=== _salem is now known as salem_ | ||
mhall119 | popey: not yet, I'm still waiting on mandel to get it to me | 01:50 |
---|---|---|
=== chriadam|away is now known as chriadam | ||
=== salem_ is now known as _salem | ||
sethj | Are there any resources on building application indicators? All the documentation seems to have been taken off developer.ubuntu.com | 02:25 |
=== fginther is now known as fginther|away | ||
alecu | man, I'm so happy with the QZ81 app... it even has the TK85 keyboard! Having the emulator for my first computer in my pocket is so so awesome! | 04:08 |
=== chihchun_afk is now known as chihchun | ||
=== dpm-afk is now known as dpm | ||
dholbach | good morning | 06:03 |
=== oSoMoN_ is now known as oSoMoN | ||
dpm | mzanetti, morning! I built upon your changes to switch between production and sandbox and I now have a branch that generates separate packages, one for each plugin: https://code.launchpad.net/~dpm/reminders-app/switch-to-production/+merge/224503 - the only place I still see we're hardcoding the UOA provider is here: http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/src/app/qml/reminders.qml#L132 - so I guess w | 07:15 |
dpm | e need to pass through the --sandbox argument to the main qml file too. What's the best way to do this? | 07:15 |
mzanetti | dpm: hi | 07:54 |
mzanetti | just reading your mail... not really following tho | 07:54 |
dpm | hi mzanetti, so in here: QString hostname = m_useSandbox ? EVERNOTE_SANDBOX_HOST : EVERNOTE_HOST; it seems we might be able to just read the hostname using the Online Accounts API, rather than checking the sandbox argument | 07:57 |
mzanetti | dpm: how would you do that? | 07:58 |
mzanetti | dpm: I've seen the "account->value()" thing.. | 07:58 |
mzanetti | but: where to get the "account" from? | 07:58 |
dpm | that's where we ask mardy for documentation on how to use the Online Accounts API in C++ :) ^ | 07:58 |
mardy | dpm, mzanetti: if you want, you can do the same from QML; or do you need C++? | 07:59 |
mzanetti | dpm: so what I mean... we're using the QML api for it... I don't think its a good idea to connect to another api from 2 different points | 08:00 |
mzanetti | mardy: oh yeah, that'd be better | 08:00 |
mzanetti | mardy: qml, that is | 08:00 |
mardy | mzanetti: http://developer.ubuntu.com/api/qml/sdk-14.10/Ubuntu.OnlineAccounts.AccountService/ | 08:01 |
mzanetti | so I get the hostname... along with the selected accounts. ok. works for me | 08:01 |
mardy | mzanetti: read the "settings" property | 08:01 |
mzanetti | mardy: ack, thanks | 08:01 |
mardy | mzanetti: something like accountService.settings.hostname (if the key is called hostname) | 08:02 |
mzanetti | yep... will figure it | 08:02 |
dpm | mzanetti, mardy, I'm not too sure how to do it via QML. We're shipping 2 plugins (sandbox, production), and their service names are different (evernote vs. evernote-sandbox). | 08:02 |
mardy | mzanetti: I think that in that case the <setting> needs to be added to the .service file, not to the .provider file | 08:02 |
dpm | we're not using the hostname from QML atm | 08:02 |
dpm | so we'd still need a way to switch services in QML | 08:03 |
mzanetti | dpm: not sure what you mean | 08:03 |
dpm | mzanetti, http://bazaar.launchpad.net/~reminders-app-dev/reminders-app/trunk/view/head:/src/app/qml/reminders.qml#L132 | 08:03 |
mzanetti | dpm: if we have 2 plugins, each of them has a url hardcoded... | 08:03 |
mzanetti | dpm: the user selects an account. we get a pointer to that account and read the hostname associated with that | 08:03 |
dpm | right, but from QML we're not accessing the host url, we specify services | 08:03 |
mzanetti | but didn't mardy just tell us a way to get to the host url? | 08:04 |
dpm | mzanetti, yes, but we're not using the host URL in QML | 08:04 |
