#fdroid-dev

/

      • BubuIIC joined the channel
      • BubuIIC has quit
      • [b__b] joined the channel
      • BubuIIC joined the channel
      • BubuIIC has left the channel
      • BubuIIC joined the channel
      • _hc joined the channel
      • mimi89999_0
        relan: Do you know Gradle well?
      • BubuIIC has left the channel
      • _hc joined the channel
      • _hc
        hi CiaranG pserwylo est31 mimi89999 uniq[m]: hi all, its meeting time
      • I'm back to being regularly online :)
      • mimi89999
      • BubuIIC joined the channel
      • _hc_roving joined the channel
      • _hc
        hi BubuIIC, you here to join in the weekly meeting?
      • _hc_roving has quit
      • BubuIIC
        _hc: when is the meeting?
      • now?
      • _hc
        now
      • started 8 minutes ago
      • so I'm still catching up on everything, I have a crazy backlog
      • uniq[m]
        hi everyone
      • _hc
        if there is anything that I should look at first, let me know
      • BubuIIC
        I'm here :-)
      • _hc
        right now, I'm thinking that getting fdroidclient 1.0 finalized is the major goal
      • and also launching the localized website
      • I was at Sneakercon at Columbia Univeristy's Journalism School
      • it was all about offline, sneakernets, nearby
      • I talked about F-Droid's swap, offline support, and decentralization
      • there were not a lot of techies, so it was interesting to be talking about this in a different setting
      • I also talked with someone inside of Google about opensourcing their Play Services/GMS libraries
      • they said there was an internal issue tracking the progress on this, and that they were definitely doing it for some of them.
      • they seemed to believe it would be most of the Play Services and Firebase libs
      • since these libs are the most common reason why fdroid rejects apps, I hope it happens.
      • then we can include a lot more apps directly, without all the work and the fighitng
      • you can see some public evidence of this here:
      • note that gms/play libs are listed under firebase
      • this will mean that we will have to check and mark a lot more things as Tracking/Ads Anti-Features
      • since they might open source their ad and analytics libs
      • that's it from me
      • who's next?
      • BubuIIC
        I'm still gathering a few issue references I wanna talk about
      • but I can continue
      • firebase/gsm will be interesting if/when it really happens
      • _hc
        indeed
      • BubuIIC
        Who has access to the webserver? _hc? There are a fee important open issues
      • _hc
        we should all think of ways to keep the pressure on
      • I have access only to forum.f-droid.org, not f-droid.org
      • BubuIIC
        a#47
      • NOTICE: [client] #47: Preferences - Update Notifications - Remember Not… - https://gitlab.com/fdroid/fdroidclient/issues/47
      • _hc
        only CiaranG has acecss to f-droid.org
      • BubuIIC
        no
      • _hc
        admin#47
      • NOTICE: [admin] #47: https://f-droid.org/jessie64.box not vailable - https://gitlab.com/fdroid/admin/issues/47
      • ah, ok. really, we should move to using the official debian base box, and not our custom one
      • but that's not trivial, it'll take some work
      • BubuIIC
        Yes, agree. I wanted to do some more tests with kvm
      • I haven't gotten time to look into it yet/
      • mimi89999
        > this will mean that we will have to check and mark a lot more things as Tracking/Ads Anti-Features
      • > since they might open source their ad and analytics libs
      • That could be easily automated
      • BubuIIC
        Next question I have is, when/how is the packge index for the website updated?
      • _hc
        mimi89999 good to know :)
      • BubuIIC: there is a page on the bottom of https://f-droid.org/docs about it
      • BubuIIC
        So it's only done on website deploy, right?
      • Does that happen automatically on a shedule?
      • because the last manual deploy was 2 weeks ago I think
      • That means we get a lot of reports about outdated packages on the website
      • Then the other thing that pops up again and again is that the website doesn't show the archive: w#57
      • NOTICE: [website] #57: App browser should list apps from other certain o… - https://gitlab.com/fdroid/fdroid-website/issues/57
      • and for some apps the links point to the repo but the app is in the archive
      • I think that's the case for all that were moved because of md5 signatures, but not sure
      • _hc
        BubuIIC: it was supposed to run after each index update, but it might not be happening...
      • BubuIIC
        lastly I get 2 months of from work starting in the middle of september. I'll be on vacation for 2 weeks, after that I'm planning to do more work on fdroid
      • _hc
        nice ;)
      • :)
      • BubuIIC: I think this is your first meeting, maybe you could say a little bit about yourself as an intro?
      • BubuIIC
        sure
      • I'm Marcus, as you see on gitlab.
