| slangasek | heh | 00:01 |
|---|---|---|
| hwilde | so is anybody looking at usb device reassignment for intrepid ? everytime I reboot the ttyUSBs are all random | 00:01 |
| hwilde | I have to run a script in rc.local to try and sort them out and create symlinks | 00:01 |
| slangasek | not that I'm aware of | 00:01 |
| hwilde | yeah I can't find any bug posts or anything even related | 00:02 |
| slangasek | in general, device names are not guaranteed to be persistent across reboots unless you use udev rules to link them | 00:02 |
| slangasek | I don't know about USB ttys though, I've never even had one of those let alone two | 00:02 |
| hwilde | right, but udev doesn't seem to know the diff | 00:02 |
| === johanbr_ is now known as johanbr | ||
| === gnomefre1k is now known as gnomefreak | ||
| hwilde | slangasek, you know udev ? | 00:11 |
| hwilde | slangasek, if you can make udev rules that figure out these usbs you would be my hero http://pastebin.com/m5ac22ce0 | 00:13 |
| slangasek | I only know udev on a cargo-cult basis | 00:13 |
| hwilde | where can I find the udev devs :) | 00:20 |
| slangasek | well, the Ubuntu udev maintainer just timed out of the channel... :) | 00:20 |
| hwilde | doh! | 00:20 |
| hwilde | hey probably read my pastebin and locked up | 00:21 |
| lucent | slangasek: ah, is there no udev/DBUS/HAL script to use to kick network-manager into gear before you login? | 00:22 |
| slangasek | lucent: the only keystore network-manager currently supports is gnome-keyring | 00:22 |
| slangasek | which makes logging in to password-protected networks ... problematic | 00:23 |
| lucent | oh gosh. | 00:23 |
| lucent | gnome isn't even there yet, so uh, I see now. | 00:23 |
| lucent | it would require some kind of "IMEH CHARGIN MY LAZERZ" followed by udev/DBUS/HAL hack | 00:23 |
| lucent | that's not a typical use case for a desktop system, is it? | 00:24 |
| slangasek | are you defining "desktop" as "running GNOME", or as "not a laptop"? | 00:25 |
| slangasek | because roaming Kerberos authentication is totally a valid use case | 00:25 |
| hwilde | the problem with udev is that it only sees the USB_BUS and USB_DEV and those swap every reboot... udevinfo does not have enough specific info to define custom rules http://pastebin.com/m5ac22ce0 | 00:25 |
| lucent | I'm defining wifi on boot as the use case | 00:25 |
| lucent | how common is that anyways? | 00:25 |
| hwilde | wifi on boot pre-login | 00:26 |
| slangasek | lucent: not as common as it ought to be, but it's *my* chosen configuration. :) | 00:26 |
| slangasek | Kerberos is pretty important at the enterprise level, both in the context of AD and for more traditional Kerberos setups | 00:28 |
| lucent | there's no sensible default for this IMO | 00:28 |
| slangasek | sure there is: have a keystore that doesn't have a password, that root is allowed to read | 00:28 |
| lucent | unless you're going to hack GDM to have a wifi selector | 00:28 |
| slangasek | they're friggin' wep keys, not state secrets | 00:28 |
| slangasek | (unless you work for the goverment, then, ok, maybe the wep keys are state secrets) | 00:29 |
| lucent | it's not just the keystore though that is a problem | 00:29 |
| lucent | it's the usability | 00:29 |
| hwilde | you could have two step login, first local to start the network support, then over AD or whatever | 00:29 |
| lucent | what if you have a wep + VPN that you log into before getting network access to the domain? | 00:29 |
| slangasek | <shrug> I don't consider "which network should we connect to?" to be a per-user setting | 00:30 |
| slangasek | so I don't think it should use per-user password stores | 00:30 |
| slangasek | hwilde: this is contrary to the goal of Single Sign-On | 00:30 |
| lucent | it's per-user because of the use of keystores | 00:30 |
| hwilde | but hten how does admin login to the secure network when you break your enterprise login :) | 00:30 |
| slangasek | hwilde: I *can* do that, but it's bloody annoying | 00:30 |
| hwilde | slangasek, consider me the devil's advocate then until I have consistent usb devices :) | 00:30 |
| slangasek | heh | 00:31 |
| lucent | slangasek: IMO the course of action is a new package which includes a boot-time configuration interface for networking | 00:31 |
| slangasek | I don't want to have to interact at all with the system in order to get it to come up on the network either, if that's what you mean | 00:32 |
| lucent | specifically "extra weird" networking like wifi on boot, or vpn tunnels | 00:32 |
| slangasek | system password store + n-m --> bring the network up without me having to touch it | 00:32 |
| slangasek | n-m actually already does this for interfaces that don't require passwords | 00:33 |
| slangasek | but if it's a password-encrypted network, you have to log in and run the applet to get anywhere | 00:33 |
| lucent | secured network access on boot without user input? That's not security! | 00:33 |
| slangasek | iz broken | 00:33 |
| slangasek | er | 00:33 |
| hwilde | might as well autologin at that point | 00:33 |
| slangasek | sure it is | 00:33 |
| slangasek | um no | 00:33 |
| lucent | seriously, hwilde | 00:33 |
| lucent | let's think this through here | 00:34 |
| lucent | how do we provide secured network access on boot, before gdm | 00:34 |
| lucent | so that gdm can login to resources on a secured network | 00:34 |
| hwilde | if this is restricted to enterprise networks, there should only be one network to connect to. | 00:34 |
| slangasek | hwilde: ahem, you already admitted above that you do the same thing on your system, only without the use of n-m | 00:34 |
| slangasek | there's no reason it would be restricted to enterprise networks | 00:35 |
| hwilde | my system is an autonomous robot so yeah it starts up automagically. | 00:35 |
| slangasek | it's restricted to machines that are part of an enterprise *environment*, which may include both enterprise wifi and public roaming | 00:35 |
| hwilde | slangasek, but there is only *one* network in that enterprise environment that you need pre-login connectivity aka where is the active directory | 00:36 |
| slangasek | i.e., if I have a company-issued laptop, I want to get Kerberos authentication at the gdm screen, whenever I'm connected to a network | 00:36 |
| slangasek | hwilde: no, why would you assume I would only use Kerberos when I'm directly connected to the network? :) | 00:36 |
| hwilde | I dunno, I guess I am missing why everyone objects to the method you have worked out | 00:37 |
| slangasek | it's my single sign-on, and I use it for everything from verifying my own password (company laptop, remember), to accessing fileshares (when I'm in range), to authenticating my email client | 00:37 |
| hwilde | if you can't login, you can't have access to the network, so why not startup networking before login? | 00:37 |
| slangasek | hmm? I don't claim that everyone objects to that | 00:38 |
| slangasek | it just doesn't work currently with n-m | 00:38 |
| hwilde | ohhhh | 00:38 |
| slangasek | because no one's implemented a system-level password store, because GNOME people think about these things upside-down. :) | 00:38 |
