How do existing dockerized visualizations get their backing data? Are there idioms we can expect outside authors to follow?
Conclusions:
- Docker documentation is often an afterthought, though there are exceptions. Often just a copy of github docs.
- Surprised by the number of desktop applications.
- Mounting a /data volume is the usual approach to getting files in there, though we also see STDIN, sockets, and providing credentials for an API.
https://hub.docker.com/r/gehlenborglab/higlass/ Data to be visualized must be in a mounted volume, and after boot ingest_tileset must be run. (I could imagine every file in the mounted volume being ingested on start up, with a mapping from filenames to uids, with the software also generating a plausible viewconf for a given set of files.)
https://hub.docker.com/r/vanessa/cogatvoxel/ Visualizing Brain anatomy? Data is baked in?
https://hub.docker.com/r/tddv/superset/ OLAP: Guess you provide connection info for DB somehow?
https://hub.docker.com/r/centurylink/imagelayers-ui/ -> http://imagelayers.io Tool for examining shared dependencies between images. Seems broken right now, but I think the docker tag is passed as a URL parameter, and then it hits the dockerhub API.
https://hub.docker.com/r/ouven/akka-visual-mailbox-visualization/ Visualize the state of a message passing system. (Reminds me of Erlang?) Web app, but another port is opened up where communication with the monitored system takes place.
https://hub.docker.com/r/jldowns/vatic-docker/ Manage mechanical turk video annotation tasks. Mounted volumes for input and output files.
https://hub.docker.com/r/inistcnrs/ezvis/ Visualization of natural language corpus. Example data hard coded?
https://hub.docker.com/r/centurylink/image-graph/ Visualizing docker images. Mount the socket file to get information about the local Docker. Can either produce as output a PNG, or run as a webapp.
https://hub.docker.com/r/nitnelave/deepvis/ Visualizing deep learning on image data. Desktop app rather than web app: I guess it works with X Window client-server, or something like that?
https://hub.docker.com/r/jupedsim/jpsvis/ Pedestrian behavior simulator. Visualize output trajectory files.
https://hub.docker.com/r/gtrafimenkov/logstalgia/ Server log analysis. Mount the directory containing the log.
https://hub.docker.com/r/gtrafimenkov/codeswarm/ Visualize software project history. Mount the repo to visualize. Output is non-interactive video, I think?
https://hub.docker.com/r/jonasrauber/c2s/ Neuroscience signal processing and visualization. Mount a data directory
https://hub.docker.com/r/funkwerk/compose_plantuml/ Generate UML from docker-compose specs. STDIN / STDOUT
https://hub.docker.com/r/3dechem/largevis/ Sci-vis: high dimensional clustering. Mount a volume with the data on start. Produces output file.
https://hub.docker.com/r/stennie/ubuntu-mtools/
https://hub.docker.com/r/sevenbridges/maftools/ An R package? Docker container installs R and libraries.
https://hub.docker.com/r/shopuz/alveo-visualization/ Unclear / Broken? I think it relies on an outside API to provide metadata on a corpus of human language data.
https://hub.docker.com/r/ecomobi/visualization/ Stub? The company is something like a web ad network for southeast Asia...
https://hub.docker.com/r/mathiask/weatherdb-visualization/ Stub?
https://hub.docker.com/r/stefan125/sofia-service-osgi-visualization-ui/ Stub?
https://hub.docker.com/r/kennethzfeng/pep-visualize/ Not sure where the data comes from.