| *** zbenjamin is now known as Guest11333 | 02:49 | |
| *** zbenjamin_ is now known as zbenjamin | 02:49 | |
| *** frinring_ is now known as frinring | 07:31 | |
| dcaliste | Hello chriadam, how are you ? | 08:08 |
|---|---|---|
| dcaliste | I've seen your notes on caldav!51. About the mReceivedStuff, it's a good question indeed. | 08:09 |
| chriadam | hi dcaliste | 08:10 |
| chriadam | it's not a real issue I suppose, since in that case it's pure additions anyway | 08:10 |
| chriadam | (the delta is most important, IMO, when there are various changes both local+remote which need to be dealt with) | 08:10 |
| chriadam | and that's handled separately from what I understand | 08:10 |
| dcaliste | Yes, but as you said, exceptions may not be counted, I need to check. At the same time, it's only on first import. | 08:11 |
| dcaliste | Exceptions are properly counted in the delta code path. | 08:11 |
| chriadam | indeed | 08:11 |
| chriadam | in principle, I have no concerns with the PR, looks good but I still need to test on device | 08:12 |
| dcaliste | Sure, no problem. | 08:12 |
| chriadam | I had meant to do so last week, but caught up chasing another bug which I'm still chasing | 08:12 |
| chriadam | some issue with mkcal date handling again unfortunately. a sync plugin which provides a VTIMEZONE definition which for some reason isn't being stored properly by mkcal, then later on the date is being returned as UTC | 08:13 |
| dcaliste | Ah, do you want to share the tracking ? | 08:13 |
| chriadam | in chasing it, I went down a rabbit hole with mkcal unit tests (e.g. some cases where it uses LocalZone in the unit tests, but the assumptions are wrong so it's enforcing the wrong behaviour at least in Brisbane tz from what I can tell) | 08:14 |
| chriadam | not much to share, currently, unfortunately. I will have PRs up soonish I hope | 08:14 |
| chriadam | the short summary: in SqliteStorage::fromOriginTime() in the case where it tries calendar specific zones | 08:14 |
| chriadam | in this case (and I haven't tracked down why this is the case, could be sync plugin's fault) the timezone field contains a valid id but the VTIMEZONE record stored in the Timezones table is partial (it looks like: "BEGIN:TIMEZONE\r\nTZID:SOME_LONG_TZ_ID_UUID\r\nEND:VTIMEZONE\r\n" without any offset information etc) | 08:16 |
| chriadam | so icaltimezone.isValid() returns true | 08:16 |
| chriadam | but it's useless for any conversion, since it doesn't store any offset info... so the .toTimeSpec(KDateTime::Spec(icaltimezone)) returns same as .toUtc() | 08:16 |
| chriadam | so "simplish" fix for that is to check to see if the icaltimezone there is "empty" (i.e. vtimezone() returns an empty-ish VTIMEZONE definition) | 08:17 |
| chriadam | but then while trying to write unit tests for various things, I ran into a bunch of issues with the unit tests... rabbit hole stuff. | 08:17 |
| dcaliste | Ah, I understand the reasoning. I don't think there are tests currently about non system time zone… | 08:18 |
| dcaliste | I'm modifying a bit some tests with mkcal!17 https://git.sailfishos.org/mer-core/mkcal/merge_requests/17 because the some assumptions were wrong or at least working only in specific cases. Which tests are you modifying ? | 08:19 |
| chriadam | let me check that one | 08:19 |
| chriadam | I'm looking at the raw events test currently | 08:19 |
| chriadam | but there were a couple of others IIRC also | 08:20 |
| dcaliste | OPk, in !17, most modifications are in allDayUTC() test. | 08:22 |
| chriadam | yes also the alldayRecurrence | 08:22 |
| chriadam | didn't think I hit issues in the alldayUTC test but maybe I'm misremembering | 08:22 |
| dcaliste | There may be several separated issues ;) | 08:23 |
| chriadam | no doubt | 08:23 |
| chriadam | every time I have to think about timezones my brain explodes ;-) | 08:23 |
| dcaliste | I agree it's a tricky problem. | 08:24 |
| chriadam | anyway, aside from that: the "honour direction" caldav MR LGTM, I was supposed to merge that but got sidetracked as above | 08:24 |
| dcaliste | Sure, merge it when you have time, no problem. | 08:24 |
| chriadam | has there been further progress from Jolla side (comments/testing etc) for the folder sync PR for emails? | 08:25 |