| lucent | I'm asking how do we provide secure network layer access before gdm so we can login | 00:39 |
| hwilde | typically they respond to popular demand | 00:39 |
| hwilde | but that is an important piece to taking market share away from windows in enterprise environments | 00:39 |
| lucent | maybe a power-on password? | 00:39 |
| lucent | USB thumbdrive as a key? | 00:40 |
| slangasek | lucent: if you really have a network that only a subset of your users are allowed to connect to, fine -- don't put that key in the system password store? :) | 00:40 |
| hwilde | what if you use an encrypted rsa/dsa key to authenticate to the network at first, then you can to kerberos login? | 00:40 |
| slangasek | and then it would work the same way that it works today | 00:40 |
| lucent | handing out physical layer access to your corporate network is not very bright | 00:40 |
| lucent | there should be a way to add a hook for VPN access too | 00:41 |
| slangasek | handing out physical access to your corporate laptop is probably not very bright either then ;) | 00:41 |
| lucent | shit gets stolen. I mean, things happen | 00:41 |
| hwilde | encrypted partition | 00:42 |
| lucent | Okay, encrypted partition, good. Now you need a password at boot to access your system? | 00:42 |
| hwilde | no the encrypted partition has the network key that gets online pre-login | 00:42 |
| slangasek | in that case, yes | 00:42 |
| hwilde | then the user can authenticate with AD or kerberos or whatever | 00:42 |
| lucent | in this scenario, having a package which implements a boot-time selectable network pre-config makes sense | 00:42 |
| lucent | without a key | 00:43 |
| lucent | instead you still need a means to be able to take user input though | 00:43 |
| lucent | like the VPN passphrase | 00:43 |
| lucent | unless you're storing VPN passphrases (not possible with OTP and not likley to be permitted by corporate) | 00:43 |
| lucent | the problem remains that you need user interaction to get network connectivity before you log in | 00:44 |
| slangasek | having to VPN in before you log in == doesn't work with Windows AD member machines either | 00:44 |
| lucent | slangasek: I am not familiar with Windows AD, what is this? | 00:45 |
| slangasek | AD == Active Directory? | 00:45 |
| hwilde | consider yourself lucky... | 00:45 |
| lucent | oh okay | 00:45 |
| hwilde | gtg, if you see Keybuk or any udevs ask them if mine is a lost cause: http://pastebin.com/m750bddea | 00:46 |
| slangasek | well, hmm, I don't /think/ you can start a VPN connection from Windows before logging in, but maybe you can run the VPN client as a system service too | 00:46 |
| lucent | The only big corp I worked for that had any level of secure login was Google, Inc. | 00:46 |
| lucent | so... not a lot of Microsoft anything | 00:48 |
| lucent | VPN pre-login would be a cool enhancement | 00:48 |
| slangasek | right, I won't ask you about what kind of VPNing they do, there are other people on the channel I can bug about that who probably have a better chance of being able to legitimately tell me :) | 00:49 |
| slangasek | but Kerberos is certainly a class of technology that's in Google's ballpark | 00:49 |
| lucent | I was a datacenter Monkey, I can't tell you any more than this :) | 00:50 |
| === cr3_ is now known as cr3 | ||
| * toresbe thinks he's found what should've been filed as an RC bug... | 01:16 | |
| toresbe | apt-get upgrade to hardy breaks on python-irclib | 01:17 |
| lucent | uh... | 01:17 |
| lucent | aren't we supposed to use the upgrade-manager? | 01:17 |
| toresbe | sorry. "upgrade-manager" breaks on python-irclib. :) | 01:18 |
| toresbe | as do subsequent invocations of dpkg --configure -a. | 01:18 |
| sistpoty | toresbe: any details? (e.g. output of "upgrade-manager" *g*)= | 01:19 |
| wgrant | toresbe: Let me guess... python-central complains about files being in the directory that aren't owned by it?" | 01:20 |
| toresbe | wgrant: won't remove local files, something to that effect | 01:20 |
| wgrant | It's like the python-opengl one I found; it needs C/R on python2.[34]-irclib | 01:20 |
| sistpoty | hey wgrant... where's fujitsu? *g* | 01:20 |
| wgrant | Right, that's right. | 01:20 |
| wgrant | sistpoty: Back in Hardy. | 01:20 |
| sistpoty | heh | 01:20 |
| wgrant | toresbe: Please file a bug, and I'll get it SRUed shortly. | 01:21 |
| toresbe | I'm in console mode trying to prepare for a reboot (my grub setup is kinda' awkward and roundabout) | 01:21 |
| jdong | whoopsies | 01:21 |
| jdong | note to self: | 01:21 |
| toresbe | wgrant: I don't know what an SRU is, but I'll do my best. | 01:21 |
| toresbe | :) | 01:21 |
| wgrant | toresbe: Stable Release Update. | 01:21 |
| jdong | do *NOT* make a loopback device on a LV and then use it to extend the LV itself | 01:21 |
| toresbe | wgrant: cool | 01:21 |
| toresbe | jdong: haha | 01:21 |
| wgrant | toresbe: For now, it's fairly easily recoverable. | 01:21 |
| jdong | well that was my attempt at perpetual motion. | 01:21 |
| toresbe | wgrant: I just removed the package, worked a treat ;) | 01:22 |
| wgrant | toresbe: apt-get remove python2.4-irclib | 01:22 |
| * lucent rolls through a Microsoft Windows XP install in kvm | 01:22 | |
| wgrant | python-central is so unbelievably fragile. | 01:22 |
| lucent | this is so difficult, versus Hardy install | 01:22 |
| wgrant | toresbe: Great. | 01:22 |
| toresbe | heh. :) | 01:22 |
| toresbe | anyway, the rest of it is working great. Thanks, everyone in here, for another neato release. | 01:22 |
| wgrant | python-central-related upgrade failures should be top priority, as they can bring down an entire upgrade very easily. | 01:23 |
| lucent | Hardy installer is missing a feature that is in Windows XP installer: It's not telling me that I'm an idiot. | 01:23 |
| lucent | :P | 01:23 |
| toresbe | haha | 01:23 |
| wgrant | lucent: Heh. Indeed. | 01:23 |
| toresbe | I'm thoroughly bummed about the Heron T-shirt being limited-edition. It's so good-looking I'd like one for myself. | 01:23 |
| wgrant | toresbe: They had Gutsy ones until very near Hardy's release. | 01:23 |
| wgrant | toresbe: I'm sure you can get one. | 01:23 |
| lucent | oh, I agree toresbe, that design is ill (really good) | 01:23 |
| toresbe | lucent: straight up, yo. | 01:24 |
| lucent | I was looking and thinking, if I had money right now, I would buy a T and wear it to burning man | 01:24 |
| lucent | I don't know, does it have "Ubuntu" words on it? I don't want the words if I'm wearing something there | 01:24 |
| lucent | the design is really cool and I like that | 01:24 |
| * toresbe wonders if he has met lucent at debconf, which would explain why he felt the need to explain "ill" to him. | 01:25 | |
| toresbe | :) | 01:25 |
| lucent | har, no *nix gatherings since a long time now | 01:25 |
| wgrant | toresbe: Wait, are you upgrading from Dapper? | 01:26 |
| lucent | I watch those things from a safe, moderate-intelligence required, distance. They're really full of people who are overwhelmingly intelligent. | 01:26 |
| toresbe | lucent: Hey, I'm an idiot, I have a great time there. | 01:26 |
| === GBGames|away is now known as GBGames | ||
