[00:11] hmm...to run boxee in lucid i had to manually add libfaad and libmysqlclient... [00:20] simba_: Faulty depends? -> file a bug ;) [00:21] arand, yes, but it was a karmic version and not from "our" repos, so i did it with boxee [00:22] arand, i just wanted to see what all the fuzz with boxee realy was [00:23] ah, okies, sorry, it was a kind of handwavey answer, I know nothing of boxee really. [00:24] arand, just read a lot about it in all the CES news...it is kind of a mediasenter [00:25] arand, have install files for win, mac and ubuntu.... [00:29] Ah, I've heard about that in a podcast sometime I think. Didn't ever dabble in mediacentres. Well I guess if the deb is from them the packaging bug should be thrown at them.. [00:30] just remember someone was having trouble getting it running the other day. [00:31] How do I need to change a previous HAL fdi file, so it gets handled in Lucid? (http://daniel.hahler.de/hal-configuration-fdi-for-kingsis-peripherals-evoluent-verticalmouse-3) [00:38] back on Lucid :) [00:39] wb dupondje [00:39] blueyed: I guess you need to change it to udev rules [00:39] start menu not translated yet ? :) [00:41] blueyed: perhaps does this give you enough hints how to transform it: http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/lucid/xserver-xorg-input-synaptics/lucid/annotate/head%3A/debian/66-xorg-synaptics.rules [00:42] thanks geser, I've found https://wiki.kubuntu.org/X/InputConfiguration#Migration%20from%20hal%20FDI%20files already [00:43] even better :) [00:44] * geser bookmarks this page on the ubuntu wiki [00:49] yay, it worked, thanks. [00:49] removing tons of packages :) [00:49] marking all manually installed packages as autoinstalled, and then select the ones you need :) [01:02] 176 packages removed :) [01:07] dupondje, what is the command for that? [01:08] simba_: auto-remove [01:09] BUGabundo, autoremove will remove them, but mark all maualy installed as autoinstalled [01:10] I still got firefox 3.0.7 languages installed, any idea how to remove them ? [01:10] purge? [01:10] I have used debootstrap to create a hardy chroot (console-only). I then upgraded that to lucid by editing /etc/apt/sources.list and "aptitude udpate;aptitude". I'm stuck with this error: http://paste.debian.net/57461/ for which google does not seem to have an answer. Do you? bug or PEBKAC? [01:11] BUGabundo: doesn't seem to have a package installed :s [01:12] /usr/lib/firefox-addons/extensions/langpack-nl@firefox-3.0.ubuntu.com [01:12] found :) [01:17] nuke it :P [01:19] Laibsch, do you need to mount sysfs and proc when installing with debootstrap? [01:19] simba_: I used to [01:19] This time proc was already mounted [01:19] And I think sysfs, too [01:20] yes, sysfs seems to have been mounted automatically [01:20] The host runs hardy, btw [01:20] the chroot has been updated from hardy to lucid [01:20] havent used debootsrap, but that was what i could think of [01:22] Laibsch, there are a few versions between hardy and lucid [01:22] I think that update-path is supported [01:22] LTS -> LTS [01:23] true, [01:26] I tend to think it's PEBKAC, but I can't rule out a bug. If it was a bug, I wonder what package it may be. [01:39] Laibsch: Upgrading Hardy->Lucid by changing the series in sources.list isn't really supported; what will be supported is upgrading via update-manager (and it's command-line brother, do-release-upgrade) [01:40] OK [01:40] I'll restart the chroot [01:41] I don't think Hardy's update-manager has been taught about Lucid yet, either. I don't *think* Hardy->Lucid upgrades are supported just yet. [01:41] I see [01:42] I'm not convinced that is the root cause of this, anyway [01:44] Laibsch, some of the pachages have a chicken-egg thing....to far leap might brake dependency's with to big leaps when installing one by one like apt-get does [01:55] Laibsch: Oh. Now that I've actually looked at your error, that'll be due to the DBus system daemon not running. [01:56] I think. Possibly it's because upstart's not running, because it's detected you're in a chroot. [01:56] Do I add medibuntu as usual or it is different for lynx ? [01:56] yes, there is something about upstart in the log [01:56] will a hardy host even be able to accomodate a lucid chroot? [01:56] Volkodav_, as usual [01:57] thanks [01:58] Laibsch: I think that some stuff just won't try to start; I'm fairly sure Lucid's udev won't work with the kernel in Hardy. [01:59] All I do in hary itself in maintain the chroot [01:59] The chroot itself shall be used for booting computers over PXE [01:59] s/hary/hardy/ [02:00] Interesting. I'm pretty sure that you should be able to get a lucid chroot happening, even under Hardy. [02:03] Laibsch, apt with --force option dont work either? [02:03] I've