/srv/irclogs.ubuntu.com/2015/09/21/#ubuntu-app-devel.txt

=== chihchun is now known as chihchun_afk
=== chihchun_afk is now known as chihchun
davidcalleMorning o/07:14
=== athairus is now known as afkthairus
lesamouraiimplementing telepathy observer client properties without .client file , any help?(have implemented AbstractClientObserver)08:49
karnizbenjamin: hey buddy. got a question. how does the qmake project translation extraction work exactly? does it just look for tr() calls? cose I realized on Friday I can't wrap strinsg for translation in tr() calls, because it's a QObject method on its own. I was wondering if you had any suggestions on the matter.10:32
zbenjaminkarni: we use that command xgettext -o $$template_pot.target --qt --c++ --from-code=UTF-8 --add-comments=TRANSLATORS --keyword=tr --keyword=tr:1,2 --keyword=N_10:34
karnizbenjamin: what is that last one, N_ ? like a C macro?10:35
zbenjaminkarni: tbh i modeled it after the cmake projects which used that, i just applied what the cmake projects do to qmake :(10:36
karnizbenjamin: but I take N_ is like a C macro? N_("string to stranslate")?10:36
zbenjaminkarni: probably like a define yes10:36
zbenjaminkarni: something like "#define N_(String) gettext_noop (String)"10:37
karnizbenjamin: cose we can't use QObject's tr() because that's the .ts not .po translations10:37
karnizbenjamin: right10:37
zbenjaminkarni: yeah right oO10:37
zbenjaminkarni: bad choice to use tr i guess10:37
zbenjaminkarni: anyhow N_ should work10:37
karnizbenjamin: what about if I need to translate a plural though? I mean, if we have tr and N_, I can't use same for plural translation, can I? (I mean the N_ one) I can't overload macros, right10:38
zbenjaminkarni: maybe i should add a possibility that you guys can pass additional arguments to xgettext10:38
zbenjaminkarni: yeah i guess thats what the tr1:2 should do *facepalm*10:39
zbenjaminkarni: clearly until you guys noone used the tr stuff in C++10:39
karnizbenjamin: right. turns out, there's a pretty common use case. :( qmake based scope (I know the SDK doesn't provide that template yet though)10:40
zbenjaminkarni: well scopes usually do not use QObjects10:40
karniactually.. the problem I have is in the push helper10:40
karniwhich uses signals/slots10:40
karnizbenjamin: yes, I take that back10:40
karnithe push helper - translation of notifiation messages10:40
zbenjaminkarni: waaaaait10:40
zbenjaminkarni: did you try to specify the global namespace?   ::tr()  ?10:41
karnizbenjamin: I have not, not sure how I go about it?10:41
zbenjaminkarni: that tells the compiled not to use the QObjects tr , but I do not know if xgettext still pulls the string10:41
karnioh10:41
zbenjaminkarni: well in your function just use ::tr("MyString")10:41
karniha, I will definitely check it out right now :)!10:42
zbenjaminkarni: lets just hope xgettext speaks proper C++ :D10:42
karni;D10:42
karnizbenjamin: how does that fit into the macro story? I currently have strings wrapped in _(...) in that QObject push helper, and #define _(value) gettext(value)10:44
karnizbenjamin: if I used ::tr(), would it still be a proper macro call, if I renamed the #define above to tr(...) ?10:45
karnizbenjamin: parallel question - is the translation of desktop files in qmake projects released already for the 15.04 SDK?10:46
zbenjaminkarni: good question with the macro, however you could have a global function named tr() :D10:46
karnilol. haha, we're taking this to the extreme. /me will try10:47
zbenjaminkarni: i'm not 100% sure, we currently try to figure out how to release that stuff .... its complicated :D10:48
om26er_zbenjamin, Hi! what's element is the handle of a slider ?11:12
om26er_I need to drag the handle for some test11:12
om26er_t1mp, Hi!11:19
=== om26er_ is now known as om26er
t1mpom26er: hello. I'm in a meeting for the next 30m11:20
om26ert1mp, ok, will wait till then.11:20