| toresbe | lucent: I do lots of stuff that require no computer skills. Last debconf, I fixed a church organ with duct tape. | 01:27 |
| * jdong moans about the word idiot's high standards these days | 01:27 | |
| toresbe | wgrant: no, Gutsy. | 01:27 |
| wgrant | Hmm, that's not what I thought it was, then. | 01:27 |
| lucent | toresbe: groovy! I'm into tinkering with other people's code and engineering work, never my own | 01:27 |
| wgrant | toresbe: Did it mention any file in particular? | 01:27 |
| toresbe | wgrant: nope... sorry, I probably should've made allowance for letting some devs poke around before I removed the package. | 01:29 |
| wgrant | toresbe: I should be able to reproduce it here. | 01:29 |
| toresbe | wgrant: cool. | 01:29 |
| * toresbe reboots, takes opportunity to stuff in a video capture card he's been wanting to try | 01:29 | |
| toresbe | up 11 days, pfah | 01:29 |
| wgrant | toresbe: Did you have python2.4-irclib installed? | 01:30 |
| toresbe | toresbe@fortran:~$ dpkg-query -f '${Status},${Package}\n' -W *irclib* | 01:39 |
| toresbe | purge ok not-installed,python-irclib | 01:39 |
| toresbe | install ok installed,python2.4-irclib | 01:39 |
| toresbe | wgrant: see above :) | 01:39 |
| wgrant | That's impressive. | 01:39 |
| wgrant | That should have been removed when you upgraded from Dapper. | 01:40 |
| wgrant | So it is the bug I thought. | 01:40 |
| toresbe | hah | 01:40 |
| toresbe | That bug is *so* three releases ago? :) | 01:40 |
| toresbe | Ugh! Sad to see that gnome-terminal hasn't gotten any less useless on this machine. Maybe it's a config file issue or something, but irssi screen redraws take several 1/10ths of a second. | 01:40 |
| toresbe | nearly a whole second | 01:41 |
| * toresbe goes back to uxterm | 01:41 | |
| toresbe | Oh, and another cute one... | 01:41 |
| wgrant | toresbe: I use gnome-terminal fine with screen here. | 01:41 |
| toresbe | it works well on my work Hardy machine with the same nVidia driver... probably just a weird issue. | 01:42 |
| toresbe | http://gunkies.org/stuff/hardy-ohdearme.png :) | 01:42 |
| wgrant | ... | 01:42 |
| wgrant | Nice. | 01:42 |
| wgrant | The /Music isn't just wrapped off the first one, is it? | 01:43 |
| toresbe | Good question. I'll kill gnome and try again | 01:43 |
| toresbe | Hrm. Could well be, but resizing is not possible. :) | 01:44 |
| toresbe | So most probably it's doing some kind of automated sizing which may or may not be to blame | 01:44 |
| toresbe | Actually, a quick question which might be construed as a user support question: Anyone here have any experiences with, and/or advice for a video4linux previewing application? | 01:46 |
| toresbe | I used to use tvtime, but this is just an svideo grabber. | 01:46 |
| LaserJock | wgrant: oh no, what happened to your nick?! | 01:52 |
| LaserJock | say it ain't so ;-) | 01:52 |
| wgrant | LaserJock: It got left in Hardy. | 01:52 |
| LaserJock | I thought maybe Colin and Tollef had pulled you to the dark side | 01:54 |
| wgrant | Not as far as I'm aware. | 01:55 |
| jdong | you'er not supposed to be aware | 01:55 |
| jdong | like that pill thingie they use in night clubs | 01:55 |
| wgrant | Hahah. | 01:55 |
| slangasek | pill... thingie? | 02:10 |
| slangasek | nevermind, I didn't ask | 02:10 |
| sladen | *babies* | 02:11 |
| sistpoty | alcohol works as countermeasure, I've heard from s.o. :P | 02:12 |
| MattJ | Don't know if this is the right place to ask, but are there plans to have the .25 kernel in the repositories at some point? | 03:52 |
| wgrant | MattJ: For Intrepid, yes. | 03:55 |
| wgrant | For Hardy, no. | 03:56 |
| MattJ | Not even in backports? | 03:56 |
| lamont | wgrant: except that I expect it'll be .26 or later for intrepid... | 03:56 |
| wgrant | lamont: But it's 2.6.25 now. | 03:56 |
| lamont | sure | 03:56 |
| wgrant | MattJ: Backporting kernels is truly evil.l | 03:56 |
| MattJ | ^^ | 03:57 |
| MattJ | iwlwifi is not working for me, and I need it to | 03:58 |
| wgrant | Getting 2.6.25 is not the soluition to that. | 03:58 |
| wgrant | Filing a bug and getting it fixed is. | 03:58 |
| MattJ | I found someone with a similar/same problem on lkml, they say it is fixed in .25 | 03:58 |
| lamont | wgrand++ | 03:59 |
| MattJ | Since the issue is not a security issue, is it likely to be fixed in Hardy? | 03:59 |
| wgrant | lamont: Pfft, I thought it would be easier to type than Fujitsu. | 03:59 |
| wgrant | MattJ: Very probably for 8.04.1 | 04:00 |
| * MattJ files bug | 04:00 | |
| wgrant | A lot of kernel regressions are scheduled for fixing then. | 04:00 |
| MattJ | :) | 04:00 |
| wgrant | 8.04.1 should happen in very early July. | 04:00 |
| pwnguin | MattJ: the proper technique there is to bisect the kernel | 04:00 |
| MattJ | and any idea of the short term solution for me? Gutsy (with ipw3945) worked ok | 04:01 |
| wgrant | MattJ: File a bug and see what the kernel devs say. | 04:01 |
| MattJ | k, will do so now | 04:01 |
| wgrant | Perhaps use the Gutsy kernel for now. | 04:01 |
| MattJ | I used to use the Gutsy kernel on Feisty, how funny :) | 04:02 |
| wgrant | MattJ: Is there anything special about your config? I've used Hardy's kernel on several ipw3945 boxes recently, and it seems to work. | 04:03 |
| MattJ | ipw was the old (pre-Hardy) driver, Hardy uses the newer iwlwifi (from the kernel) | 04:04 |
| MattJ | Nothing special, this is a fresh install | 04:04 |
| MattJ | A lot of people have had problems even detecting networks, but using the hardy backports modules seems to help those | 04:05 |
| MattJ | My problem is that it won't set up an ad-hoc network | 04:05 |
| MattJ | which I use every day | 04:05 |
| MattJ | The hardest part of filing a bug is knowing which package to file under | 04:10 |
| wgrant | MattJ: linux, probably. | 04:10 |
| MattJ | That warns me about filing a bug for upstream linux, etc. | 04:12 |
| wgrant | MattJ: The linux package. https://launchpad.net/ubuntu/+source/linux/+filebug | 04:15 |
| MattJ | Ah, thanks :) | 04:15 |
| sladen | MattJ: I filed this as high about 6weeks before release | 04:19 |
| sladen | MattJ: but the more dups the better | 04:19 |
| MattJ | sladen: Oh, where? | 04:19 |
| MattJ | I couldn't find any with my specific problem | 04:20 |
| sladen | MattJ: file _your_ bug | 04:20 |
| sladen | if it's a dup, it'll get duped, but at least there will be an independent report in your words | 04:21 |
| MattJ | How do duplicates help? :) | 04:21 |
| MattJ | If you insist | 04:21 |
| dsas | MattJ: It helps persuade someone that the problem is important | 04:21 |
| MattJ | If it were my bug tracker I would be annoyed at someone knowingly filing a duplicate :) | 04:22 |
| sladen | some people do | 04:22 |
| sladen | for me (personally), a good indicator of the relative size of a problem are the number of dups | 04:22 |
| sladen | and with them you get separate individual reports (multiple angles onto a problem) and not just misleadings "me toos" are actually turn out to be something else and need weeding out | 04:23 |
| sladen | MattJ: [after you've filed yours], start at bug | 04:25 |
| sladen | MattJ: [after you've filed yours], start at bug #205390 and recursively follow the dups | 04:26 |