already removed the chroot [02:03] and have restarted [03:12] upraded to lucid, wired connection not working. ifconfig shows i have an ip address and my nfs mount is working. [03:13] i can also ssh to this machine from another machine.. [03:27] prefrontal: can you define 'not working'? You can't access the internet? [03:28] what does 'ping google.com' in a terminal tell you? maybe the DNS setup is wrong [06:30] hi. when i install recommended nvidia driver from jockey-gtk i got an error msg like this http://is.gd/6UWB8 and then i try to restart the computer but it seems that i've installed wrong driver like this http://is.gd/6UX1p [06:38] and this is the the output from jockey.log http://pastebin.ca/1763693 [10:22] how can i troubleshoot locale-problems? [10:43] Lucid just feels a bit better then Karmic :) [10:43] nice improvement [10:46] everythign feels a bit better then the previous [10:47] except my terminal -.- [10:48] :) the user settings programm seems fucked up, I set my user to administrator, but nothing changes when I click ok, it keeps telling 'custom' :) [10:58] firefox 3.6 coming when? [11:04] 501. By Alexander Sack 19 hours ago [11:04] releasing version 3.6+nobinonly-0ubuntu1 [11:04] they are working on it :D [11:05] sauce? [11:07] i got it from the ppa :p [11:07] which [11:07] anyone able to shed light on why ubuntu karmic refuses to boot at all? [11:07] (the release version, not hg) [11:29] somebody should push new versions of synce to Lucid :) [11:38] hi, i'm using alpha2 and have an ati card. is it possible that the ati drivers aren't compatible with the latest xorg-server? because my X always complains that is runs in safe mode [11:46] junkY_San: in release notes there is a warning about that [11:47] The fglrx binary driver for ATI video chipsets does not yet support the X server in Lucid. As a workaround, users should use the open source -ati driver instead. (506656) [11:48] ah thanks [12:53] hi. when i install recommended nvidia driver from jockey-gtk i got an error msg like this http://is.gd/6UWB8 and then i try to restart the computer but it seems that i've installed wrong driver like this http://is.gd/6UX1p and this is the the output from jockey.log http://pastebin.ca/1763693 [13:06] hi, i have texlive in status held back for almost a week now [13:06] is it save to dist-upgrade? it wants to remove texlive-base-bin, dvipdfmx and libkadm5srv6 [13:08] meh, petsounds gone already... [13:10] knittl: texlive-base-bin got replaced with texlive-binaries so it's removal is ok [13:10] libkadm5srv6 is unreleated to TeX [13:12] ok, i'll dist-upgrade then. thanks geser [13:13] need to get 500 mb of archives. texlive is HUUUGE [13:19] finaly got the radeon driver working again with 3d acceleration. there were some strange libgl packages installed [13:34] pidgin icon went fucked :p [13:34] its now with a big white border ... [13:34] any idea ? :) [13:46] sometime in the last week gnome shell got installable [13:53] hmm but apparently not usable [14:27] hey guys I am on lucid and am somewhat a newbie at installation fom source. [14:27] I need the latest version of a program called tucan so I did [14:27] svn co https://forja.rediris.es/svn/cusl3-tucan/trunk [14:27] in terminal after installing svn [14:27] now what is the next step I got the long list of output alongwith [14:27] Checked out revision 1396. [14:27] now what to do? [14:28] where are the files downloaded to? how to compile? please soemone help [14:37] cd trunk && sudo make install [14:40] i have an existing karmic installation, and a free partition that i would like to use to test lucid. if i install it there, and choose not to install a boodloader, will karmic's GRUB pick it up when i next run update-grub? [14:42] diverse_izzue: I don't think so [14:43] penguin42, how would you parallel install the two? [14:44] diverse_izzue: I typically install in a vm for testing, but I think the otherway might be to install the bootloader in the partition that you are installing lucid into and then tell the first grub to chainload the 2nd - I don't know the syntax on gnome2 to do that though [14:45] penguin42, that's what i was thinking also, but i know that grub2 has some scripts to incorporate other linuxes, that's why i ask [14:45] diverse_izzue: It might, I've not tried to do taht - but I don't think so [14:46] chris|: that seemed to have worked [14:46] next what do I do I mean how do I run it. I had earlier installed tucan from the repos so now is that version itself updated? [14:46] can I run this latest version (from source) by using the applications menu icon of tucan [14:55] diverse_izzue, just install it side by side it`ll pick up the other linux [14:56] duffydack, thanks, that's what i wanted to hear [14:57] diverse_izzue, I removed mine ealier (lucid). rebooted into karmic, removed the partition and dod sudo grub-install --root-directory=/ /dev/sda to restore grub [14:57] dod=did [14:58] duffydack, so you had lucid's grub installed into the MBR? [14:59] diverse_izzue, yes. it was the default boot distro [14:59] duffydack, i plan to not install a bootloader for lucid and use karmic's for the time being [15:01] diverse_izzue, I guess update-grub will invoke os-prober and handle it for you.. never tried [15:01] diverse_izzue, I wanna try some of this. http://www.omgubuntu.co.uk/2010/01/make-grub-themes-beautiful-look-nicer.html [15:02] looks sweet, but this is my primary machine, playing so much with the boot loader is asking for trouble *g [15:03] diverse_izzue, thats why ive got a clonezilla backup of everything sat on my usb hd :) [15:06] diverse_izzue, could use wubi I guess? [15:12] i do regular backups too, still, one easily looses a lot of time restoring [15:41] sudo lshw messed up my screen (all turned green), had to do , to fix it...anyone else have this? [15:44] simba_: It's OK here [15:44] all fine here too [15:45] what gfx caru u use? === Adys_ is now known as Adys [15:45] i have a intel... [15:45] nvidia 9800gtx+ [15:45] mine is radeon [15:46] https://launchpad.net/~ubuntu-desktop/+archive/ppa <-- added this ppa updated and nothing is listed to upgrade. Any ideas? [15:46] anyone using intel i915 driver? [15:47] xguru: packages should already be in lucid, then. [15:48] http://www.clickonf5.org/linux/ubuntu-lucid-windows-aero-style-transparent-look/6038 <-- desktop doesn't look like this...is there something i have to enable? === yofel_ is now known as yofel [17:07] is the daily build installer working for 32 bit intel? [17:24] where to put x11vnc line in lucid for being able to connect at login? [17:26] simba_, /etc/init.d but you need to make it into a script [17:27] simba_: Do you mean so that you can login via vnc or do you mean so that when you do login it immediately starts? [17:28] the two are quite different [17:28] BluesKaj, it used to be just a line in gdm.conf just afther path.... [17:29] then try that [17:29] penguin42, so i can use vnc when gdm gives login screen [17:29] simba_: are you sure you want x11vnc? Another way to do the same thing is to run a VNC Xserver and then have gdm start on that X server as well as your default one? [17:31] yeah, that might be better, just used to x11vnc sins it shows the same as the screen....but i dont need it now sins it will be a remote box in time [17:31] having said that, I'm not sure what the right way to get gdm to start that is [17:32] if i am lucky the install script might do the job *hoping* === Floops[w]1 is now known as Fringe [17:39] any idea why my pidgin icon is messed up ? :) its now in a white box .. :) [17:59] this works for lucid to for x11vnc: http://flukylogs.blogspot.com/2009/11/setting-up-vnc-for-gdm-login.html [18:45] simba_: so that gets you a vnc shared with the display you can see on your real display? [18:49] penguin42, yes [18:51] simba_: It is possible to make them actually seperate displays by getting the gdm to start on an Xvncserver - but I can't find a modern recipy for that [18:53] i still don't have umlauts in my gnome-terminal [18:53] but i think i'm closer to the solution [18:53] some site mention to edit /etc/environment, but i wanted to ask if there is another solution? [18:54] it's only since karmic [18:54] und X/gdm seem to reset the locale, because my ttys work perfectly [19:06] yay, kde4.4 rc2 has a sane KDM backround again :) [19:31] how can i remove the debug stuff from 10.04? [19:32] what debug stuff? [19:32] koud: notify-osd? [19:32] for the osd forexample [19:32] yeah [19:32] eheh [19:32] I have a bug for it [19:32] you need to change a flag [19:33] where? [19:33] is deadline for compiz the partner upload deadline or the feature freeze? [19:33] [19:33] Open "/usr/share/dbus-1/services/org.freedesktop.Notifications.service" [19:33] with a text editor and remove "DEBUG=1". [19:35] thanks [19:40] Is it possible to install lucid onto an SD card? [19:40] I'd like to test it, but don't want to have it on my HD [19:41] when do they fix plymouth for nvidia ... [19:59] distro[Ubuntu "luiid" 10004] [19:59] bug in xchat ;) [20:02] luiid is a secret distro created by a shadowy agency that's drawing it's plans against us [20:04] CosmiChaos: compiz is unrelated to partner [20:04] don't know where the heck it gets luiid from :P [20:04] Codename: lucid [20:04] what is it related to then? [20:04] patches accepted [20:04] it parses this, and outputs luiid :) [20:05] CosmiChaos: bug fixes have no deadline [20:06] CosmiChaos: feature additions have the deadline of FeatureFreeze [20:08] crimsun: today my sound started clipping... it was so nice two days ago :( [20:10] define clipping [20:13] my sound skips also if I play from share, but it looks like Microsoft can't even have their cifs crap bugfree :( [20:17] bjsnider: you still have your 64bits flash ppa? [20:17] bjsnider: clipping: as in sound is so high it disorces [20:17] *distorces [20:18] BUGabundo, if you want to use it can enable it for a brief period [20:18] just upgraded to lucid. i have an ip address, my nfs mount is working and I can ssh to this machine, but it is not connected to the internet. ping doesn't work for example [20:19] route ? [20:19] ? [20:20] whats output of route ? [20:20] bjsnider: was looking for a more "permenant" ppa, so I can give it way to ppl having trouble with frlash wrapper [20:21] dupondje, http://pastebin.ca/raw/1764349 [20:21] BUGabundo, get written permission from adobe for me and i'll do that [20:21] :\ [20:22] right [20:22] we can redistribute it [20:22] :( [20:22] F*** [20:22] get me a document that will stand up in court [20:22] dupondje, here is the route output for a jaunty machine right next to this lucid one that is working correctly: http://pastebin.ca/raw/1764350 [20:23] seems like you get weird ip ? :p [20:24] these machines have dedicated ip addresses IIRC [20:26] default gateway should be a router's ip address. instead you've got a .c file or something [20:27] is that to me? a .c aka config file? [20:28] wait, the default gateway is not an ip address on the machine that works [20:29] `locate muen-gw-clipr.c' brings up nothing [20:33] well, th config file should contain the necessary ip address [20:35] routers for the past few years have been able to do a combination of static and dhcp, so you don't need to configure that on the client [20:35] you can use dhcp on the cient and configure the router to give it the same ip address every time [20:35] unless you have an old router [20:36] we have a dozen or so identically configured workstations. the rest of them are running jaunty, this new one is running lucid [20:36] i don't think i need to reconfigure anything on the router as the rest of them are working [20:38] it does seem plausible that a config file was overwritten on the workstation during do-release-upgrade. the question is, which one? [20:40] the muen-gw-clipr.c file exists on the lucid system? [20:41] it doesn't even exist on the jaunty system [20:41] muen-gw-clipr.c is the default gateway for the jaunty system [20:42] there are files in /etc/network [20:43] i'll bet if you compare them with the jaunty systems you'll find out what's wrong [20:44] i still think it would be better to configure the router, just simplify each client to use dhcp and back up the router settings to a jump drive [20:52] found it fixed it. somehow the dns nameserver was misconfiged [20:52] thanks bjsnider [20:53] Is couchdb installed by default on ubuntu? tried checking the dependency but that doesnt list ubuntu-desktop [20:53] acicula: ubuntu one uses it [20:54] hmm ok [20:56] not sure if i ever installed that, but i guess something pulled it in, thanx yofel [20:56] acicula: actually I might be wrong [20:57] seems like evolution depends on it [20:57] over evolution-couchdb [20:59] ah yeah, that makes sense [21:10] Does "Switch User..." works for anyone? [21:18] alex_mayorga: with what GPU and drivers? [21:26] hmm, I don't get a working plymouth. [21:26] http://users.csc.calpoly.edu/~dgoyette/bootcharts/EliteBook-lucid-20100124-4.png [21:30] any one those? [21:37] grmbl [21:37] * yofel wants qt 4.6.1 [21:38] No, you can't have a pony :P [21:38] lol [21:46] \0 = NULL ? [22:06] no, NULL is a pointer while \0 is a char [22:10] Has anyone had libvirtd break for you? If so, how did you solve it? [22:10] (updating in the last month... I had it working fine in December) [22:26] lol, there has been a bug in xchat-xsys for 2 years [22:26] and nobody noticed it :) [22:27] xchat is almost a dead project, that's why [22:28] not to say people don't use it, because they do. but i don't think anybody's developing it [22:29] http://codepad.org/G2fwwjSS => the bug is here, you have the last strcpy is wrong, but don't know how to solve it 100% correctly) [22:29] dupondje: perhaps try quassel [22:31] Has anyone tried libvirtd with the current version? It seems to have broken. [22:31] BUGabundo, how can I tell? [22:31] º? [22:31] context please [22:32] I have an nvidia card, but I don't really know what driver is in use [22:32] Sunday January 24 2010, 15:09 - alex_mayorga: Does "Switch User..." works for anyone? [22:32] oh [22:32] now I remember [22:33] I'll have to test it later [22:33] I do know I can have two accounts loged int [22:33] but its via FreeNX [22:33] argh, virt-manager