zbenjaminom26er: thats a question for t1mp :)11:29
aquariusAny QML experts around to explain how to do schema version upgrades for LocalStorage sqlite databases? I don't think I understand it.11:34
om26ert1mp, Hi! are you free :)11:58
t1mpyeah11:59
t1mphmm12:00
t1mpom26er: it is not good to refer to internal implementation in the tests, normally you have to use the autopilot custom proxy objects for that12:00
t1mpI see there is no CPO for slider... :(12:00
t1mpom26er: I'm checking the slider implementation, but the thumb doesn't have an objectName that you can use to select it...12:02
t1mpom26er: the proper solution is that we add an autopilot CPO for Slider which has a dragTo(value) function12:06
t1mpom26er: can you report a bug for that on https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit12:06
t1mpom26er: because you know your specific requirements best12:06
om26ert1mp, i'll report the bug but whats the element name of the thumb?12:09
om26ert1mp, right now if i have the slider then I can definitely select the thumb and rely on it12:09
t1mpom26er: it is an UbuntuShape, but it does not have an objectName12:11
om26ert1mp, oh, well maybe i can rely on its colors as a secondary parameter ?12:11
t1mpom26er: http://bazaar.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/trunk/view/head:/src/Ubuntu/Components/Themes/Ambiance/1.3/SliderStyle.qml12:12
t1mpom26er: do you need to use the thumb for the test? It is also possible to just set the value of the slider12:13
t1mpom26er: it is not ideal, but the Slider.__styleInstance.thumb Item may work for you? but only until we have the proper CPO12:14
t1mpbecause it depends very much on the internals of Slider which may be changed12:14
om26ert1mp, i need the thumb because I want to slide incremental, relative to the previous position12:15
t1mpom26er: what I propose is to report the bug, and in your test use Slider.__styleInstance.thumb but add a note there referring to the bug. So if we change the SliderStyle implementation and the test breaks it is easy to track down the problem12:17
t1mpom26er: or for now use slider.value = slider.value + dv12:17
om26ert1mp, or the simples is to just add the objectName for the thumb ?12:19
om26er*simplest12:19
om26ert1mp, will this work slider.select_single('__styleInstance').select_single('thumb')  ?12:21
t1mpom26er: you cannot select by id, so we normally add an objectName string12:22
t1mpom26er: the flaw with adding an objectName is that if we would change the implementation not to have that object any more, then tests will break12:22
t1mpso having a function in autopilot like Slider.dragThumbTo(value) would be best12:22
* t1mp bbl, lunch.12:23
=== chihchun is now known as chihchun_afk
mhall119bzoltan: oSoMoN: the ubuntu-sdk isn't installable on wily at the moment because of some unmet oxide dependencies12:58
mhall119liboxideqt-qmlplugin : Depends: liboxideqtcore0 (= 1.9.1-0ubuntu1) but 1.9.2-0ubuntu0.15.04.1 is to be installed Depends: liboxideqtquick0 (= 1.9.1-0ubuntu1) but 1.9.2-0ubuntu0.15.04.1 is to be installed12:59
aquariusSaviq, do you understand how DB upgrades in QML LocalStorage are meant to be done? I don't think i get it.13:12
Saviqaquarius, never did, but looking at the docs... you open the db as usual in the lower version, call .changeVersion(from, to, callback(tx)) on it, and in the call back you perform any sql you need on the transaction passed13:15
Saviqaquarius, seems it's Web API, so http://www.w3.org/TR/2009/WD-webdatabase-20091029/#introduction has some examples13:16
aquariusSaviq, yeah. What I don't think I get is... how this builds up. Like... I do openDatabaseSync(name, version), but what happens if it's not *at* the version I supply?13:16
aquariusah!13:16
aquariusthat's useful, thank you pal!13:16
Saviqaquarius, right, it raises, so you probably need to try and catch the exception13:17