| ubotu | Launchpad bug 205390 in ubuntu "8.04 beta; wireless broken after upgrade (ipw3945/iwl3945/ifrename/udev) (dup-of: 183968)" [Undecided,New] https://launchpad.net/bugs/205390 | 04:26 |
| ubotu | Launchpad bug 183968 in udev "mac80211 "master" interface matches existant persistent network rules" [Critical,Fix released] https://launchpad.net/bugs/183968 | 04:26 |
| MattJ | sladen: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/222302 | 04:37 |
| ubotu | Launchpad bug 222302 in linux "iwl3945 can't set parameters for ad-hoc network" [Undecided,New] | 04:37 |
| MattJ | I don't have the wlan0_rename issue other people have | 04:38 |
| MattJ | I'll investigate installing the Gutsy kernel tomorrow | 04:43 |
| MattJ | Thanks for the help :) | 04:43 |
| hdevalence | has anyone ever heard of an idea where you would use a torrent to download debs for version upgrades instead of having everybody upgrade veeerrrryyy slowly when a new version is released? | 04:51 |
| hdevalence | because torrents of course scale opposite to http/ftp downloads, so if you did that upgrading would go much faster | 04:52 |
| mgolisch | ? | 04:53 |
| mgolisch | how would that be faster? | 04:53 |
| mgolisch | i mean you would need a dedicated torrent for each file | 04:53 |
| mgolisch | that would unmanageable | 04:53 |
| hdevalence | mgolisch: no | 04:53 |
| hdevalence | no, you don't | 04:54 |
| mgolisch | how else? | 04:54 |
| hdevalence | you can make a torrent of all the deb files and then only download the ones you want | 04:54 |
| mgolisch | if it where just one you would have to wait for the whole thing to complete | 04:54 |
| mgolisch | or did i get something wrong? | 04:55 |
| mgolisch | i mean you cant control in what order the chunks are downloaded | 04:55 |
| hdevalence | then how is it in KTorrent I can proitize files and choosenot to download them? | 04:56 |
| hdevalence | ick | 04:56 |
| hdevalence | prioritize files and choose not to | 04:56 |
| mgolisch | hm no idea maybe iam wrong | 04:56 |
| hdevalence | basically I'd be interested to know if it's been done, because it's something I think would be interesting to work on but I don't want to put in lots of effort if it's already made | 04:58 |
| awalton__ | hdevalence, it's been done, but never very successfully. a new implementation would be nice. | 05:01 |
| awalton__ | hdevalence, apt-torrent: http://sianka.free.fr/ | 05:02 |
| hdevalence | i was thinking it would be neat to try getting it to use the alternate install cd torrent | 05:03 |
| hdevalence | but I don't know if that's the best way | 05:04 |
| hdevalence | anyways, I need to sleep | 05:05 |
| nixternal | or you can use Conary and download only the files that have changed :) | 05:05 |
| === thegodfather is now known as fabbione | ||
| LaserJock | anybody around running Hardy and have gnumeric installed? | 05:35 |
| beuno | LaserJock, I can probably install it. What seems to be the problem? | 05:36 |
| LaserJock | pi doesn't worh | 05:36 |
| LaserJock | *work | 05:37 |
| LaserJock | bug #222062 | 05:37 |
| ubotu | Launchpad bug 222062 in gnumeric "functions don't work in formulas" [Undecided,New] https://launchpad.net/bugs/222062 | 05:37 |
| beuno | LaserJock, installing. Might take a minute or two while the repos try not to die with all the hammering they're getting these days | 05:38 |
| beuno | LaserJock, what are you doing exactly? =PI() in a cell? | 05:41 |
| LaserJock | hmm | 05:42 |
| LaserJock | I just used PI | 05:42 |
| beuno | just type in PI? | 05:42 |
| LaserJock | ah, PI() does work | 05:42 |
| beuno | it doesn't for me :) | 05:42 |
| beuno | oh | 05:42 |
| beuno | yes it dows | 05:42 |
| beuno | er, does | 05:43 |
| beuno | I went for pi() because I saw the word function somewhere :p | 05:43 |
| LaserJock | well, pi isn't a function | 05:43 |
| LaserJock | it's a constant | 05:43 |
| LaserJock | but ... hmm | 05:43 |
| beuno | right, I wouldn't of tried that if the bug didn't mention "functions" | 05:44 |
| beuno | LaserJock, http://www.gnome.org/projects/gnumeric/doc/gnumeric-PI.shtml | 05:44 |
| LaserJock | well, we need to figure out if it's a change in behavior | 05:45 |
| beuno | LaserJock, http://www.mailarchives.org/list/gnumeric-list/msg/2005/00002 | 05:47 |
| beuno | tehre are emails from 2005 | 05:47 |
| beuno | of users using pi() | 05:47 |
| beuno | (yes, my typing sucks today) | 05:47 |
| LaserJock | hmm | 05:47 |
| LaserJock | so what went wrong for the bug reporter I wonder | 05:47 |
| LaserJock | beuno: thanks for the help, I've followed up on the report | 05:49 |
| beuno | LaserJock, np | 05:49 |
| beuno | sometimes it just takes someone totally unaware of everything to look at it differently :) | 05:50 |
| beuno | I force random people at the office to look at code when I'm stuck | 05:50 |
| beuno | and it usually is something terribly obvious | 05:50 |
| edugonch_ | Hello, why I get compiler errors in this statement ----> this->_txt_Password->signal_insert_at_cursor().connect( | 06:15 |
| edugonch_ | sigc::mem_fun(*this, &LoginForm::onInsert_txt_Password)); | 06:15 |
| === asac_ is now known as asac | ||
| DaBonBon | someone told me that snd-hda-intel was kind of disabled in hardy, to convenience the other sound card drivers. is that true? | 07:15 |
| DaBonBon | specifically, --with-cards=hda-intel was not put during alsa configuration | 07:15 |
| infinity | DaBonBon: Can't imagine how that's possibly true, given than my hda_intel stuff is working great. | 08:11 |
| wgrant | Mine's working better than ever as well. | 08:14 |
| DaBonBon | strange, because my card is working really pathetic | 08:29 |
| wgrant | DaBonBon: Can you please be more descriptive? | 08:30 |
| DaBonBon | wgrant: yes, first of all, sound sometimes works, and sometimes doesn't. headphone automute doesn't work. i get too many channels in alsamixer | 08:30 |
| DaBonBon | all this used to happen pre 1.0.15 release of alsa, but after installing ubuntu-backport-modules in gutsy, which had alsa 1.0.15, all the problems were solved | 08:31 |
| wgrant | DaBonBon: Well, there are so many types of HDA cards that it's not surprising. | 08:31 |
| DaBonBon | now again, in hardy, i get those problems | 08:31 |
| wgrant | Please file a bug. | 08:31 |
| DaBonBon | wgrant: no, this card if fully supported | 08:31 |
| wgrant | Apparently not. | 08:31 |
| DaBonBon | wgrant: well, someone told me that snd-hda-intel has been deliberately disabled to facilitate other sound cards | 08:31 |
| DaBonBon | wgrant: yes, because it works perfectly on alsa 1.0.15, atleast in gutsy and other distros | 08:32 |
| wgrant | That sounds very odd, as Intel HDA is one of the most common types these days... | 08:32 |
| DaBonBon | exactly! | 08:32 |
| wgrant | And how on earth would disabling something facilitate support for other card? | 08:32 |
| * laga hands out tinfoil hats | 08:33 | |
| DaBonBon | i guess i'll file a bug | 08:33 |
| DaBonBon | wgrant: here is a part of the logs -- | 08:33 |
| DaBonBon | http://rafb.net/p/LQQ7ff80.html | 08:34 |
| DaBonBon | i don't know till how much extent is it true | 08:34 |
| wgrant | Ah. | 08:35 |
