Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    sure one second
    seems ok to me
    Operations to perform:
      Apply all migrations: accounts, admin, auth, contenttypes, photos, refresh_token, sessions
    Running migrations:
      No migrations to apply.
    Damian Moore
    @damianmoore
    Yeah looks fine
    :(
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    lol yeah i agree
    Damian Moore
    @damianmoore
    I know it's annoying but could I suggest you try running the whole example docker-compose file (maybe in a VM) first to confirm that works? https://raw.githubusercontent.com/photonixapp/photonix/master/docker/docker-compose.example.yml
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    yeah i'll give that a go in a bit, i have a RP4 lying around somewhere i could use as a dev box
    i'm also going to try and blow away all of the backend pieces and start again from scratch to see if that does anything
    you recommend me putting the admin credentials in the compose file or leave them out to see if i can get to the initial setup screen?
    Damian Moore
    @damianmoore
    Cool.
    I'd say leave out the admin creds to make sure the onboarding screens come up.
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    perfect
    really appreciate you taking the time, i'll ping you with results either way!
    Damian Moore
    @damianmoore
    No problem. Sorry it didn't work first time. I'd be really interested to know if you figure out what the problem was you I can add it to the docs if helpful.
    Good luck!
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    absolutely, will make sure to let you know and put up a PR if i can figure out what's up and can update the docs for you accordingly. seems really weird honestly because im going through the logs in dozzle and can't see a single thing that's failing clearly besides that 400 response from the POST request
    and don't feel the need to apologize, you're the one donating time and effort to some stranger so that i can maybe pull my wife away from GCloud so i really appreciate it!
    Damian Moore
    @damianmoore
    Haha. Thanks. Our logging could be improved but hopefully that's not the issue.
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    agreed

    actually this is the only other thing i could find that failed

    today at 2:57:22 PM WARNING:root:Limited tf.compat.v2.summary API due to missing TensorBoard installation.
    today at 2:57:22 PM WARNING:root:Limited tf.summary API due to missing TensorBoard installation.

    this a problem that may result in what i'm seeing?

    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    oh crap and i forgot to mention, i'm behind cloudflare's DNS
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]

    ok, so i got rid of everything and started from scratch. saved the initial log file before i tried to load up at all so i have a copy of that initial log to share. as soon as i went to the website to try and load up for the initial, i'm met with the login screen and that same error starts to fly into the log over and over

    today at 3:32:04 PM {GW IP} - - [23/Jun/2021:19:32:04 +0000] "POST /graphql HTTP/1.1" 400 154 "https://{domain}/login" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.114 Safari/537.36"

    so it looks like it's having a problem as soon as I try to load up for some reason. i also took out all of the middlewares for traefik just to make sure there wasn't some issue i was missing from that perspective on routing but that didn't have any effect. i'll let you know how the stock run on the RPi4 goes, i apparently now have to fight with my orbi routers as the firmware they're on is borked and i cant keep them from randomly restarting throughout the day

    wahoo!
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    morning @damianmoore - went ahead and ran the instructions on the RPi4 and it worked as anticipated (forwarded to the onboarding page) - it loaded up the same login screen i normally see but eventually routed over to the right spot. Got me thinking - is there an issue with that redirection when i'm coming in from a reverse proxy url as opposed to the direct ip address? im running my normal stack off of a headless server so i've been coming in that way instead of doing the localhost approach i just took
    Damian Moore
    @damianmoore
    Hi @zutart:mysmatrix.zutart.xyz , sorry for the delayed response - it's been a busy week. I don't think there is anything that would require an IP address over a host name. I'm sending my docker-compose.yml that I use on a home server. (I've only removed some irrelevant other services and replaced my home server domain name with mydomain.com. This successfully uses Traefik with LetsEncrypt, I access it via domain name and it redirects to /onboarding as I haven't started using it fully yet. Hopefully this gives you something else to try.
    version: '3.3'
    
