Anthony Shaw
@anthonypjshaw.bsky.social
📤 1822
📥 338
📝 413
Python obsessive at Microsoft. PSF Fellow. Creator of VS Code Pets
I no longer begrudge my US colleagues for getting paid 50% more than me. I just went grocery shopping and everything costs… 50% more 💁🏻♂️ (Except beer and chocolate which is cheaper)
1 day ago
3
5
0
Where else
www.bbc.co.uk/news/article...
loading . . .
New 'Lucifer' bee with devil-like horns found in Australia
The native bee was found near a rare wildflower, prompting scientists to call for the area to be protected.
https://www.bbc.co.uk/news/articles/c051yjv90dmo
1 day ago
1
1
0
Day 1 jet lag challenge. Getting up at 6am and staying awake all day to hopefully reset the body clock for day 2 🥱 would normally go for a run but I forgot to pack shoes
2 days ago
1
0
0
All this talk of travel chaos, SFO is empty 🤷🏻♂️ walked through CBP in 5 minutes
3 days ago
0
0
0
Last minute travel plans to Seattle. I’ll be there all week and SF next week
3 days ago
1
1
0
I used to listen to this episodes whilst mowing the lawn. The episodes keep getting longer and longer. Now we’re at 4 hours I have to find bigger yard jobs to cover a whole episode
add a skeleton here at some point
5 days ago
0
7
0
@mkennedy.codes
@brianokken.bsky.social
@scott.hanselman.com
does this guy roam the streets of Portland?
add a skeleton here at some point
5 days ago
1
9
0
How did I miss putting print statements in tests 🤦🏻♂️
add a skeleton here at some point
10 days ago
0
3
0
Claude AI's Pytest anti-patterns: 1. Using test classes in pytest tests for no reason 2. Putting module fixtures in conftest instead of in the modules 3. Creating long doc strings for test functions 4. Unnecessarily fragile assertions
11 days ago
4
7
1
0x67
14 days ago
1
3
0
TFW you’ve been running a benchmark for an hour, all your CPU cores are at 90+% but the overall benchmark progress is 0% Kill it or wait?
17 days ago
1
0
0
Ok, penalise Lewis Hamilton but nothing for the Dutch Lawnmower?
#mexico
17 days ago
1
3
1
Going to report a bug literally any Python project on GitHub to find
@hugovk.dev
already got there before you, reported it and fixed it. Every time.
20 days ago
1
30
4
Explaining to someone outside of this industry how we meet deadlines by refining the “definition of done” sounds pretty ridiculous when you say it out loud
20 days ago
0
4
0
One thing I love about Open Source is that you can’t hide behind your BS. Don’t tell me it has that feature. I can read the code. It doesn’t. Don’t tell me “we support that” when I can see it clearly doesn’t.
21 days ago
1
8
2
I’ll be live in an hour with Stephen McCullough from NVidia to talking about Open models on serverless GPUs. We’ll be diving into the tech, demos and how to setup autoscaling.
www.youtube.com/live/GvZJHHC...
loading . . .
Run open models on Serverless GPUs (APAC)
YouTube video by Microsoft Reactor
https://www.youtube.com/live/GvZJHHCk244?si=cLaFxljgRPQ_E96Z
22 days ago
0
3
2
Partner: can you clean the shower & drain whilst I’m out TIL it is possible to clean the shower with a pressure washer but it makes more mess than it cleans.