| wgrant | Well, file a bug, anyway. | 08:35 |
| DaBonBon | wgrant: what package should i file it against? | 08:38 |
| wgrant | DaBonBon: I think it's in linux-ubuntu-modules-2.6.24, but possibly linux. | 08:40 |
| DaBonBon | ok, i must leave.. thanks for the help, wgrant .. | 08:48 |
| Trewas | is hardy-proposed supposed to be completely empty (I am looking at bug 208666 and fixed audacious)? | 10:39 |
| ubotu | Launchpad bug 208666 in xmms-crossfade "audacious crashed with SIGSEGV in g_type_check_instance_cast()" [Medium,Fix committed] https://launchpad.net/bugs/208666 | 10:39 |
| jeromeg | Trewas: why would it be supposed to be empty ? | 10:43 |
| Trewas | well, it seems to be empty :) | 10:43 |
| jeromeg | Trewas: what mirror are you using ? | 10:44 |
| Trewas | I was lookign at http://archive.ubuntu.com/ubuntu/dists/hardy-proposed/ | 10:45 |
| === azeem_ is now known as azeem | ||
| jeromeg | Trewas: https://launchpad.net/ubuntu/+source/xmms-crossfade/0.3.14-1build1 | 10:47 |
| jeromeg | it has been built only a few ours ago | 10:47 |
| jeromeg | it should be available soon in the archive | 10:47 |
| Trewas | ok, I noticed that the last message in that bug was only a few hours old so I wouldn't have been surprised if audacious was not yet there, but as hardy-proposed is completely empty I wondered if some switch had not yet been flipped or something | 10:50 |
| jeromeg | Trewas: maybe it's because hardy was released two days ago ;) | 10:52 |
| jeromeg | hardy-proposed is to handle SRU | 10:52 |
| Trewas | well yes, but the bug mentions that people should test the new package from hardy-proposed :) | 10:52 |
| jeromeg | yes, yes | 10:52 |
| jeromeg | it will be available soon | 10:53 |
| jeromeg | or it should at least ;) | 10:53 |
| Trewas | hehe, I'll wait and see | 10:53 |
| beniamino | is there *any* documentation on update-manager? /usr/share/doc contains only a one-line readme on update-manager-core | 10:55 |
| === andrew_ is now known as PixelSmack | ||
| === cprov is now known as cprov-out | ||
| emgent | heya | 12:34 |
| twi_ | what happened to binary-ppc on the mirrors? | 13:34 |
| geser | wasn't ppc moved to ports.ubuntu.com? | 13:36 |
| twi_ | ah | 13:37 |
| crimsun | asac: a lot of people seem to be clamoring for simply enabling nspluginwrapper for i386. What are your thoughts on doing that for 192888 instead of adjusting pulseaudio? | 14:10 |
| crimsun | asac: (for hardy-proposed) | 14:10 |
| ion_ | ubotu: A link to bug #192888 for the lazy bastards among us, thank you. | 14:23 |
| ion_ | Okay then. | 14:23 |
| Amaranth | bug 192888 | 14:28 |
| ubotu | Launchpad bug 192888 in pulseaudio "firefox crashes on flash contents when using libflashsupport" [High,In progress] https://launchpad.net/bugs/192888 | 14:28 |
| Amaranth | I would love to get nspluginwrapper on i386, flash screws up all the time even without pulseaudio | 14:29 |
| ogra | not for me ... with the first proposed fix to drop hal | 14:30 |
| ion_ | amaranth: <metoo/> | 14:30 |
| ogra | (pulse-hal stuff indeed) | 14:30 |
| johanbr | Is https://launchpad.net/bugs/211205 marked private or something? The error message I get is not very informative. | 16:48 |
| stgraber | johanbr: this bug contains a backtrace so only QA members can read it | 16:49 |
| johanbr | Apparently it is marked private. At least the bot just told me so. | 16:49 |
| johanbr | stgraber: Maybe it'd be a good idea for launchpad to actually say that. | 16:49 |
| johanbr | And should private bugs really be referenced in a public changelog? | 16:50 |
| tuxice | features in intrepid? | 16:51 |
| tuxice | !intrepid | 16:51 |
| ubotu | Intrepid Ibex is the code name for Ubuntu 8.10, due October 2008 - For more info, see: https://wiki.ubuntu.com/IntrepidIbex | 16:51 |
| ion_ | It supports keyboards. | 16:51 |
| tuxice | huh? | 16:53 |
| ScottK | johanbr: It's not private any more. | 16:56 |
| johanbr | ScottK: thanks | 16:58 |
| ScottK | You're welcome. | 16:59 |
| kennethr | when I click on the clock in gnome-panel, gnome-panel freezes. How do I debug? | 17:10 |
| kennethr | How do I troubleshoot a hanging gnome-panel on ubuntu? | 17:13 |
| === kitterma is now known as ScottK2 | ||
| johanbr | kennethr: Change the session property for the panel from respawn to normal, then kill the panel and start it under gdb or something. | 17:18 |
| kennethr | johanbr: thanks, I'll start there. | 17:19 |
| kennethr | johanbr: how do I make use of the gnome-panel-dbg package? Is it usefull in this case? | 17:22 |
| johanbr | Yes. You should have that installed to get a proper backtrace. | 17:22 |
| johanbr | kennethr: And probably also libgtk2.0-0-dbg and libglib2.0-0-dbg at least. | 17:23 |
| kennethr | johanbr: when I get into gdb and say run, then I reproduce the hang, how do I get back to the gdb prompt to get a backtrace? | 17:25 |
| johanbr | Try just ctrl-c. | 17:25 |
| kennethr | perfect...thanks | 17:26 |
| scorcher7 | Hi, I just uploaded a patch to launchpad that fixes launchpad bug #173772 in atomix. I have never written a patch before and the wiki says to find a dev. to review the patch for inclusion. | 17:30 |
| ubotu | Launchpad bug 173772 in atomix "about dialog won't close" [Undecided,Confirmed] https://launchpad.net/bugs/173772 | 17:30 |
| scorcher7 | Is that something one of you can help me with? | 17:30 |
| hunger | scorcher7: You have come to the right place, but I guess most devs are out for the weekend. | 17:36 |
| scorcher7 | hunger: Should, I just come back on Monday and ask the same question? | 17:39 |
| hunger | scorcher7: You might get lucky and somebody might come around, but your chances are way better during the week. | 17:39 |
| scorcher7 | hunger: Thanks for the tip. | 17:41 |
| hunger | scorcher7: You are welcome. | 17:43 |
| gustavold | hi, I dont know if it is offtopic... I've upgraded from gutsy to hardy... now every time I push a button in the numpad the X crashes | 19:36 |
| Ng | mdke: does the doc team want to know about mistakes in gnome documentation, or should I go and hassle the gnome folks? :) | 19:38 |
| dsas | Ng: Hassling the gnome folks is appreciated. You can always file a bug on the gnome documentation packages and it *may* get fixed here. (at least I think that happened in at least one release) | 19:43 |
| Ng | dsas: okidoki | 19:43 |
| dsas | perhaps gutsy we did a little, temporary fork | 19:43 |
| * Ng wonders how keen they will be to update docs for a point release | 19:45 | |
| ffm_ | What's the proper way to make a spec? | 19:54 |
| ffm_ | Should I just add a page on the wiki, or should I start a forum thread about it? | 19:54 |
| === ffm_ changed the topic of #ubuntu-devel to: Ubuntu 8.04 LTS released! | Development of Ubuntu (not support, not application development on Ubuntu) | #ubuntu for support and general discussion for dapper/edgy/feisty/gutsy, #ubuntu+1 for intrepid | #ubuntu-motu for getting involved in development | http://wiki.ubuntu.com/UbuntuDevelopment | See #ubuntu-bugs for http://wiki.ubuntu.com/HelpingWithBugs | ||
| LaserJock | ffm_: a wiki page is good | 19:56 |
| ffm_ | LaserJock, Care to look at the one I've made? FontInstallSpec. (I'm not probably able to implement it, I only know Python ATM) | 19:57 |
