[03:14] <ToddBrandt> asac: you around?
[09:21] <dholbach> good morning
[10:24] <Mithrandir> hiya Daniel
[02:22] <guardian> hi
[02:23] <guardian> from what i saw on the last image i made for my q1, ubuntu-mobile is only using matchbox-keyboard
[02:23] <guardian> does someone works on putting hildon input method framework in, now that it's released ?
[02:23] <guardian> s/works/work
[02:23] <Mithrandir> your observation is correct.
[02:24] <Mithrandir> the hildon input framework looks interesting, but I don't think anybody has started looking at it yet.
[02:31] <mdamt> Still you need a plugin which sends normal keypresses otherwise it doesn't work in browser.
[02:31] <mdamt> You could make a plugin which embeds the mb-keyboard for example.
[02:32] <Mithrandir> that's a slight show-stopper. :-P
[02:32] <mdamt> Yeah, it works only with Gtk apps.
[02:32] <guardian> that's not what i understood from the hildon input method framework
[02:32] <guardian> the plugin sends messages to the framework
[02:32] <Mithrandir> well, midbrowser is basically like firefox, so that ought to work?
[02:32] <guardian> then it's up to the framework to give the stuff to the apps
[02:32] <mdamt> Yes, the framework is currently for "normal" gtk.
[02:33] <mdamt> It doesnt work for Qt apps for example.
[02:33] <guardian> as far as i remember, the terminal on the nokia n800 receives key presses from the input method
[02:33] <guardian> but maybe they wrapped it in a gtk app
[02:35] <guardian> is there anything else evaluated ? scim ? uim ?
[02:42] <agoliveira> Mithrandir: Hi. Isn't the latest image-builder supposed to get the packages for lpia already? I didn't complete the process yet but so far it's downloding the packages for i386.
[02:43] <Mithrandir> agoliveira: I have not changed the default configuration, no.
[02:43] <agoliveira> Mithrandir: Ah, ok.
[02:43] <Mithrandir> and I'm trying to get IS to apply my patch so we'll have dailies built for LPIA
[02:44] <agoliveira> Mithrandir: But this will also apply to the standalone image-breator I suppose?
[02:44] <Mithrandir> "this" = ?
[02:45] <Mithrandir> I'm not changing the default configuration and won't for a little while
[02:45] <agoliveira> I see.
[02:45] <agoliveira> Thanks
[02:54] <seb128> Mithrandir: do you need the bluez-utils gstreamer plugin for mobile?
[02:54] <Mithrandir> seb128: I don't think so, no.
[02:54] <Mithrandir> why?
[02:55] <seb128> Mithrandir: because we got rhythmbox and totem crashers due to it and I'm wondering if that's worth keeping it for gutsy or if we should rather delay it to next cycle
[02:55] <Mithrandir> the alsa plugins works for me
[02:55] <seb128> if it's creating bugs on the desktop and not used anywhere I would suggest not building it for now
[02:55] <Mithrandir> agreed
[02:56] <seb128> another solution would be to binary split it
[02:56] <seb128> which might be the cleaner way
[02:57] <Mithrandir> I'll be fine removing it for gutsy and we can spend time on it for hardy.
[03:00] <seb128> ok
[03:41] <smagoun> Anyone else having trouble getting to launchpad?
[03:41] <Mithrandir> smagoun: yes, the canonical data centre is having some connectivity problems.
[03:42] <smagoun> ok, thanks.
[03:46] <Mithrandir> hiya mdz
[04:11] <agoliveira> Using the Q1 I have to enter my passphrase for the wireless conection twice and it's not being recorded, I have to enter it again everytime I boot the device. Does anyone have the same issue?
[04:12] <agoliveira> Once it's connected, it works well tough.
[06:16] <HappyCamp> asac, ping
[06:18] <asac> HappyCamp: ola
[06:19] <HappyCamp> asac, hello :)
[06:19] <asac> i git tree updated?
