andersmurphy

7 ms·

91.
by andersmurphy 5mo ago
Is the main use case for this for languages that only have access to process based concurrency? Struggling to see why you would otherwise need this in java/go/clojure/C# your sqlite has a single writer, so you can notify all
92.
by andersmurphy 6mo ago
Yeah, there's always Fil-C (Rust isn't memory safe in practice).
93.
by andersmurphy 6mo ago
Yeah JS is closer to lisp/scheme than C (I say this as someone who writes JS, Clojure and the occasional C).
94.
by andersmurphy 6mo ago
In some ways Mythos (like many AI things) can be used as the ultimate accountability sink. These libraries/frameworks are not insecure because of bad design and dependency bloat. No! It's because a mythical LLM is so powerful that
95.
by andersmurphy 6mo ago
Absolutely love the one on Forth. Issue 5 I think? Has a build it yourself modem for under 50$ in it.
96.
by andersmurphy 6mo ago
That's the irony of Mythos. It doesn't need to exist. LLM vibe slop has already eroded the security of your average site.
97.
by andersmurphy 6mo ago
For what it's worth blob types and application functions make it pretty straight forward to implement your own datatypes. I often have one for EDN and one for bigdecimal.
98.
by andersmurphy 6mo ago
The bigger change is those that do write by hand have stopped open sourcing their work. The age of sharing innovation for free is quickly coming to an end. I also expect a bunch of open source/open core projects to go closed source in
99.
by andersmurphy 6mo ago
Recommend this series if you want to dabble in assembly/forth: https://tumbleforth.hardcoded.net/
100.
by andersmurphy 6mo ago
Python from scratch now means writing python without LLMs? It's like wild swimming all over again. The shark is being jumped.
101.
by andersmurphy 6mo ago
I've been diving into aarch64 assembly recently building a small Forth. It's honestly really refreshing. If you're prepared to forgo some portability and pick an architecture assembly opens up a lot if options. Things like co
102.
by andersmurphy 6mo ago
Coffi [1] built on java 22 (project panama) C FFI. Makes binding to C from Clojure a lot of fun. - [1] https://github.com/IGJoshua/coffi
103.
by andersmurphy 6mo ago
We don't know what anthropic's true costs are. So pricing that in is at best a guess.
104.
by andersmurphy 6mo ago
Yup. Also regardless of price they need to spend more and more as the project collapses under the inevitable incidental complexity of 30k lines of code a day. It's similar to how if you know what you're doing you can manage a simp
105.
by andersmurphy 6mo ago
> What happens is that weak models hallucinate (sometimes causally hitting a real problem) So the bigger models hallucinate better causally hitting more real problems?
106.
by andersmurphy 6mo ago
That's a good point. I'll give that a go and do a write up at some point. The choice to use SQLite for me isn't actually about speed (LMDB is way faster). I just get tired of people saying switch to Postgres for speed/sc
107.
by andersmurphy 6mo ago
Clearly they didn't use enough hammer. Screws always require the most hammer. Common knowledge to any certified practitioner of Hammer™[1]. - [1] Get 20% off your Hammer Master™ certificate with referral code THUMBPAIN
108.
by andersmurphy 6mo ago
This is a great point! Reminds me of Agentic software development. When it doesn't work out it's only evidence that you could have used more agents. You can never use enough tokens.
109.
by andersmurphy 6mo ago
Depends how you do your backups. If you do them by replicating. They are both. See litestream [1]. With SQLite this is even more obvious as a database is just a file (or three in the case of WAL). Which means you can replicate to not just a
110.
by andersmurphy 6mo ago
> How about lisp? I was wondering why lisp (and forth) were omitted from the initial list of languages named in the post. I guess Scheme is in the list has ok macros.
111.
by andersmurphy 6mo ago
Right and that window is bigger for RDS by the looks of it.
112.
by andersmurphy 6mo ago
Sqlite supports nested transactions with SAVEPOINT so each client can have their own logical transaction that can be rolled back. The outer transaction just batches the fsync effectively. So an individual client failing a transaction doesn&
113.
by andersmurphy 6mo ago
An interactive transaction works like this in pseudo code. beginTx // query to get some data (network hop) result = exec(query1) // application code that needs to run in the application safeResult = transfor
114.
by andersmurphy 6mo ago
Not specific books per say. Though I'd advise starting with some constraints. As that really helps you focus. Your reading/learning material can spin out of those constraints. So for me my recent constraints were: 1. Multiplayer&#
115.
by andersmurphy 6mo ago
Litestream [1] is quick to set up and has point in time backup to the second. - [1] https://litestream.io
116.
by andersmurphy 6mo ago
> for inserts only into singe table with Actually, there are no inserts in this example each transaction in 2 updates with a logical transaction that can be rolled back (savepoint). So in raw terms you are talking 200k updates per second
117.
by andersmurphy 6mo ago
If anything backups are the key to high availability. Streaming replication lets you spin up new nodes quickly with sub second dataloss in the event of anything happening to your server. It makes having a warm standby/failover trivial
118.
by andersmurphy 6mo ago
That only solves read latency not write latency. Unless you don't care about consistency.
119.
by andersmurphy 6mo ago
Last I checked RDS uploads transaction logs for DB instances to Amazon S3 every five minutes. Litestream by default does it every second (you can go sub second with litestream if you want).
120.
by andersmurphy 6mo ago
Backups, litestream gives you streaming replication to the second. Deployment, caddy holds open incoming connections whilst your app drains the current request queue and restarts. This is all sub second and imperceptible. You can do fancier

More ›