| Nafallo | hmm | 19:57 |
| Nafallo | #ubuntu have support for hardy surely? | 19:57 |
| Nafallo | but maybe not edgy. | 19:58 |
| ffm_ | Nafallo, What? | 19:58 |
| LaserJock | Nafallo: yeah, that should be updated | 19:58 |
| Nafallo | yay! edgy is EOL today :-) | 19:59 |
| === ffm_ is now known as ffm | ||
| Kim^J | Nafallo: Yo! | 20:15 |
| dmsuperman | What kind of libraries/examples exist that I could use to learn to write a keylogger/keystroke sender app in either C#, PHP, or (preferably not) C++ | 20:17 |
| Kim^J | dmsuperman: The question is, why? | 20:18 |
| Kim^J | And I think you need to use some Xlib. | 20:18 |
| dmsuperman | Have you ever used AutoHotKey in windows? | 20:18 |
| Kim^J | dmsuperman: Yes. | 20:19 |
| dmsuperman | I basically want to emulate that, at least the text expansion portion | 20:19 |
| dmsuperman | there aren't any good existing solutions | 20:19 |
| dmsuperman | at least that I could find after much googling | 20:19 |
| Kim^J | dmsuperman: Check out xbindkeys source code and check what they use. | 20:19 |
| Kim^J | I think you WILL need to learn some C/C++ and use that. | 20:19 |
| mdke | Ng: it's probably helpful to file a bug on the ubuntu gnome-user-docs package too though, so that we can fix it in ubuntu then pass the fix upstream | 20:20 |
| dmsuperman | Alright. I learned the basic syntax of C++ quite a while ago, and I know C# now, so it shouldn't be too much of a leap I hope | 20:21 |
| Kim^J | Heh... | 20:21 |
| Kim^J | How much in C# do you know? (The more, the worse.) | 20:21 |
| dmsuperman | Not much | 20:22 |
| dmsuperman | I'm mainly a PHP guy, but I'm taking a few classes in C# | 20:22 |
| Kim^J | Good, smaller leap then. | 20:22 |
| dmsuperman | but over the years I've dabbled in various compiled languages, so i'm no stranger | 20:22 |
| dmsuperman | yeah | 20:22 |
| Kim^J | Well, get to work then. :) | 20:22 |
| dmsuperman | I take it C++ is the common denominator language in linux programming? | 20:22 |
| Kim^J | dmsuperman: Naaah, C is. | 20:23 |
| dmsuperman | ah | 20:23 |
| LaserJock | C/C++ | 20:23 |
| Chipzz | dmsuperman: errr, xlib programming is mostly C programming | 20:23 |
| Kim^J | LaserJock: Well, moslty C. | 20:23 |
| Chipzz | not much C++ involved | 20:24 |
| Chipzz | if any at all | 20:24 |
| LaserJock | Kim^J: there's an awful lot of C++ out there | 20:24 |
| highvoltage | python must count for something too? or are you excluding interpreted languages? | 20:24 |
| LaserJock | and it's very similar to C | 20:24 |
| dmsuperman | I'd love to do it in PHP but I doubt it's very possible | 20:24 |
| Kim^J | LaserJock: I know, I code C/C++ every day. :) | 20:24 |
| Kim^J | highvoltage: For the thing he's doing, Python might not be the right choice. :) | 20:25 |
| LaserJock | so I usually put them together in any case | 20:25 |
| LaserJock | Kim^J: why not? | 20:25 |
| LaserJock | there are a lot of python libraries | 20:25 |
| Kim^J | Well, it looks like there's a xlib for Python. | 20:28 |
| LaserJock | python would be more likely than PHP or C# anyway :-) | 20:28 |
| dmsuperman | I don't know python though :( | 20:28 |
| dmsuperman | is it difficult to pick up coming from something like PHP? | 20:28 |
| LaserJock | it's not hard to pick up | 20:28 |
| Kim^J | dmsuperman: Good for you! (I think so, yes.) | 20:29 |
| dmsuperman | Kim^J, You say good because I _don't_ know it? | 20:29 |
| LaserJock | python is probably about the easiest programming language to learn, as a gross generalization | 20:29 |
| Kim^J | dmsuperman: Yes, because the less people know Python, the less programs get programmed in Python. Which I think is great. | 20:29 |
| ffm | Wait, is this a VHLL/HLL flamewar? | 20:30 |
| * ffm wants in. | 20:30 | |
| dmsuperman | no | 20:30 |
| dmsuperman | wait | 20:30 |
| dmsuperman | stop | 20:30 |
| Kim^J | VHLL/HLL? | 20:30 |
| dmsuperman | I'm gonna learn python now haha | 20:30 |
| ffm | Kim^J, Very High Level Language (Python) vs High Level Language (C) | 20:31 |
| Kim^J | Very-High-Level-Language/High-Level-Language? | 20:31 |
| Kim^J | ffm: Ah.. :) | 20:31 |
| ion_ | Python, ew. :-) </flame> | 20:31 |
| dmsuperman | psh | 20:31 |
| dmsuperman | I compile my code on paper | 20:31 |
| Kim^J | Naah, it's just that me and Python doesn't get along, I miss alot of things from C++, things doesn't behave the way I want it to, etc. | 20:31 |
| ffm | How hard is it to add functionality to a C gui app using python? | 20:31 |
| ffm | I know very little C, but have been meaning to learn it so I can contribute more to Ubuntu.. | 20:32 |
| Kim^J | I don't what to contribute too. | 20:32 |
| ion_ | jdong: Your hostname is shouting at me. :-) | 20:33 |
| highvoltage | how rude. | 20:33 |
| ffm | Kim^J, Well, you can implement my new spec... | 20:33 |
| LaserJock | ion_: he's an elitist MITer apparently ;-) | 20:33 |
| ffm | ;) | 20:33 |
| Chipzz | python indeed is very easy to learn - but make sure that you use consistent indentation settings across editors of you use more than one | 20:33 |
| Kim^J | ffm: Where? | 20:33 |
| dmsuperman | holy crap, my alternate install cd iso just peaked at 28mbps. I have advertised 15mbps connection 8-) | 20:34 |
| ffm | Kim^J, https://wiki.ubuntu.com/FontInstallSpec . It's very new (not even in launchpad) | 20:34 |
| Kim^J | Chipzz: There's only one editor! (But I won't tell since it's going to start a flamewar.) | 20:34 |
| Chipzz | Kim^J: no need to do so - we all know it's vim ;) | 20:34 |
| highvoltage | *nod* | 20:34 |
| * Chipzz ducks and runs away :)P | 20:34 | |
| LaserJock | emacs forever!! | 20:35 |
| jdong | ion_: lol I'm sorry to hear that ;-) | 20:35 |
| dmsuperman | Eclipse ftw | 20:35 |
| LaserJock | oh wait, I'm in vim | 20:35 |
| Kim^J | Chipzz: Yep | 20:35 |
| * dmsuperman runs | 20:35 | |
| Kim^J | ffm: Well, I don't know to much about fonts and stuff, where can I look for more info? (You simply want to move the file right?= | 20:35 |
| Kim^J | r) | 20:35 |
| highvoltage | gooooogle | 20:36 |
| ffm | Kim^J, Basically, I'd like to add a button that says "Add Font" to gnome-appearance-properties, opens up a file picker dialog, and copys the file to ~/.fonts (or the system-wide equivalent) | 20:37 |
| Kim^J | ffm: Ok ok, well, I could do that, but I'm thinking of a download the ttf, double click it and it asks you to add that font... | 20:38 |
| Kim^J | Well, could do both. :) | 20:38 |
| ffm | Kim^J, Yeah. | 20:39 |
| LaserJock | wouldn't that kind of project be more appropriate in Gnome itself? | 20:40 |
| Kim^J | LaserJock: Well, maybe, we can create it for Ubuntu, give it to Gnome. | 20:40 |
| Kim^J | That is: Ubuntu --> Gnome is possible, it mustn't be Gnome --> Ubuntu all the time. .) | 20:40 |
| mjg59 | Kim^J: Projects tend to be happier if you work with them, rather than just presenting them with a finished product | 20:41 |
| Kim^J | Ok... | 20:41 |
| LaserJock | well, I don't want to make a big deal about it, but a consistent critique of Ubuntu is that it does stuff and then they never end up going upstream | 20:41 |
| Kim^J | Tell that to ffm, not to me. :) | 20:41 |