Saviqaquarius, OTOH it will have to be at that version since you need to open it first, and you open a particular version13:17
aquariusand then loop until I find what version it *is * at? gordon bennett.13:18
aquariusOnce I've worked this out I think I may need to write a blog post explaining it all :)13:18
Saviqaquarius, so IIUC, you try and open at the highest version, then downgrade until the open call succeeds, and run the upgrade there13:18
oSoMoNmhall119, indeed there’s no oxide 1.9.2 in wily, only in the vivid overlay PPA13:18
aquariusSaviq, that makes sense13:18
aquariusthe documentation is utterly, utterly unclear about this :)13:19
oSoMoNbzoltan, any idea why the ubuntu SDK tries to pull in a version of oxide that’s not available in wily?13:19
Saviqaquarius, db.changeVersion('', 'foo'...) seems to be used to initialize the db13:19
oSoMoNchrisccoulson, see ^^, it would be good to push oxide 1.9.2 to wily13:19
Saviqaquarius, hum hum, actually, open...() will not raise, but create the database13:20
bzoltanoSoMoN: No idea .. some dependency magic in the deep13:20
aquariusSaviq, now I'm way confused13:20
Saviqaquarius, yeah, me too13:20
aquariusSaviq, this is why I was hoping that someone had done this and documented it :)13:21
Saviqaquarius, but I'm starting to feel like you need to call changeVersion('', '$expected') to see if the database is fresh13:21
aquariushttp://www.gajdos.sk/ubuntuapps/qml-sqlite-upgradedb-change-db-version-using-changeversion/ is about the only example I can find, and it doesn't explain anything, because he opens the DB with an empty version number13:21
aquariusand if that's what you're *meant* to do to make anythnig work, then I don't understand why you'd bother designing an API which takes a version number parameter at all :)13:22
Saviqaquarius, and if so, initialize it, otherwise catch and go down your expected schema numbers13:22
mhall119oSoMoN: bzoltan: I assume that the SDK is landing in both the overlay PPA and wily at the same time, so any of it's dependencies should be as well13:22
mhall119oSoMoN: is there any reason not to get oxidd 1.9.2 into wily?13:23
Saviqnow I wonder how do you open a db that's in current schema... changeVersion($current, $current...)??13:23
oSoMoNmhall119, no other reason than "it hasn’t happened yet"13:23
Saviqaquarius, http://stackoverflow.com/questions/18052225/db-changeversion-doesnt-work-as-expected13:23
bzoltanmhall119: oSoMoN: i do not really land on wily... as UITK does not compile on wily due to a gcc black magic13:23
mhall119bzoltan: we're a month away from release, we should probably make sure the SDK is installable and working :)13:24
aquariusah, that looks useful. And by rpadovani too! ping :)13:24
bzoltanmhall119: there is UITK in wily .. but not as new as on the phone releases13:25
mhall119bzoltan: ok, that needs to be fixed then13:25
bzoltanmhall119:  and since wily will _never_ land as an upgrade on any phone/tablet device I have hard time to beleive that it si super urgent to bring up the UITK to the level of the OTA releases on Wily13:26
mhall119bzoltan: wily is going to be a supported release on the desktop, it will be supported when vivid id EOL, it needs to have the UITK and SDK13:26
bzoltanmhall119: It would be super nice to see a synced UITK, but the build failure  is way beyond my knowledge13:26
aquarius"Hi! I'm an Ubuntu app developer, and I'd like to be able to test my application on my desktop before running it on the phone, because it's one thousand times easier to do it that way. Can you explain why the SDK doesn't work on my desktop? I thought we were one Ubuntu now?"13:27
aquariusthere's a reason :)13:27
mhall119^^13:27
bzoltanaquarius: we went through this zillions of time... there is no OS what provides you such developer experience. It is not a sustainable development model13:28
mhall119bzoltan: no, we went through this in the opposite direction, and we agreed you didn't have to support new development from *old* releases of Ubuntu13:28
