SQL or NoSQL!? 👩‍💻 #code #programming #technology #tech #software #developer #programmer #coding

[sohaibmalik]

Here's an example: For my billing system I used SQL database due to its advanced queries of calculations and many other stuff. However for my movie app I used NoSQL, due to that key value factory and large data like 500 words movie description etc.

[shapelessed]

Nah... The best *sequel* was The Dark Knight Rises, thank you very much.

[MK-bsdq]

Perhaps horizontal scaling (sharding) in SQL exists long time ago.

[kkiimm]

Use SQL until you are experienced enough to recognize situations where it is unnecessarily hard, then start experimenting with nosql. You're most likely going to need SQL over nosql in most jobs, so prioritize SQL until you're good.

[rami]

Friendly reminder that nosql stands for not ONLY sql, depending on your needs, using nosql while still having defined schemas can be useful.

[luukw]

Currently using a combination of SQL Server and ScyllaDB. They both shine in their respective areas/strengths.

[Simon-xitb]

NoSQL is just a bad word. It doesn't make sense anymore. It just doesn't fit to what is actually avlable now. I could imagine having non relational db that uses sql, or db that has ACID guarantees but is not relational etc etc..

[Dev-Siri]

Currently I have two production apps deployed.
One uses SQL (PostgreSQL) which is the app's primary database. (with Redis cache)
And there are two other apps. A survey/feedback app and a chat (something like a support), which uses NoSQL (MongoDB).

I feel like NoSQL is easier to work compared to SQL. But SQL is WAY better for related data. Sometimes I tried to do some relational stuff with NoSQL but always ended up shooting myself in the foot. I know there are ways like sub collections and stuff, but they have been a pain to work with compared to the SQL.

[yutubl]

Yes, combine both related cherry picking each strength: a) NoSQL = flexible changing data structures not scaling with big data as fast than b) SQL: fix table-formatted hoarding big data of constant structured not deep nesting data that don't change over time (measurement values, money).

[nefrace]

I'm not using "sequel". I'm using "squeal".

[curiousnerd]

Not all sql db have to be vertically scaled, sqldbs like Microsoft SQL support sharding

[sonychandel]

SurrealDB: I see this as an absolute win

[blender_wiki]

In nowadays we have many solutions to scale SQL db horizontally depends on the kind of service that uses the db.
And for the future don't forget to use a vector database to store embeddings 😉

[JustinShaedo]

I'm an amateur who doesn't understand why people keep saying SQL schemas are inflexible; yet there's nothing that can't be changed in a SQL schema. I get that, for example in Mongo you can have individualized data in a way that SQL can't but that's not a lack of flexibility. Again, I'm an amateur so I understand this is my failing to understand here.

[BarisPalabiyik]

With Vitess you can horizontally scale the MySQL which I prefer, you are losing some MySQL features but they working on it. Youtube also uses it.

[ckydtvg]

Can you do a video on a possible use case that benefits from using sql and nosql together?

[-maxipoo-]

Wait it’s pronounced ‘sequel’?!?! I’ve always says ‘S-Q-L’…

[lunedefroid]

SQL as a primary data storage is a perfect option with zero drawbacks. NoSQL are only good as complementary storage for edge cases

[fluffybunny]

I recommend postgresql using some library is good to work with, otherwise Mongodb and casendra is better

[tryoasnafi]

SQL good if your data structure clear, need complex query, and need ACID. NoSQL good for large and unstructured data, and less update