/srv/irclogs.ubuntu.com/2009/04/11/#ubuntu-devel.txt

=== Keybuk is now known as Guest97642
=== Keybuk_ is now known as Keybuk
ScottKjdstrand: Sounds good.  Got it.02:45
gopogojaunty has same look and feel as the 10 ubuntu before it :-(02:52
gopogomakes it boring02:53
Miloszit's in the details02:54
directhexgopogo, all three new themes are the same as Human?02:54
directhexhttp://webupd8.blogspot.com/2009/03/new-themes-in-ubuntu-904-jaunty.html02:55
Miloszalso consistency is better than useless change (where something can still change and be consistent, but you don't want to reinvent the entire thing each time)02:55
gopogospecially when mark said that OS X usability was the gold standard02:55
gopogoi dont like the dark themes makes it difficult to read menu items02:57
=== Snova_ is now known as Snova
gopogonew notification system is really great congrats guys !03:00
jdongyeah, it's pretty :)03:00
jdongand I like the idea of the "notification center" too for messaging03:01
ScottKjdong: How are you on not getting notified for up to a week about new updates?03:01
gopogois it inspired by Growl in Os X03:01
MiloszIt seems like an own invention to me and it's great03:02
gopogoonly thing I not found working was wep support in wifi03:02
jdonggopogo: by "inspired" you mean a clone of, right? ;-)03:02
jdongbut yeah Growl is the first thing that came to mind.03:03
MiloszCan't be praised enough that they did this, maybe you can argue whether the system itself as it turned out is good or not (I think it is, but just as not to take a stance here), but the fact they did it based on an own idea is fantastic03:03
jdongScottK: I don't like it; that's definitely going to get some reworking here locally....03:03
ScottKjdong: Yes.  Fortunately it's just a gconf change.03:03
jdongScottK: I'm still disappointed that unlike Kubuntu, the GNOME stack still doesn't use packagekit for updating03:03
jdongprimarily I'm interested in PolicyKit03:04
jdongbeing able to preauthorize users to do system updates03:04
ScottKI'd prefer something more mature myself, but none of the KDE options were particularly appealing.03:04
ScottKNeither Adept (the other choice) nor Kpackagekit complain about unverified packages which IMO makes them unsuitable for actual use, but I got outvoted.03:05
jdongthat's disappointing.03:05
jdongI guess we NEED someone to launch a package poisoning attack before people get the point...03:05
ScottKThe KDE3 version of Adept did, so it's a significant regression in my opinion.03:06
ScottKAgreed.03:06
jdongconsidering that deb packages are instant root access, I would expect that to be considered a pretty serious problem03:07
AmaranthRemember the wallpaper change?03:08
AmaranthI thought people would get the message then03:08
SnovaHmm?03:08
gopogoi think altest the icons should be changed in default theme03:09
directhexgopogo, are you volunteering to make replacement icons for everything in tango?03:09
gopogoyeah why not03:10
AmaranthSomeone made this big list of repositories to get interesting unofficial packages and one of the guys included in the list didn't want his packages used by such people. He made a package that forced the GNOME wallpaper to be set to a skull and crossbones03:10
gopogonice ;-)03:10
ScottKThat's a warning about random third party repositories.03:10
ScottKCryptographic package verificatio wouldn't help that one.03:11
AmaranthPeople apparently didn't realize a package for something like your panel could change other things on your system at the same time03:11
AmaranthScottK: Still, it's the general "packages from other places can harm you" thing03:12
ScottKSure.  The problem with the lack of cryptographic verification is packages you think are from places you trust can harm you.03:13
* Amaranth wonders wtf happened to his IO performance03:14
gopogothis ico n set looks great -> http://www.gnome-look.org/content/preview.php?preview=2&id=32146&file1=32146-1.jpg&file2=32146-2.jpg&file3=32146-3.jpg&name=Glass+Icons+Theme03:14
Amarantheep, no way03:16
AmaranthHuman looks so much better03:16
Miloszyeah I give it a -- too03:16
Miloszthe details are not right03:16
gopogo326990 users like ......its one of the most downloaded in gnome look :-;03:18
Miloszwell they have no taste :P03:21
Miloszit's quite surprising btw how many people think mediocre stuff is good03:21
pwnguinthis the tyrrany of "no" ;)03:21
Miloszhence you can not really let the majority take a decision when it's about style.. it sounds a little harsh but is IMO the reality03:21