26 days ago
0
5
0
29 days ago
1
4
0
Ant vs Ant 🐜 . I think I lost this round 🥊
about 1 month ago
1
1
0
If you demo an AI agent, it must be a weather demo. If you demo a multiagent, it must be a travel booking demo. This is written in law. Do not question it.
about 1 month ago
1
6
0
Hacktoberfest 2025 : Getting Claude to implement features like normal but it's someone else's credits and they get a free t-shirt.
about 1 month ago
0
3
0
🎃It's October! New release of vscode-pets 1.34 - New Autumn (Fall) theme with falling leaves - New squirrel and skeleton pets (thank you Chris) I've added several up-for-grabs issues for
#hacktoberfest
to the backlog.
about 1 month ago
1
6
0
On a
#PyConJP
adventure with
@tiangolo.com
to Osaka and Himeji
about 2 months ago
1
25
3
TFW you catch your little toe on coffee table but can’t say the f word because there are kids around
about 2 months ago
0
4
0
Choo-Choo. Obvs at 300km/h you don’t need seatbelts 🤷🏻♂️
about 2 months ago
1
6
0
Solar export $ is practically $0 here so I’m looking at doubling the system generation size then adding a 24kWh battery. Have 9 months of data in Home Assistant to do some modelling. Shame there isn’t a nice API to fetch it into Pandas. 🐼
about 2 months ago
1
1
0
Finally got some VSCode pets stickers on my laptop
about 2 months ago
2
6
0
Wow,
#hacktoberfest
spam is starting early this year.
about 2 months ago
1
3
0
"The majority of staff who voluntarily go into the office say they are happy to travel to the office; therefore, we extrapolate that the people who work from home would be happier in the office" 🫠
about 2 months ago
0
17
0
reposted by
Anthony Shaw
Tim Kellogg
2 months ago
Hallucinations are accidentally created by evals They come from post-training. Reasoning models hallucinate more because we do more rigorous post-training on them The problem is we reward them for being confident
cdn.openai.com/pdf/d04913be...
8
65
17
reposted by
Anthony Shaw
James Bennett
2 months ago
I do not want him on the Board. I do not think he struck a chord. I would not trust him with a groat. I did not trust him with my vote. I don't want Python in his hands. I did not vote for Angry Franz.
1
5
3
hard is concurrency of second rule Python is The
add a skeleton here at some point
2 months ago
0
13
1
The first rule of Python is there are no rules
2 months ago
2
7
1
reposted by
Anthony Shaw
✍️New Post ✍️ Python's async support is 10 years old this year. Python 3.14 has "proper" parallelism now -- free-threading and subinterpreter pool workers, so does that make async redundant? I look into the use cases and challenges of asyncio.
tonybaloney.github.io/posts/why-is...
loading . . .
Python has had async for 10 years -- why isn't it more popular?
A deep-dive into the challenges and misconceptions surrounding async programming in Python
https://tonybaloney.github.io/posts/why-isnt-python-async-more-popular.html
2 months ago
2
24
5
✍️New Post ✍️ Python's async support is 10 years old this year. Python 3.14 has "proper" parallelism now -- free-threading and subinterpreter pool workers, so does that make async redundant? I look into the use cases and challenges of asyncio.
tonybaloney.github.io/posts/why-is...
loading . . .
Python has had async for 10 years -- why isn't it more popular?
A deep-dive into the challenges and misconceptions surrounding async programming in Python
https://tonybaloney.github.io/posts/why-isnt-python-async-more-popular.html
2 months ago
2
24
5
My UMXFP1 paper keeps getting rejected-- any suggestions?
2 months ago
1
1
0
My GitHub Inbox always has one unhinged item 🐿️ 🐿️
2 months ago
2
15
1
Please don’t put him in the second RedBull seat. Let us enjoy seeing him succeed a little while longer
add a skeleton here at some point
2 months ago
1
6
0
Writing one of those blog posts where you start with a bunch of opinions and as you research them you discover you were actually wrong. I've learnt a lot today and this one is gonna take a bit longer.
3 months ago
0
7
0
reposted by
Anthony Shaw
Brett Cannon
3 months ago
We have an opening for a principal software engineer on the Python tools team. If you're interested in debuggers, language server protocol, and AI then you might find the position interesting!
https://jobs.careers.microsoft.com/global/en/job/1854048/Principal-Software-Engineer
0
4
9
I've been experimenting with timing latency and throughput on different LLM and Embedding models. It ended up as a plugin for
@simonwillison.net
's LLM CLI so that you can compare any of the supported models with any combination of options. Here's a writeup--
tonybaloney.github.io/posts/perfor...
loading . . .