doesn't allow ARM VMs. [22:34] when I click switch user I see the user list, but when the pointer moves or a ke is pressed I get the password prompt for the user currently logged in [22:34] And none of the ubuntu ARM stuff works for the stuff qemu can emulate. [22:35] DanaG, is virt-manager working for you at all in 10.04? [22:36] It would connect, but didn't offer any button but "cancel" on the new-vm wizard. [22:36] Oh.. [22:36] libvirt-bin is the daemon -- had to be installed separately. [22:44] alex_mayorga, that seems like there is some sort of permissions issue that got messed up [22:44] I can't confirm it though, because I am only using a LiveCD version right now [22:45] I wonder what's the component I should bug report against [22:45] can anyone else reproduce? [22:46] BUGabundo: how do I tell what's the video driver in use? [22:51] alex_mayorga: check out jockey [22:52] lsmod [22:52] check for nv, nouveau or nvidia [22:52] of course the module could be loaded but not in use [22:54] thanks, let me check [22:56] none of those show, so it's happening with the vesa driver I guess [23:01] is anyone using xchat? [23:02] I have it usable. I switch between xchat and weechat right now. [23:03] charlie-tca, are your hilighted tabs acting up? [23:03] yup [23:03] ok thought it was just me [23:03] is there a bug filed? [23:03] bih yup [23:03] bug 508297 [23:03] Launchpad bug 508297 in gtk+2.0 "[lucid] xchat channels do not change color anymore" [High,Confirmed] https://launchpad.net/bugs/508297 [23:04] That's why I thought I would try weechat. Xchat is driving me nuts [23:05] tree view works fine still [23:05] BUGabundo: I'm not sure what that means, but you definitely need to give me alsa-info.sh output when that symptom occurs, because I need both the HDA codec dump and the mixer settings [23:07] BUGabundo: here are general steps to troubleshoot: 1. check PA's volume. 2. check the amp val for your stereo-out mixer widget. 3. check the verbose PA log. [23:07] so no-one here switch users, ever? [23:07] Hmm. [23:07] It *appears* that if you use virt-manager, and have a CDROM drive with no ISO connected, [23:07] the VM will not start. [23:07] its broken with plymouth right now alex_mayorga [23:08] crimsun: thanks. I check both 1st. I usually do, before reporting [23:08] If you remove that CDROM drive, it will. Last month's version did not do that :) [23:08] Sarvatt, Thanks! No need to report then? Is there a bug #? [23:08] wanna now the funny part, no longer happening ... it _fixed_ it self.... I hate this bugs [23:08] disable splash from your kernel command line and it'll work normal, or use log out applet for a few days [23:08] yeah there are a bunch but i dont have any handy at the moment, i'm sorry :( [23:09] will get back to you when i get home if you havent found one [23:09] Yep, getting rid of uninserted CDROM drives makes libvirt work again [23:09] BUGabundo: please, just get me the requested info when it next occurs, because I can't troubleshoot lack-of-symptom [23:10] crimsun, just letting you know... that system bell bug report got upstreamed by one of the reporters [23:10] I know I know Daniel... [23:10] WeatherGod: as it should have been. [23:10] WeatherGod: I don't care much for GNOME bugs, as that lies outside my resource allocation. [23:10] Sarvatt: thanks, much appreciated [23:11] alex_mayorga: if you use the log out applet in gnome instead of the indicator-applet-session's switch user it'll work even with plymouth for now though [23:11] crimsun, yeah, kinda figured, but this one was so messy, it was hard to determine where the fault lies [23:11] do you get kicked to a VT with the mouse still functional? [23:11] WeatherGod: it won't even be remotely resolvable until Ubuntu gains a 2.6.34 kernel [23:11] Sarvatt, unfortunately I don't have much space for yet another applet :) [23:12] crimsun, so, that will be 10.10? [23:12] Sarvatt, no I get kicked to the user logged unlock screen [23:12] WeatherGod: no idea [23:12] over and over [23:12] WeatherGod: I have no idea what sorts of regressions will be in whatever 10.10 ships [23:12] ah and it doesn't go farther? [23:13] DanaG: how's your crusade against notify going? [23:13] crimsun, guess we will have to wait and see [23:13] eh, I just use notification-daemon. [23:13] alex_mayorga: it might not be the same bug then, what GPU are you using? the plymouth one is only really affecting intel [23:14] Though, oddly enough, brightness control doesn't work in gnome-power-manager. [23:14] The ubuntu devs must've done something to break it when not using notify-osd. [23:14] Sarvatt: the laptop has an nvidia chip, but I don't think the driver is in use [23:14] you dont see the splash when you boot up do you? [23:15] black screen