| chriadam | if not, I will ask pvuorela or flypig to take another look at that - I understand that some hours were allocated for that internally | 08:25 |
| dcaliste | No, and I didn't look at it neither, mainly working on mkcal and kcalcore recently. | 08:25 |
| chriadam | cool np | 08:26 |
| dcaliste | About TimeZones, memory calendars are supposed to cache incidences per dates (as extended calendar is doing), so I'm proposing a patch upstream to properly handle this : https://phabricator.kde.org/D25066 | 08:27 |
| dcaliste | Currently no reply, but I'll try to find some more reviewers. | 08:27 |
| dcaliste | If accepted, I'll backport it to Mer kcalcore, so extended calendar can properly inherit memory calendar. | 08:28 |
| chriadam | excellent, thank you | 08:29 |
| dcaliste | Because incidences added to a calendar can be given a different time zone than the calendar itself, and the date of the event be different than the date seen by the calendar. | 08:29 |
| dcaliste | mkcal extendedcalendar is doing this properly, but not memory calendars. | 08:29 |
| dcaliste | As I told you also last week, I've a MR (https://git.sailfishos.org/mer-core/buteo-sync-plugin-caldav/merge_requests/52) removing the Q_FOREACH macro from caldav plugin. | 08:33 |
| chriadam | Yep, I haven't had a chance to take a look at that one yet, unfortunately, but I suspect it's mechanical and straight-forward. | 08:33 |
| chriadam | so, LGTM once I've tested it | 08:33 |
| dcaliste | I'm supposed to have taken good care for constness issue to avoid copying, but it's better if checked a second time ! | 08:33 |
| chriadam | yep :-) | 08:34 |
| chriadam | I don't have anything else to discuss today - sorry again for being slow with your CalDAV MRs | 08:35 |
| dcaliste | pvuorela: about https://git.sailfishos.org/mer-core/python3/merge_requests/11, there is this problem with duplicating the variable definitions between the two specs and the fact that %include is not working (or that I cannot make it work). Can you comment when you have a piece of time ? | 08:36 |
| chriadam | was there anything else you wanted to discuss this week, or things you'd like me to poke people about? | 08:36 |
| dcaliste | chriadam: no problem, you're busy, that's fine. | 08:36 |
| dcaliste | Nothing else to discuss this week I think. I'm working on the purge API for mkcal, MR should land next week or the one after I think. | 08:37 |
| pvuorela | dcaliste: sure. i was supposed to earlier but then got too many other things :/ | 08:38 |
| chriadam | sounds good! | 08:38 |
| dcaliste | pvuorela: ok, no problem if it's still on the radar ;) | 08:39 |
| dcaliste | pvuorela: (and thanks for the quick fix in account qsetting packaging yesterday) | 08:39 |
| pvuorela | sure :) | 08:39 |
| adantes | Ive been using sailfish 2.1 community edition in Gemini PDA for a while, and the impression cannot be better | 11:08 |
| adantes | Went to Android, because of the spice cliente, and immediatly drained 7% of battery, just because | 11:08 |
| adantes | this is just amazing. | 11:09 |
| adantes | Sailfish sells sailfish X for gemini PDA. Does anyone find it worthful to make the leap? | 11:10 |
| Dakon | anyone at QtWS2019? | 11:11 |
| lachs0r | hm, sadly the connman patch didn’t improve battery life on the jolla tablet much | 11:41 |
| lachs0r | it did last slightly longer than it usually does on wifi, but the difference isn’t nearly as huge as on the xperia x | 11:42 |
| adantes | anyone using gemini pda? | 12:08 |
| nyctea | yes, me | 12:23 |
| nyctea | but I feel like you've asked earlier and I replied the same | 12:23 |
| leinir | Yup! ...but i'm terrible and don't run sailfish on it :P | 12:24 |
| nyctea | yeah, you (or someone with your nick) did ask this a few days ago, you had the wifi problem | 12:25 |
| nyctea | leinir, you're terrible | 12:25 |
| leinir | Basically, nay literally the worst | 12:25 |
| nyctea | adantes, I also paid for Sailfish, yes, but I'm still waiting for android app support ;-) | 12:26 |
| lachs0r | I’m not waiting for better android support, I just write apps when something is missing (reverse engineer if I have to) | 12:34 |
| lachs0r | admittedly there hasn’t been a lot of that | 12:34 |
| nyctea | I haven't tried writing them myself, yet | 12:35 |