| Chipzz | LaserJock: that works the other way around to imo | 20:43 |
| Chipzz | ie, upstream not looking at what has been done in ubuntu | 20:43 |
| Chipzz | PackageKit for example | 20:43 |
| LaserJock | well, yes, there is that kind of thing | 20:44 |
| LaserJock | but in general upstreams are supposed to be upstream | 20:44 |
| Chipzz | I wonder if hughise even bothered to look at what's been done in ubuntu | 20:44 |
| ffm | Kim^J, As I said, I can't implement it, I can just draft. Want to work on it here, or upstream? | 20:44 |
| Kim^J | ffm: I can work on it here. | 20:44 |
| ffm | LaserJock, The whole point of FOSS is that we can make changes, and the decentralized bit helps. | 20:44 |
| Kim^J | Just need to clear out some details on Nautilus and fonts. :) | 20:45 |
| ffm | Kim^J, Thanks. | 20:45 |
| LaserJock | ffm: well, I don't want to debate the whole point of FOSS too much, but I rather think its that we are free to join with upstream, not fork everything | 20:46 |
| ffm | LaserJock, We arn't forking, exactly... | 20:46 |
| LaserJock | ffm: you may get a lot more help upstream | 20:46 |
| ffm | LaserJock, If we make a successful feature, we can tell upstream about it. Plus, I like launchpad. | 20:47 |
| ffm | LaserJock, Ok, then what type of specs should be made in ubunu? | 20:47 |
| === ryu2 is now known as ryu | ||
| LaserJock | ffm: well, I'm not saying you can't | 20:47 |
| Kim^J | LaserJock: What's the point of arguing this? We ARE going to tell Gnome about it. | 20:47 |
| LaserJock | but in general, packaging and integration changes | 20:47 |
| LaserJock | things that are Ubuntu-specific | 20:47 |
| LaserJock | it's not telling Gnome about it | 20:48 |
| LaserJock | it's about working with Gnome on it | 20:48 |
| Kim^J | Why? So Ubuntu must wait even longer to get it? | 20:48 |
| LaserJock | why would it? | 20:48 |
| LaserJock | it won't take any longer | 20:48 |
| Kim^J | Do you think SuSE works WITH Gnome all the time? No, but the things gets to Gnome anyway. | 20:48 |
| Kim^J | s/gets/get/ | 20:49 |
| LaserJock | yes, openSUSE does work with Gnome all the time | 20:49 |
| LaserJock | that's why Novell employees Gnome people | 20:49 |
| LaserJock | *employs | 20:49 |
| Kim^J | I think it goes MUCH faster to develop on own platform, test, use, extend, then make it upstream. | 20:50 |
| LaserJock | but why not do the *exact* same thing upstream? | 20:50 |
| LaserJock | it seems like a waste of effort | 20:50 |
| Kim^J | What? | 20:50 |
| Kim^J | It's being sent to Gnome anyway, the whole thing, done and ready to compile. | 20:50 |
| LaserJock | but why not work *with* Gnome is the point | 20:51 |
| LaserJock | you can get more testers, more knowledge, more help | 20:51 |
| Kim^J | But we are going to. | 20:51 |
| LaserJock | seems like a win-win situation | 20:51 |
| LaserJock | only after it's done | 20:51 |
| LaserJock | perhaps the Gnome guys have a much better way to implement it, etc. | 20:52 |
| LaserJock | or perhaps they've started work on something similar, who knows | 20:52 |
| LaserJock | but it's probably worth a look | 20:52 |
| Kim^J | Whatever... | 20:53 |
| ffm | LaserJock, The people on #gnome on GIMPnet are dead. | 20:53 |
| ffm | 140 ppl in chan, noone responds. | 20:54 |
| Kim^J | No life in ##gnome @ Freenode... | 20:54 |
| LaserJock | ffm: it's a weekend | 20:54 |
| Kim^J | LaserJock: So? Weekends = More time to IRC. | 20:54 |
| ffm | LaserJock, And I'm in school on weekdays. | 20:54 |
| LaserJock | for working people it's often a break | 20:54 |
| LaserJock | ffm: there's an interesting new technology called email, you may have heard of it ;-) | 20:55 |
| ffm | LaserJock, Bah! | 20:55 |
| uyann | word of advice, make sure users are adequately warned before using the entire disk for ubuntu during installation | 21:00 |
| ffm | uyann, You mean removing the windows partitions? | 21:01 |
| ffm | uyann, Or filling it to the brim? | 21:01 |
| uyann | I had both windows and ubuntu installed on my machine | 21:01 |
| uyann | I decided to install the latest ubuntu today | 21:02 |
| uyann | I was happily clicking on next next etc... | 21:02 |
| pwnguin | so in your state of clicking next, do you think you would have read and understood a warning? | 21:02 |
| uyann | because the default was set to use the entire disk I clicked on next | 21:02 |
| uyann | yes if the button was not "Next", I may have stopped | 21:02 |
| uyann | and if a red warning sign appeared.. absolutely | 21:03 |
| ffm | uyann, You'd be surprised. | 21:03 |
| Kim^J | uyann: Naaah, normal Windows-users would just click on that too without reading. ;) | 21:03 |
| uyann | I understand that Ubuntu is being made to be installed as easy as possible.. but it's really dangerous... | 21:03 |
| uyann | :/ | 21:03 |
| ffm | uyann, Some people will only do so if required to do something like type "I understand all data on my disk will be lost" | 21:04 |
| LaserJock | uyann: do you remember if it had anything else before formating the drive? | 21:04 |
| uyann | yes | 21:04 |
| uyann | I remember the button "Advanced" | 21:04 |
| pwnguin | ffm: i recall a packages somewhere requiring that sort of interaction | 21:04 |
| uyann | and then I clicked on next and it started repartitioning | 21:04 |
| LaserJock | uyann: did it give you a summary of what it was going to do? | 21:04 |
| LaserJock | it's been a while since I used the GUI installer | 21:05 |
| pwnguin | uyann: maybe if we forced users to back up their data before installing | 21:05 |
| highvoltage | uyann: word of advice, if you boot from /any/ CD, try to be careful with just clicking on 'next' ;) | 21:05 |
| uyann | pwnguin, no, do not force users to backup | 21:05 |
| pwnguin | then there's no good solution | 21:05 |
| uyann | yeah, I understand it's my fault.. | 21:05 |
| pwnguin | things can and will (in a small number of cases) go wrong | 21:05 |
| pwnguin | best practice is to make backups | 21:06 |
| pwnguin | i think the installer even says at the beginning to make a backup | 21:06 |
| LaserJock | I think a nice little thing with a stop sign or something that says "The entire hard drive is going to be erased" or some such | 21:06 |
| uyann | 4 seconds just passed when it was repartitioning | 21:06 |
| uyann | I believe I can still rescue the partition | 21:06 |
| uyann | LaserJock, I agree | 21:06 |
| pwnguin | LaserJock: so what if you resize the partition? | 21:07 |
| LaserJock | I think that's a bit more self explanatory | 21:08 |
| uyann | 'data may be lot by proceeding with this operation' or something to that effect | 21:08 |
| LaserJock | but if I remember right the installer says "Use entire hard disk" or something like that | 21:08 |
| uyann | yes | 21:08 |
| uyann | it was set to the default | 21:09 |
| LaserJock | "Use" does not imply "Erase" | 21:09 |
| hdevalence | i think the warning needs to be big, red, and scary | 21:09 |
| LaserJock | which is where I can see confusion | 21:09 |
| hdevalence | otherwise it'll be ignored | 21:09 |
| pwnguin | there's no situation i can imagine where "data may be lost by proceeding with this installation" SHOULDN'T be shown under this argument | 21:10 |