[06:19] <HappyCamp> Okay, I think I fixed the duplicate issue a few seconds ago, fingers crossed.
[06:19] <asac> hehe :) let me know when build is through and lintian run
[06:19] <HappyCamp> I just removed build-essentials from the build-depends list
[06:20] <HappyCamp> Was that the correct thing to do.  I'm kinda guessing based on the lintian output.
[06:20] <asac> yes
[06:20] <asac> build-essentials is always installed
[06:20] <HappyCamp> good :)  Okay the build should be done in about 5-10 minutes.
[06:20] <asac> did you do anything about the versioned lib links?
[06:21] <HappyCamp> No, what does that mean exactly?  All the (>= ##.##.##)  ???
[06:23] <HappyCamp> asac, or are you talking about package-has-a-duplicate-relation depends
[06:23] <asac> no ... you install libxyz.so.1.2.3 + libxyz.so ... but the plain .so link should not be shipped outside a -dev package
[06:24] <HappyCamp> Oh, I think that has been fixed.  I don't see it in my lintian check
[06:24] <asac> ok
[06:30] <asac> HappyCamp: you should definitly add a the interpreter to the network-customize script
[06:31] <HappyCamp> asac, huh?  what network-customize script?  ToddBrandt is the actual package owner
[06:32] <HappyCamp> So I don't know a great deal about moblin-applets.
[06:32] <asac> HappyCamp: ah ok ... and why do you need to explicitly add the depends? (you get package-has-a-duplicate-relation in lintian) ... those can just be removed from control
[06:32] <HappyCamp> asac, Is there a better way to do the depends?  By the way, you do have comit rights to the package.
[06:33] <asac> HappyCamp: as you can see in lintian they are automatically detected and added
[06:33] <asac> HappyCamp: so just don't add them ... ${shlibs:Depends} will contain them
[06:33] <HappyCamp> asac, don't add any of the ones that lintian complains about.  Delete them from the depends line.  correct?
[06:34] <asac> HappyCamp: yes
[06:34] <HappyCamp> will do
[06:34] <asac> HappyCamp: ok
[06:35] <asac> HappyCamp: please add #!/bin/sh to the network-customize script ... then we should be ready to go
[07:06] <HappyCamp> asac, will do
[07:08] <HappyCamp> asac, maybe I'm blind, but I can't find a script with the title "network-customize".
[07:08] <asac> me neither ;) ... maybe its generated during build somehow?
[07:08] <HappyCamp> Ah, okay a question for ToddBrandt 
[07:08] <asac> ToddBrandt: ^^
[07:08] <asac> ?
[07:08] <ToddBrandt> yea, I see your comment
[07:08] <HappyCamp> He is the owner of moblin-applets
[07:09] <asac> ToddBrandt: ok fine.
[07:09] <ToddBrandt> I'm changing the network-customize script
[07:09] <HappyCamp> Thanks ToddBrandt 
[07:09] <asac> thanks
[07:12] <ToddBrandt> asac: about the library version issues, basically I'm just letting automake handle the installation, how would you change the Makefile.am files to only install the .so and not the .so.x.x.x links?
[07:12] <asac> ToddBrandt: do we need those libs in /usr/lib at all?
[07:12] <asac> why not install them in pkglibdir ?
[07:13] <asac> ah you already do that
[07:14] <ToddBrandt> asac, I install like 12 libs into /usr/lib/hildon-control-panel, no choice on that, and then there are two common libraries that only moblin-applets will use, libhcpcommon and libemap, and I just changed those to go to /usr/lib/moblin-applets instead of /usr/lib
[07:15] <ToddBrandt> Is that enough or do I need to change it further?
[07:16] <asac> ToddBrandt: aren't the /usr/lib/hildon-control-panel/ libs technically modules?
[07:16] <asac> ToddBrandt: like ... http://www.gnu.org/software/automake/manual/html_node/Libtool-Modules.html#Libtool-Modules
[07:16] <asac> ?