| nyctea | I generally don't like programming GUI's | 12:35 |
| lachs0r | but I’m working on a public transit app that uses OneBusAway’s API (mostly puget sound area since I’ll be staying there for the remainder of the year) | 12:35 |
| nyctea | and the only thing I actually miss is signal | 12:35 |
| adantes | nyctea: it was me | 12:35 |
| adantes | nyctea: sorry, just forgot your nick | 12:36 |
| lachs0r | heh, my approach to IM is use IRC and get people to use tox when privacy is required | 12:36 |
| adantes | the wifi prob was solved | 12:36 |
| adantes | KILL IT | 12:36 |
| lachs0r | nyctea: oh you should give it a try, qml/silica is extremely nice to work with | 12:36 |
| lachs0r | especially compared to literally every other mobile platform out there | 12:36 |
| lachs0r | almost everything that is just an interface to some web api doesn’t even require writing any c++ | 12:37 |
| adantes | lachs0r: I'm following the same path... bu this is a noreturnpath | 12:38 |
| lachs0r | and very few lines of js — the rest is just declarative UI markup | 12:38 |
| nyctea | bleh, js | 12:38 |
| adantes | I guess last time I used IRC activly was about 20 years ago | 12:38 |
| adantes | mIRC | 12:39 |
| nyctea | hey I started using irc actively about 20 years ago | 12:39 |
| adantes | msn messenger an icq turned out to be great | 12:39 |
| adantes | and better chicks used it | 12:39 |
| adantes | :-D | 12:39 |
| lachs0r | nyctea: very basic js. the most repulsive thing you’ll encounter is xmlhttprequest | 12:40 |
| adantes | have you found the upgrade worthy? | 12:41 |
| adantes | this community edition is prety ood | 12:41 |
| nyctea | well it didn't make a huge difference | 12:42 |
| nyctea | can't really recall any, actually | 12:42 |
| adantes | If i'm going for the upgrade is for one reason solely: support the devs | 12:42 |
| nyctea | yeah, that's why I bought it | 12:42 |
| adantes | even though I've bought for my XA2 | 12:42 |
| nyctea | supporting development and hoping for more features | 12:42 |
| nyctea | future features | 12:43 |
| adantes | mehhh.. don't know if thats happening | 12:43 |
| adantes | this is very very nieche market | 12:43 |
| adantes | niche | 12:43 |
| adantes | whatever the tv wconnomy nerds call it | 12:43 |
| lachs0r | if you want more features, organize and implement them | 12:47 |
| lachs0r | that’s my opinion at least | 12:48 |
| lachs0r | relying solely on jolla to solve all of our problems for us seems like a Bad Idea that’s only going to diminish the user base | 12:49 |
| adantes | lachs0r: couldnt agree more | 12:50 |
| lachs0r | which is why I think it’s really damn important we get a decent OBS setup already | 12:50 |
| adantes | lachs0r: If I hadnt got a family to sustain | 12:50 |
| adantes | :-P | 12:50 |
| lachs0r | so openrepos can rest in peace | 12:50 |
| adantes | I guess that -- and this is my oppinion, for whats worth -- even though I earn my life making android and ios software for mass public | 12:53 |
| adantes | fuck users | 12:53 |
| adantes | we need something hardcore, reliable and secure | 12:53 |
| lachs0r | it’s just kind of sad to see a community that is as tech-savvy as sfos’ not organizing more | 12:54 |
| lachs0r | and I regret not getting involved more in the past | 12:55 |
| adantes | no need for shitty compoonents an layouts | 12:55 |
| lachs0r | I even got the damn tablet ffs | 12:56 |
| lachs0r | been doing way too little given how much I care about the platform | 12:56 |
| adantes | something may be done, and I wish I had the time to contribute, learn qml an stuff | 12:57 |
| lachs0r | not much to learn about qml to be able to use it | 12:58 |
| lachs0r | there’s a page on common UI pitfalls in silica’s documentation, but other than that it’s all pretty intuitive once you get past hello world | 12:58 |
| adantes | when you see you lost hundresds of hours doing boring stuf like custom dropdowns, and other custom bs, that wasgreat in android gingerbread, and then they ruined it all jus by moving forward an deprecating design guidelines | 12:59 |
| lachs0r | android UI is a fucking disaster anyway | 12:59 |
| adantes | because users wanted to... funny things with images | 12:59 |
| adantes | bahhh... | 12:59 |
| adantes | I really enjoy the looks in sailifsh | 12:59 |
