HackerTrans
TopNewTrendsCommentsPastAskShowJobs

matheusalmeida

no profile record

Submissions

Is Linux disk I/O fast enough?

floedb.ai
10 points·by matheusalmeida·قبل 4 أشهر·0 comments

Why JSON Isn't a Problem for Databases Anymore

floedb.ai
7 points·by matheusalmeida·قبل 5 أشهر·0 comments

Two Bits Are Better Than One: making bloom filters 2x more accurate

floedb.ai
184 points·by matheusalmeida·قبل 5 أشهر·35 comments

Databases should contain their own Metadata – Use SQL Everywhere

floedb.ai
46 points·by matheusalmeida·قبل 5 أشهر·33 comments

[untitled]

1 points·by matheusalmeida·قبل 5 أشهر·0 comments

Floecat: A Catalog of Catalogs for the Modern Lakehouse

floedb.ai
8 points·by matheusalmeida·قبل 5 أشهر·0 comments

Making geo joins faster with H3 indexes

floedb.ai
173 points·by matheusalmeida·قبل 5 أشهر·69 comments

comments

matheusalmeida
·قبل 5 أشهر·discuss
The key difference is that it's not just about schema metadata (tables, indexes, views, columns, etc...). PostgreSQL is fabulous regarding this. Even native types are part of the catalog (pg_catalog).

Things are great in your DB... until they aren't. The post is about making observability a first-class citizen. Plans and query execution statistics, for example, queryable using a uniform interface (SQL) without the need to install DB extensions.