4 ms·

Boring technology is so crucial. As a private sector startup company you are already living a fraught existence - why risk it even more with unproven tech? As

Boring technology is so crucial. As a private sector startup company you are already living a fraught existence - why risk it even more with unproven tech? As an example, mongodb inexplicably became a de facto DB for a bit between 2013 and 2016. It seemed like 1/3 of startup companies switched from MySQL or Postgres to Mongodb. What an absolute unmitigated shitshow clusterfuck moronic disaster

kaptainscarlet 3y ago
I remember the days Mongo or NoSQL in general. It's all in memes these days
- coffeebeqn 3y ago
  Every company I’ve worked at uses Dynamo. It counts as boring. Most also use SQL - it’s really not difficult to choose the right tool for the job
  - monero-xmr 3y ago
    NoSQL in general is a niche tech. Data is inherently relational. It is very hard for me to imagine a situation where basing everything on Dynamo or Mongo is the correct move, vs. only using NoSQL in specific problems that make sense. Dynamo and mongo generally smell of developers enamoured with specific tools over solving problems efficiently.
pastacacioepepe 3y ago
Literally every client I've had in the past months uses MongoDB, even though their data is relational. I can't imagine why would someone choose mongo in that case except to be slightly faster in the first few days of development, since you don't have to create a schema and update it. The problem is that you pay back that initial advantage with a lot of interests once your data becomes more complex and you need stuff like BI. Anyway, even if your data isn't relational, JSON in postgres still works better than mongo.
- redwood 3y ago
  What makes data relational in your view?
  - pastacacioepepe 3y ago
    Data with multiple entities that have a relationship between each other. Take a website that has users, posts and comments, for example. Users have many posts and many comments, posts have many comments and one user.
    - redwood 3y ago
      Got it: I consider different objects or entities stored, indexed, and queried separately but with references to each other in the form of identifiers to really just be sound design. Relational data models typically come with more constraints that inhibit classes or online data mutations and scalability choices. I think it's easy to conflate multiple sets of tradeoffs. The deepest question is whether it's useful to be able to model the world with first class structure in the form of embedded arrays and embedded sub structures that can be richly indexed. Does this involve certain forms of denormalization? Sure but do tables really represent how you think about the world?
      - monero-xmr 3y ago
        I always struggle to think of you would ever want NoSQL to represent data. The benefit of NoSQL is... lack of schemas? Why would you want this except with unstructured user-generated content (which then will need to be structured later to be useful)?
        redwood 3y ago
        I think there are may be some misconceptions here. There's always a schema. It's about flexibility and heterogeneity of the schema shape and evolution over time
ericjohannsen 3y ago
Yep. Chasing shiny new tech can cause a company to focus on learning the tech with all its early stage issues rather than focus on solving the problems the business faces. Some new tech fades away. Some is here to stay. SQL was the shiny thing near the start of my career. I pushed my company hard to abandon network databases in favor of SQL and that worked out quite well. Same with object-oriented programming over classic C. Some tech never lives up to its promise and becomes an albatross for its adopters. There was a time when it was too early to responsibly incorporate LLM technology. There will soon be a time when the board will ask why you didn't. (Maybe for external use cases, certainly for internal ones). One challenge for CTOs is making the call on new tech. When is boring tech the best thing for the company, when is it acceptable to introduce early stage tech, and when is new tech such an accelerator that you must adopt it?

kaptainscarlet 3y ago

I remember the days Mongo or NoSQL in general. It's all in memes these days

coffeebeqn 3y ago
Every company I’ve worked at uses Dynamo. It counts as boring. Most also use SQL - it’s really not difficult to choose the right tool for the job
- monero-xmr 3y ago
  NoSQL in general is a niche tech. Data is inherently relational. It is very hard for me to imagine a situation where basing everything on Dynamo or Mongo is the correct move, vs. only using NoSQL in specific problems that make sense. Dynamo and mongo generally smell of developers enamoured with specific tools over solving problems efficiently.

pastacacioepepe 3y ago

Literally every client I've had in the past months uses MongoDB, even though their data is relational. I can't imagine why would someone choose mongo in that case except to be slightly faster in the first few days of development, since you don't have to create a schema and update it. The problem is that you pay back that initial advantage with a lot of interests once your data becomes more complex and you need stuff like BI. Anyway, even if your data isn't relational, JSON in postgres still works better than mongo.

redwood 3y ago
What makes data relational in your view?
- pastacacioepepe 3y ago
  Data with multiple entities that have a relationship between each other. Take a website that has users, posts and comments, for example. Users have many posts and many comments, posts have many comments and one user.
  - redwood 3y ago
    Got it: I consider different objects or entities stored, indexed, and queried separately but with references to each other in the form of identifiers to really just be sound design. Relational data models typically come with more constraints that inhibit classes or online data mutations and scalability choices. I think it's easy to conflate multiple sets of tradeoffs. The deepest question is whether it's useful to be able to model the world with first class structure in the form of embedded arrays and embedded sub structures that can be richly indexed. Does this involve certain forms of denormalization? Sure but do tables really represent how you think about the world?
    - monero-xmr 3y ago
      I always struggle to think of you would ever want NoSQL to represent data. The benefit of NoSQL is... lack of schemas? Why would you want this except with unstructured user-generated content (which then will need to be structured later to be useful)?
      - redwood 3y ago
        I think there are may be some misconceptions here. There's always a schema. It's about flexibility and heterogeneity of the schema shape and evolution over time

ericjohannsen 3y ago

Yep. Chasing shiny new tech can cause a company to focus on learning the tech with all its early stage issues rather than focus on solving the problems the business faces. Some new tech fades away. Some is here to stay. SQL was the shiny thing near the start of my career. I pushed my company hard to abandon network databases in favor of SQL and that worked out quite well. Same with object-oriented programming over classic C. Some tech never lives up to its promise and becomes an albatross for its adopters. There was a time when it was too early to responsibly incorporate LLM technology. There will soon be a time when the board will ask why you didn't. (Maybe for external use cases, certainly for internal ones). One challenge for CTOs is making the call on new tech. When is boring tech the best thing for the company, when is it acceptable to introduce early stage tech, and when is new tech such an accelerator that you must adopt it?