| adantes | its more than enough | 13:00 |
| adantes | clear, well arranged | 13:00 |
| lachs0r | the UI is what really sets sfos apart. the way interaction works is a whole lot better than anything on other platforms | 13:00 |
| adantes | and pretty aswell | 13:00 |
| lachs0r | most of all, consistent | 13:00 |
| adantes | yes | 13:00 |
| adantes | I needd to choose between 10 things. Use a drop. done | 13:01 |
| lachs0r | android doesn’t even have a consistent push-and-hold behavior | 13:01 |
| lachs0r | or gestures that make sense | 13:01 |
| adantes | users way of thinking, i need to choose between 10 things: god! those are too many things! whydon;t they make a graph with diferent screens and explanations and animations, an videos | 13:02 |
| lachs0r | and every app has dozens of tiny buttons | 13:02 |
| lachs0r | you never know what they do | 13:02 |
| lachs0r | some trigger actions, some take you to different screens | 13:02 |
| adantes | uck users | 13:02 |
| lachs0r | some are so close to the scroll bar you can barely use them | 13:03 |
| lachs0r | sfos gets a lot of it right, and continues to improve | 13:03 |
| lachs0r | also it’s really nice how easy it is to patch the system’s UI | 13:04 |
| adantes | sfos gets a lot of things right. Its my main OS for proessional and personal use | 13:04 |
| adantes | It's very very clear | 13:04 |
| lachs0r | it’s also real linux. with a real libc, real package management, applications that adhere to freedesktop specs, etc. | 13:05 |
| adantes | exactly | 13:05 |
| lachs0r | because it’s not designed for lock-in | 13:05 |
| adantes | once I put my sfos behind a wireshark | 13:06 |
| adantes | and my android | 13:06 |
| adantes | an winX | 13:06 |
| adantes | sfos, was silent | 13:06 |
| adantes | android an winx, even in idle, wireshark looked like a christmas tree | 13:07 |
| adantes | since then it is my first choice in personal phone | 13:07 |
| lachs0r | sfos pretty much only communicates with jolla-owned domains to verify the licensing status and to check for ipv4/v6 connectivity iirc | 13:07 |
| adantes | yes; more than acceptable | 13:08 |
| adantes | it reminds me a lot of my home setup using arch | 13:08 |
| adantes | just clean clean clean | 13:08 |
| adantes | dwm, and no bs | 13:08 |
| lachs0r | google android also uploads ALL of your media without asking unless you opt out | 13:08 |
| adantes | thats a huge problem, you know | 13:09 |
| lachs0r | doesn’t seem to bother people | 13:09 |
| lachs0r | because it’s all just reduced to a matter of privacy | 13:09 |
| lachs0r | when there’s so much more to it | 13:10 |
| adantes | it almost makes it worth to hang around with a raspberry pi in your pocket just acting like proxy/frewall for your google android and ios devices | 13:10 |
| adantes | but that would be dumb if you just use vanilla sfos | 13:10 |
| lachs0r | all the data big tech collects is being used to manipulate people (because that’s literally what targeted advertising does — of course it’s being used for things other than advertising as well) | 13:10 |
| adantes | yes, thats a huge concern | 13:11 |
| adantes | ssh for life :-D | 13:11 |
| adantes | well, must go | 13:11 |
| lachs0r | I hate sounding like a conspiracy theorist but these things are happening right now | 13:11 |
| adantes | nice chat | 13:11 |
| lachs0r | facebook literally bragged about how they can target depressed teens | 13:12 |
| lachs0r | people are extremely vulnerable to begin with, but they don’t even realize just how much psychological firepower is aimed at them when they use those products | 13:12 |
| lachs0r | oh well, guess I’m just preaching to the choir in this channel ;) | 13:13 |
| adantes | lachs0r: it's users fault | 14:25 |
| lachs0r | no it isn’t | 14:26 |
| adantes | if it wasnt for the need to implement communication/cameras/chat/gamming portable devices so massivly in the market, this wouldnt happen | 14:26 |
| lachs0r | and who creates this need? | 14:27 |
| adantes | I remember having a psion revo | 14:27 |
| lachs0r | the companies who build all these platforms to begin with | 14:27 |
| adantes | it was mind blowing | 14:27 |
| adantes | with a modem, Id use it again | 14:27 |
| adantes | users. | 14:27 |
| lachs0r | and no one is immune to manipulation | 14:28 |