[07:17] <ToddBrandt> asac: Well they're dynamically loaded by hildon-control-panel and executed, I suppose that makes them modules
[07:17] <ToddBrandt> lemme take a look at that doc
[07:18] <asac> right ... so try to add the -module LDFLAGS
[07:18] <ToddBrandt> ok
[07:22] <asac> ToddBrandt: now sport ... bbl
[09:23] <ToddBrandt> asac and HappyCamp: I added the -module LDFLAG to each Makefile.am, however it doesn't seem to have changed anything.
[09:23] <ToddBrandt> except the lintian warning no longer seems to appear
[09:53] <asac> ToddBrandt: did you push to branch already?
[09:53] <ToddBrandt> asac: yea
[09:54] <asac> ok let me pull
[10:06] <asac> ToddBrandt: yes should be fine ... only thing I would prefer is to not use the lib prefix for the modules: e.g. hcplib_LTLIBRARIES = background.la
[10:06] <asac> but its not that important
[10:08] <ToddBrandt> asac: oh, yea I thought that weas the standard for libraries
[10:10] <asac> ToddBrandt: if you see the automake manual above they drop the lib prefix as well ... as in the end modules are not libs.
[10:10] <asac> ToddBrandt: let me know if you want to change that ... since they are in not in libdir we can also rename them later i guess
[10:12] <ToddBrandt> I can change it, just let me make sure hildon-control-panel doesn't have some dependency on that prefix
[10:14] <asac> ToddBrandt: right ... please test the package as well ... i will only look for packaging issues ... not test functionality.
[10:47] <HappyCamp> asac, Mithrandir agoliveira who do I talk to about the rsync server not being on for ports.ubuntu.com?
[10:47] <HappyCamp> ] # rsync ports.ubuntu.com::
[10:47] <HappyCamp> rsync: connection unexpectedly closed (0 bytes received so far) [receiver] 
[10:47] <HappyCamp> rsync error: error in rsync protocol data stream (code 12) at io.c(359)
[10:47] <HappyCamp> mdz would you know?  above question
[10:48] <agoliveira> HappyCamp: Probably someone from IS but I can't point you to someone. I'll have to leave that to the old timers to answer.
[10:49] <agoliveira> HappyCamp: But I can ask around, of course.
[10:49] <HappyCamp> agoliveira, Thanks.  It was working and now it isn't.  We really kind of need to mirror the site.  Otherwise development is very slow :(
[10:51] <agoliveira> HappyCamp: I just remembered that our DC had some problems today. Maybe that's the reason of your problem.
[10:51] <HappyCamp> agoliveira, the http access is working though???
[10:52] <agoliveira> HappyCamp: Yes, it seems to be ok.
[10:52] <HappyCamp> http://ports.ubuntu.com/ubuntu-ports/
[10:57] <agoliveira> HappyCamp: Someone is looking at it already...
[10:57] <HappyCamp> agoliveira, Thanks!
[11:04] <agoliveira> HappyCamp: I was informed that we're just overloaded at this moment. The IS people is asking to hold on for some time and try again later but can't promisse any timeframe right now.
[11:04] <HappyCamp> agoliveira, okay.  Thanks for the update.  Is overloaded mean because of the DC issues or overloaded because too much bandwidth is being used?
[11:06] <agoliveira> HappyCamp: Looks like some syncing procedures were started some time ago and that's putting the rsync daemon to the limit. I think it's not a bandwidth problem.
[11:06] <HappyCamp> okay.  thanks.  I may bother you tomorrow if I am still seeing it :)
[11:08] <agoliveira> HappyCamp: No problem.
