| *** frinring_ is now known as frinring | 02:13 | |
| dcaliste | Hello chriadam. | 05:06 |
|---|---|---|
| chriadam | dcaliste: hi :-) | 05:08 |
| dcaliste | chriadam, I've looked at my issue with secret UI. It comes from collectionkeysmodel.cpp#135 (called from line 115 on component complete). There are two possibilities: | 05:09 |
| dcaliste | - either I've misinterpreted StoredKeyIdentifiersRequest when collectionName is empty, | 05:09 |
| dcaliste | - or I don't understand what should be retrieved in collectionkeysmodel.cpp in case of specifyCollectionName = false. | 05:09 |
| chriadam | when collectionName is empty, it should retrieve the identifiers of all keys from any unlocked collection | 05:10 |
| dcaliste | That's my interpretation, indeed. | 05:10 |
| chriadam | (but will not automatically attempt to unlock any locked collection, in order to return keys from those) | 05:11 |
| dcaliste | But then, look at ui-secrets/src/collectionkeysmodel.cpp line 135. | 05:11 |
| chriadam | ooi is this the behaviour implemented by the sqlcipher plugin? | 05:11 |
| chriadam | let me check the collectionkeysmodel | 05:11 |
| chriadam | right, so if a collectionName is provided in the request, that WILL trigger an unlock flow (but only the keys for that specific collection will be returned) | 05:13 |
| dcaliste | Yes, but coming from the QML part, we have repeater for every collection in the plugin. | 05:13 |
| chriadam | so I think that code is saying: if the client wants the identifiers of the keys from one specific collection, then it will set the collectionName property in the request, and perform it (which will trigger an unlock flow) | 05:13 |
| chriadam | it would only call requestKeys(true) in the onClicked handler | 05:14 |
| dcaliste | and then for each repeater, we ask all keys because collectionName is not set at line 135. | 05:14 |
| chriadam | ah, I see what you mean | 05:15 |
| chriadam | in the "false" case, it will request the keys for all collections. yes. but it should then filter the results, right? | 05:15 |
| dcaliste | Ah, didn't catch the filter part. Where is it? | 05:15 |
| chriadam | I assumed it would in the lambda below | 05:16 |
| chriadam | but it doesn't seem to... | 05:16 |
| chriadam | what is the actual issue, though? is it that we're doing unnecessary work, because we request all keys N times? | 05:17 |
| dcaliste | Ah, you mean the part with old and new identifiers? The std::transform() call? It's some C++ syntax I don't know about. Will investigate further then. | 05:18 |
| chriadam | (each collection delegate should only display a given key if the key.collectionName matches the delegate collectionName, I assume | 05:18 |
| dcaliste | I mean doing listing of all keys N times is slow, yeh. But more over all keys are displayed also for each collection. | 05:19 |
| chriadam | dcaliste: I think that code you mention is just ensuring that the delta is applied correctly (new results vs cached results) | 05:19 |
| chriadam | dcaliste: do you reproduce this with the sqlcipher plugin? | 05:19 |
| dcaliste | No, that's why I think it's my misundertsanding of something in the plugin calls. | 05:20 |
| chriadam | dcaliste: other possibility is that collections are created in sqlcipher locked by default with AccessRelock semantics, which would prevent the plugin from returning keys from any other collection, normally | 05:21 |
| chriadam | i.e. the bug exists but is "hidden" due to the AccessRelock semantics | 05:21 |
| dcaliste | Yep, in GnuPG plugin all keys are always unlocked in the sense of sailfish-secret, since unlocking is done externally by the pinentry and not by any sailfish-secret unlock flow. | 05:22 |
| dcaliste | But going back to the code in collectionkeymodel.cpp, I'm sorry, but I don't see any filtering in the lambda. On sucess all identifiers (soo all keys of all collections in our case), are compared to stored ones. Not listed are removed from old ones and new are added. | 05:26 |
| chriadam | I agree | 05:30 |
| chriadam | I guess any filtering should happen in the UI (that is the delegate would have something like: visible: key.collectionName == delegate.collectionName | 05:31 |
| chriadam | or similar) | 05:31 |
| chriadam | but I'm not too familiar with the secrets-ui code. rainemak and Venemo did most of the development there. | 05:32 |
| dcaliste | IMHO, one should add new identifiers to list only when i.collectionName() == m_collectionName if m_collectionName is not empty. | 05:33 |
| dcaliste | I mean in the cpp part around line 182. | 05:34 |
| dcaliste | chriadam, I can propose a PR in bitbucket and invite venemo and rainemak to discuss the issue there, what do you think? | 05:36 |