| adantes | dumb like hell. If you do sw, you feel the urge to dumbify your software, r the users wont buy it | 14:28 |
| lachs0r | the role of advertising is to create a perceived need for a product or service | 14:28 |
| lachs0r | nobody really needs facebook, signal or whatsapp | 14:29 |
| adantes | lachs0r: thats my point | 14:29 |
| lachs0r | yeah but you’re blaming the wrong people | 14:29 |
| adantes | I run a business, and do most of my needs in f#$ SSH | 14:29 |
| adantes | spreadsheet included! | 14:30 |
| adantes | fuck useres | 14:30 |
| lachs0r | it’s the users who can do something about it, but it’s not the users who are responsible for the way bad software dominates the market | 14:30 |
| adantes | they buy stuff when its dumb enough | 14:30 |
| adantes | :-) | 14:30 |
| lachs0r | they buy software when it’s advertised enough | 14:30 |
| adantes | yep | 14:30 |
| lachs0r | even when it’s overly complex and unwieldy | 14:31 |
| adantes | you build an amazing piece of software, that links to everything, runs all apis smoothly, but what is the key for the user to use it? if it has baby drawings, and a 3 click interface | 14:31 |
| adantes | even if it has multiple flaws, and obliterates your privacy:they care baby drawings, and 3 click user interface | 14:32 |
| lachs0r | not really true. if you look at a lot of commercially successful UIs, they’re all absolutely hideous garbage that’s way more difficult to use than it has any right to be | 14:33 |
| lachs0r | but what they all have in common is good branding | 14:33 |
| adantes | remember the huge difference in paradigm between apple and MS? Apple adapt to ther target; microsoft demanded users to adapt to their software | 14:33 |
| lachs0r | take discord as an example. or even in open source software, mastodon | 14:33 |
| lachs0r | they’re both trash but the branding has made them super successful | 14:34 |
| adantes | yes, I agree | 14:34 |
| lachs0r | or telegram | 14:34 |
| adantes | signal | 14:34 |
| adantes | sooooo much better that whatsapp | 14:34 |
| adantes | telegram has bot anf | 14:35 |
| adantes | and game features | 14:35 |
| lachs0r | these are all trash | 14:35 |
| adantes | whatsapp is a case study | 14:35 |
| adantes | and so is skype | 14:36 |
| lachs0r | decades later and despite all the flaws, IRC is still the pinnacle of instant messaging, and that’s just infuriating | 14:36 |
| lachs0r | the only thing that comes close to being a decent, modern alternative is tox | 14:36 |
| adantes | no that easy to use, for the commonuser #fuckusers | 14:36 |
| lachs0r | xmpp is too busy circlejerking to do anything useful (like a reference implementation with a C API) | 14:37 |
| lachs0r | as a matter of fact, corporations love xmpp because every bit of the standard is designed to benefit their use case: a nice walled garden | 14:38 |
| adantes | never cared hat much about xmpp, because theres a client for evrything in thre | 14:38 |
| lachs0r | interoperability is just not a thing they do | 14:38 |
| lachs0r | you can’t even get two instances of prosody or ejabberd to reliably talk to each other | 14:38 |
| adantes | ive implemented successfyllu in a business i was administrating, and everyne was using it | 14:38 |
| lachs0r | xmpp only works where you’re in control of all deployments | 14:39 |
| adantes | android, ios, web | 14:39 |
| adantes | everything was smooth | 14:39 |
| adantes | pidgin wherever possible | 14:39 |
| lachs0r | oh god libpurple | 14:40 |
| lachs0r | no no no | 14:40 |
| adantes | why? plugins for everything | 14:40 |
| lachs0r | none of which work well | 14:40 |
| tadzik | hrm, it always worked well for me | 14:41 |
| lachs0r | in xmpp’s case it doesn’t support message carbons and omemo properly (and even with plugins to extend it, things are quite broken) | 14:41 |
| tadzik | in its limited scope at least | 14:41 |
| lachs0r | unstable too. as most code involving glib ist. | 14:41 |
| lachs0r | -t | 14:41 |
| adantes | thats too bad | 15:32 |
| adantes | it worked great whenever i needed it | 15:32 |
| adantes | with open fire | 15:32 |
| adantes | but, indeed it was used by users, and they're dumb as hell... no special needs needed #fuckusers | 15:32 |
Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!