| pwnguin | repartitioning can cause data loss | 21:10 |
| pwnguin | reformatting the entire drive will cause data loss | 21:10 |
| pwnguin | the warning doesn't solve anything | 21:10 |
| pwnguin | backups do | 21:10 |
| LaserJock | no, I don't think we necessarily need a "your data may be lost" | 21:11 |
| LaserJock | but users need a reasonable expectation of what's gonna happen to their disk | 21:11 |
| LaserJock | with perhaps special attention to defaults | 21:11 |
| LaserJock | for the "Next" clickers ;-) | 21:11 |
| * uyann pour moi | 21:12 | |
| pwnguin | whats going to happen is theres a chance for data loss, and even if warned, they'll blame anyone but the only person who could have backed it up | 21:12 |
| LaserJock | if you take the time to pick something other than the default then I assume you can read ;-) | 21:12 |
| pwnguin | heh | 21:13 |
| pwnguin | maybe switch next and back on that frame | 21:13 |
| pwnguin | so if they just click without looking, it'll be a loop | 21:13 |
| LaserJock | hah | 21:13 |
| LaserJock | I always use the Manual partitioning so I have no idea what the guided partitioning does | 21:14 |
| pwnguin | on a more serious note, why can't the ubuntu live cd provide at least an optional "back drive up" system? | 21:14 |
| LaserJock | s/can't/has nobody yet written/ | 21:15 |
| pwnguin | i tried guided a few times, it's messy and sorta fluctuates. at one point the resize and install to a new partition guided thing was removed because it was prone to breakage | 21:15 |
| highvoltage | pwnguin: you can create a spec for something like that | 21:16 |
| ffm | I think there should be no default. | 21:16 |
| ffm | You should have to choose one of them. | 21:16 |
| highvoltage | ffm: wouldn't that, in itself be a default? :) | 21:16 |
| LaserJock | ffm: yeah, that would make sense actually | 21:16 |
| ffm | A) Wipe disk and use sensible defaults <bold> ALL DATA WILL BE LOST</bold> | 21:16 |
| pwnguin | highvoltage: maybe if there was an Ubuntu SoC it could have been a pick | 21:16 |
| LaserJock | uggg | 21:17 |
| LaserJock | that would mean it would never get done ;-) | 21:17 |
| pwnguin | colorfilter got done | 21:17 |
| pwnguin | its interesting though how debian never seems to get "this should be easier / safer" complaints | 21:18 |
| ffm | pwnguin, They don't bill themselves as designed for normal humans either. | 21:18 |
| LaserJock | they don't? | 21:18 |
| LaserJock | I'm pretty sure I've seen that kind of thing on debian-devel | 21:19 |
| uyann | <ffm> I think there should be no default. <<< I agree, a simple change that may work.... | 21:19 |
| ffm | Why don't we bug the people in #ubuntu-installer? | 21:19 |
| ffm | LaserJock, Do I have to make a spec just for that? | 21:20 |
| LaserJock | ffm: I'm guessing not a lot of people are around in #ubuntu-installer because of the weekend | 21:20 |
| LaserJock | I would think just for the no-default thing that would be a bug report | 21:20 |
| LaserJock | safer handling of partitioning UI in general would probably be a spec | 21:20 |
| LaserJock | but you'd want to probably talk to Colin about it first | 21:21 |
| pwnguin | i still like my circular reference of doom idea ;) | 21:21 |
| pwnguin | if there's no default, how will our poor, mildly informed user make a choice? | 21:22 |
| uyann | pwnguin, hehe.. "Do you wish to go back an change a setting?" "Next" | 21:22 |
| uyann | s/an/and | 21:22 |
| pwnguin | the default is the default for a reason -- it's the most likely to work | 21:22 |
| uyann | it'll work, ofcourse | 21:22 |
| uyann | but it'll erase the entire disk | 21:22 |
| LaserJock | well, it's also the most damaging | 21:22 |
| pwnguin | (or does the most undamaging ;) ) | 21:23 |
| LaserJock | I doubt you'd say that if you accidentally wiped a drive you cared about ;-) | 21:23 |
| pwnguin | if i had drives i cared about | 21:24 |
| pwnguin | i'd back them up | 21:24 |
| LaserJock | maybe | 21:25 |
| LaserJock | but a great majority of users don't | 21:25 |
| pwnguin | then we should help them | 21:25 |
| uyann | yup, a majority of people don't backup and don't read instructions | 21:25 |
| LaserJock | and it's a rather lame excuse for failure | 21:26 |
| uyann | I trusted Ubuntu. I'm now scarred for life. I feel like I've been raped by it. | 21:26 |
| pwnguin | as long as failure is possible, not pushing backups is a failure itself | 21:26 |
| uyann | lol | 21:26 |
| LaserJock | that's like a care mechanic saying "oh, you have a backup car right?" when he destroys your engine | 21:26 |
| LaserJock | sure, it's good to have one | 21:27 |
| pwnguin | when he destroys your engine he gets you a new one | 21:27 |
| LaserJock | but you should be able to reasonably expect that the mechanic isn't going to destroy your car | 21:27 |
| pwnguin | there is no new data copy unless you make it | 21:27 |
| pwnguin | a warning sign doesn't change your argument | 21:27 |
| LaserJock | sure | 21:28 |
| LaserJock | if we can lower the number of people who run in to problems then that's good | 21:28 |
| pwnguin | going back to the mechanic "i destroyed your engine; you read the sign on the door right?" | 21:28 |
| LaserJock | *if* there was a sign on the door then it would be my problem, yes | 21:28 |
| LaserJock | hence why people put signs up like that all the time | 21:29 |
| pwnguin | but you can't assess that risk | 21:29 |
| pwnguin | the mechanic can | 21:29 |
| LaserJock | so we (the mechanic) has a responsibility to reasonably warn users | 21:29 |
| LaserJock | *have | 21:29 |
| LaserJock | that's my point | 21:30 |
| pwnguin | and where possible, mititage the risk entirely | 21:30 |
| LaserJock | sure | 21:30 |
| LaserJock | but that doesn't exist right now | 21:30 |
| LaserJock | so it's rather mute | 21:30 |
| pwnguin | wow, i misspelled that pretty badly | 21:30 |
| pwnguin | mitigate | 21:30 |
| _hdevalence | has it been considered to use something like debtorrent for distribution upgrades? | 21:31 |
| LaserJock | I believe it has been considered to use debtorrent period, I believe | 21:32 |
| _hdevalence | I have to say, it's REALLY annoying when the servers are swamped with everyone trying to upgrade | 21:33 |
| LaserJock | uh, yeah | 21:33 |
| _hdevalence | and I'm sure that server bandwith for that many upgraders is not cheap | 21:33 |
| LaserJock | sure | 21:34 |
| pwnguin | well, everyone and their mom mirrors | 21:34 |
| _hdevalence | well, using my local mirror, I was getting speeds from 30-60 KB/s | 21:34 |
| LaserJock | I haven't really noticed all that much of a slow down this time | 21:34 |
| LaserJock | it was waay worse in the past I think | 21:35 |
| LaserJock | but yeah, it would help to have things like debtorrent, and package diffs | 21:35 |
| pwnguin | i wonder about package diffs | 21:36 |
| pwnguin | seems like you'd need to keep several of those around | 21:36 |
| _hdevalence | I think debtorrent would be good enough. package diffs are more complicated | 21:37 |
| LaserJock | pwnguin: it's a lot of load on the servers I think any way you do it | 21:37 |
| === ryu2 is now known as ryu | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