[11:16] <ToddBrandt> asac: automake complains if I take the lib prefix from hcpcommon and emap, those two are actually just pulled in by all the other modules, so I'll leave those two as alone
[11:18] <asac> ToddBrandt: yes ... they are libs so that is sane
[11:19] <ToddBrandt> I still get this:
[11:19] <ToddBrandt> dpkg-shlibdeps: warning: could not find any packages for libhcpcommon.so.0
[11:19] <ToddBrandt> dpkg-shlibdeps: warning: unable to find dependency information for shared library libhcpcommon (soname 0, path libhcpcommon.so.0, dependency field Depends)
[11:19] <ToddBrandt> How do I get automake to stop its whining
[11:25] <ToddBrandt> ok, well, I just took out dh_shlibdeps since I don't intend to share any of those libs with other packages
[11:34] <HappyCamp> agoliveira, it is working again :)
[11:34] <agoliveira> HappyCamp: Cool.
[11:40] <jbinder> Hi.
[11:41] <jbinder> So this won't be too good for the Nokia 770? :|
[11:43] <agoliveira> jbinder: Nope.
[11:44] <jbinder> :(
[11:44] <jbinder> It looked so cool.
[11:48] <asac> ToddBrandt: better keep it .. and just ignore the whining
[11:52] <ToddBrandt> asac: well if I'm going to formalize moblin-applets I might as well go the whole 9 yrads
[11:52] <ToddBrandt> It doesn't seem to be needed anyway
[11:52] <asac> ToddBrandt: it is ... you have explicit depends on libs in your Depends: line ... which is bad
[11:52] <asac> ToddBrandt: those should be dropped
[11:53] <asac> and be detected by dh_shlibdeps
[11:53] <ToddBrandt> oh ok, 1 sec
[11:53] <asac> ToddBrandt: don't confuse: dh_shlibdeps and dh_makeshlibs
[11:54] <asac> ToddBrandt: i see that you dropped those explicit depends already ... so without that your package should lack depends now
[11:54] <ToddBrandt> ok, so the debian/control file should have what removed?
[11:55] <asac> ToddBrandt: i think this was already done by HappyCamp a few hours ago ... let me pull latest
[11:55] <ToddBrandt> Depends: ${shlibs:Depends}, ${misc:Depends}, hildon-control-panel, evolution-data-server
[11:55] <ToddBrandt> that's the latest line
[11:55] <asac> yes thats good ... previously you listed some libs there
[11:55] <asac> but you need dh_shlibdeps now
[11:56] <asac> which you still have in git
[11:56] <ToddBrandt> ok, building, 1 moment
[11:56] <ToddBrandt> I think it's ready, I just have to make sure everything still runs
[11:56] <asac> yes please do
[11:57] <ToddBrandt> basically the two libs in /usr/lib/moblin-applets still have the lib prefix, and all the ones in /usr/lib/hildon-control-panel all have the lib prefix removed
[11:57] <asac> thats ok
[11:58] <ToddBrandt> lintian only displays the "lack of a man file for the binaries" warning
[11:58] <asac> if the libs in /usr/lib/moblin-applets might be of use for other applications as well we might want to make a real -dev package for them at some point
[11:58] <asac> ToddBrandt: good
[11:58] <ToddBrandt> should I create a man page with a quick blurb about the three binaries I install? moblin-settings-daemon, moblin-applets, and network-customize?
[11:58] <ToddBrandt> or will anybody care
[11:58] <asac> ToddBrandt: better no manpage than a bad one imo
[11:59] <asac> there are lots of binaries without manpage ... its bad, but not critical
[11:59] <asac> if you can come up with something reasonable ... even better
[11:59] <ToddBrandt> hmm, well I'm curious about it, should the man page just install with the moblin-applets package or does it have to be split off into a doc package?