mzanetti | well, we're not right now... but if I get that information I'd change my branch in a way to use that instead of the bool useSnadbox flag | 08:05 |
dpm | mzanetti, right, and I'm not debating that. What I mean is that in QML selecting the host name will not be enough to switch providers, as the QML API needs to specify "service" to load the values from the right .service file (we install both evernote.service and evernote-sandbox.service) | 08:07 |
mzanetti | dpm: yeah sure... you select an account | 08:07 |
mzanetti | dpm: and we'll use the hostname associated with that account | 08:07 |
mzanetti | we're not letting the user to select the hostname in the reminders app | 08:08 |
mzanetti | maybe I'm still misunderstanding :/ | 08:08 |
dpm | mzanetti, yeah, yeah, I get that, but regardless of the host name, what's not clear to me is how we select an account in QML (evernote vs.evernote-sandbox) | 08:09 |
mzanetti | dpm: we have the 2 plugins installed. When you open the accounts settings page to add a new one, you'll see it twice in there | 08:10 |
dpm | mzanetti, yes, that's clear. What I don't know is how we switch between them in the app | 08:10 |
mzanetti | dpm: ah... if you have multiple accounts, we already pop up a dialog which lets you select the account | 08:10 |
mzanetti | iirc. | 08:10 |
mzanetti | so when the reminders app starts up and there are multiple possible accounts, you'll get a list to select which account you want | 08:11 |
mzanetti | and that list will contain sandbox and production server accounts | 08:11 |
mzanetti | Oh... I guess I see your point... we can't get both account types in one list | 08:12 |
mzanetti | is that what you mean? | 08:12 |
dpm | mzanetti, exactly: http://pastebin.ubuntu.com/7704704/ | 08:13 |
mzanetti | ok.. got it now | 08:13 |
mzanetti | mardy: any chance there's a way to get an AccountsServiceModel containing accounts from multiple services? | 08:14 |
mzanetti | dpm: right... worst case we'll just go with the -s switch | 08:14 |
mzanetti | as you already wrote | 08:14 |
dpm | mzanetti, yeah, that might be an option | 08:14 |
dpm | mzanetti, I've replied to your comment on https://code.launchpad.net/~dpm/reminders-app/switch-to-production/+merge/224503 - mardy, do you have any ideas on that (see last comment)? | 08:25 |
mardy | dpm: commented | 08:32 |
dpm | thanks mardy | 08:35 |
mzanetti | dpm: yep, what mardy wrote would be my suggestion too. "building" the same code twice with different flags to replace key and hostname at build time | 08:37 |
dpm | mzanetti, mardy, that sounds sensible to me too, and I understand how the key and hostname can be replaced. However, we still need two sets of files to be installed, right? http://pastebin.ubuntu.com/7704802/ | 08:40 |
mardy | dpm: yes, I think so | 08:40 |
mzanetti | dpm: yes. having it installed twice is ok. but not having to maintain the code twice | 08:40 |
mzanetti | mardy: hmm... the settings object seems to be empty here | 09:01 |
mzanetti | at least JSON.stringify(accountService.settings) gives me just {} | 09:02 |
mardy | mzanetti: weird! Can you push the branch somewhere? | 09:02 |
mzanetti | mardy: yep. lemme just do one more check first | 09:03 |
mzanetti | mardy: https://code.launchpad.net/~mzanetti/reminders-app/use-production-servers/+merge/224499 | 09:07 |
mzanetti | mardy: line 42 | 09:08 |
mzanetti | afaiu it should contain things like "RequestEndpoint" and such | 09:09 |
mzanetti | is that correct? | 09:09 |
mardy | mzanetti: so, the fact is that the <template> element is not present in http://bazaar.launchpad.net/~mzanetti/reminders-app/use-production-servers/view/head:/src/account-plugin/data/services/evernote.service | 09:11 |
mardy | mzanetti: so, the accountService object gives you a subview of the account | 09:11 |
mardy | mzanetti: there is an accountService object which uses the .provider file for the settings, and as many accountService objects for each .service file | 09:12 |