Miloszs/take/make/03:22
pwnguinMilosz: so what you're saying is that style is making a decision that the majority of people will disagree with03:23
gopogothere should be public voting about themes/ icon and users should be allowed to contribute ..........some kind competintion03:23
Miloszbut it works the other way around - when someone who's proven of himself to have an eye for style decides something for most people and they "just" use it because it's let's say the default, it will be better for them even on a usability level (when it comes to Human for example usability and style are both a concern pertaining to the same thing), even if they don't notice it consciously03:23
Miloszpwnguin, not really ^03:23
gopogo~ competition03:24
Miloszthey will not disasgree if you give them something that is in the end effect better03:24
Miloszbut they are to say it carefully possibly unable to make that decision in the first place03:24
gopogowho made human theme any way  some guy  in canonical ?03:24
Miloszi.e. something better will have a better effect on them but they're not able to choose it for themselves (sometimes, with some things)03:24
Miloszit's style dictatorship and it's generally good03:24
pwnguinMilosz: and how do you reach this conclusion, that the general population cannot find good "optimimum" designs?03:25
Miloszpwnguin, usability studies03:26
gopogowell Human theme sucks .........its dull and boring ..........i rather use os x icons03:26
pwnguinMilosz: like the Wichita State studies?03:27
gopogohahaha03:27
Miloszit's superficially boring but it has properties that make it good for a default theme03:27
Miloszpwnguin, I don't know these03:27
pwnguinMilosz: they're the only ones i know of.03:27
Miloszpwnguin, I worked at a company who devised its own studies03:28
Miloszpartially ordered by clients, partially in the FLOSS usability work we've been doing03:29
gopogowhat happened to that company03:29
Milosznothing, but I don't work there anymore03:29
gopogooks is it still there03:29
pwnguinoh, i guess there's the nielsen group studies03:31
pwnguinbut they generally trust users03:32
gopogoas developer you might to use VIM or editor , w3m for browsing web  mutt for checking email but general users may not03:34
Miloszjust got to be observant03:35
Miloszand somewhat merciless03:35
Miloszif a user does not accomplish the task, then the interface or design has failed03:36
calcslangasek: ping, ispell-uk03:37
pwnguinMilosz: if that's the objective metric, where's it published?03:40
gopogoi am testing jaunty on my Dad .........who is Phd with 5 degrees ....Sr fellowship ....... who finds using checking his mail etc to be some kind of regrious routine  ......... he had tried 3 systems (windows xp -> moved bcos too many virus attacks ,  kde 3,59 and yesterday i installed jaunty for him)03:40
Miloszpwnguin, what is an objective metric?03:42
pwnguinMilosz: something used to settle disputes, mainly03:42
Miloszno03:42
gopogohe immediately asked me change ubuntu background ...wall paper etc03:42
Miloszthat was not what I meant to ask03:42
MiloszI meant what in this context are you calling an "objective metric"?03:43
Miloszand by this context I mean the current discussion03:43
pwnguinbroadly, usability03:43
pwnguinspecifically you mentioned accomplishing a task or not03:44
MiloszThen I don't understand to what exactly that I've said you're referring03:44
MiloszAnd it certainly doesn't make sense to me to ask where "it" is published because usability is being published everywher03:44
Miloszwhere*03:44
pwnguin"if a user does not accomplish the task, then the interface or  design has failed"03:44
Miloszpwnguin, that's usability testing 1-0-103:44
Miloszthat's nothing specific to any specific company, specific way of proceedings or specific type of test  (if anything, than that, but even that not really because it generally applies to all usability tests, just that the definition of 'failed' may differ from the usual meaning)03:45
pwnguinim not under the impression that any testing guides decision making03:45
gopogoyeah some kind Human Interface Guide line by apple03:46
Miloszno, but you're understanding accomplishment in a task-focused sense, and by task you understand a process that you do from start to completion with a goal in mind03:46
gopogoif dont have it why not follow apple's03:46
Miloszbut this is not how you read these terms in a usability testing context03:46
