thangchung
@thangchung.bsky.social
OSS contributor, software developer, and architect. Microsoft Azure MVP. Newbie Rustaceans. Opinions are my own.
1. LLM-generated code tries to run code from online software packages. Which is normal but 2. The packages don’t exist. Which would normally cause an error but 3. Nefarious people have made malware under the package names that LLMs make up most often. So 4. Now the LLM code points to malware.
David D. Levine@daviddlevine.com · last yr.LLMs hallucinating nonexistent software packages with plausible names leads to a new malware vulnerability: "slopsquatting."
⚛️📝 New on Overreacted: JSX Over The Wire overreacted.io/jsx-over-the...
JSX Over The Wire — overreacted
Turning your API inside-out.
overreacted.io
Deploying Distributed Apps (Spin + Dapr) on SpinKube (AKS) #spin #spinkube #aks #dapr /cc @technosophos.bsky.social @daprdev.bsky.social @fermyon.com Check it out: dev.to/thangchung/d...
Deploying Distributed Apps (Spin + Dapr) on SpinKube (AKS)
Introduction In the last part, we walk through the journey of how to build and run Spin...
dev.to
Orchestrating Distributed Apps (Spin/Rust-WebAssembly and .NET/C#) with .NET Aspire/Dapr Check it out: dev.to/thangchung/o... @technosophos.bsky.social @davidfowl.com #dotnet #aspire #webassembly
I didn't know about Neosync, but I might start mentioning it to some prospects. It's an open-source platform that for data anonymization, synthetic generation, and environment synch - very useful for when you can't use real data for a PoC. Other interesting ones too: dev.to/bytebase/dat...
Database Tools 🛢️🔧 in 2024: A Year in Review 🗓️
Database tools are the critical enablers that bridge the diverse needs of developers, database...
dev.to
>My nervous system developed wrapped around a computer.
Tejas Kumar@tej.as · 2y agoI’ve been coding since I was 8 years old. I often code on nights and weekends and I just… love it. I also have had a hard time feeling wrong for doing so because I keep hearing about work/life balance from folks and how I shouldn’t. I wrote a blog post about this tension. tej.as/blog/coding-...
What *is* Flink CDC? Here's my updated mindmap with some clarity added around source connectors vs pipeline connectors. tl;dr: Flink CDC Pipelines are *very* cool, but with limited support for connectors _currently_. Stay tuned for detailed blog later this week! #dataBS
Look at this! 🤩 @pydantic.bsky.social for AI Agents 🤖🚀
Samuel Colvin@sa.muelcolvin.com · 2y agoPydanticAI is here! An Agent Framework designed for production, from the team who created and maintain @pydantic.bsky.social. As some of you will know, I've been working on this for some time, can't wait to see what people build with it. ai.pydantic.dev
That stuff again? I have written about it in the past hauleth.dev/post/beam-pr...
Hauleth's blog - How much memory is needed to run 1M Erlang processes?
How to not write benchmarks
hauleth.dev
Did you know that a semantic layer is not something entirely new? ↠ Is there any other semantic layer through the evolution that comes to mind?
The practical .NET Aspire 9 builds on the coffeeshop app business domain, check it out 👉 github.com/thangchung/p... #aspnetcore #dotnet
GitHub - thangchung/practical-dotnet-aspire: The practical .NET Aspire builds on the coffeeshop app business domain
The practical .NET Aspire builds on the coffeeshop app business domain - thangchung/practical-dotnet-aspire
github.com
The disaggregated wal blog.schmizz.net/disaggregate...
The disaggregated write-ahead log
The traditional way replicated systems are architected is to physically co-locate the write-ahead log (WAL) on the nodes where the state is being maintained. Then a consensus protocol like Paxos or Raft is used to make sure the log on each replica ag...
blog.schmizz.net
I'm currently exploring #SchemaEvolution. Very early WIP, but do you see other pointers that relate to the history of schema evolution? #databs
Simon Späti 🏔️@ssp.sh · 2y agoschema evolution I did all my professional life, maybe we didn't have a name, but the schema was changing all the time :). Data contracts are just a word picked up from socials.
New post! Materialized View turned one last week. 🎂 To celebrate, I wrote about what’s good and bad about DuckDB.
DuckDB Is Not a Data Warehouse
DuckDB is a tool, not a product.
materializedview.io
So you want to roll your own application platform. All you need is: Linux Docker Kubernetes Istio Prometheus Fluentd Grafana Jaeger Harbor Open Policy Agent Vault GitHub Actions and Argo CD Oh, almost forgot, you're also going to need servers, people, and glue. Bring lots of glue.
👀 mooncake.dev launched today. Supports both read/write for data lake in PostgreSQL (and cross table joins). Lots of these lately (pg_analytics, pg_duckdb, pg_parquet), and still very much a WIP. They're already on Neon, though. ⚡
GitHub - Mooncake-Labs/pg_mooncake: Columnstore Table in Postgres
Columnstore Table in Postgres. Contribute to Mooncake-Labs/pg_mooncake development by creating an account on GitHub.
github.com
Just leaving this here for reference if someone wants to try gist.github.com/sspaeti/31d5...
Reading bsky posts with DuckDB example.
Reading bsky posts with DuckDB example. GitHub Gist: instantly share code, notes, and snippets.
gist.github.com
In the database world the idea of OLTP vs. OLAP is a common one, but if you're not a database person what does that actually mean. While running the risk of oversimplifying there are two parts to this, the way data is stored and the type of queries you are running against it. More details in 🧵
a favorite part of the pre-print published today is how few of these lines are dashed. account migration and PDS federation are big, for sure, but much of the network is already open! nothing stopping somebody from running PDS+Relay+AppView today, w/ content from bsky PDS arxiv.org/abs/2402.03239