[12:00] <Mithrandir> it should be in the main package
[12:00] <ToddBrandt> ok, I'll write a man page, I;ve never done that before
[12:04] <ToddBrandt> oh, snap, gnome-control-center has a man page too, I'll just use that as a template
[12:07] <asac> ToddBrandt: maybe look at docbook-to-man as well
[12:07] <ToddBrandt> k
[12:08] <AaronL2> ToddBrandt:  I think that the keybindings stuff might be a little two aggressive
[12:09] <AaronL2> ToddBrandt:  I was running emacs on the Samsung, and when I hit Ctrl-s to search, instead, it popped open the Tasks applications
[12:10] <ToddBrandt> AaronL: hah, didn't think about potential conflicts with custom package installs
[12:10] <AaronL2> ToddBrandt:  Also, it seems doubtful that actual devices are going to have a Ctrl key
[12:10] <ToddBrandt> You can change the keybinding to be something else, also the defaults can be less common
[12:10] <AaronL2> ToddBrandt:  they may not even have a USB port for a keyboard
[12:11] <AaronL2> ToddBrandt:  plus, the default in Keybindings says CTRL-S, not CTRL-s
[12:11] <ToddBrandt> AatonL: well they have to have buttons somewhere, my last task is to enable hardware buttons to be bound
[12:13] <AaronL2> ToddBrandt:  I seriously doubt that there will be a hardware button provided for the express purpose of starting the dictionary application, for instance
[12:13] <AaronL2> it's one thing to have one for the Web browser and perhaps the e-mail client
[12:13] <ToddBrandt> also I'm not sure it delineates between ctrl and ctrl-shift, or even if it's capable of doing so on the samsung with it's odd button setup
[12:13] <AaronL2> but calculator, terminal?
[12:13] <AaronL2> ToddBrandt:  I tried it, it actually shows "Shift" if you have the Shift key pressed
[12:14] <ToddBrandt> I would have calculator hooked up to a key, what if you want to do a quick math check in the middle of writing a message?
[12:14] <AaronL2> I think there will be very limited hardware keys for this sort of thing
[12:14] <AaronL2> may depend on the device though
[12:14] <ToddBrandt> well I'm just thinking about the samsung at the moment
[12:15] <ToddBrandt> both it and the Zi9 had 2 or three superfluous keys that could be put to use
[12:15] <AaronL2> for this to be really useful, I imagine a dedicated hardware key just for the Web browser
[12:15] <AaronL2> that has a web look on the actual button
[12:15] <AaronL2> users may have trouble remembering semi-arbitrary key combos like Ctrl-S
[12:16] <ToddBrandt> so what are you suggesting, that the defaults be unmapped, or that the functionality be removed altogether?
[12:16] <AaronL2> no, but perhaps only a few apps show up
[12:16] <AaronL2> not sure
[12:16] <AaronL2> e-mail, web, perhaps that is it, perhaps there is an advanced dialog to control other things
[12:16] <AaronL2> and the other applications, by default, are unmapped
[12:16] <Mithrandir> ToddBrandt: please don't just randomly grab keys like that; If the device has a keyboard, C-s will most likely be "save".
[12:16] <AaronL2> maybe unmapped is best
[12:16] <ToddBrandt> The gnome-settings-daemon had email, www, terminal, calulator, help, logout, power, and the volume knobs
[12:17] <ToddBrandt> I went ahead and added options for mapping all of the default applications
[12:17] <AaronL2> unmapped is probably best--an OEM might want to use a unique hardware key for specific apps
[12:17] <ToddBrandt> makes sense
[12:18] <AaronL2> okay, onto the next thing
[12:18] <ToddBrandt> so what are the core apps that could potentially need shortcuts straight out: www? 
[12:18] <ToddBrandt> I could leave that mapped
[12:18] <Mithrandir> or make it use XF86XK_WWW and so on
[12:18] <AaronL2> right, forgot about those special keys
[12:19] <AaronL2> still, it would be weird to see "XF86XK_WWW" in a user application
[12:19] <ToddBrandt> wait, what is XF86XK_WWW?
[12:19] <ToddBrandt> is that a gconf key?
[12:19] <Mithrandir> no, it's a X keysym.