aquariusI get that I, a person who wants to stick on the LTS, get "supported" by having to have a device to test on, but I didn't think that applies to everybody everywhere -- that desktop development of apps is officially declared impossible and not supported...13:28
mhall119but the Ubuntu SDK *must* be installable and working in the *latest* stable release of Ubuntu13:28
rpadovaniaquarius, pong?13:28
bzoltanmhall119:  aquarius -> https://developer.ubuntu.com/en/blog/2015/03/18/everything-you-always-wanted-know-about-kits-were-afraid-ask/13:29
aquariusrpadovani, got a few minutes to talk about version upgrades to QML LocalStorage databases? You seem to be the expert, looking at http://stackoverflow.com/questions/18052225/db-changeversion-doesnt-work-as-expected which is the best documentation anywhere on the internet ;)13:29
mhall119bzoltan: I can't use a Kit if I can't even install the SDK!13:30
bzoltanmhall119:  there is a fully functional and working UITK + IDE on Wily13:30
mhall119bzoltan: no, there isn't, that's what I'm saying13:30
mhall119apt-get install ubuntu-sdk *fails*13:30
popeymhall119: in the archive or the ppa?13:30
mhall119popey: archive + sdk-team PPA13:30
popeyerk13:31
popeywell that's not good13:31
mhall119following instructions on https://developer.ubuntu.com/en/start/ubuntu-sdk/installing-the-sdk/13:31
bzoltanmhall119: what is the failure?13:31
rpadovaniaquarius, sure thing, it's  a long time I don' t take a look, but feel free to ask13:31
mhall119bzoltan: the dependencies issue with oxide I originaly pinged about13:31
bzoltanmhall119: could you please pastebin the logs?13:32
aquariusMy objection is that even if the SDK is installable, that exact document says " You simply change from the LTS Desktop Kit to the 15.04 based armhf target and you have a whole different compiler toolchain and API set at your service." and should say "don't bother to install any sort of LTS Desktop Kit because you can't use our stuff to develop desktop applications at all; we do phones" :-) But mhall119's issue13:32
aquarius is way more important here!13:32
popeymhall119: seems I dont have that ppa enabled, and sdk works for me (currently) on wily :S13:32
mhall119bzoltan: http://paste.ubuntu.com/12514263/ is the root cause13:32
aquariusrpadovani, I don't think I understand how it's all meant to work. Which version number do I pass to openDatabaseSync, for example?13:33
mhall119priority #1 is to get the package deps fixed so I can isntall the SDK13:33
mhall119but we still need to have the latest supported UITK working in wily, it's crazy to not have the latest supported version of our own toolkit not working in the latest supported version of our own distro13:34
bzoltanmhall119: I have  no idea what has caused the dependency problem. It is  not something I have changed ... actually we do not change anything there.13:35
rpadovaniaquarius, that's a good question. I leave "" usually, because it seems it's not used13:35
bzoltanmhall119: Well.. the UITK issue is related to the toolchain upgrade13:36
rpadovaniaquarius, like LocalStorage.openDatabaseSync("app", "", "", 5000);13:36
mhall119bzoltan: ok, and that's not in your wheelhouse, fair enough, but who is working on fixing it?13:36
bzoltanmhall119: but keep in mind that the Desktop environment is _NOT_ supported as runtime environment quite yet.13:36
bzoltanmhall119:  somebody who made that change and broke the sdk13:37
mhall119bzoltan: do you know who that is and are you tracking their work on fixing it?13:37
bzoltanmhall119: I have heard about it just now13:37
bzoltanmhall119:  I wonder who landed '1.9.2-0ubuntu0.15.04.1' oxide on Wily ...13:38
mhall119bzoltan: nobody did, I believe that's the problem I hit13:39
bzoltanmhall119:  I will check with the CI folks ...13:39