gopogowhats test group ?03:48
Miloszwhat a test group is?03:51
gopogowhats test group for usability testing for ubuntu ?03:53
Amaranthgopogo: We have a HIG03:53
AmaranthWell, GNOME does03:53
gopogoubuntu HIG == Gnome HIG ?03:54
pwnguinwell, probably not03:54
pwnguinfirstly because KDE is in "ubuntu"03:54
ScottKXubuntu too03:54
pwnguinyes, apologies03:54
gopogokde 4 ...ithink they got there own guidlines03:55
pwnguinsecondly because Canonical's been developing some new GNOME tech03:55
pwnguinthat might redo some of the guidelines if successful03:55
gopogolike ?03:55
pwnguinnotifications03:55
gopogois different from kde 4 notifications03:56
MiloszI worked on the KDE 4 HIG in that company incidentally03:58
MiloszBut I code myself using Gtk+ :)03:58
ScottKYes and the Canonical stuff is very unlikley to get accepted by KDE, so even if they do produce a KDE version, it will be Kubuntu specific.03:58
gopogoi was always liked KDE ...its much much flexible and feature rich04:00
pwnguinin fact, i dont see anything in the GNOME HID a all about notifications04:01
Amaranthpwnguin: long standing bug04:05
pwnguinall of which supports my statement that ubuntu's acting guidelines are not == GNOME's published HIG :)04:06
* ScottK suspects this would be a much better discussion for #ubuntu-offtopic04:08
gopogomaybe a minor difference04:28
calcgar i have to fix 7 dictionary source packages :\05:10
calcat least that was all that was buggy out of ~ 5005:10
=== Zic is now known as Guest87256
=== Guest87256 is now known as Zic
=== Yasumoto_ is now known as Yasumoto
=== Zic is now known as Guest48812
=== Guest48812 is now known as Zic
=== jefferai is now known as ferai
calchmm is requestsync broken?07:00
calchmm i just dropped the -d sid from it and it worked07:02
=== ferai is now known as jefferai
savvascan you disable the packagekit error notification that another package manager is using apt?09:39
savvasor is that deliberately shown?09:39
=== asac_ is now known as asac
cjwatsonkirkland: you assigned bug 236640 to me for jaunty, but I'm not sure there's much more I can do with it at this point beyond what I've already uploaded ...11:54
ubottuLaunchpad bug 236640 in open-iscsi "iSCSI install fails under hardy" [High,In progress] https://launchpad.net/bugs/23664011:54
cjwatsonI guess I can ask for testing11:54
=== thunderstruck is now known as gnomefreak
=== korn_ is now known as c_korn
=== murdok is now known as pablog
pochuslangasek: hi, bug #354475 should be fixed before Jaunty IMHO, could you look at the debdiff I've attached and maybe approve it?13:48
ubottuLaunchpad bug 354475 in libproxy "Ubuntu patch breaks libproxy" [High,In progress] https://launchpad.net/bugs/35447513:48
=== thunderstruck is now known as gnomefreak
jtismecjwatson, how would i go about downloading the  kde_ui.py code that R. Shtylman modified  for side by side fix, so i can test it14:37
=== jtisme is now known as jtholmes
=== thunderstruck is now known as gnomefreak
cjwatsonjtholmes: http://bazaar.launchpad.net/~ubuntu-installer/ubiquity/trunk/revision/3210 - but if you don't know how to apply that by hand, just wait until we upload it and a daily build is available16:01
RicardoPerezasac: Hi. I'm online, if you need me.16:35
asacRicardoPerez: hi. could you try to add some debug printf's in gdm code to see the font it actually uses?17:42
RicardoPerezasac: Mmhm... I can try it, but I'm not a developer, so I don't know if I'll be able to do that :)17:43
asacRicardoPerez: do you know how to rebuild deb/ubuntu packages?17:43
RicardoPerezasac: I'm using "dpkg-source -x" & "fakeroot debian/rules binary" in the past17:43
RicardoPerezs/I'm using/I used/g17:44
asacRicardoPerez: ok. wait a bit please17:47
RicardoPerezasac: sure17:47
asacRicardoPerez: hmm. seems i didnt put the debug output to the right code paths ... will check after dinner. if you are gone i will ask on the bug18:21
RicardoPerezasac: right! see you soon, then :)18:22
RicardoPerezasac: and thanks a lot18:22
RicardoPerezasac: BTW, what is suppossed to be the right default font? Bitstream Vera or DejaVu?18:25
jtholmescjwatson, thanks i had to go away for a few hours, i think i know how to apply will try any way19:00
philsfhi, I need help debugging jaunty's python packages. I can't open many GUI packages (jockey, apport, ubuntu-bug, software properties). I was thinking maybe I could reinstall some key packages, but don't know where to begin19:38