    services:
      traefik:
        image: "traefik:v2.2"
        restart: always
        container_name: "traefik"
        command:
          - "--log.level=DEBUG"
          - "--api.insecure=true"
          - "--providers.docker=true"
          - "--providers.docker.exposedbydefault=false"
          - "--entrypoints.web.address=:80"
          - "--entrypoints.web.http.redirections.entryPoint.to=websecure"
          - "--entrypoints.web.http.redirections.entryPoint.scheme=https"
          - "--entrypoints.websecure.address=:443"
          - "--certificatesresolvers.myresolver.acme.tlschallenge=true"
    #      - "--certificatesresolvers.myresolver.acme.caserver=https://acme-staging-v02.api.letsencrypt.org/directory"
          - "--certificatesresolvers.myresolver.acme.email=webmaster@mydomain.com"
          - "--certificatesresolvers.myresolver.acme.storage=/letsencrypt/acme.json"
        ports:
          - "80:80"
          - "443:443"
          - "8080:8080"
        volumes:
          - "/var/run/docker.sock:/var/run/docker.sock:ro"
          - "/volumes/letsencrypt:/letsencrypt"
    
      postgres:
        container_name: postgres
        image: postgres:11.1-alpine
        restart: always
        ports:
          - '5432:5432'
        environment:
          POSTGRES_DB: photonix
          POSTGRES_PASSWORD: password
        volumes:
          - /volumes/postgres:/var/lib/postgresql/data
    
      redis:
        container_name: redis
        image: redis:3.2.8
        restart: always
        ports:
          - '6379:6379'
    
      photonix:
        container_name: photonix
        image: photonixapp/photonix:latest
        restart: always
        ports:
          - '8888:80'
        environment:
          ENV: prd
          POSTGRES_HOST: postgres
          POSTGRES_DB: photonix
          POSTGRES_USER: postgres
          POSTGRES_PASSWORD: password
          REDIS_HOST: redis
          ALLOWED_HOSTS: '*'
        volumes:
          - /volumes/photonix/photos:/data/photos
          - /volumes/photonix/raw-photos-processed:/data/raw-photos-processed
          - /volumes/photonix/cache:/data/cache
          - /volumes/photonix/models:/data/models
        links:
          - postgres
          - redis
        labels:
          - "traefik.enable=true"
          - "traefik.http.routers.photonix.rule=Host(`photonix.mydomain.com`)"
          - "traefik.http.routers.photonix.entrypoints=websecure"
          - "traefik.http.routers.photonix.tls.certresolver=myresolver"
          - "traefik.http.services.photonix.loadbalancer.server.port=80"
    The TensorBoard warnings you were seeing are nothing to worry about BTW.
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    @damianmoore: absolutely no worries, my second kid is due in 2 weeks so I totally get the busy week thing. I did manage to pin down the problem (i think) - when I use docker secrets for all of the environment variables, the site doesn't work as expected. if I set the variables directly in the docker compose file, it comes up as anticipated. still can't figure out exactly why it
    's not coming up with the secrets based on the logs and the little bit i was able to look at the code, but i figured i'd send that along in case there's something that jumps out to you
    zutart
    @zutart:mysmatrix.zutart.xyz
    [m]
    nevermind - just tried again and it didn't work even with the variables directly in. at this point i'm certain there's something i'm completely missing on my end with the traefik/cloudflare/whatever build i have. if i actually ever figure out what's going on i'll let you know, but until then i'll stop bugging you!
    Damian Moore
    @damianmoore
    No worries @zutart:mysmatrix.zutart.xyz . Thanks for the update. If you find something interesting we should add it to the docs.
    osterbergsimon
    @osterbergsimon
    Hey! Trying to set up photonix, looks and feature set is great, thanks for your hard work! However, I can't figure out how to choose which library photos are imported to. What feels intuitive (and seems to be the case, according to the source) is that library depends on the path photos are uploaded to. However, the create_library command seems to set all library paths to 'data/photos/' for some reason
    Is there something I'm missing?
    thesilversverker
    @thesilversverker:matrix.org
    [m]

    But sure if I'm missing something obvious - running a test in docker-compose, site works fine. Onboards correctly, but after creating additional users via cli, the login page begins failing to display correctly. Redeployed and the issue happened again. Dev mode on the browsers shows a couple js errors about a.environment is null.

    Anyone seen something similar/thoughts on workaround?

    Ritesh Raj Sarraf
    @riteshsarraf:matrix.org
    [m]
    Hey! Just dropped by ensure I can see where the community prefers for live talk. And thank you for the work on Photonix. Good stuff.
    Damian Moore
    @damianmoore
    Hi @riteshsarraf:matrix.org , thanks for dropping by and welcome. Yes this is the official chat channel for Photonix. I try to pop in fairly regularly but the community here is still fairly small right now.
    Raphaël
    @cirdanmiriel:geekio.fr
    [m]

    Hi @damianmoore and the dev team 🙂 I am thrilled that you guys are building up what I was searching for for a very long time!
    Even if it's still in early stages, this app seems pretty advanced so far!
    Unfortunately, I'm currently facing 2 issues, trying to make it work:

    1. I can't make the Android app to work, if crashes as soon as I launch it ☹️ (xiaomi mi 9 if it can be of any help)
    2. I deployed the app via docker, in my current environment and wanted to use my existing Redis server. The problem is that this redis server currently needs a password to connect to, and it doesn't seem in the docs that there is a way to specify a redis password (or am I missing something?)