aquariusrpadovani, ah, so if you want to support upgrades, you *have* to open with a blank version? got it.13:40
bzoltanand back to aquarius :) my favourite developer :) so... when you are developing an android app on Win10 .. do you expect the app to be running on Windows? I guess no :) When you develop and iOS app OSX, do you expect the app to run on OSX? I think no...because runtime env and devtime env are different.13:41
mhall119bzoltan: falrse comparison, android never claims to be capable of that13:41
bzoltanaquarius: and actually we are quite lucky that at certain level we can use the dev environment as runtime13:41
mhall119we advertise that we do13:42
aquariusbzoltan, indeed I do not. Because they're running different OSes. We've been very clear that Ubuntu phone and Ubuntu desktop are not different things and it's all one Ubuntu, from IOT devices to phones to tablets to desktops to supercomputers.13:42
bzoltanmhall119:  and so we should not do either .. because we are not13:42
rpadovaniaquarius, maybe :P13:42
bzoltanaquarius: that is the target of the convergence ambitions we have. We are moving to that direction, but we are not there13:42
mhall119bzoltan: that has been our strategy and goal for several cycles now13:42
aquariusif that's not the case -- if Ubuntu phone and Ubuntu desktop are fundamentally different -- then the message needs changing :)13:42
rpadovaniaquarius, I never studied the thing, I just found a working way13:42
bzoltanmhall119: and are we there yet? :)13:42
mhall119bzoltan: we were in vivid13:43
mhall119and in utopic before that13:43
aquariusrpadovani, ok, cheers :)13:43
bzoltanmhall119:  precisely13:43
mhall119and we should be in wily upon release13:43
mhall119otherwise we're regressing13:43
bzoltanmhall119:  and vivid env is not the same as wily or OTA613:43
seb128bzoltan, if you don't break forward compat then the wily version of the uitk should be able to run old code just fine13:44
mhall119I understand the rationale behind using vivid+overlay for phones, rather than upgrading them to wily, but that doesn't give us a pass on supporting wily on the desktop13:44
Laneydo you have some PPAs enabled?13:45
LaneyI don't see an oxide 1.9.2 in the archive13:45
bzoltanseb128:  it is perfectly compatible that way13:45
mhall119Laney: it isn't in wily currently13:45
Laneydo you have some PPAs enabled?13:46
Laneyerm, sorry13:46
bzoltanmhall119:  it is not about giving pass or not... the UITK does not build on wily since few weeks due to the toolchain change.13:46
Laneywhere's the uninstallability then?13:46
mhall119Laney: the ubuntu-sdk package depenency chain as some point required oxide =19.113:48
mhall1191.9.113:48
Laneybut you have a PPA which is giving 1.9.213:48
Laneyhow's that a problem within wily?13:48
mhall119not >= 1.9 or >= 1.9.113:48
mhall119I have 1.9.2 installed from vivid before I upgraded13:49
Laneynot vivid13:50
Laneymaybe some PPA like the "stable phone overlay"?13:50
seb128mhall119, https://launchpad.net/distros/ubuntu/+source/oxide-qt13:50
mhall119Laney: ah, that could be13:50
seb1281.9.1 is what is on the different series13:50
mhall119looks like uninstalling oxide allows ubuntu-sdk to install properly, so just conflicts between the vivid overlay PPA and upgrading to wily14:01
bzoltanmhall119:  apt-cache policy can tell you where your package is from and what is the competition between installation candidates14:01
mhall119bzoltan: yup, got it installing now14:01
bzoltanmhall119: I doubt that overlay PPA is meant or tested for/on desktop14:02
mhall119bzoltan: it isn't, I installed it to get some latest something14:02
mhall119bzoltan: so the failure ot install ubuntu-sdk was entirely my fault, not yours :)14:02
mhall119but still, getting the UITK fixed on wily is important, so glad that things are happening on that front14:03
=== chihchun_afk is now known as chihchun