with a ubuntu logo in the middle [23:15] DanaG: might want to ping them during the workweek in #ubuntu-desktop, then. [23:15] yeah doesn't seem like the same bug, sorry to throw you off alex_mayorga [23:15] DanaG, you're on a crusade to stop notify-osd before it kills again? [23:15] (I doubt there's any sort of conspiracy, but people love to make up all sorts of FUD ;) [23:16] Sarvatt: no splash, I see text [23:18] my angst with notify-osd is I don't get *any* configuration controls, but they keep saying thats a feature not a bug :S [23:18] I am actually dealing with a bug report right now about the notify-osd... [23:18] the guy is convinced that having the osd disappear on hover is a bug [23:18] bjsnider: not that I like to support DanaG, I do fell that notify-osd (that I didn't like that much when it came) is now taking tooooo much visual space, and not transparent enough [23:19] point the OR to the design spec for notify-osd, and have him/her argue with macslow, then. [23:19] (now that would be an interesting exchange) [23:19] My argument is that it's supposed to make display time proportional to amount of text... but it doesn't. [23:19] hmm, that's a good idea, maybe [23:19] DanaG: Which is a bug, obviously. [23:20] A 3-word notification and a 30-word notification, both sent via notify-send, will both show for 10 seconds. [23:20] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=544750 [23:20] Debian bug 544750 in notify-osd "notify-osd: Unable to lower notification's expiry time than ten seconds" [Normal,Open] [23:20] not sure why that report is on debian. [23:20] C isn't *that* hard to learn; implement that part of the spec :) [23:20] https://bugs.launchpad.net/debian/+bug/423314 [23:20] Ubuntu bug 423314 in notify-osd "Unable to lower notification's expiry time than ten seconds" [Wishlist,Invalid] [23:21] I say, they should stop ignoring the time they're asked to display the thing for! [23:21] At least within reason. [23:21] Have you had that discussion with the ayatana folks? [23:21] well, about the most I have is the comment on that bug report. [23:21] And what does that name "Ayatana" mean, anyway? [23:22] WeatherGod: I tend to agree, usually when I move the mouse I'm trying to achieve a target [23:22] and that blurring... augh, eye torture. [23:22] yeah, and when using something like pidgin, and you have someone IM'ing you, the bubble stays up and gets in the way [23:23] stress-test: while (true); do notify-send this sucks; done [23:23] that'll block notify-osd for 500 seconds. [23:23] DanaG: Another part of the spec that isn't implmented, yes. [23:23] 50 is the maximum queue length. [23:23] 10 seconds is the duration. [23:23] heh [23:24] DanaG: the blur SUCKs [23:24] mouse over does NOTHING [23:24] Run that on MacSlow's computer some time when he's not watching. =þ [23:24] Then ctrl-c it after a moment. [23:24] what blur? [23:24] specially when the users come from win32 and they're so used to click the msn "toaster" to act on it [23:25] DanaG: While there's certainly stuff of value in what you're saying, bitching about it in #ubuntu+1 isn't really going to do anything other than make you feel better. [23:26] And the bugtracker seems ignored, as well. [23:26] Unfortunately, bug reports do nothing also [23:26] DanaG, let's go to #ayatana I'll back you up :) [23:26] seems like the group is pretty sure, they're doing it right [23:26] Wish you more luck than I had with ayatana [23:26] yeah! I tried and failed once too [23:27] eh, I don't feel like arguing with anyone right now... have homework to do. [23:27] folks, hint: patches (or git/bzr branches) speak louder than any amount of bitching on irc. [23:27] yay, raytracing. NO recursion yet. [23:27] oops, didn't mean to double-caps. [23:27] crimsun, so true [23:27] show those notification to a win32 grandma and try to explain they'll dissipate when she tries to attain them with the mouse [23:28] but, why should they try and click on them? [23:28] because that's the way they're used to interact with notifications [23:29] I *think* you'll find that the way they're used to interacting with notifications is to look at them and ignore them. [23:29] I'm not arguing with the noninteractivity, per se... just with the ludicrous minimum time. [23:29] also when I click on the envelope and click gwibber it's window won't raise [23:30] alex_mayorga: Buggy behaivour isn't a good critique of the specification :) [23:31] as I don't think I have the skill set to branch that right now and these being an ubuntu only thing I'm eying other distros, I guess it's easier [23:34] RAOF, last time I checked I think plenty of people had spoken on their wiki, mail list, etc. [23:34] but even when put against a clear user case they refuse to