your data out of the encrypted area, then ensure the user is not logged in, so that the encrypted layer is unmounted, and "mv .ecryptfs .ecryptfs_disabled" (as root, obviously) - "Le Rob" misspelt the name of the thing. Take note that some of the files required to mount private area are in fact stored in /var/lib/ecryptfs/USERNAME (to which ~/.ecryptfs is a symlink) [18:13] maxb, appreciate it! [18:14] stupid question about encrypted home dirs -- does root have access to them? [18:14] not if user isn't loged in [18:15] but yes, when they are [18:15] has anyone ever heard of an opensource timesheet program [18:33] jmedina: you around? [18:33] Sam-I-Am: yeap [18:34] so i've figured out that using launchpad PPAs is near impossible [18:34] they dont allow binary uploads [18:34] lol [18:34] point of ppa is to build source [18:34] it makes my dependencies a huge mess which means each thing i'm building almost needs its own PPA [18:35] ivoks: i viewed it as a place to host source and binary packages [18:35] not compiled somewhere else where i can't monitor exactly how they're getting built [18:35] for example, i had to manually install some libraries (.deb) from jaunty on hardy... i didnt build them, they just installed fine. [18:36] however, launchpad wont know what to do [18:36] so it looks like i also need to build all those libraries [18:36] i build in chroots so my packages don't wind up having interdependencies... which means i'll need to upload all the same source dependencies for everything i'm building (or something like that) [18:37] jmedina: so if you want these packages either it'll have to wait until i can wrangle launchpad into working... or i'll put up a web server and you can just grab them there. [18:40] Sam-I-Am: Having hardy packages depend on binaries from jaunty kinda defeats the whole purpose of a PPA. If you need to backport jaunty versions to make it work, then do that in the PPA as well. [18:40] infinity: thats what its looking like [18:40] (Nevermind defeating the purpose of a PPA, having an older suite depend on a newer one defeats the purpose of suites...) [18:40] well, for example, theres no libcap2 in hardy [18:40] but the jaunty one installs fine [18:41] And what's wrong with building against libcap1? [18:42] I suspect it's just a matter of s/libcap2-dev/libcap-dev/ in your Build-Depends, and boom, it builds on hardy. [18:42] well, some things work like that [18:43] an example of one that doesnt is samba and openldap [18:43] they hork with older libraries [18:43] Things that don't work like that require backporting either the code to work with older libraries, or the newer libraries to the older suites, yes. [18:43] Again, that's kinda the point. [18:43] Telling me that I can only use my 2003 car with gas released in 2006 isn't helpful. :) [18:44] i try to minimize backporting dependencies [18:44] thats why i'm using chroots... so something that builds find against oldlibversion doesnt magically find newlibversion and compile against it. [18:48] just talked with some folks in #launchpad ... think i figured out how i need to work this... might just be more complex than initially determined. [18:55] Installing binaries built for jaunty into a hardy system is somewhat dodgy. Sometimes it'll work, yes, but each case needs careful individual verification [18:56] i usually test those first [18:56] its few and far between [18:56] Generally people prefer to rebuild the jaunty source in a hardy environment (with a suitably ~suffixed version number) [18:57]