Performance Benchmarking LLM Models
Using the LLM CLI to compare latency and throughput of multiple LLMs
https://tonybaloney.github.io/posts/performance-benchmarking-llm-models.html
3 months ago
0
3
2
DocumentDB, the open-source NoSQL database built on top of Postgres is now part of the Linux Foundation.
www.linuxfoundation.org/press/linux-...
loading . . .
Linux Foundation Welcomes DocumentDB to Advance Open, Developer-First NoSQL Innovation
Vortex Project Joins LF AI & Data Foundation
https://www.linuxfoundation.org/press/linux-foundation-welcomes-documentdb-to-advance-open-developer-first-nosql-innovation
3 months ago
0
2
0
reposted by
Anthony Shaw
Gwyneth Peña-Siguenza
3 months ago
@marlene.bsky.social
and I started a new show: Python on Azure! We had to go BIG for our first episode: our guest is the one and only
@mitsuhiko.at
- Could he vibe-code Flask? - Is the terminal the ideal UX for AI coding tools? ...and more!
youtu.be/bpWPEhO7RqE?...
loading . . .
Agentic Coding: Armin Ronacher
YouTube video by Microsoft Azure Developers
https://youtu.be/bpWPEhO7RqE?feature=shared
0
18
3
Finally stopped raining so we joined the Great North Walk to hike this funny looking “mountain”
3 months ago
0
3
0
Released a small update to generate boxplots for the 4 main metrics (time, time to first chunk, length, chunks/sec). Nice simple way to quickly compare a handful of models when latency matters
add a skeleton here at some point
3 months ago
0
1
1
今年もPyCon JPに参加するよ。広島で開催されるイベントのチケット、さっき予約したばかり✨ 楽しみにしてます!→
pyconjp.connpass.com/event/359523/
それと、@tiangolo.com さんのキーノートもめっちゃ気になる
loading . . .
PyCon JP 2025 (2025/09/26 09:00〜)
# PyCon JP 2025 PyCon JPは、日本最大のPythonコミュニティイベントで、Pythonコミュニティメンバーが主体となって運営する非営利の国際的なカンファレンスです。 一般社団法人PyCon JP Associationが主催しています。昨年の東京開催では650人が参加し、アジアや欧米など、世界各国から、Pythonのコア開発者を含む、エキスパートや愛好者が集まりました。...
https://pyconjp.connpass.com/event/359523/
3 months ago
0
1
0
reposted by
Anthony Shaw
Hugo van Kemenade
3 months ago
Just released: UltraJSON 5.11.0! 🚀 ⌨️ Inline type stubs 🐍 Support for Python 3.14 & PyPy3.11 🛞 Windows ARM64 wheels (thanks
@anthonypjshaw.bsky.social
!) 💧Drop EOL Python 3.8 & PyPy3.8-PyPy3.10 ➕ And more!
github.com/ultrajson/ul...
#Python
#release
#ujson
#UltraJSON
loading . . .
Release 5.11.0 · ultrajson/ultrajson
Added Inline type stubs (#674) @MarcoGorelli Add support for Python 3.14 (#680) @hugovk Add support for PyPy3.11 (#658) @hugovk Add Windows ARM64 wheels (#663) @tonybaloney Changed Migrate to sr...
https://github.com/ultrajson/ultrajson/releases/tag/5.11.0
1
2
2
Today I worked on a little profile plugin for
@simonwillison.net
's LLM that runs the same prompt against as many models as you please. Including from different providers. You get timing statistics for streams. Useful for throughput and expected length benchmarks
github.com/tonybaloney/...
3 months ago
2
47
4
Spent the morning building and testing a rope swing in the garden. It was the perfect spot for one.
3 months ago
2
5
0
TIL you can pipe raw markdown to `python -m rich.markdown -` and it'll render it beautifully on the command line
3 months ago
0
35
7
Load more
feeds!
log in