The page "Falsehoods Programmers Believe About Names" covers some of the ways names are hard to deal with in programming. This post will ignore most of those complexities, and deal with the problem of matching up loose user input to a database of names.

Handcrafted interactions focused on utility & beauty.

LLM’s suck at complex OCR, and probably will for a while. LLMs are excellent for many text-generation or summarization tasks, but they falter at the precise, detail-oriented job of OCR—especially when dealing with complicated layouts, unusual fonts, or tables. These models get lazy, often not following prompt instructions across hundreds of pages, failing to parse information, and “thinking” too much.

LLMs process images through high-dimensional embeddings, essentially creating abstract representations that prioritize semantic understanding over precise character recognition

Consider a simple table cell containing "1,234.56". The LLM might understand this represents a number in the thousands, but lose critical information about:

Exact decimal placement
Whether commas or periods are used as separators
Font characteristics indicating special meaning
Alignment within the cell (right-aligned for numbers, etc.)

https://news.ycombinator.com/item?id=42966958

Markdown extraction is just the first step. For documents to be effectively used in RAG pipelines, they must be split into smaller, semantically meaningful chunks.

Recent studies have shown that using large language models (LLMs) for this task can outperform other strategies in terms of retrieval accuracy. This intuitively makes sense - LLMs excel at understanding context and identifying natural boundaries in text, making them well-suited for generating semantically meaningful chunks.

The problem? Cost. Until now, LLM-based chunking has been prohibitively expensive. With Gemini Flash 2.0, however, the game changes again - it's pricing makes it feasible to use it to chunk documents at scale.

https://news.ycombinator.com/item?id=42952605

(disclaimer I am CEO of llamaindex, which includes LlamaParse)
Nice article! We're actively benchmarking Gemini 2.0 right now and if the results are as good as implied by this article, heck we'll adapt and improve upon it. Our goal (and in fact the reason our parser works so well) is to always use and stay on top of the latest SOTA models and tech :) - we blend LLM/VLM tech with best-in-class heuristic techniques.

Some quick notes: 1. I'm glad that LlamaParse is mentioned in the article, but it's not mentioned in the performance benchmarks. I'm pretty confident that our most accurate modes are at the top of the table benchmark - our stuff is pretty good.

2. There's a long tail of issues beyond just tables - this includes fonts, headers/footers, ability to recognize charts/images/form fields, and as other posters said, the ability to have fine-grained bounding boxes on the source elements. We've optimized our parser to tackle all of these modes, and we need proper benchmarks for that.

3. DIY'ing your own pipeline to run a VLM at scale to parse docs is surprisingly challenging. You need to orchestrate a robust system that can screenshot a bunch of pages at the right resolution (which can be quite slow), tune the prompts, and make sure you're obeying rate limits + can retry on failure.

Cabbage with eggs is better than pizza! Easy, quick and very delicious recipe! Delicious, easy, quick and very delicious breakfast, lunch, dinner or quick snack recipe. You won't believe how tasty it is – even more delicious than pizza, or okay definitely healthier 😅 It's so easy to make, and you'll enjoy every step of the cooking process. Try this easy and very delicious dish and let me know in a comments "How do you like it?". 🤩🥰😋

INGREDIENTS:
1/2 small cabbage (250g / 9 oz)
2 eggs
Season with salt
1 tbsp Olive oil
1 tbsp pizza sauce
mozzarella
basil

Cabbage with eggs tastes better than meat! Easy, quick and very delicious dinner recipe. Delicious and quick recipe for cabbage with onions! You won't believe how tasty it is – even more delicious than meat! I can't believe I didn't know about this recipe before, but now it's become one of my favorites. It's a simple recipe that only requires cabbage, onions, carrot, eggs and spices. If you have these ingredients at home, you can easily whip up this delectable cabbage pie in a pan. The taste of the cabbage in this pie is absolutely incredible, and your whole family will devour it. Whether you're looking for a quick lunch, dinner, or snack, this recipe is perfect. It's so easy to make, and you'll enjoy every step of the cooking process. So go ahead, grab some cabbage and onions, and let's cook this delightful dish together! 🤩🥳🥰😋

INGREDIENTS:
2 tbsp Olive oil
1 medium Onion
1 Carrot
2 lb / 900g white cabbage
5-6 Eggs
1 tbsp All purpose Flour (optional)
Manchego cheese (optional)
Fresh parsley (optional)

Season with Salt and black pepper
1/2 tsp Oregano
1/2 tsp Sweet Paprika
1/2 tsp Garlic powder
1/2 tsp Onion powder
1/2 tsp Dry Parsley

The best way to learn Eleventy is to build something with it.

This course was originally authored by Andy Bell and hosted at learneleventyfromscratch.com. This fork is now maintained by uncenter.

I am looking for help replacing the timer on my 9 year old dishwasher. The timer stops when the timer knob is pointing down (almost 6 o'clock) in the wash cycle. If the knob is advanced...

About Roel: https://pixelambacht.nl

Depending on whether the information is technical or not, you might want different font features.

29:20 - "Never grab a Monotype font"

font-variant-numeric: tabular-nums;

@font-face {
size-adjust: 110%;
}

https://wakamaifondue.com/
(what can my font do?)

https://pixelambacht.nl/

Java Regular Expressions

IronCalc is an open-source spreadsheet engine and ecosystem. Try it now and help us build the future of spreadsheets!

A simple, lightweight, and flexible note-taking template for Eleventy.

Welcome to Eleventy Notes, an opinionated template for publishing your personal notes, documentation, or even your blog! Built with Eleventy, the powerful static site generator, Eleventy Notes is packed with features you'll love.

When I made these blocks to attach the solid-wood top to a dining table (“English-Inspired Trestle Table,” FWW #310)) I was aiming for the blocks’ shape to be in sympathy with the design of the corbels that act to stiffen the connection of the foot and the brace to the trestle’s post. It’s an example of echoing the overall design even in the smallest of ways.

** Could be used for poker racks??

Se Refléter Dans (“Reflected”) is a lively and lovely survey of 30 years of work by artist Colette Bresilla pairing multimedia sculptures with paintings large and small.

Picking your general-purpose AI

Also:

https://www.oneusefulthing.org/p/doing-stuff-with-ai-opinionated-midyear

In this video Paul Sellers shows how to make a simple shooting board from pine. This is an important piece of equipment for woodworking in the hand tool shop as it allows you to make sure the end of a board is either exactly square or exactly forty-five degrees. It also allows you to plane the end of a board straight across without 'break-out' on the far side.

At 1:47 the middle measurement should be 12 1/2" or 320mm. Sorry for the confusion.

The main difference between MR11 and MR16 are as follows:

1. Size The MR-16 bulbs diameter is 2 inch (50mm)
   The MR-11 bulbs diameter is 1.375 inch (35mm)

2. Base the most common base for MR16 bulbs is GU5.3 or GX5.3 base - 2 pin base with 5.3mm between pins the most common base for MR11 bulbs is GU4 - 2 pin base with 4mm between base

Roth IRA Conversions

Taxable income tends to increase for many retirees at 73, when those born between 1950 and 1959 are required to take withdrawls from traditional retirement accounts.