thinkredstone

5 ms·

1. 1.▲
   by thinkredstone 2y ago

   Also outperforms flink's setup time, and by a bigger factor
2. 2.▲
   by thinkredstone 3y ago

   Like I wrote in the article, the engine uses `rust_decimal` to represent arbitrary precision floating point numbers- this is how the engine supports SQL types likes postgres's `decimal`. We support using ieee754 floating point numbers
3. 3.▲
   by thinkredstone 3y ago

   As part of importing my markdown into medium, I created a github gist with the markdown of the post- after I imported the post into markdown, I deleted the gist, but it seems medium keeps this tag in the page HTML.
4. 4.▲
   by thinkredstone 3y ago

   With `isize` you can support up to 2^63 duplicate rows, which means that if a row is as short as one byte a single diff can represent several exabytes of rows
5. 5.▲
   by thinkredstone 3y ago

   I actually wrote the post in markdown, so even getting it into medium was superficially complicated process
6. 6.▲
   by thinkredstone 3y ago

   You might even say I have an incredible taste in books.