bzoltanmhall119: I have checked ... Wily is on rev1627 and the UITK staging trunk is on 1642 ... so the diff is not significant. The  Wily release is two weeks old. So an eary September release from the upstream trunk would not be a shame for 15.10 .. but let's hope that the gcc 5.3 fix will land and Wily gets a UITK update.14:12
* bzoltan is not happy with the new toolchain :(14:13
mhall119thanks bzoltan, does 1642 contain Ubuntu.Components 1.3?14:13
bzoltanmhall119:  the 1.3 is already available in Wily14:13
mhall119cool, that's the most important thing14:13
bzoltanmhall119:  1.3 was released with rev1510 back in May14:14
mhall119especially since the new API docs will be pulled from their wily packages14:14
bzoltanmhall119:  since then we just add more APIs and fix bugs + we do convergence features14:14
karnizbenjamin: I had to change -g template to -g $${UBUNTU_TRANSLATION_DOMAIN} near the intltool-update call14:57
karnizbenjamin: also, I'm seeing weird thing that the msgid "Telegram" from all over the app appears twice in the pot file. the second instance is for the desktop file, msgid "Telegram" (app name)14:58
karniwhich causes fatal error when calling intltool-update14:58
zbenjaminkarni: hmm14:59
zbenjaminkarni: weird, you see that i just use the commands to update that file14:59
karniright, I'm looking for where the issue is14:59
=== chihchun is now known as chihchun_afk
=== karni is now known as karni-afk
=== karni-afk is now known as karni
lesamouraii want to run a shell command as root with c++ , tried using system() function , got this ERROR: sudo: no tty present and no askpass program specified16:27
lesamourai.16:27
ogra_lesamourai, that is correct ... root is locked down and you would have to provide a UI to sudo to recieve the PIN/PW16:29
ogra_and even then it would break app confinement i guess ... an app cant actually do anything outside of its installation dir16:30
ogra_jdstrand, ^^^ right ?16:30
ogra_(i guess not even via sudo)16:31
lesamouraiits running in unconfined mode , so is it possible?16:32
lesamouraiassuming password and uname are known already16:33
jdstrandthe confinement restricts escalating to root16:34
jdstrandwhether via system installed sudo or something else16:35
lesamouraiits running as 'unconfined'16:35
jdstrandoh16:36
jdstrandif it is unconfined there is nothing that would prevent that-- that is a normal running sudo from within your app issue (which I don't have the answer to, but others here might)16:36
ogra_well, you still need a UI app as askpass command to hand over the password/PIN16:37
lesamouraiI want to run a script as sudo from c++ , username and password to be incorporated withing the script.16:37
ogra_or dynamically create a shell snippet to point the ASKPASS variable to that echos the PW/PN into sudo16:38
ogra_lesamourai, see the manpage of sudo (specifically the -A option ) and read up about askpass scripts16:38
lesamouraiOhk16:39
ogra_you can ship a shellscript that simply echos the PW/PIN and point to it in the SUDO_ASKPASS env var ... that you need to export to sudo before calling it with -A16:40
ogra_http://bazaar.launchpad.net/~phablet-team/phablet-tools/trunk/view/head:/citrain see line 96/97 and the sunsequent lines16:41
ogra_*subsequent16:41
lesamouraiSorry I am not very pro at that, can you give a tiny example16:43
ogra_i just did16:43
* ogra_ wonders how to be more detailed than that 16:44
lesamouraiI don't understand this part 'point to it in SUDO_ASKPASS en var' , yes it was detailed16:46
ogra_see what the code above does16:46
ogra_line 100 is a good example hos you call sudo in that case16:47
ogra_(everything after "adb shell" )16:47
ogra_line 96 creates a shellscript that echos the password to stdout16:47
ogra_line 97 makes that script executable -...16:48
ogra_line 100 uses sudo with that script to obtain the password from the script16:48
lesamouraiohk16:49
ogra_the "script" doesnt need to be shell ... you could as well write a mini C++ thing that prints the PW to stdout16:49
