social.tchncs.de is one of the many independent Mastodon servers you can use to participate in the fediverse.
A friendly server from Germany – which tends to attract techy people, but welcomes everybody. This is one of the oldest Mastodon instances.

Administered by:

Server stats:

3.8K
active users

#sqlite

5 posts5 participants0 posts today

As a stress test I loaded 8M bookmarks into a little SQLite backed bookmark saving/sharing web app I'm building. The slowest part turned out to be the `SELECT count(1) FROM bookmarks` query that's used as part of the pagination UI. Not what I was expecting but known/expected behaviour from what I've read. It takes 400ms on Snapdragon X Elite CPU, which is fine in practice—I'd be surprised if anyone actually stored more than 100k items in it.

Subway Stories: Building an Visualization w/ Transit Data - Jediah Katz & Marc Zitelli -nyhackr Feb

"Every year, New Yorkers take more than a billion trips on the subway. Using data from the MTA, we mapped out how riders flow between stations at every hour. Each story explores a slice of city life."

Brilliant deep dive into prototyping, building and deploying the New York Subway Stories website.

youtube.com/live/83s_MWc5Yvs?s

subwaystories.nyc/

I'm a database extremist. I default to #SQLite for pretty much anything.

As an independent software consultant, I need time tracking and CRM like anyone else. Guess what? It's just a single SQLite database file. :D Has worked wonderfully for years using just a database viewer app. (Tableplus, if you're wondering.)

But now, with this wonderful tool from Rune Kaagaard, I can hook up my database to Anthropic's Claude, and get insights into my time spent as well as my connections, just by chatting with Claude Sonnet 3.7.

Wild! :D

github.com/runekaagaard/mcp-al

GitHubGitHub - runekaagaard/mcp-alchemy: A MCP (model context protocol) server that gives the LLM access to and knowledge about relational databases like SQLite, Postgresql, MySQL & MariaDB, Oracle, and MS-SQL.A MCP (model context protocol) server that gives the LLM access to and knowledge about relational databases like SQLite, Postgresql, MySQL & MariaDB, Oracle, and MS-SQL. - runekaagaard/mcp-alchemy

Anyone using #litestream for an #sqlite DB

I have a 3GB database, and Litestream is replicating it to another server. However, it seems to be running out of control with disk space. On the original server (where the live DB is) is a
generations folder under the hidden litestream folder. It is full of WAL files. Currently over 37,000 files consuming 118GB of disk space.

My config includes
retention: 24h and retention-check-interval: 1h with a snapshot-interval: 4h. I understood it would delete anything older than 24 hours. Am I doing something wrong? These wal files go back to the day I started using litestream

#GoToSocial #gts @gotosocial@thefedi.forum

I did not expect my clumsy announcement of #BadgeFed to make waves, and now I have a few followers in the @fediverse

It is only less than a dozen, but still I want to honor the follow LOL, and my database for the poc was a #sqlite. Now I need to download and backup that db, otherwise I will lose the pub/priv keys.

Continued thread
Brought to you by Mozilla #Firefox using dozens of #SQLite databases and clearly not batching the transactions.

Did no one stop to think for a second of the performance implications of dozens of close-together fsyncs|dsyncs on HDDs?

Also, why in the world so many databases? It's completely unnecessary and even though they *could* be used as attached databases, that means one can only use them in non-WAL mode (yay, blocking readers) to maintain proper global transactionality (https://sqlite.org/lang_attach.html https://sqlite.org/wal.html).

Did they even read the docs before doing this nonsense?
sqlite.orgATTACH DATABASE