| chriadam | I think that would be best, thank you | 05:36 |
| dcaliste | Ok, thanks for the discussion and new lighting on the code ;) | 05:36 |
| Thaodan | can anyone help me with this spec file? | 09:24 |
| Thaodan | https://build.merproject.org/package/live_build_log/home:Thaodan:patchmanager/sailfishos-improved-notification-preview/sailfishos_latest/armv8el | 09:24 |
| Thaodan | %qmake5 doesn't expand | 09:25 |
| Thaodan | however in the sdk it does | 09:25 |
| elros | maybe you need some qt5 package in BuildRequires | 09:35 |
| coderus | Thaodan: why you qmake files project? :) | 09:42 |
| Thaodan | automate packging and transaltions | 09:43 |
| Thaodan | however its not mine | 09:43 |
| Thaodan | I got asked to build it | 09:43 |
| r0kk3rz | yeah BR qt5base should do it | 09:45 |
| Thaodan | thanks | 10:07 |
| Thaodan | fixed it | 10:07 |
| Thaodan | I have a changes file but the file isn't included what can be the reason of this? | 10:52 |
| fledermaus | hmph. GPS doesn't seem to work very well since the latest update :( | 10:56 |
| tadzik | how so? | 10:56 |
| tadzik | today I needed to start gpsinfo in the background, otherwise whogo maps wouldn't keep track of me at all | 10:57 |
| fledermaus | never seems to get past 3 satellites except briefly, never seems to pick my actual location, takes forever to pick up any satellites | 10:57 |
| r0kk3rz | using agps? are you inside? | 10:58 |
| fledermaus | before this was intermittent behaviour which was fixed by a reboot but now it's permanent. | 10:58 |
| fledermaus | r0kk3rz: dunno about agps, insode, outside, all the same. | 10:58 |
| r0kk3rz | check your location settings then | 10:58 |
| fledermaus | it used to lock ok outside my flat, where I start my run, but now it only ever thinks I'm a few streets over in either direction and always loses lock during my run | 10:59 |
| fledermaus | I have. | 10:59 |
| fledermaus | I have tried ebery combination. | 10:59 |
| fledermaus | every* | 10:59 |
| fledermaus | all the same. | 10:59 |
| tadzik | fledermaus: wild guess, but try starting gpsinfo while using whatever you're using to track your run | 10:59 |
| tadzik | your symptoms sound like mine today and that helped instantly | 11:00 |
| fledermaus | hmmm | 11:00 |
| fledermaus | where do I get gpsinfo? | 11:00 |
| fledermaus | store? | 11:00 |
| fledermaus | green icon w/cross-hair? | 11:01 |
| r0kk3rz | what app are you using? | 11:02 |
| fledermaus | maps & spotra | 11:02 |
| r0kk3rz | is that a native thing, an androidy thing? | 11:02 |
| leszek | tried poormaps on my balcony (otherwise no signal). Got instantly located | 11:03 |
| leszek | so I don't think its a general issue | 11:03 |
| fledermaus | r0kk3rz: native | 11:03 |
| leszek | gpsinfo is a native app. Should be on openrepos I guess | 11:03 |
| leszek | maybe even jolla store | 11:03 |
| fledermaus | leszek: yeah I got it now | 11:03 |
| fledermaus | it agrees w spotra about # of satellites. | 11:04 |
| fledermaus | I'm indoors right now but I'm by a wondow at the outer wall of my building. | 11:04 |
| fledermaus | behaviour was the same this morning outdoors though | 11:04 |
| tadzik | fledermaus: yes, that one | 11:08 |
| fledermaus | eventually grabbed 4 satellites while was outside but location on mapp was still quite off. | 11:27 |
| pykape | hi | 11:30 |
| pykape | how did one fix this error in the IDE?: 'Error on file "/home/deploy/installroot/usr/share/applications/*.desktop": No such file or directory' | 11:30 |
| *** bugzy_ is now known as bugzy | 15:35 | |
| potter | hello there, i want to build sailfish os for potter where should i start? | 16:33 |
| potter | please anybody there | 16:36 |
| r0kk3rz | the hadk document is where you start | 16:36 |
| r0kk3rz | and join #sailfishos-porters | 16:36 |
| potter | ok | 16:37 |
| gena2x | Hi. trying to undersatand whats wrong with tethering here. | 18:18 |
| gena2x | device connects, there is internet on sailfish | 18:18 |
| gena2x | i can ssh from notebook to jolla via wifi | 18:19 |
| gena2x | actually my question is | 18:19 |
| gena2x | i logged in to sailfish via ssh. doing route. | 18:19 |
| gena2x | there is exactly one route, to local wifi network. | 18:20 |
| gena2x | ping google.com says network not available | 18:20 |
| gena2x | in same time browser works, both in android and native jolla | 18:21 |
| gena2x | so my Q: how in the hell? | 18:21 |
| gena2x | some container? | 18:22 |
| gena2x | i see masquerade rule (double) in iptables | 18:31 |
| *** Blizzzek is now known as Blizzz | 20:42 | |
| *** feodoran is now known as Guest88379 | 23:57 | |
| *** feodoran_ is now known as feodoran | 23:57 | |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!