site logo
Notes on the margins
Rost Glukhov. Personal site and technical blog
  • Documentation tools
  • Hardware
  • LLM Hosting
  • LLM Performance
  • RAG
  • Observability
  • Coding
  • DevOps
  • CookBook
  • Hardware News
  • AI
  • Ollama
  • Cheatsheets
  • HowTo
  • Offline
  • About
Building a Dual-Mode AWS Lambda with Python and Terraform

Building a Dual-Mode AWS Lambda with Python and Terraform

Step-by-step example

Here we have a Python Lambda example of SQS Message Processor + REST API with API Key Protection + Terraform script to deploy it for serverless execution.

Comparison: Qwen3:30b vs GPT-OSS:20b

Comparison: Qwen3:30b vs GPT-OSS:20b

Comparing Speed, parameters and performance of these two models

Here is a comparison between Qwen3:30b and GPT-OSS:20b focusing on instruction following and performance parameters, specs and speed.

Writefreely Federated Blogging Platform - selfhosting vs managed costs

Writefreely Federated Blogging Platform - selfhosting vs managed costs

Approximate hosting cost estimations vs subscription.

Here’s a quick info on Write.as / WriteFreely - how it fits into the fediverse, where to get managed hosting, what the usage trend looks like, and how to self-host (plus rough costings).

Integrating Ollama with Python: REST API and Python Client Examples

Integrating Ollama with Python: REST API and Python Client Examples

+ Specific Examples Using Thinking LLMs

In this post, we’ll explore two ways to connect your Python application to Ollama: 1. Via HTTP REST API; 2. Via the official Ollama Python library.

Proxmox in 2025: A Practical, All-In-One Virtualization Stack

Proxmox in 2025: A Practical, All-In-One Virtualization Stack

What is Proxmox?

Proxmox Virtual Environment (Proxmox VE) is an open-source, type-1 hypervisor and datacenter orchestration platform built on Debian.

Configuring Multi-Language Website SEO with Hugo

Configuring Multi-Language Website SEO with Hugo

What to do with `canonical`, `lang` and `hreflang`

Translating your Hugo website can improve your Google and Bing ranking - but only if the translated content is high quality, localized, and properly structured for multilingual SEO.

Privacy-Oriented Browsers: Practical Guide to Safer Web Browsing

Privacy-Oriented Browsers: Practical Guide to Safer Web Browsing

Including intallation instructions for Windows, Linux and Mac

Online privacy is a hot topic. From targeted ads to silent trackers, many mainstream browsers share more about you than you’d like. Fortunately, several privacy-oriented browsers exist to help you stay safe and anonymous.

NVidia RTX 5080 and RTX 5090 prices in Australia - October 2025

NVidia RTX 5080 and RTX 5090 prices in Australia - October 2025

AI-suitable Consumer GPU Prices - RTX 5080 and RTX 5090

Let’s compare prices for top-level consumer GPUs, that are suitable for LLMs in particular and AI in general. Specifically I’m looking at RTX-5080 and RTX-5090 prices. They have slightly dropped.

Building, Publishing and Monetizing Alexa Skills

Building, Publishing and Monetizing Alexa Skills

How to develop Amazon Alexa Skill - instructions.

In this article, we’ll take a deep dive into Alexa skill development, testing, and publishing. We’ll also cover design principles, technical considerations, and best practices to ensure your skill provides an engaging and natural user experience.

Ollama GPT-OSS Structured Output Issues

Ollama GPT-OSS Structured Output Issues

Not very nice.

Ollama’s GPT-OSS models have recurring issues handling structured output, especially when used with frameworks like LangChain, OpenAI SDK, vllm, and others.

Structured output comparison across popular LLM providers - OpenAI, Gemini, Anthropic, Mistral and AWS Bedrock

Structured output comparison across popular LLM providers - OpenAI, Gemini, Anthropic, Mistral and AWS Bedrock

Slightly different APIs require special approach.