listen [23:34] I think you can turn off a lot of those notification popups [23:35] That said, they are incredibly stupid. In Pidgin or whatnot, there should be a right click option "Notify me when this person is online" [23:35] and it should NEVER show me one in any other case [23:35] but what do a user know I guess they're the usability masters for a reason [23:35] 99% of people, you really don't care [23:36] alex_mayorga: I suspect they'd say that anything that requires an action isn't a notification, so it's out of spec for notify-osd. [23:36] It comes from the belief that 'more information is better'. One page versus five hundred page manual, some people will say the latter is better because it has more information. They should not be allowed near any GUI code :) [23:37] Zer, your suggetion sounds more like a wishlist item for Pidgin [23:37] alex_mayorga: Can you remember the use-case you were thinking of? [23:37] Perhaps. All notifications are like that though. :) [23:37] You can do some fairly simple hacking of notify-osd http://mupuf.org/blog/article/22/ if you look around there's min_fade_time as well, still doesn't [23:37] People who think what their app has to say is so worthwhile that it should interrupt the flow of my work [23:37] RD @TravisB: EVERYONE, IN AN HTML5-CAPABLE BROWSER, DO THIS! http://www.chromeexperiments.com/detail/browser-pong/ [23:37] EVERYONE, IN AN HTML5-CAPABLE BROWSER, DO THIS! http://www.chromeexperiments.com/detail/browser-pong/ [23:38] gaaack. once is enough. And lay off the cruise-control. =þ [23:38] oops... sorry for the double post [23:38] =þ [23:38] slow pc [23:38] I wonder if it is the game :\ [23:38] RAOF, maybe is language barrier, but in Spanish or in my mind at least I get notified because I need to do something [23:38] a fire alarm is a notification and you must act on it [23:39] Would it kill them to make those #defines into gconf keys? [23:39] heh... yeah, I am gonna want to walk away from the computer during a fire alarm [23:40] I also wish they'd unbreak gnome power manager. [23:40] alex that is exactly right [23:40] It seems to be waiting for god-only-knows what, before actually changing brightness. [23:40] BUGabundo, great now I lost my train of thinking :) [23:40] ahahahaha [23:40] DanaG: Not sure, I'm guessing it's less than trivial to migrate them though, having to read and store them, etc... [23:41] the brightness controls have always been messed up [23:41] and they don't let me pick what I get notified of [23:41] Used to work fine before they tied it into notify-osd. [23:41] DanaG: At least in part they're not gconf keys so that there's motivation to actually get the right behaviour. [23:41] alex_mayorga: But notifications are transitory; if the notification means that you need to *do* something, and you miss it, what happens? [23:41] I oughtta' try building g-p-m with those notify-osd patches disabled. [23:42] RAOF, they get conveniently stored in their envelope icon [23:42] RAOF: I think you hit the nail on the head [23:42] alex_mayorga: So why does the notification itself need actions if they're always available from the envelope? [23:43] my beef is that most of the times I'm there to respond so why should I wait for it to disappear and get "filed" [23:43] It's to let you know the icon just appeared [23:43] oh, so THAT'S what that envelope is for? [23:43] alex_mayorga: A notification, in the sense of notify-osd (and notification-daemon, really) is “Hi! This thing that you're hopefully interested in happened” [23:43] RAOF: Well, it needs an option "I'm not interested, go away and never come back" because developers think their apps matter more than they do [23:44] On Windows nobody likes system tray icons that I've met [23:44] because things lurking there tend to harass you about this or that [23:44] Zer, that should be controled from the apps [23:44] or if you're feeling angry, it's "**** off and die". [23:44] As I say about audio ads. [23:44] WeatherGod: Have you ever watched someone use a PC, and the popup comes up, and they struggle to figure out how to make it go away? Often those have a checkbox "Don't ask me again" but nobody even _reads_ that [23:44] Zer: That's an idea that I'd guess the DX team would be receptive to. As long as you could implement the configuration in a sensible way. [23:44] They just get irritated about it, day in and day out, for years [23:45] for Windows, you mean... yes, because it keeps coming back and requires my interaction to make it go away [23:45] notifications are tempoarary, though, and always goes away [23:46] I wonder if it would be possible to allow a three-grade verbosity of notifications or something... [23:46] Well, most people use Windows, but the principle there is the