@simonw I look at this and I see is that the higher the thinking effort the higher the un requested additions you get. Prompt never asked for sun and cloud. In code you ask for a feature and you get a complex system. That's exactly what I tried to cover in https://t.co/GSl90bU38s
Detailed plans can make coding agents look aligned while hiding guesses.
Architectural probes do the opposite: fake software that reveals structure before implementation, then evolves step by step.
https://t.co/GSl90bU38s
#AI#CodingAgents#vibecoding#programming#software
@PythonTorino purtroppo non potrò tenere talk in futuro finché l’evento sarà presso @tlbxcoworking vista la mancanza di un comportamento civile da parte dello staff verso le persone che sono in visita dai loro ospiti. Spero incontrare tutti al Keynote della @pyconit
I have been finally able to get to work on OSS projects those past weeks, and wow… I really missed that. Did in 2 weeks more than I could get done in the past year
Nanoarrow release 0.5.0 with major improvements to the Python implementation, nanoarrow keeps getting better as the lightweight solution to exchange Apache Arrow data https://t.co/yM23usBPt4… #python#arrow#data
Recently found out that Perspective ( https://t.co/8nXzWpVaKS ) internally uses a delta format based on Arrow when streaming data to the visualization part.
Arrow FlightSQL protocol provided a 10x speed up in Apache Doris for #datawarehouse, and it can be leveraged via the #python ADBC driver. See https://t.co/PPSGyAxze6…
The growth of PyArrow, surpassing 120M downloads/month, is a good indicator of how the composable data ecosystem is actually working in the real world already #python#arrow#data
Substrait-Python ( the interoperable query format ) 0.17 has support for query plans in JSON which makes more convenient to share query plans https://t.co/4Gimbtipxn #python#analytics
Version 16 of #python#arrow PyArrow https://t.co/UIbQGVzsu7 adds a very interesting feature for AI and ML: the support for converting Tables and RecordBatches to Tensors.
This makes very fast and convenient to use Dataframe like structures with ML libraries
I finally found the time to cut the release, so I'm happy to share that Beaker #session#caching#locking management framework for #python version 13.0.0 is now out with support for redis clusters as backend. https://t.co/eEOh2NqhQH
@ionos_ayuda_es I have been searching around the website for a while and I can't find a way to open a support ticket nor a way to upgrade a server... Those seem to be basic things in a cloud vps service
I was finally able to find the time to cut releases of my #Python projects Depot ( https://t.co/6Gfjp3P0CF ) and DukPy ( https://t.co/jdnmVA053P ).
So sorry it took me so long to make them available with Python 3.12 support
@Jonny_Waffles@IbisData It would be interesting to hear what problems you are finding with the pyarrow api. If there are opportunities to make them better it's always good to hear 👍
Feel free to report your experience at https://t.co/MO1HrsyV31
I was pleased to discover that the Web Browser Engineering book ( https://t.co/6xSfkSQ90q ) on how Web Browsers work actually uses my DukPy project ( https://t.co/BQXT3q0SW7 ) to implement Javascript support in the custom web browser built through the book.
Does anyone know if there is any pure python implementation of https://t.co/ieU3QSJ6QR ? I'd like to avoid having to maintain my on implementation and I need to be able to generate new co_linetable