Here’s a side-by-side support comparison of structured output (getting reliable JSON back) across popular LLM providers, plus minimal Python examples

Constraining LLMs with Structured Output: Ollama, Qwen3 & Python or Go

Constraining LLMs with Structured Output: Ollama, Qwen3 & Python or Go

A couple of ways to get structured output from Ollama

Large Language Models (LLMs) are powerful, but in production we rarely want free-form paragraphs. Instead, we want predictable data: attributes, facts, or structured objects you can feed into an app. That’s LLM Structured Output.

PostgreSQL Cheatsheet: A Developer’s Quick Reference

PostgreSQL Cheatsheet: A Developer’s Quick Reference

Quick reference to postgresql

A quick reference for day‑to‑day PostgreSQL work: connections, SQL syntax, psql meta-commands, performance, JSON, window functions, and more.

Converting Word Documents to Markdown: A Complete Guide

Converting Word Documents to Markdown: A Complete Guide

Using pandoc, python, or online tools for convertion to MD

Converting Word documents to Markdown format is a very common task for technical writers, developers, and content creators who want to move their content to platforms with Markdown (like GitHub, GitLab, static site generators like Hugo).

Comparing Go ORMs for PostgreSQL: GORM vs Ent vs Bun vs sqlc

Comparing Go ORMs for PostgreSQL: GORM vs Ent vs Bun vs sqlc

A practical, code-heavy look on ORMs in GO

Most prominent ORMs for GO are GORM, Ent, Bun and sqlc. Here is a little comparison of them with examples of CRUD operations in pure GO.

Jabber / XMPP userbase and popularity

Jabber / XMPP userbase and popularity

A bit of Jabber / XMPP statistics

Jabber (now commonly referred to as XMPP) is an open-source communication protocol and suite of client applications that enable instant messaging, voice/video calls, conferencing, desktop sharing, and presence features for individuals and organizations.

« 8/19 »

Recent Posts

  • Browser Automation in Go: Selenium, chromedp, Playwright, ZenRows
  • How to Configure Desktop Launchers on Ubuntu 24 with Standard Icons
  • Create AWS CloudFront on Pay-as-You-Go (not the Free Plan)
  • Browser Automation in Python: Playwright, Selenium & More
  • Terminal UI: BubbleTea (Go) vs Ratatui (Rust)

Categories

  • AI
  • Architecture
  • Cheatsheet
  • Coding
  • Community
  • Cookbook
  • Dev
  • DevOps
  • Documentation
  • Hardware
  • Howtos
  • LLM Hosting
  • LLM Performance
  • Observability
  • Offline
  • Privacy
  • RAG
  • Research
  • Security
  • Self-Hosting

Tags

AI AI Coding Anaconda Android API Architecture AWS AWS Amplify Backup Bash Benchmarks Cheatsheet Claude Cloud CloudFront Coding Community Conversion Cookbook Cpu Database DeepLearning Dev DevOps DGX Spark Digital Detox Docker Documentation Embeddings Filofax Fine-Tuning Flutter Food Git Gitea GitHub Go Golang Gpu Grafana GraphQL Hardware Hosting Howtos Hugo Images Inference Infrastructure JavaScript K8S Kubernetes LabelStudio Latex Linux LLM LLM Performance Machine Learning Mainroad Markdown MCP Melbourne Memory Microservices Minio MMDetection Monitoring Node.js NVidia ObjectDetection Observability Offline Ollama Open Source Pdf Performance Perplexica Photos PostgreSQL Printing Privacy Prometheus Python PyTorch RAG Reranking Rust S3 Security Self-Hosting SEO Serverless Social Media SQL Terraform Testing TypeScript Ubuntu Vector Database Vector Databases Vllm VSCode Web Hosting Windows

Social

Twitter
LinkedIn
root@@@glukhov.au
Mastodon
Mastodon
rost @ lemmy.world
rosgluk @ github
rosgluk @ bluesky
gluk @ reddit
rosgluk @ Medium
rosgluk @ blogspot
rosgluk @ tumblr

Languages

  • EN English
  • RU Русский
  • DE Deutsch
  • ES Español
  • FR Français
  • IT Italiano
  • JA 日本語
  • KO 한국어
  • PL Polski
  • PT Português
  • NL Nederlands
  • SV Svenska
© 2026 Rost Glukhov. Hugo and Mainroad.