      • living in Berlin, working as an embedded linux developer
      • I only ever used f-droid as an app store for my phone, so my interest is basically to have a viable android ecosystem without using google stuff
      • not sure what else you wanna know :)
      • _hc
        sounds like you're in the right place ;)
      • BubuIIC: w#121
      • NOTICE: [website] #121: include timestamp of index-v1.jar used to generat… - https://gitlab.com/fdroid/fdroid-website/issues/121
      • I'm in Vienna, and make it up to Berlin probably once a year
      • speaking of Berlin, the Reproducible Builds summit will be there at the end of October:
      • BubuIIC
        _hc: are you planning on attending?
      • _hc
        I can't, but maybe uniq[m] can
      • or maybe mimi89999 is also interested? They should be able to cover travel costs
      • uniq[m]
        Not sure, I've commited helping with one of FSFEs free your android workshops arount the same time
      • BubuIIC
        i have to leave for now. (dentist appointment)
      • Re: access to f-droid.org: is anyone in contact with CiaranG?
      • _hc
        a few of us are
      • he's got limited bandwidth these days
      • we'll add that index updating trigger when we work to launch the fully localized website
      • uniq[m]
        me next?
      • BubuIIC
        (off, I'll read the log)
      • uniq[m]
        I'm still working on s#291 and s#334 (index-v1 update for apks with fdroid and developer signing keys)
      • _hc
        go!
      • NOTICE: [server] #291: include APK signatures in build metadata file - https://gitlab.com/fdroid/fdroidserver/issues/291
      • uniq[m]
        Most of the work is actually done.
      • NOTICE: [server] #334: make index generation fully reproducible - https://gitlab.com/fdroid/fdroidserver/issues/334
      • I'm still figuring out some issues with off by one hour timestamps in the added and lastUpdated.
      • The code assigning those values is a bit confusing. :(
      • One question came up for me which here where I'd like some input.
      • I'm building a file containing a list of all app-singing-key-fingerprints in our keystore when running publish
      • This file has to be available when running update, for identifiying which apks were singed by FDroid.
      • Should this file go into the stats folder. (currently I've placed in repo)
      • The important thing is, it has to get coppied from the signing server back to the server which runs update.
      • BubuIIC has quit
      • _hc
        uniq[m]: hmm, I think we'll have to get CiaranG to set up a custom copy rule for the offline signing server setup
      • so might as well leave it in repo
      • but it shouldn't be deployed to the final website
      • uniq[m]
        I'm speculating ciaran just sync the entire repo folder so he might won't have anything to update
      • _hc
        uniq[m]: w!110
      • NOTICE: [website] !110: docs: add how to create fdroid base.box from defa… - https://gitlab.com/fdroid/fdroid-website/merge_requests/110
      • mimi89999 what about s!320 vs s!321, should I merge both? or do they conflict?
      • NOTICE: [server] !320: New gradle-plugin 3.0 output apk location - https://gitlab.com/fdroid/fdroidserver/merge_requests/320
      • NOTICE: [server] !321: handle gradle-plugin 3.0 output apk location - https://gitlab.com/fdroid/fdroidserver/merge_requests/321
      • CiaranG
        uniq[m]: So that file is generated by publish (from scratch each time?) and needs to be available when an index is created, is that right?
      • uniq[m]
        CiaranG: yes
      • CiaranG
        Ok, as hans said, it will need to be handled in a custom way for the production setup, but that will be easy.
      • It should go in stats rather than repo
      • Will `update` just work as before if the file isn't there?
      • uniq[m]
        yes
      • sounds consistent, I'll put it into stats then
      • CiaranG
        Great, so there's no need to synchronise the code change and the custom setup for moving that file.
      • If someone pings me when this change is in master (assuming I don't notice myself) I'll make the last bit work.
      • uniq[m]
        Will do. Since no apps are using this feature yet, this is also not time critical.
      • mimi89999
        _hc: They are 2 solutions for the same issue
      • _hc
        uniq[m]: could you look at s!320 and s!321 when you have a moment. You probably know that code better than me
      • NOTICE: [server] !320: New gradle-plugin 3.0 output apk location - https://gitlab.com/fdroid/fdroidserver/merge_requests/320
      • NOTICE: [server] !321: handle gradle-plugin 3.0 output apk location - https://gitlab.com/fdroid/fdroidserver/merge_requests/321
      • basically, choose one of them
      • right mimi89999?
      • mimi89999
        _hc: In mine I forgot a to lower, but after adding it, it will be OK
      • Yes, somebody has to choose one of them.
      • est31_ joined the channel
      • BubuIIC joined the channel
      • BubuIIC has quit
      • BubuIIC joined the channel
      • relan joined the channel