manyfold3d / manyfold Goto Github PK

Would it be possible to create an Unraid docker template for this to simplify the installation? I'm not sure what would be required (sorry).

Depends on #30

I made a library that I don't want, and I'd like to delete it, please.

Many miniature proxies are the incorrect scale. It would be great to be able to store a scale factor for a model or part at least as metadata, and perhaps even rescale the files themselves to fix the models.

So that we can have a single container, without any need for docker-compose to set up postgres and redis.

Add, remove, and create new tags manually.

So that one day we can auto-organise the files on disk, we first need a standard, or a few options for standards.

We need to decide:

• How to name model folders
• How to name part files
• How to structure folders within a model
• How to structure folders within a library, that contain models

I would love to get some examples from the community on how they structure their folders.

Hi guys,

I'm having an issue where the stl previews don't seem to be rendering correctly, they just show as a grey box as seen below.

I can download the files fine and the .png file preview seems to work. the files inside the container are root:root 775.
Sometimes when i move from page to page i will see the stl files render for a split second but them disappear again. could it be maybe something different about my stl files thats causing the renderer to crash? they are all exported from tinkercad if that makes any different? here's an example stl
wyse_bottom.stl.zip

checked the /logs URL but don't see any specific errors that stand out
logs.zip

running from the latest container image on ubuntu 20.04.03 lts, 5.4.0
docker ver 20.10.2:2291f61

my compose

version: "3"
services:
  app:
    image: ghcr.io/floppy/van_dam:latest
    container_name: 3dp
    ports:
      - 3214:3214
    volumes:
      - "/var/lib/docker/containers/3dp/libraries:/libraries"
    environment:
      DATABASE_URL: postgresql://username:[email protected]/van_dam?pool=5
      SECRET_KEY_BASE: 
      GRID_SIZE: 260
    depends_on:
      - db
      - redis

    networks:
      new-macvlan-network:
        ipv4_address: 172.16.1.75

  db:
    image: postgres:13
    container_name: 3dp_postgres
    volumes:
       - "/var/lib/docker/containers/3dp-db/data:/var/lib/postgresql/data"
    environment:
      POSTGRES_USER: username
      POSTGRES_PASSWORD: password
    ports:
      - 5432:5432
    restart: always
    networks:
      new-macvlan-network:
        ipv4_address: 172.16.1.76

  redis:
    image: redis:6
    container_name: 3dp_redis
    restart: on-failure
    networks:
      new-macvlan-network:
        ipv4_address: 172.16.1.77

volumes:
  3dp_db_data:
networks:
  new-macvlan-network:
    external:
      name: new-macvlan-network

Any other logs I can look at to help narrow down the issue? or things to test out?
Thanks.

If pagination is enabled, the tag list only shows the tags for the models on the current page. As the title says, really.

It's a bit dizzying. It's useful sometimes, but not constantly. I'd like to be able to toggle it on or off, or only have it spin when the image is focused.

I was looking for a hosting solution to share stls for me and my friends of our local printing-group and just found this awesome peace of software and set it up immediately. It's realy great!

But i got a big request: Is it possible to add multi-user support (or one password protected account) with the possibility to upload files? I then could it make "public" to my friends and share the credentials.

We would appreciate it very much!

Greetings and thank you so much for sharing your work!

Thingiverse, Cults3d, CGTrader, MyMiniFactory

Add the tag and sha in the footer. We can get these with CI_REF_NAME and CI_SHA from https://github.com/FranzDiebold/github-env-vars-action, and then inject them into env vars in the Docker container. Then we just display the env var contents in the UI.

Given a thingiverse ID or URL, could we automatically import a model?

I have a library setup called "mirror" and within this path on the disk I have several sub-folders which I use like categories. I use this to differentiate between STLs I have created, and STLs I didn't create, but I'm mirroring incase they disappear from the original source.

I went to add a new STL file to one of the subfolders, and I pressed the "Rescan" button in the interface, but it never adds the new file, the permissions are the same, but it never appears.

Given #486, we could add the ability to keep an entire collection in sync with local disk. You go to Cults3d, add something to a collection, or like it, and bingo, it's automatically downloaded and added into VanDAM.