[12:19] <AaronL2> I think it is an actual X key event
[12:19] <Mithrandir> it's what gets sent if you press the web browser button
[12:19] <AaronL2> on some keyboards, there is a Web button
[12:19] <ToddBrandt> oh, wierd
[12:19] <AaronL2> but, there should probably be a user-friendly name for this sort of key
[12:19] <AaronL2> I wonder what happens when it is pressed
[12:20] <ToddBrandt> ok, in theory that should already be usable, it's just a matter of figuring out the text designation of that key event
[12:20] <Mithrandir> the name of my email button is XF86Mail, and it just works out of the box.
[12:20] <AaronL2> yeah, I suspect that the application already has an appropriate key designation
[12:21] <AaronL2> Todd:  I think just Web and e-mail mapped initially, to the keys that Mithrandir suggested
[12:21] <ToddBrandt> ok, that's a great suggestion, I'll look up those keys and try to make them the defaults
[12:21] <AaronL2> cool :-)
[12:21] <ToddBrandt> ok, and just www and email
[12:21] <AaronL2> Mithrandir:  another issue, this one for you :-)
[12:21] <ToddBrandt> and I'll map the "f" key to logout....    just kidding ;)
[12:21] <AaronL2> Mithrandir:  you made a change recently to the xorg.conf file for samsungs
[12:21] <AaronL2> changed mouse2 to mouse1
[12:22] <Mithrandir> AaronL2: yes, without that it didn't work on my Q1.
[12:22] <AaronL2> Mithrandir:  However, as a result of this change, when I click with the stylus, it results in a double-click
[12:22] <Mithrandir> yeah, I noticed that.  If you have a better config, please commit it to git and prod me to upload.
[12:22] <AaronL2> Mithrandir:  for testing, I switched it back to mouse2, and stylus clicking started working properly
[12:22] <Mithrandir> that is, I noticed it today.
[12:22] <Mithrandir> hm
[12:23] <ToddBrandt> oh yea, I noticed that too, things pop open with a single click now
[12:23] <AaronL2> it's very annoying when you are trying to use menus
[12:23] <AaronL2> you click on the menu and it immediately opens and closes
[12:23] <AaronL2> a click and hold works though
[12:23] <AaronL2> you have to hold down for a second
[12:24] <Mithrandir> yup, please commit a fix and prod me to get it uploaded
[12:24] <AaronL2> what's the fix?
[12:24] <AaronL2> switching back to mouse2 will just result in the mouse not working again
[12:25] <Mithrandir> if that works for you, it might just have been something else on my end.
[12:25] <Mithrandir> but, I'm going to sleep now, just finished my beer. :-P
[12:25] <AaronL2> no drinking on the job! :-P
[12:26] <AaronL2> although, no one would know if you are working from home....
[12:26] <AaronL2> ... in your pajamas ....
[12:26] <ToddBrandt> well the typing might get worse than usual, better not tempt fate ;)
[12:27] <AaronL2> little hard to revert a change when I don't have an svn account
[12:27] <AaronL2> Todd:  how should I go about getting such an account?
[12:28] <ToddBrandt> moblin uses git, which svn account are you talking about?
[12:28] <AaronL2> oh, guess I would need a git account then
[12:29] <ToddBrandt> HappyCamp is the gatekeeper, and he must be appeased before entrance is allowed... $50 bucks is usually enough
[12:29] <AaronL2> ah right, HappyCampy
[12:29] <AaronL2> oops
[12:29] <AaronL2> I meant HappyCamp, really!
[12:29] <ToddBrandt> heh
[12:38] <ToddBrandt> asac: ok, the changes are all checked in with the exception of the man page, I'm writing that now
[12:40] <asac> ToddBrandt: thanks ... will look tomorrow and upload if all looks good. 
[12:41] <agoliveira> HappyCamp: Is a cheap bouncer :)
[12:50] <ToddBrandt> Does anybody know the best way to run utilities that need root access without bothering the user? e.g. on moblin, network configuration won't open when running as user ume, so I imagine sudo needs to be called somewhere while opening the controlpanel, so if there is no root password will it still query the user?