mzanetti | mhm... | 09:12 |
mzanetti | that means I'm using the wrong accountService object | 09:13 |
mardy | mzanetti: maybe. Or you add some settings to the .service file | 09:14 |
mzanetti | yeah, I guess that'd be easier | 09:14 |
mzanetti | dpm: fyi: ^ | 09:15 |
mzanetti | dpm: I've updated my branch to read the hostname from the accounts service... but we'd need to add the hostname setting to the .service files | 09:15 |
JamesTait | Good morning all; happy Chocolate Pudding Day! :-D | 09:16 |
t1mp | JamesTait: sounds good :) | 09:17 |
dpm | mzanetti, I'm on it | 09:17 |
JamesTait | t1mp, my kind of day, for sure. ;) | 09:17 |
dpm | mzanetti, https://code.launchpad.net/~dpm/reminders-app/switch-to-production/+merge/224503 - however, I don't think this will work yet. How to just rebuild the accounts plugin binary package instead of building all of the binary packages twice (which is a bit of human and LP builders waste of time) exceeds my packaging-fu | 09:21 |
dpm | mzanetti, also, I've called the setting "HostName" to be consistent with the naming in the XML file | 09:22 |
mzanetti | dpm: ack, will update my hostname code | 09:22 |
mzanetti | dpm: I would probably leave debian/rules alone but add additionall install targets to src/account-plugin/CMakeLists.txt | 09:26 |
mzanetti | dpm: so you just duplicate every install target in there to install it once for production, once for sandbox | 09:27 |
mzanetti | just change the configure_file() command for the two targets to configure the keyfile we want | 09:27 |
mzanetti | dpm: makes sense? | 09:28 |
dpm | mzanetti, yeah, that's what I was thinking when I failed to do it over debian/rules. Let me give it a go. | 09:28 |
mzanetti | dpm: ok. updated my branch | 09:29 |
dpm | but let me find out if it's possible to do it over debian/rules to have all the options on the table | 09:29 |
mzanetti | dpm: well, it sure is, but you'd need to patch cmake files nevertheless to allow building only the plugin | 09:30 |
dpm | oh, I didn't realise it'd require patching | 09:30 |
dpm | I'd rather not do it, so I'll go for the cmake-only approach | 09:31 |
mzanetti | dpm: yeah... just inject all the keys/hostnames/whatnot in one single cmake -D... call | 09:36 |
mzanetti | dpm: then in the accounts-plugin's CMakeLists.txt use the ones you need for each install step | 09:36 |
dpm | I still need to wrap my head around it, let me see what I can come up with | 09:37 |
=== ondra- is now known as ondra | ||
=== chriadam is now known as chriadam|away | ||
=== chihchun is now known as chihchun_afk | ||
nik90 | zbenjamin: hey in the manifest.json file, what should the hooks include? Does it require the project-name? | 11:48 |
zbenjamin | nik90: no the hook just needs the desktop file path and the apparmor file path | 11:49 |
zbenjamin | you appid is like: packagename.appname.version | 11:49 |
nik90 | zbenjamin: true but what about the hook name itself? | 11:49 |
zbenjamin | sorry packagename.hookname.version | 11:49 |
zbenjamin | nik90: well the hook specifies a application/scope inside your click package. A click package will be able to have more than one app , or maybe a app and a scope | 11:50 |
nik90 | for some reason it always replaces it with my bzr branch name...anything other than that fails | 11:50 |
=== MacSlow is now known as MacSlow|lunch | ||
zbenjamin | the hook name? | 12:02 |
zbenjamin | also how does it fail? | 12:02 |
zbenjamin | nik90: ^ | 12:02 |
nik90 | zbenjamin: I gave up..I spent the past several hours trying to fix the cmake files to work on the device. I cannot take this anymore :/ .. I will wait to see if someone can fix this next week | 12:06 |
zbenjamin | nik90: what does clock need? does it have a c++ backend yet? | 12:08 |
zbenjamin | nik90: i wonder if i could help you | 12:08 |