ogra_"mini"16:49
ogra_:)16:50
ogra_important is that you call sudo with -A and that the SUDO_ASKPASS var points to your script/tool/C++ program that echos the PW16:51
lesamouraiI see16:51
ogra_also be aware that you cant really upload unconfied apps to the store ... (well, you can, but it will never be accepted)16:52
lesamouraiohk16:53
lesamouraithank you for helping me understand the concept :)16:53
ogra_:)16:53
rpadovanipopey, is your branch attached to this bug ready to be merged?17:30
rpadovanihttps://bugs.launchpad.net/reminders-app/+bug/145292417:30
ubot5Ubuntu bug 1452924 in Ubuntu Notes app "Conflict resolution wording unclear" [Medium,Triaged]17:30
* popey proposes17:33
rpadovanity17:35
lesamouraiogra , prob stays unsolved17:37
lesamouraiscript runs fine in terminal , but in C++ ERROR is17:38
lesamouraisudo: no tty present and no askpass program specified17:38
=== afkthairus is now known as athairus
rschrollt1mp, zsombi: Do either of you feel like fielding questions on drag-and-drop with the new ListItems?  Or anyone else?17:58
mhall119rschroll: can you put them in askubuntu and then post the links here? I'm sure the answers you get will be useful to a lot of developers18:03
popeyit doesn't help that tim and zsomb i are in inconvenient timezones for rpadovani18:05
popeyer rschroll18:05
mhall119that too18:08
rschrollWhoever I am.18:13
rschrollSome of them are possible bugs -- I'm not sure about the desired behavior.18:14
rschrollIs it better to post questions or bugs about them, if I can't check with the devels first?18:14
rschrollmhall119 or popey ^^18:17
mzanettipopey, rpadovani: https://code.launchpad.net/~rpadovani/reminders-app/integration/+merge/271846/comments/68503818:17
rpadovanimzanetti, so set up reminders to accept things from content-hub? We need to set up it also for links18:18
rpadovanihttps://bugs.launchpad.net/reminders-app/+bug/139842618:18
ubot5Ubuntu bug 1398426 in Ubuntu Notes app "Share menu never shows anything but Messages" [Medium,Confirmed]18:18
rpadovanicould work for me18:18
mzanettirpadovani, it can do most of the things already18:18
mzanettiIIRC I implemented support for text already18:18
mzanettirpadovani, right, links is missing18:19
mzanettiactually no18:19
mzanetticontenthub.json says:     "destination": [18:19
mzanetti        "pictures",18:19
mzanetti        "links",18:19
mzanetti        "text"18:19
mzanetti    ]18:19
mzanettirpadovani, ah right... I remember what the issue with that bug is:18:20
mzanettirpadovani, the browser supports only "share" but not "export"18:20
rpadovanioh, I see18:20
mzanettiand as the notes app is not sharing anything with other people, it's only destination18:20
rpadovanimzanetti, about my fix: with 'Text' type is possible to set a title?18:21
rpadovaniI'm thinking about exporting an email from Dekko: user expects the note to have the same name of the mail18:21
mzanettidon't think... it will just append the text to a note18:21
mzanettimhm18:21
mzanettiwe could do that, yes18:21
mzanettirpadovani, ContentItem has a name field18:21
mzanettiif dekko sets the subject to that, we could use it for the note title18:22
rpadovanidinner time, sorry, talk to you later18:22
mzanettisame here18:22
mzanettibbl18:22
mhall119rschroll: if you can phrase it as "How do I ..." then use askubuntu, but if it's "... doesn't work as expected" file a bug18:23
rschrollOk.  I'll make my best judgement.18:24
rschrollBut what I expect and what others expect may not be the same things... :)18:24
mhall119rschroll: that's fine, others can re-post it somewhere more appropriate if they want18:25
rschrollFirst question: http://askubuntu.com/questions/676750/how-do-you-detect-the-end-of-a-live-drag-with-the-new-listitem18:32
rschrollSecond question: http://askubuntu.com/questions/676755/how-do-you-make-the-background-of-a-dragged-listitem-opaque18:41