Currently, a library is a single list of models. It might be good to group models into collections, which then are part of libraries or other collections. In fact, a Library just becomes a collection.

Don't use latest

All the source links as in #41, but also Patreon, Instagram, etc.

We should be able to use a multistage build to strip the container down in size quite a lot.

And move all files appropriately

It would be really nice to show STLs while they are inside a Zip archive (or any other archive for that matter, but mostly will be in .zip, at least for me).

Bit of background to better understand the request / feature (I hope):

I'm creating and organizing a repository of STLs files, currently uploaded and hosted through Nextcloud, but the files themselves are on a shared network drive.
Most of these are not singular files but instead an archive of multiple STLs.

Would it be possible to show (in this case without either editing or else, just showing the preview) of the files inside the archive?

Thanks!

• Add supported flag for file
• Guess if supported when scanned
• #1851
• update UI appropriately (in above PR)
• #1876

When scanned, remove stuff that's not there.

At the moment there is no way to exclude certain directories, if you use a have a folder you put all junk files in, it would be useful to be able to exclude it from scan/rescan.
Also there are certain folders you really don't want to scan: node_modules, pycache , log etc, this would speed up scanning too.
I know you can just create more entry points further down the directory tree, but being able to fine tune a rescan by excluding directories you know are not ever needing to be touched would be a nice addition.

The app container fails in docker. I'm using docker in portainer.
The db and redis containers both spin up fine and wait for connections.

The log file from the app container

Preparing database...
bundler: command not found: rails
Install missing gem executables with bundle install

The ability to keep a list of things you want to print.

This has always been the case, but for some reason the detailed view rendering is low resolution; implies something wrong in the calculations in the renderer setup, probably?

It would be good to be able to store license metadata perhaps? At the very least, commercial vs free.

NLNet Milestone 1.4

At the moment, if you use a library path that doesn't exist on the file system, it silently fails and behaves as if it's an empty library. I'd have liked it to either validate the path when I created the library or show some sort of warning if I open a library that doesn't have a corresponding path.

For example, I'd like to have everything in 3MF - it's smaller, better for 3d printing, enforces manifold geometry, etc. It would be great to be able to standardise files into a single format.

TODO:

• #1683
• #1684

NLNet Milestone 1.8

Hi,

If you have more than 15 or 16 objects in your library, when you scroll down past the 15th or 16th preview it will start flushing out the previews from previously rendered thumbnails as seen below

Get this in the console;

This only happens on chrome and safari, firefox can render them all fine.
I can add an argument into the chrome shortcut to allow for more contexts that fixes the problem as seen this post - https://bugs.chromium.org/p/chromium/issues/detail?id=771792
OS: macOS 12.1 + Windows 20H2:19042
Browser: Chrome Version 99.0.4844.84 + Safari Version 15.2 (17612.3.6.1.6)

Van_dam is a great app, thanks for sharing it.

On my mac, continuous scroll starts to lock up Chrome if I scroll through more than a couple dozen models.

Threejs also emits a lot of warnings like this:

three.module.js:20481 WARNING: Too many active WebGL contexts. Oldest context will be lost.

If we defaulted to pagination we could avoid this but still provide the option for continuous scroll to those that want it.
This seems to me like a more usable default setup, but what do you think @Floppy ?

We can use a better way of clustering parts into sections automatically based on their name. Some sort of string clustering or tree approach.

Hi.
I was thinking it would be awesome if this software would create a queue of newly imported stl-files and then prompts a user to fill in possible information such as tags etc and perhaps would then move files to folders accordingly. This way it would be easy to import new files and keep the library tidy.

Add the ability to synchronise local files with model sources like Thingiverse, Cults3d, etc. If the remote model changes, automatically pull in the change to the local copy.

We now have a few settings in the app, which it would be good to be able to set interactively. https://github.com/ledermann/rails-settings looks like a nice way of doing it so we don't roll our own.

I have a large intricate model which gets picked up as the default preview, so I manually change it to a different model, but when re-scanning, this gets lost and reset.