philsfI was directed in #ubuntu+1 to ask here19:38
slangasekcalc: what about ispell-uk?19:41
siretartyay! the intel 2.7 driver fixes resume in UXA on my 945GM hardware!19:51
ScottKsiretart: How about crashiness?19:54
siretartScottK: I've just installed it in /opt/somewhere and rebooted. uptime ~10 mins, but it already survived the first resume19:57
siretartperformance is way better than UXA in pure jaunty, but still not at EXA in intrepid. but its getting closer19:57
ScottKsiretart: I see.  Jaunty is pretty unuable for me as is (with 945GM), so it wouldn't take much for me to think we should update.19:58
ScottKunuable/unusable19:58
siretartwell, planet-penguin-racer is still unplayable for me19:59
siretartlet's try another suspend/resume19:59
siretartyay, survived again19:59
siretartScottK: btw, this upgrade required a newer libdrm as well20:00
siretartso it is of course a nogo for jaunty, IMO20:00
ScottKWe did a new libdrm like a week ago ...20:00
siretartwell, jaunty has libdrm 2.4.520:01
siretartthe intel 2.7 driver requires 2.4.620:01
siretartat some point, they seem to have botched with the version numbers, I think yesterday they released 2.4.920:01
siretartthat's what I compiled for my laptop20:02
ScottKFor me the downside risk seems nil.  Once Jaunty is released and I don't need it for testing anymore, I'm dumping it as it is.20:02
siretartfor karmic or intrepid?20:02
ScottKProbably Debian, but I'm not for sure yet.20:02
siretartI see20:02
ScottKIntrepid kernel is very problematic for me.20:03
=== ScriptRipper_ is now known as ScriptRipper
jdongsiretart: does 945GM crash X on resume from suspend for you?20:16
siretartjdong: it did in jaunty. it doesn't anymore with intel 2.7 for me20:16
jdongah; what does getting 2.7 consist of?20:16
jdong(it's driving me nuts in Jaunty right now)20:17
siretartgetting libdrm 2.4.9 and intel sources, installing build-dependencies, installing it to /opt/somewhere, add ModulePath directive to your /etc/X11/xorg.conf20:17
siretartluckily, I didn't need to touch any kernel modules..20:18
jdongokay, I was primarily worried about kernel land; that sounds straightforward20:18
jdongI'll have to put that on my TODO list20:18
siretartfor intel the kernel parts are in the kernel tree.20:19
siretartfor other drivers (radeon, etc) you need to build from the libdrm tree20:19
jdongcool; I was just worried the jaunty versions wouldn't suffice20:20
=== maco_ is now known as maco
calcslangasek: er can it get approved?20:30
calcslangasek: you asked me about the changes a week or two ago so i had documented them20:30
slangasekcalc: approved where?  I don't see a bug that ubuntu-release is subscribed to, nor an upload in the queue?20:31
calchmm i thought it had been20:31
calcperhaps you unsubcribed it when you asked for more details?20:32
calcit just needs sync not uploading20:32
slangasekoh, you subscribed ubuntu-archive for the sync20:32
calcah so i need to subscribe ubuntu-release as well?20:33
slangasekno20:33
slangasekyou need to subscribe ubuntu-release if you want an exception, but you said it doesn't need an exception20:33
calcwell it esentially isn't a big change even though it is a new upstream20:33
slangasekhmm, did you actually look at the diff for those bits that aren't just data?20:33
calcyes, they seemed to be fine20:34
slangasekok20:34
calcthe ooo xcu for example is just for building OOo extension with it20:34
slangasekthen yeah, it's fine to sync (though you should also set the bug back to 'confirmed' :)20:34
calcok20:34
calci also need syncs for espa-nol and ipolish (already filed bugs for them as well)20:35
calci found out just yesterday that the dictionary issue affected more packages than i thought, rene had just filed bugs on a few that he felt like doing20:35
calcso last night i went through all the dictionary packages in ubuntu (afaik anyway) and found the ones affected, only 7 packages total, the other 4 i still need to fix in debian first then sync20:36
calci'll try to get those done monday morning20:36
calcnot too bad out of ~ 50 source though20:37
slangasekI still think this should have been fixed in OOo.20:38
slangasekhaving to provide exhaustive country maps in order to use a dictionary for a language is wrong20:38
calcslangasek: agreed... however as upstream doesn't use system dictionaries at all it won't get fixed quickly, and i don't understand what is it doing since it is convoluted between their extensions and system support :\20:56