nik90 | zbenjamin: the new clock app has a sample C++ plugin at the moment which I plan to implement the QTimezone class. | 12:09 |
nik90 | I used the qtcreator ubuntu template with c++ plugin. | 12:10 |
zbenjamin | and you have my new packages installed? | 12:10 |
nik90 | yeah I installed them yesterday | 12:10 |
zbenjamin | when i tried i only got a black screen from the clock app, but it was creating and installing the click package | 12:10 |
nik90 | somehow qtcreator keeps on changing the hooks to " | 12:10 |
nik90 | fix-cmake-stuff": { | 12:10 |
nik90 | "apparmor": "fix-cmake-stuff.json", | 12:10 |
nik90 | "desktop": "fix-cmake-stuff.desktop" | 12:10 |
nik90 | } | 12:10 |
zbenjamin | for the bazaar branch name, i guess there is a rule in the cmake file that does that | 12:11 |
nik90 | well previously I set that to "@CMAKE_PROJECT_NAME@" so that it would get the correct name | 12:12 |
nik90 | and in the CMakeLists.txt it is declared as com.ubuntu.clock | 12:12 |
* nik90 is away | 12:15 | |
zbenjamin | nik90: are you working on the branch you gave me yesterday? | 12:20 |
zsombi | nik90: t1mp and I ended up to a conclusion that we need to put the PullToRefresh to be available only in UbuntuListView | 12:22 |
t1mp | zsombi: if we add the pullToRefresh only to UbuntuListView, it won't work for the clock-app right? | 12:22 |
zsombi | nik90: there are few issues with the topMargin handling... | 12:22 |
t1mp | zsombi: it is not a conclusion yet, but an idea ;) | 12:22 |
zsombi | t1mp: well, half conclusion :) | 12:23 |
t1mp | does the clock-app work with a flickable or a listview? in case of a listview it should be fine to use UbuntuListView | 12:23 |
zsombi | nik90: I guess it is a Flickable, right? | 12:23 |
zsombi | t1mp: the layouts there have different items | 12:23 |
zsombi | nik90: R U with us? :) | 12:24 |
t1mp | zsombi: I prefer to have the pullToRefresh for UbuntuListView only tomorrow, instead of landing nothing at all | 12:24 |
zsombi | t1mp: agree... | 12:24 |
t1mp | zsombi: seems like nik90 got frustrated with cmake and left ;) | 12:25 |
zsombi | t1mp: not envying him :D | 12:26 |
=== fginther|away is now known as fginther | ||
dpm | mzanetti, I think that should be it: https://code.launchpad.net/~dpm/reminders-app/switch-to-production/+merge/224503 - that branch already includes your changes, and I'll see if I can actually test the thing now | 12:35 |
=== MacSlow|lunch is now known as MacSlow | ||
=== renato_ is now known as Guest98529 | ||
mzanetti | dpm: not saying you have to change it, but if you want you could clean it up by doing: | 12:45 |
mzanetti | macro(add_evernote_account_plugin EVERNOTE_HOSTNAME EVERNOTE_KEY ... ) | 12:46 |
mzanetti | // do all the set(), configure_file() install() things once | 12:46 |
mzanetti | endmacro() | 12:46 |
mzanetti | add_evernote_account_plugin(http://sandbox.evernote.com ...) | 12:47 |
mzanetti | add_evernote_account_plugin(http://evernote.com ...) | 12:47 |
dpm | oh, I see, neat. I can do that, yes | 12:47 |
mzanetti | dpm: check out lp:unity8, in the directory cmake/modules/ you'll find a lot of examples on macros if you need them | 12:48 |
=== _salem is now known as salem_ | ||
dpm | cool, thanks | 12:50 |
pmcgowan | mzanetti, our silo with your fix failed, any ideas? cant find a lander right now | 12:54 |
mzanetti | pmcgowan: no... was wondering too why it wouldn't build in the silo. it does build in the CI job tho... | 12:55 |
pmcgowan | it failed fetching the source tarball | 12:55 |
mzanetti | yeah... seemed like a temporary network failure, so I triggered it again today morning, same error, while the CI job was able to fetch it | 12:56 |
mzanetti | I've just kicked it again now, maybe we were just really unlucky | 12:56 |
pmcgowan | weird | 12:56 |
mzanetti | pmcgowan: nope, still running into a timeout when trying to fetch the upstream tarball | 12:59 |
=== Guest98529 is now known as renato__ | ||
