[07:39] good morning === cjwatson_ is now known as cjwatson === cjwatson_ is now known as cjwatson === cjwatson_ is now known as cjwatson === cjwatson_ is now known as cjwatson === cprov is now known as cprov-lunch [15:17] moblin-applets in hardy depends on libhildon1>=2.0.1, which isn't in hardy (hardy instead has libhildon-1-0. That dependency isn't explicit in the debian/control file, so presumably a builder did the wrong thing? What's the right way to fix? === cprov-lunch is now known as cprov [15:46] Fenario, let me know if you get my swap day request, it complained when I submitted it [16:02] StevenK: ^^^^^^ it's only 0300 there, why aren't you still at work?? :P [16:29] Mithrandir: ^^^^ re: moblin-applets [16:48] Hi gents. I trust I'm right in my belief ubuntu-mobiles not made it onto the HTC tytn II aka Kaiser aka AT&T Tilt [16:48] ?? === robr_ is now known as robr [17:34] ComradeHaz: you are correct [17:36] ta [17:36] sucks to be right ;) [17:42] heh [17:57] does a phone stack mostly use AT commands over a serial port? [18:06] smagoun: file a bug; we need to adjust the shlibs and rebuild moblin-applets. [18:21] Mithrandir: thanks, I'll file a bug. The problem's actually in libhildon's rules file, I guess that's worth a bug too. [18:23] Mithrandir, Do you think it is ok to add an entry in NetworkManager's dbus config to allow the user ume access? [18:24] Mithrandir, By doing so it fixes the nm-applet bug we talked about last week === Sciri_ is now known as Sciri [18:38] tonyespy, you there? [18:40] tonyespy_, you there? [18:40] raji: yes [18:40] tonyespy_, how did the testing go with marvel v9 driver? [18:42] raji: I didn't get much further... right now I'm trying to get our modified network-manager and nm-applet into the mobile ppa for hardy [18:42] tonyespy_, I see , then I will test that today.. [18:43] raji: test the v9 driver? [18:43] raji: on gutsy? [18:43] tonyespy_,yes [18:43] tonyespy_, yes [18:44] raji: OK... [18:45] raji: FYI, we're trying to switch to Hardy this week and it appears the v9 drivers aren't in Hardy. I just cc'd you on an email to robr and crew... [18:46] tonyespy_, Another question, while rootcausing the 'create new network', I see that iw_set_ext in nm_device_802_11_wireless_set_mode is returning error, I used wext driver option for wpa_supplicant, [18:46] tonyespy_,ok [18:46] raji: with v8 or v9? [18:47] tonyespy_ with ath_pci [18:47] raji: that's the name of the module, not the version [18:48] if you're using v8, you have to use the marvell-specific driver module in wpa_supplicant [18:48] I'm not sure it's worh testing with v8 [18:50] tonyespy_, I am testing with ath_pci on my samsung q1 ,not on our release hardware menlow... [18:51] tonyesp_, it is ath_pci 0.9 [18:51] raji: Oh... ath_pci is the low-level module of the madwifi stack. madwifi has very poor support for adhoc, so you probably won't get too far. [18:52] raji: you have to manually switch the mode of the driver from infrastructure to adhoc-demo [18:52] raji: even then it may not play with other non-Atheros wifi cards [18:52] tonyespy_, I changed the wireless interface to use adhoc mode and changed the code to use 'wext' instead of madwifi [18:52] raji: you would be *much* better off trying to test using the v9 driver on a crown beach... [18:53] raji: I'm pretty sure "adhoc" mode doesn't work, however the separate "adhoc-demo" mode does.... [18:53] tonyespy_, I am not aware of that 'adhoc_demo', is it a mode [18:54] raji: I haven't worked with madwifi for awhile... so most of this is from what I recall on the madwifi mailing list. [18:55] tonyespy_ will try that on crown beach.. [18:55] raji: again, I think you're better off working with the Marvell / Crown beach combination cause that's what our customers will be using. adhoc is tied very tightly to the driver being used. [18:56] raji: keep me posted [18:57] tonyespy_, another question , networkmanager 0.6.5 dont have ability to configure static ip , is it worth trying to root cause this as it anyway fails in the stage 5 [18:59] raji: right now, there's no plan to do anymore work on nm 0.6.5. If I were testing adhoc, I'd do so first without NM or wpa_supplicant and just try and use iwconfig/iwpriv, etc... [18:59] raji: I'm not even sure adhoc works for *any* drivers in 0.6.5. [19:02] tonyespy_, 0.6.5 doesnt support adhoc with 'any' driver, then I can just make sure if driver supports it or not . [19:03] tonyespy_, ok , I will try it on crownbeach and see what happens..I will let you know.. [19:04] raji: correct. If the driver doesn't support it, then it's pointless to try it with NM and wpa_supplicant. Good luck! [20:04] raji: 'ahdemo' mode is not 802.11 standard so it won't work with other cards unless they support it [20:05] raji: people like ahdemo (as opposed to ad-hoc) because there are no beacons, and the stations do not need to be associated in order to receive [20:25] bfiller__: I pushed a fix to bzr, but it doesn't show up on launchpad yet. Do you know if there's supposed to be a delay before my revision shows up on LP? [20:28] StevenK: Please rebuild libhildon for Hardy to fix LP: 189041 [20:30] smagoun_: yes, there will be a slight delay [20:31] StevenK: please build hildon-desktop as well as the patches are done [20:32] StevenK: Please rebuild moblin-applets too (LP: 189058) [21:44] smagoun, bfiller__: The simplest way is to either push the changes to bzr, and then ping me a lot to upload it, or just tell me what I screwed up and let me fix it. I'll fix libhildon, and wait for it to get published, and then rebuild moblin-applets, and then I'll look at hildon-desktop. [21:46] StevenK: I pushed the libhildon change to bzr already, it just needs uploading [21:47] smagoun: I saw that, thanks so much. [21:47] smagoun: And to answer your implied question about your push not showing up on Launchpad, it does take a little while to appear. [21:47] StevenK: I can give you a .dsc or a debdiff for the moblin-applets rebuild. I wasn't sure whether it's easier for you to use one of those or just do the rebuild yourself. [21:48] StevenK: thanks. I figured it out after a couple minutes of F5F5F5F5 :) [21:49] smagoun: To be completly honest, it's very easy to trigger a rebuild (it's only a new changelog entry), so I'll just fix it. [21:49] ok, that's what I figured [21:49] smagoun: Maybe we want to think about putting moblin-applets into bzr? [21:50] StevenK: we could do that, sure. [21:51] smagoun: Right, libhildon uploaded. It will take until :04 or so to show up as published, and then another publisher cycle for the binaries. I have to wait until that publisher run is done before uploading moblin-applets or hildon-desktop. [21:52] StevenK: thanks much. Our nightly build will pick up the changes automatically at about 0800UTC [21:52] smagoun: About moblin-applets, let's play wait and see -- see how much stuff we need to change in it, and how much work we have to do together on it. [21:53] smagoun: At 0800UTC I should have all three (and *hopefully* libosso) sorted out. [21:53] ok, that works too. We're fine working with patches for now. [21:54] * StevenK kicks tomboy. [22:09] StevenK: BTW I filed bug 189058 for the moblin-applets rebuild [22:09] Launchpad bug 189058 in moblin-applets "Rebuild moblin-applets to fix bad libhildon dependency" [Undecided,New] https://launchpad.net/bugs/189058 [22:10] StevenK: for some reason, debuild still fails even after fetching fresh hildon-desktop unless I use -i.bzr. What can I do to fix this? [22:10] StevenK: did I forget to commit something? [22:12] bfiller__: No, the reason it fails is because it can't represent the changes to the files under .bzr [22:13] StevenK: how do I fix that? [22:13] bfiller__: Export it from bzr as opposed to pulling -- you won't be able to push things into bzr from that one, but you won't have to add -i.bzr [22:15] StevenK: ok, thanks === Sciri_ is now known as Sciri [22:22] ChickenCutlass_: no, we should fix it so that when the desktop session starts, pam_foreground knows you're on the console. [22:24] smagoun: Still around? [22:24] StevenK: yup [22:25] smagoun: Can you check git for m-i-c and see if fsets for m{ccaslin,enlow}-hardy-whatever still mention moblin-chat? [22:26] StevenK: Only the mccaslin-lpia and menlow-lpia fsets mention moblin-chat. I think this is correct. [22:27] Sounds good to me. [22:27] smagoun: When did you want to get the new m-i-c into Hardy? [22:28] smagoun: I have to upload it to fix mccaslin (since I'm a bozo), but if you're ready for the new shiny upstream, we can push that in. === rzr is now known as rZr [22:29] StevenK: anytime. I expect it might need to be coordinated with Mithrandir; I don't know how the nightly build gets its MIC, and the configs changed incompatibly at some point recently. [22:29] nightly build=the Ubuntu nightly; we have our MIC built for gutsy here [22:30] StevenK: If you want to push your fix now we can deal with shiny upstream later on. I don't want to lose out on MOTU points though! :) [22:33] smagoun: Right, I've pushed my fix. [22:40] StevenK: OK. What was your fix? do I need to redo the 0.40 package? [22:40] smagoun: My fix was to take the changes I made for menlow's fsets to the mccaslin fset. [22:40] smagoun: Which was in the diff I sent you [22:41] ok, I think I pushed those into all the fsets save the moblin ones for 0.40, so we should be all set. [22:43] Yay! [22:44] StevenK: I'm heading home, send me email if you need anything === robr_ is now known as robr