slangasekheh20:57
calcit might have a chance at getting fixed due to some other bugs in the dictionary support for multiple dictionaries of the same language20:58
calceg if you have a medical dictionary as an extension it currently completely overrides the system dictionary for the language20:58
calcso the only correct spelled words are the medical dictionary ones :\20:58
calcerm the only ones it thinks are correct20:59
* calc bbl, gone to easter stuff21:07
philsf hi, I need help debugging jaunty's python packages. I can't open many GUI packages (jockey, apport, ubuntu-bug, software properties). I hope I could reinstall or reconfigure some key packages, but don't know where to begin21:58
philsfsoftware-properties-gtk error http://paste.ubuntu.com/149125/22:00
philsfjockey-gtk error http://paste.ubuntu.com/149126/22:00
philsfapport ubuntu-bug error http://paste.ubuntu.com/149127/22:01
* Snova is looking at the software-properties-gtk one22:04
philsfthanks22:04
SnovaAnd... I have no idea, I just try not to let people feel they're ignored. :)22:04
philsfI have just uninstalled python2.4 and python2.522:05
SnovaLooking at the package files, the files are there...22:05
philsfSnova: this is the result of a bad upgrade, but I hope to save it without re-installing22:05
SnovaOh dear... check that this file exists: /usr/share/pyshared/softwareproperties/gtk/SoftwarePropertiesGtk.py22:06
philsfI don't believe they are really bugs in the packages, but maybe someone here understands what's going on, and has some spare time to help me22:06
Nafalloconsidering it's not only a weekend, but also the easter holiday...22:06
philsfSnova: it exists22:07
philsfNafallo: you're right, bad timing22:07
SnovaOpen a Python interpreter and paste the output of this:22:07
Snovaimport sys22:07
Snovaprint sys.path22:07
philsf['', '/usr/lib/python2.6', '/usr/lib/python2.6/plat-linux2', '/usr/lib/python2.6/lib-tk', '/usr/lib/python2.6/lib-old', '/usr/lib/python2.6/lib-dynload', '/usr/lib/python2.6/dist-packages', '/usr/lib/python2.6/dist-packages/PIL', '/usr/lib/python2.6/dist-packages/gst-0.10', '/var/lib/python-support/python2.6', '/usr/lib/python2.6/dist-packages/gtk-2.0', '/var/lib/python-support/python2.6/gtk-2.0', '/usr/local/lib/python2.6/dist-packages']22:08
SnovaNot to mention I'm not running on Jaunty...22:14
philsfSnova: http://paste.ubuntu.com/149136/ I just tried the following: force-remove python26 (and -minimal, and libpython26), and reinstall them22:14
philsfImportError: No module named shutil22:14
philsfthis should be simple, now, right :)22:15
SnovaEr, not so much.22:15
SnovaThat's supposed to be included with the standard install... if even *that* won't work, there are problems.22:16
savvasphilsf: is it a desktop install or a server?22:20
philsfsavvas: desktop22:21
savvasjaunty right?22:21
philsfI can install python2.4 and 2.5 without a problem22:21
philsfjaunty, yes22:21
savvasphilsf: can you send this to pastebin: apt-cache policy python python2.6 python2.6-minimal22:23
savvasphilsf: also this, sorry: apt-cache policy ubuntu-desktop22:24
philsfsavvas: http://paste.ubuntu.com/149138/22:25
savvasok try: sudo aptitude reinstall                 if not self.arguments['mode'] == 'file':dpkg-reconfigure22:26
savvas                    self.usage("ERROR: -i should be used with -f <filename>")22:26
savvas                self.arguments['in-place'] = True22:26
savvasoops22:26
savvaswait :)22:26
MiloszWhen will my sources show up in the PPA?22:26
MiloszI thought I can at least see immediately that they were uploaded correctly22:26
savvasMilosz: try #launchpad - I think it takes about 15 minutes22:27
philsfmaybe if I remove or purge all python packages that depend on py2.6, reinstall py2.6, then reinstall the deps from ubuntu-dekstop22:27
Miloszok that's good because I thought already something's wrong22:27
Miloszand, switching to #launchpad, thanks22:27
savvasphilsf: try this: sudo aptitude reinstall python2.6; sudo dpkg-reconfigure python2.622:27
savvasphilsf: anything?22:32
philsfsavvas: just a sec, I made things worse. trying to get back to the point where I had python-support installed :)22:33
savvasok :p22:33
philsfsavvas: will you be here in 30-60 min from now. I have to leave for now22:36
savvasphilsf: no idea, but try to reinstall everything if you know how to do it :)22:38
philsfsavvas: I will, thanks anyway22:38

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!