same in most cases... if they queue up I will have a hard time seeing under them etc [23:46] right, which is why they now disappear on hover [23:46] I'm not saying 'because it requires user interaction', it's that when people are using a program and something else pops up, whatever it is, the instant reaction is 'go away' [23:46] arand: Absolutely; there are currently three grades of notification urgency - low, high, critical (IIRC). [23:46] Not 'ooh information'. It won't even get read [23:47] I love it for Pidgin on my UNR [23:47] I can still see messages to me, even though the chat window isn't in focus [23:47] RAOF: Yea, true, but that only affect the ordering, right? [23:47] I am just saying, all notifications, individually, should be explicitly opt-in, unless it's something horribly, horribly important [23:48] For instance, a checkbox 'Notify me when X logs in'. Never on by default [23:48] That way, the person reading them actually expects them, and reacts positively since they asked for it [23:48] Zer, I can set Pidgin to not notify me for all users [23:48] arand: There's no reason why you couldn't have a system-wide setting to only show “high” & “critical” when you're busy; notify-osd currently does some of this, with presentation mode. [23:48] you are asking for a wishlist, I suggest you file it [23:48] it is a good suggestion [23:49] RAOF: yes they do but they chose the importance for me [23:49] Hmm I'll go look. By the way, I noticed that the Ubuntu OpenSSL version is before that 'renegotiation bug' was discovered/patched/whatever. Does it still suffer from it? [23:49] it might be nonsensical, but a twitter mention is critical to me [23:50] Zer: more context, please. Which Ubuntu release? [23:50] alex_mayorga: You could potentially have application-level options for that, but I think that use-case is basically a huge outlier. [23:50] All of them, really [23:50] RAOF: exactly... hm I didn't know about presentation mode, is that *supposed* to make it not brute-force harass fullscreen apps? [23:51] arand: I'm not sure what you mean by that, but presentation mode should kick in when you're running a fullscreen app, yes. [23:51] Actually I take that back, 10.04 has a recent one [23:51] 9.10 is still back at g. alex: That's what I mean, it ought to, in all cases, be something you told it to notify you of [23:51] distro[Ubuntu "lucid" 10.04] [23:52] hell yea :P [23:52] Zer: which CVE? 2009-4355? [23:52] RAOF: but as I read the specification configurability is not even planned [23:52] RAOF: Dunno if it's improved in KK/LL but way back here in jaunty, fullscreen apps jumps and twists whenever notifications kick in.. [23:53] alex_mayorga: The ability to, in some central place, set the importance of every notification you have ever seen? No, I don't think that's likely to be implemented. [23:53] I know caring about twitter mentions is nonsense, but that's how most people contact me these days [23:53] crimsun: I'm not a security researcher, so I don't know the specifics, but I remember reading on some site that there was an issue with renegotiation [23:53] The individual apps are perfectly capable of setting the importance of their notifications; that's where such options should go. [23:54] Zer, yeah, it was a MITM attack [23:54] and it was definitely more recent than 0.9.8g, though, that might be patched I suppose [23:54] Zer: I'd check the changelog; the security team is generally pretty good with stuff like that. [23:54] RAOF: That just adds another option. Then, on both the notify end and on the program end, the user has to specify an importance, instead of just 'tell me' [23:55] Zer, you could intercept a SSL connection and splice a relay to continue it or something like that [23:55] Zer: Both the notify end & program end? I was thinking just on the program end. But I agree; I don't think it's a particularly valuable option. [23:55] RAOF, so for now I should be bugging gwibber and not notify-osd? [23:56] Yeah I agree, programs ought to just behave [23:56] alex_mayorga: I don't see why notify-osd would ever implement “let me override the importance for $THESE notifications”. [23:56] Zer: Right. [23:58] anybody know where I should file a UNR-only bug? [23:58] There may be some situations where the application having an option to mark a particular class of notifications as “important” - for your pideon example, maybe there'd be an “important” modifier to the “show me when this user logs on” option, that'd set the priority and additionally stick a message in the messaging-menu, too. [23:59] That might be a bit ugly. [23:59] To what end? [23:59] Either they need to know or they don't [23:59] Checkbox :) [23:59] Actually, the modifier would be “Show this even when I'm busy”.