mzanetti | hmm... seems it uses a different url in the ci-train builder... | 13:01 |
pmcgowan | mzanetti, asking in ci channel | 13:03 |
pmcgowan | pmcgowan: But something seems to be setting a bogus version number | 13:04 |
pmcgowan | <cjwatson> bzr: ERROR: Unable to find the needed upstream tarball for package qtdeclarative-opensource-src, version 5.3.0+14.10.20140626. | 13:04 |
pmcgowan | Which is not what https://code.launchpad.net/~mzanetti/kubuntu-packaging/qtdeclarative-opensource-src/+merge/224517 says | 13:04 |
mzanetti | true... | 13:05 |
pmcgowan | mzanetti, can you join #ubuntu-ci-eng, cjw is learning some things | 13:08 |
mzanetti | sure | 13:08 |
=== chihchun_afk is now known as chihchun | ||
t1mp | mzanetti: I'm making the changes in uitk for exposing headerconfiguration and headerstyle in 2 small steps, here is the first: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/102-PageHeadConfiguration/+merge/224621 | 13:25 |
mzanetti | t1mp: you trying to confuse me with switching channels? :D | 13:26 |
t1mp | mzanetti: I though you didn't read my previous message :) | 13:26 |
t1mp | mzanetti: did I manage to confuse you? ;) | 13:27 |
mzanetti | t1mp: no, but you were close | 13:27 |
mzanetti | slightly change the link next time and you're good | 13:28 |
t1mp | mzanetti: if you look at the MR, you see that I change the name of the component and I move the documentation around :) | 13:28 |
t1mp | maybe that will complete the confusion | 13:29 |
mzanetti | t1mp: ok, so I'll drop this one completely, right? https://code.launchpad.net/~mzanetti/ubuntu-ui-toolkit/expose-NewHeaderStyle/+merge/224582 | 13:30 |
t1mp | mzanetti: yes | 13:30 |
dpm | mzanetti, hm, the macro is much neater, but it seems not to quite work when building the package, it complains about the provider file not being available at install time. I suspect it might have something to do with overwriting variables: http://bazaar.launchpad.net/~dpm/reminders-app/switch-to-production-cmake-improvements/revision/177 | 13:31 |
dpm | mzanetti, I think for now I'll leave the MP as it is | 13:31 |
t1mp | mzanetti: NewHeaderStyle is also no good name (but it was fine internal), I'll change it to PageHeadStyle | 13:31 |
dpm | so that we get the migration done | 13:31 |
dpm | I can look at making the cmake better later | 13:31 |
mzanetti | dpm: ok | 13:32 |
=== chihchun is now known as chihchun_afk | ||
mzanetti | t1mp: updated my unity8 branch to use your latest one. confirming it works fine | 13:52 |
t1mp | mzanetti: thanks | 13:53 |
t1mp | zsombi: ^ care for another review? https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/102-PageHeadConfiguration/+merge/224621 | 13:54 |
mzanetti | dpm: https://launchpad.net/~ci-train-ppa-service/+archive/landing-003/ | 14:13 |
dpm | mzanetti, it seems it's building arm now, and it's finished amd64 and others. Does it mean success, then? | 14:16 |
mzanetti | dpm: I hope so | 14:17 |
dpm | mzanetti, cool. I've also kicked off a build of Reminders app in my personal PPA (didn't put it in the core apps PPA so that Jenkins does not rewrite it). I'm going to test the new production plugin on the emulator now. https://launchpad.net/~dpm/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=utopic | 14:18 |
mzanetti | dpm: looks good :) | 14:19 |
zsombi | t1mp: get back to me tomorrow with all the reviews you want | 14:23 |
t1mp | zsombi: sure. There are two at the moment, I'll assign you so you can get started before I am around ;) | 14:25 |
zsombi | t1mp: kewl | 14:25 |
zsombi | t1mp: you will get the pullToRefresh tomorrow morning | 14:26 |
t1mp | zsombi: okay | 14:26 |
t1mp | mzanetti: here is the MR exposing the style: https://code.launchpad.net/~tpeeters/ubuntu-ui-toolkit/104-PageHeadStyle/+merge/224636 | 14:26 |
t1mp | mzanetti: if you test any of my branches, please reply on the MR so that the main reviewer (zsombi) knows it was tested | 14:27 |