[01:05] <AaronL2> Todd:  modify /etc/sudoers
[01:05] <AaronL2> possibly
[01:07] <ToddBrandt> hmm, good call, thanks!
[01:07] <AaronL2> Todd:  what exe would you add to /etc/sudoers in this case?
[01:09] <ToddBrandt> /usr/bin/controlpanel, basically that would 
[01:09] <ToddBrandt> solve it
[01:09] <AaronL2> yeah, but it would result in giving all control panel plug-ins superuser privileges
[01:09] <AaronL2> that's a problem
[01:09] <ToddBrandt> exactly what I need
[01:10] <AaronL2> but, it's possible to install additional control panel applets
[01:10] <AaronL2> some user software may do this
[01:10] <AaronL2> and such control panel applets could do malicious things, possibly
[01:11] <ToddBrandt> yea, but they all run within a single process, is it possible to delineate priveleges within different modules of the same executable?
[01:11] <AaronL2> probably not, the way to go might to use inter-process communication
[01:11] <AaronL2> and have another EXE started up by the network applet
[01:11] <AaronL2> that is added to /etc/sudoers
[01:11] <AaronL2> 'course, that's more work :-)
[01:13] <AaronL2> ToddBrandt:  what can you do from the network config applet that requires superuser privileges?
[01:13] <ToddBrandt> Hildon desktop doesn't work that way, I actually wish it did, in gnome-desktop all those applets are full blown applications with their own execution space, they can be more easily controlled
[01:14] <AaronL2> sure you can do that will hildon desktop
[01:14] <ToddBrandt> it edits files in the /etc directory basically
[01:14] <AaronL2> you would use something like exec to start a process
[01:14] <AaronL2> from the network control panel applet library
[01:14] <AaronL2> and communicate with it using inter-process communication of some sort
[01:14] <AaronL2> perhaps you could use dbus as well
[01:15] <AaronL2> but, perhaps the network config applet doesn't need superuser privileges
[01:15] <AaronL2> for instance, the user can easily use /sbin/ifconfig -a
[01:15] <AaronL2> without superuser privileges
[01:15] <AaronL2> if it's just for providing information, all that info should be accessible without su privileges
[01:15] <ToddBrandt> yea I think maybe the easiest solution is just removing the superuser dependency
[01:15] <AaronL2> :-)
[01:16] <AaronL2> probably also the safest from a security standpoint
[01:16] <ToddBrandt> maybe add the network configuration files to a group that user ume is a member of
[01:16] <AaronL2> well, or make them readable by all
[01:17] <AaronL2> which files are these?
[01:17] <AaronL2> I would suspect that they are already readable by all
[01:17] <ToddBrandt> writeable actually, it needs to be able to edit connection settings
[01:17] <AaronL2> why?
[01:17] <ToddBrandt> it's the network configuration applet
[01:17] <AaronL2> true
[01:17] <AaronL2> but it could be just used for providing information
[01:17] <ToddBrandt> configuration implies writing :D
[01:17] <AaronL2> what is in there that the user needs to edit
[01:17] <AaronL2> no, not necessarily
[01:18] <AaronL2> that's not how it works on the Nokia Internet tablets
[01:18] <AaronL2> some of the control panel applets are just used to provide information
[01:18] <ToddBrandt> WEP keys, static ip settings, DNS servers, DHCP or no DHCP
[01:18] <AaronL2> hmm
[01:18] <AaronL2> I think that Nokia got around this security issue by having a completely separate application for connectivity settings
[01:18] <ToddBrandt> basically it's the manual configuration dialog, under 99% of people's use cases they'll never touvch it, they'll just use nm-applet's nifty little auto interface
[01:19] <AaronL2> see, it's perhaps more complicated than global settings
[01:19] <AaronL2> On the Nokia, you can have different, saved configurations for each connection