Stream of Consciousness

I wish you could change the theme of vscode depending on what kind of file you were looking at

Car Shipping Cost

The average cost for auto shipping within the continental US is around 1,000 USD, with the lowest flirting near 700-750 USD and the highest average price being in the 1,200-1,300 USD neighborhood. In general, the per-mile average for automobile shipping is around 0.60 USD/mile for distances over 1,000 miles. Shorter routes cost more per mile.
American Auto Shipping - How Much Does it Cost to Ship a Car

Car Depreciation

• Driving the car off of the lot generally results in about 10% depreciation.
• After that fast initial drop the depreciation progresses a bit more slowly - often by about 10 to 15 percent per year. According to some experts, mileage depreciation is about $0.08 a mile, but depreciation is more commonly measured in years than miles.

Foreign Keys Reminder

A foreign key constraint specifies that the values in a column (or a group of columns) must match the values appearing in some row of another table. We say this maintains the referential integrity between two related tables.
PostgreSQL Docs

• You need to add ON DELETE CASCADE when you want the referencing row to be deleted when the referenced row is deleted.
• If the two tables are independent, then you can add ON DELETE RESTRICT or ON DELETE NO ACTION

Creating Search Reference

1. Create a table to hold what you want to search.
2. Add a document column (the column which holds the text that you want to search for).
3. Add a column search tsvector GENERATED ALWAYS AS (to_tsvector('english',document) || ' ') STORED NOT NULL
4. Add a column for embedding if you want to include embeddings.

CREATE TABLE <search_table> (
	...
	document TEXT NOT NULL,
	search tsvector GENERATED ALWAYS AS (
		to_tsvector('english',document) || ' '
	) STORED NOT NULL,
	embedding vector(1536) NOT NULL, -- number inside vector() should be the length of the vector
	...
);

Creating Indexes

1. To create an index for string similarity matching on the document column, use a GIST index or GIN index - these are recommended by pgtrm
2. To create an index for the search column, add a GIN index to speed up searches.
3. To create an index for the embedding column, add a hnsw index for the embedding column.
4. 1. The <option> below can either be:
   2. 1. vector_ip_ops: Inner Product
      2. vector_cosine_ops: Cosine Distance
      3. vector_l1_ops: L1 Distance
      4. bit_hamming_ops: Hamming Distance
      5. bit_jaccard_ops: Jaccard Distance

CREATE INDEX text_sim_index ON <search_table> USING GIST (document);
CREATE INDEX search_index ON <search_table> USING GIN (search);
CREATE INDEX embedding_index ON <search_table> USING hnsw (<embedding_column_name> <option>);

Lexical Reminder

How Lexical copies Lexical Nodes:

• Copy:
• • copyToClipboard: Lexical content serialized with $getLexicalContent
• Paste:
• • $insertDataTransferForRichText: if the DataTransfer data has application/x-lexical-editor data, then that data is JSON.parsed, nodes are generated with $generateNodesFromSerializedNodes, and inserted with $insertGeneratedNodes

PostgreSQL

• When you add a FOREIGN KEY, make sure that you include ON DELETE CASCADE if you intend the row in the child table to be deleted when the referenced row is deleted.

CSS Reminder

• Whenever using position: sticky, you probably need to increase the z-index of the component and set opacity: 1 so that the sticky content is not partially covered up by content on the page.

Test Stream of Consciousness Post with Lexical Editor