mzanetti | t1mp: ack | 14:27 |
balloons | dpm, re: reminders I left comments about things on https://code.launchpad.net/~elopio/reminders-app/fix_with_account/+merge/224539. That's what I hoped we'd land yesterday. | 14:27 |
balloons | dpm, still working on the evernote sdk module merge as well, but the pep8 one landed at least :-) | 14:27 |
dpm | nice, thanks balloons | 14:28 |
dpm | balloons, what do you mean by isolate in your comment? | 14:29 |
rpadovani | mzanetti, hey :-) | 14:40 |
mzanetti | rpadovani: o/ | 14:40 |
rpadovani | mzanetti, couple of questions: I started to look to upgrade to Oxide, and I see there is a comment about a bug with flickable with some notes. Do you have any note, so I can see if the bug is present also in Oxide? | 14:41 |
rpadovani | well, the second isn't a question: here all proprieties of Oxide: http://paste.ubuntu.com/7706211/ I have to find a sobsitute to navigatorQtObjectEnabled and onMessageReceived, I'll try in next days ;-) | 14:42 |
mzanetti | rpadovani: without that flickable the webview was behaving really badly, if you remove it with the old webview you'll notice immediately | 14:42 |
rpadovani | ok, I try | 14:42 |
mzanetti | rpadovani: right... I think I've seen it somewhere already... the property is not the same any more, but there is some other mechanism to do message passing between qml and the webview engine | 14:43 |
rpadovani | mzanetti, yes, sure, I'll read the manual and I'm sure I find a way | 14:44 |
balloons | dpm, by isolation I mean the attempts to create a fake environment to run the tests in. Namely setting HOME | 15:01 |
mzanetti | t1mp: here's your 104 branch build output: http://paste.ubuntu.com/7706323 | 15:02 |
t1mp | mzanetti: thanks... fixed it | 15:03 |
* mzanetti kicks another build | 15:03 | |
balloons | dpm, basically the tests fail if you have an account on the phone already, and attempts to launch the app in a isolated environment cause the app to not load when run on the phone (upstart app launch doesn't seem to like it) | 15:03 |
dpm | ok | 15:05 |
zsombi | timp: the pullToRefresh is coming in ~10 mins :) | 15:08 |
mhall119 | \o/ | 15:28 |
dpm | mardy, I'm testing the account plugin for Evernote production, but I'm getting some strange behaviour: if I go to System Settings and choose the Evernote account, it then shows me the login page. After entering my credentials, it seems to work (Evernote asks me if I want to re-authorize the app to access this account). However, after hitting the button to re-authorize, it then hangs forever like this: http://i.imgur.com/TZi6P4u.png | 15:44 |
dpm | any ideas on how to get debug output? | 15:45 |
balloons | dpm, you can use account-console to see what's up.. do 'account-console list', then grab the id in question and account-console show ID | 15:48 |
dpm | balloons, that's a very good tip. However, I never get to create the account, so account-console list shows no accounts :( | 15:50 |
balloons | dpm, :-) | 15:50 |
mardy | dpm: try running it like this: | 15:53 |
balloons | elopio, are you about? | 15:53 |
mardy | export OAU_LOGGING_LEVEL=2 | 15:53 |
mardy | export OAU_DAEMON_TIMEOUT=9000 | 15:53 |
elopio | balloons: I'm here. | 15:53 |
mardy | online-accounts-ui --desktop_file_hint=/usr/share/applications/online-accounts-ui.desktop | 15:54 |
balloons | elopio, did we want to ping mardy about our findings last night as well? | 15:54 |
mardy | dpm: after killing any running instances; then open it from the system settings, and you'll find some logs | 15:54 |
nik90 | zsombi: In the clock app I am using a Flickable instead of listview since I need to show the clock, some labels and then finally the delegate listitem standards. | 15:55 |
zsombi | nik90: ok I thougth so ^ t1mp | 15:55 |