cwaynehas anyone else found the delete action in a listview has a really small touch target on mx4?18:42
=== freeelectron is now known as Guest11571
rpadovanimzanetti, if we want to continue talking I'm available again :-)18:51
rschrollFirst bug: https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/149813818:56
ubot5Ubuntu bug 1498138 in ubuntu-ui-toolkit (Ubuntu) "z-order weirdness when dragging ListItems" [Undecided,New]18:56
rschrollSecond bug: https://bugs.launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/+bug/149814319:03
ubot5Ubuntu bug 1498143 in ubuntu-ui-toolkit (Ubuntu) "Separator missing when last ListItem dragged into middle of ListView" [Undecided,New]19:03
popeycwayne: yes19:58
popeycwayne: nik90 had a bug about it on the tookit19:58
cwaynepopey: ah great, thanks20:03
dobeyhmm20:09
dobeyis there an easy way to build multiarch clicks now?20:09
popeydobey: no, i build all three arches and then smoosh them together with a script20:17
dobeyoh :-/20:21
GreatDantonHeya, if anyone is familiar with Ubuntu sdk and Pyqt5, I need a little bit of help here:  http://ubuntuforums.org/showthread.php?t=229567720:36
mzanettirpadovani, still around?20:37
rpadovanimzanetti, sure thing :-)20:37
mzanettirpadovani, heh :)20:37
mzanettirpadovani, so, what do you think about the suggestion to use ContentItem's title property for the note title?20:37
rpadovanimzanetti, seems a very good idea, less code on our side and a more strong implementation20:38
mzanettirpadovani, I feel like transmitting whole note content through an URI seems abusing it.20:38
mzanettirpadovani, also feels fragile when it comes to special chars, escaping etc20:39
rpadovaniagree on all20:39
mzanettirpadovani, only odd thing is that we can't share text and image in one go through contenthub (afaik, didn't ever really try)20:40
mzanettiotherwise we could even include attachments20:40
mzanettibut the uri-handler wouldn't give us that either20:41
mzanettirpadovani, anyhow, if you want to give it a go, the only missing thing should be to actually make use of the title20:41
rpadovaniwell, but we can't neither via URI20:41
rpadovaniso we aren't missing a functionality20:41
rpadovanimzanetti, ok, in next days I'll take a look20:41
mzanettirpadovani, this can share text: https://code.launchpad.net/~mzanetti/+junk/contenthubtest20:41
mzanettirpadovani, but you probably need to include the title property thing in the contenthub test tool too20:42
rpadovanimzanetti, okay, if I'm able I'll do a patch20:42
mzanettiI'm sure you are :D20:42
rpadovanimzanetti, about https://bugs.launchpad.net/reminders-app/+bug/1485948 I think the only way to fix it I found is deleting note's cache every time user saves it - otherwise, we have to parse the note and calculate a diff - or, again, implement a way to delete from the disk the attachment and then check if the attachmetn has been deleted20:43
ubot5Ubuntu bug 1485948 in Ubuntu Notes app "NoteDelegate preview caches images even when they are removed from the note" [High,Triaged]20:43
rpadovaniso, since deleting the cache when a note is saved is one line change20:43
rpadovaniI think is the better solution atm :D20:44
mzanettirpadovani, not happy about your suggestion... but I understand this is a tricky one...20:49
rpadovanimzanetti, I understand, other possible approaches?20:49
mzanettirpadovani, pretty sure this can be fixed by emitting a changed signal in the right place20:50
mzanettiwe need to make the tagline refresh20:50
rpadovaniokay, I'll try to take a look on this direction - but I promise nothing20:51
nemomcphail: hey, if you're around, could I have the contents of your ppas.sh ?21:36
mcphailnemo: http://termbin.com/d86l , if that helps. Won't have time to look at this further for quite a while, though21:51
nemomcphail: 'k :-/22:11
nemowell... maybe I can rope Locutus into trying this22:12

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