Why use online databases?

For example, MAL, IMDB, Letterboxd

At a higher level. Its all:

• metadata
• community

Reasons to use online databases:

Granularly:

• Share your list with others~ and be social!
• You (hopefully) get a nice interface which shows who images, descriptions, etc.
• Some of the benefits of a RDBMS, where you can filter/see who worked on what/search (depending on how much the website chooses to display)
• [IF AN API EXISTS, AND IS FREE] API access to the information, information on episodes/images (e.g. Trakt API info to TVDB)
• Discover new entries (in general), and based on recommendations that the system/other users give

For anyone who is serious about exploring media, online databases are impractical to avoid. You’re either going to be using one (and perhaps contributing back to one), or using the metadata from one to create your local index of media to check out based on different creators/genres/rankings.

Unfortunately, databases by the nature catalogue a specific thing, and there will always be media that lie in the between the focus of those databases. Exclusively using one site means you miss out on media, using multiple sites means there’s overlap and its not clear whether to log entries on one place/everywhere (possibly to gain recommendations on one site). Using your own local system means there’s no restrictions on what is allowed, but that means theres no restrictions on what is allowed. You have to personally impose rules, else you risk your system becoming too disorganized to function.

Reasons to not use online databases:

• Downtime
• If no API, no way to query your list. However, your own local list/spreadsheet means without a bunch of work, no access to metadata to filter your personal list
• Using a website in general:
  • Not owning your information (especially since this is a personal list, and not being able to export it easily means if the site disappears, your list/community disappears)
  • General advertising/tracking concerns
  • No way to add a feature you may want

Using a personal system (like CSV/google sheets)

• If no metadata, can’t do fun things like:
  • query by genre
  • connect by people who worked on two projects
  • anything else metadata and an RDMBS lets you do
• share/compare you lists to others (but how much do I care about this… ? Could just create a SSG against local data and host that on my website instead of investing in another content silo). Or… get/maintain a list of entries/IDs and back-fill info into the databases based on local changes. Local First, then update remotely to maintain the community aspect.

My general sentiment is, I want to move closer to what I did with my music/albums solution, but using local CSV/Sqlite files instead of google sheets. Create static site generators based on the local databases, and run those on my server whenever a git repo gets updated

databases I use

Generally Good Metadata, have APIs which I use:

• Trakt - for https://sean.fish/feed
• Discogs - for https://sean.fish/s/albums (my google sheet/SQL system using discogs metadata)

Generally good metadata:

• https://letterboxd.com
• https://www.grouvee.com (this uses giantbomb internally)

Average metadata, tons of downtime, generally not a great experience:

• https://myanimelist.net/ - however, im personally invested a lot and have a lot of the domain knowledge of how the site works here, and am friends with a lot of the power users. Moving away from MAL is something Ive tried to do a few times, but haven’t for one reason or another.