nik90 | zsombi: it is fine to only support the UbuntuListView since I have a working implementation in the clock app at the moment | 15:55 |
elopio | balloons: shouldn't we ping jdstrand instead? As the problems were with the home dir and the vars. | 15:55 |
elopio | I haven't seen the problem on delete account anymore. | 15:56 |
balloons | elopio, I was trying to remember if we resolved the fact we saw differences between accounts-console and python. I guess you are right, we didn't end with that problem | 15:56 |
zsombi | nik90: I got it fixed for a generic use :) | 15:56 |
dpm | mardy, if I run online-accounts-ui --desktop_file_hint=/usr/share/applications/online-accounts-ui.desktop from the terminal, then online accounts never starts, it stays forever in the 3 dots animation | 15:57 |
nik90 | zsombi: awesome. looking forward to using it in other places | 15:57 |
elopio | balloons: ah, you are right. mardy can help with the problem where with a temporary home we don't get the id#1 on the accounts | 15:57 |
mardy | dpm: yes, you need to start it from the system settings | 15:58 |
dpm | mardy, ok, on it. And where should I find the logs? | 15:59 |
zsombi | t1mp: nik90: the final updates on the MR: https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/pullToRefresh/+merge/219090 | 15:59 |
mardy | dpm: on the terminal where you lanched that command | 16:00 |
t1mp | zsombi: is it ready for a review? | 16:00 |
t1mp | zsombi: I'll have a look after dinner | 16:00 |
balloons | elopio, yes ok :-) indeed. It's coming back now. We need to understand about signond and some other things | 16:00 |
zsombi | t1mp: sure it is | 16:00 |
zsombi | t1mp: me left now :) | 16:00 |
dpm | mardy, ah, yes indeed. I was looking for a file in .cache/upstart | 16:01 |
balloons | mardy, ok so if you are done with dpm, elopio and I have a puzzling question. We're trying to run a series of tests back to back. Before each test, we want to setup a clean isolated environment. So we fake HOME and XDG_HOME_DIR and then use account-console to setup a fake account. | 16:07 |
balloons | This is all the work you helped elopio with in the paste | 16:07 |
balloons | The trouble is mardy we are seeing the account ids for the subseqeuent test runs not be id 1. That seems to indicate we're not setting up clean environments. So we started stopping signond before setup, which didn't seem to change anything. Is there anything else we need to do to ensure we have a blank slate for each test? | 16:09 |
mardy | balloons: libaccounts is using XDG_CONFIG_HOME | 16:25 |
balloons | mardy, ahh yes I see elopio has it right in the test, XDG_CONFIG_HOME | 16:32 |
balloons | mardy, should we need to stop signond? Is there any other service we need to account for? | 16:32 |
mardy | balloons: well, signond doesn't write to the accounts DB, it has its own DB (which is also under $XDG_CONFIG_HOME/signond/), but that doesn't affect the account numbering | 16:33 |
balloons | mardy, is there a reason to believe timing could play an issue if we ran tests in quick succession? | 16:35 |
dpm | mzanetti, I've tested reminders with the new plugin and left a comment with the result. It seems an account cannot be created because something is failing when requesting the user name: https://code.launchpad.net/~dpm/reminders-app/switch-to-production/+merge/224503 | 17:35 |
mzanetti | dpm: btw. silo 003 is built | 17:38 |
dpm | \o/ | 17:38 |
mzanetti | dpm: I haven't had the time yet to check it out | 17:38 |
mzanetti | mind giving it a test run? | 17:38 |
dpm | sure | 17:39 |
mzanetti | t1mp: http://design.canonical.com/2014/03/new-apps-header/ specifies a search behavior of the header. Can't find that in that code. I assume its not implemented yet, or am I just missing it? | 17:44 |
dpm | mzanetti, do you know why a 'sudo apt-get update && sudo apt-get upgrade' after having added that ppa does not offer the silo packages for upgrade? Do I need to upgrade each one individually? | 17:47 |
