Jorge Pablo Franetovic 🎩
@jpfraneto.eth
normalize databases as .txt files
2 replies
0 recast
9 reactions
artlu is on lunchbreak!
@artlu
I'mma get Inigo Montoya to say it lol https://www.essentialsql.com/database-normalization/
1 reply
0 recast
1 reaction
artlu is on lunchbreak!
@artlu
I'm being silly. In all seriousness, you might like Mongo after the dataset becomes too large for text files. It's an easier jump to NoSQL than to a relational schema. What you gain is that the computer doesn't have to pick up and look at every item of data, as database logic lets it do lookups efficiently using database tech
1 reply
0 recast
1 reaction
Jorge Pablo Franetovic 🎩
@jpfraneto.eth
@breck wdyt?
1 reply
0 recast
1 reaction
Breck Yunits
@breck
Indexes and fast query engines are easy to write. Using plain text files backed by git via ScrollSets is the future. Binary storage engines were needed before fast SSDs, git, et cetera, but for 99.9% of people ScrollSets will be a better choice.
1 reply
0 recast
2 reactions
Breck Yunits
@breck
You do need all the features from SQL (joins, indexes, foreign keys, etc), but we'll add all that as we go, all while keeping your data completely pristine.
1 reply
0 recast
2 reactions
artlu is on lunchbreak!
@artlu
ah, so OP is not just slurping an adhoc structured .txt file. I guess it leverages git to ensure consistency across concurrent writes to the same space? looks like this datastore will develop in an interesting space, starting close to YAML and approaching NoSQL over time.
0 reply
0 recast
1 reaction