    Many thanks for the answers, and kudos to you for this project, can't wait to try it!!! 🙂

    Raphaël
    @cirdanmiriel:geekio.fr
    [m]

    :point_up: Edit: Hi @damianmoore and the dev team 🙂 I am thrilled that you guys are building up what I was searching for for a very long time!
    Even if it's still in early stages, this app seems pretty advanced so far!
    Unfortunately, I'm currently facing 2 issues, trying to make it work:

    1. I can't make the Android app to work, if crashes as soon as I launch it ☹️ (xiaomi mi 9 if it can be of any help)
    2. I deployed the app via docker, in my current environment and wanted to use my existing Redis server. The problem is that this redis server currently needs a password to connect to, and it doesn't seem in the docs that there is a way to specify a redis password (or am I missing something?) => I opened photonixapp/photonix#314 concerning this problem

    Many thanks for the answers, and kudos to you for this project, can't wait to try it!!! 🙂

    Damian Moore
    @damianmoore
    Hi @cirdanmiriel:geekio.fr . Welcome and thanks for the support with what we're building. In response to your two issues:
    1. That definitely doesn't seem good - I've not seen that happen before. Could you provide more details about your Android device (Android version and whether it's the MiUI or custom ROM etc.)
    2. Thanks for creating an issue in GitHub. I should be able to add that fairly quickly.
    Raphaël
    @cirdanmiriel:geekio.fr
    [m]
    Wow, thanks for the fast answer! Regarding your answers:
    1. No problem at all, I am running on Android 10 on a Xiaomi Mi 9, Custom ROM ( Paranoid Android Quartz 5 )
    2. That would be awesome, I will wait for it (I'ld rather not have x Redis servers running ^^)
    Damian Moore
    @damianmoore
    @cirdanmiriel:geekio.fr
    1. Thanks for the info. It's probably going to be quite difficult to replicate so I'm thinking it's probably better use of time for me to try and add some crash detection so user can see and send error report from the app. I'll be in touch again once this is done and hopefully you can update the app and get the details of the crash.
    2. Cool I'll let you know when this setting is added.
    Raphaël
    @cirdanmiriel:geekio.fr
    [m]
    @damianmoore:
    1. Indeed, it could be a lot easier, if I could be of any help regarding this problem, don't hesitate to ask 🙂
    2. That's great, thanks!
    ShiniGandhi
    @ShiniGandhi
    Hey, is there a way to delete photos through the Android app (or the web ui for that matter)?
    Damian Moore
    @damianmoore
    Hi @ShiniGandhi, thanks for your question. Currently Photonix does not yet support this but the feature is being worked on right now. You can follow progress here: photonixapp/photonix#299
    ShiniGandhi
    @ShiniGandhi
    Okay, thanks!
    Ryan Foster
    @PhasecoreX
    Hello! Is it possible to reset the application and have it rescan everything? I know I could just drop all of the tables and run the initial setup, but I would like to keep the face training data and library setup path.
    I'm having an issue where some deleted photos are still showing up in the thumbnails list (but clicking them doesn't show the photo, as expected) and the newly released "images with no time stamp being sorted in the timeline correctly" feature doesn't seem to affect existing photos. Figured I would just rescan everything, didn't know if it was possible.
    thesilversverker
    @thesilversverker:matrix.org
    [m]
    I get too - I have to purge files with .trashed filenames. Check if you have any with ls -a in your photo directory?
    Ryan Foster
    @PhasecoreX
    No, there aren't any hidden files in the photo directories. I deleted some photos and it hasn't detected the deletion I suppose. But regardless I'd like a "rescan all" thing to reorganize the photos with no time stamps that are all at the top of the timeline.
    Miguel Ruivo
    @miguelpruivo
    Hi, I've been trying to get this up and running on Docker in my Mac and I've been experiencing multiple issues — no pictures or whatsoever despite being the folder, sometimes the sign-in doesn't work, sometimes I completely remove it but there is no new onboarding after new install, the settings don't save even if I change it. What am I missing? Thank you!