mzanetti | dpm: dist-upgrade | 17:47 |
dpm | same result | 17:48 |
mzanetti | hmm... that shouldn't be then | 17:48 |
dpm | ah, no wait | 17:48 |
dpm | dist-upgrade works | 17:48 |
mzanetti | ah ok | 17:48 |
dpm | ok, upgrading now | 17:48 |
mzanetti | *drumroll* | 17:48 |
* bluezone doesn't understand the purpose of the install file | 17:48 | |
dpm | mzanetti, so *more drumroll*... | 17:51 |
dpm | it works! :) | 17:51 |
mzanetti | *\o/* | 17:51 |
bluezone | are you in the twilight zone | 17:51 |
bluezone | nothing ever works :O | 17:51 |
mzanetti | my patches do! | 17:51 |
mzanetti | :P | 17:51 |
mzanetti | actually its not even mine :D | 17:51 |
dpm | :) | 17:52 |
mzanetti | dpm: cool stuff. can you do some more testing of other stuff, to make sure we didn't break anything else? | 17:52 |
mzanetti | i.e. walk through the unity8 test plan | 17:52 |
mzanetti | dpm: https://wiki.ubuntu.com/Process/Merges/TestPlan/unity8 | 17:53 |
dpm | sure | 17:53 |
dpm | mzanetti, so all the steps in that test plan work, except for installing apps (I think the store might still be broken?) and I've not ran the unity8 autopilot tests | 17:59 |
mzanetti | dpm: ok... that's good enough, given that its the exact same package as before + that one patch only in listview | 17:59 |
mzanetti | dpm: is it known that the store is broken? | 18:01 |
dpm | beuno, I'm still having a similar problem installing apps from the store. Was the issue from yesterday solved? What happens now is that when I click on Install it shows me 100% download, and then it fails showing me a dash page with a download error message. This is on image 98 | 18:01 |
beuno | dpm, it was totally fixed | 18:01 |
beuno | sbeattie, it's likely the image at that point | 18:01 |
beuno | er | 18:01 |
beuno | sorry sbeattie, I meant dpm | 18:01 |
sbeattie | heh, no worries | 18:02 |
dpm | beuno, ok, let me upgrade to the latest (100, I think) and try again | 18:02 |
beuno | dpm, the 100% download is how I get rid of all blame! | 18:03 |
beuno | once I give you 100% of the file | 18:03 |
beuno | NOT MY FAULT | 18:03 |
dpm | beuno, ah, I don't need that, I generally blame popey directly ;) | 18:04 |
popey | hmm? | 18:04 |
beuno | good, we all have the same strategy | 18:04 |
popey | I hate you all. | 18:04 |
dpm | some things never change | 18:05 |
* dpm hugs popey | 18:05 | |
popey | Consistent if nothing else. | 18:05 |
dpm | mzanetti, so I've re-run the unity8 test plan now with image 100 with the same results ^ | 18:25 |
mzanetti | dpm: store still broken... hmm... does it work without installing silo 3? | 18:26 |
dpm | beuno, I still can't install apps from the store, so I'll poke alecu :) | 18:26 |
dpm | let me try on the emulator | 18:26 |
dpm | mzanetti, strangely enough, it works well on the emulator. But I had the same issue yesterday without having installed the silo packages | 18:29 |
=== dpm is now known as dpm-afk | ||
kalikiana | if anyone happens to be familiar with pinyin, please have a look, I need help testing this https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1324955 | 19:05 |
ubot5 | Ubuntu bug 1324955 in Ubuntu UI Toolkit "Automatically commit text in AbstractButton" [High,Confirmed] | 19:05 |
=== boiko_ is now known as boiko | ||
alecu | dpm-afk: I'm testing #99 on mako, and I can install apps with no issues. | 19:18 |
=== renato_ is now known as Guest82399 | ||
=== Guest82399 is now known as renato__ | ||
=== salem_ is now known as _salem | ||
=== t1mp_ is now known as t1mp | ||
rpadovani | popey, I took a look to calc bugs, unfortunately no one is easy to solve for hackdays, they are all related to swype textfields or to advanced functions or to performance. As we discussed today, we need to do some work on this :( | 23:31 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!