mg

Hmm... the way I use it is that when I put the mouse on an item, that is the one I am looking at. So it is fine that some others are hidden. And when I want to see all items again, I move the mouse into an empty area (usually right next to the item I just looked at) so the popup goes away.

Also, I usually use the filters first. Say for laptops, I set the screen size to >=12inch and the weight to <=3pounds. So there ain't that many items left on the screen.

Do you use it differently?

Affiliate commissions and license fees from companies who want to use the interface for their use-cases.

I am developing this project, which replaces product lists with what I call "product charts":

https://www.productchart.com

The idea is to sort products not by one parameter (like price or release date) but by two - which creates an x/y chart. The product info is displayed dynamically - by default only the image is show. On hover, more info is displayed in a tooltip. And when you click "details", all data is shown.

This way, 300 products easily fit on the screen.

You need to watch it on a monitor to see the chart interface. On mobile, I just display a normal list.

Which LLMs have you tried?

The way I "ask" is that I really ... ask!

I ask the LLM "Can you find anything in this file(s) that can be made shorter or more logical?"

And then as I said, I like less than 10% of the ideas the LLM comes up with. But it is so fast to read through 10 ideas (A minute or so) that it is well worth it.

Refactoring is indeed one of the areas where LLMs already shine bright. And I love it.

I have been coding since I am 12 years old. I always loved writing code. But I also always loved reading code. I don't know, for me code is a kind of art. Never met anyone else who sees it like this. When a friend was hiring for his startup a while ago, I was happy to sit down for multiple hours and read all the code the applicants wrote and gave him advice on whom to hire.

So for me, the new times are paradise. I try to not touch code directly anymore. I write prompts that would enable a really good developer to implement features and then let various LLMs work on it. Afterward, I rate the results so I have an overall score for each LLM. I pick the best solution for my codebase and manually finetune it to perfection.

After each commit, I also ask the LLMs if they can find anything in the files that can be refactored to make the code shorter or more logical. The result is that the codebase becomes better and better. Because the LLMs often find stuff to improve. They usually come up with 10 ideas I dislike, but also one idea that I like. And so the codebase becomes better and better over time. Instead of worse and worse like in the past when you had to keep a balance of refactoring for the sake of beautiful code and building new features. Nowadays, refactoring becomes more and more free.

Only more expensive in terms of the initial purchase price. Maintenance and energy costs are much lower than for a gasoline bus.

Electric buses are not just about the environment.

They are also so much more comfortable.

Less noisy, less rattly.

I would think it makes society healthier to have quieter, less annoying ways to commute.

The nature of technological progress is that it makes formerly uninhabitable areas inhabitable.

Costs of buses are mostly the driver. Which will go away. The rest is mostly building and maintaining them. Which will be done by robots. The rest is energy. The sun sends more energy to earth in an hour than humans use in a year.

Would you be ok if instead of 97% of earth being empty 94% is empty and your rent is cut in half? Another plus point of the future: An electric autonomous bus is at your disposal every 5 minutes, bringing you to whatever nice lonely place you wish.

One thing nobody seems to discuss is:

In the future, we will do a lot more.

In other terms: There will be a lot more work. So even if robots do 80% of it, if we do 10x more - the amount of work we need humans to do will double.

We will write more software, build more houses, build more cars, planes and everything down the supply chain to make these things.

When you look at planet earth, it is basically empty. While rent in big cities is high. But nobody needs to sleep in a big city. We just do so because getting in and out of it is cumbersome and building houses outside the city is expensive.

When robots build those houses and drive us into town in the morning (while we work in the car) that will change. I have done a few calculations, how much more mobility we could achieve with the existing road infrastructure if we use electric autonomous buses, and it is staggering.

Another way to look at it: Currently, most matter of planet earth has not been transformed to infrastructure used by humans. As work becomes cheaper, more and more of it will. There is almost infinitely much to do.

Hopefully, over time, we will get better and better in creating the world of software from small interoperable pieces that can be maintained by small teams or even solo entrepreneurs.

I think it is a balance. My current view on this is:

For the frontend, everything but plain HTML, CSS, JS is too much.

For the backend, an opinionated framework like Django is too much.

Even a less opinionated framework like Flask is too much.

Using modules that Flask uses - like Jinja templates - is the right choice for me. This way, I don't have the overhead of a full framework, but also don't have to write basics like a template engine myself. And I have the benefit that other developers (and AI) easily understand my code, as it has many similarities to a Flask project.

So instead of going for Flask when I need templates, all I do is

    import jinja2

    env = jinja2.Environment(
        loader=jinja2.FileSystemLoader('templates'),
        autoescape=jinja2.select_autoescape(['html', 'xml'])
    )

    def template_data(template, data):
        tmpl = env.from_string(template)
        return tmpl.render(data)

And I have a function I can use to template my data.

I would say the even bigger problem is that the pieces you stitch together to make a web app will rot rather quickly.

He mentions supabase and vercel. It will only take so long, until the supabase code or the vercel service will have a breaking change. And those breaking changes add up throughout the stack. Making the cost of running a web app much higher than it would have to be.

The site should have an RSS feed.

I sometimes visit Berlin and if I could add their news and events to my Berlin feed, that would be nice.

The persistence of Marginalia's creator is impressive.

I run a web search comparison site (Gnod Search) and every time I saw Marginalia, I played around with it for a while and thought "What is this for? Does anybody actually use this?" and decided against adding it to Gnod Search. Thinking this project would soon give up in the face of the giant competitors it has.

But it kept coming up over and over again :)

Then recently I was looking for blog posts about experiences with Linux on tablets. I had a hard time finding something with Google & Co. I remembered that this could be a good use-case for Marginalia. I tried it, and indeed I found some relevant posts by Linux enthusiasts.

I then immediately added it to the default search engines:

https://www.gnod.com/search/

I keep it in Yaml because that makes it easy to build upon it.

For example it lets you easily make your own project based on mine which renders the Yaml file in a way that is suitable for reading. Even if I add more attributes to each entry, your version will always stay user friendly by rendering only the attributes that you deem right for your interface.

Maybe. But this way, the project layout is better suited for future enhancements.

Over the last months, I have been compiling a list of state of the art in Mandelbrot implementations that run in the browser:

https://github.com/no-gravity/WorldWideMandelbrot

Every solution has its pros and cons. I still have not seen one that allows deep zooms (The parent only allows up to 48x zoom for example), is written in JS (parent is written in Rust and then compiled to JS), is fast, has a UI that nicely works on mobile and allows to save arbitrarely high resolutions. If there is none, I'm planning to write one myself.

Thinking about it, this could also be an interesting test on how far AI has come. When will I be able to hand the above text to an LLM and get out an HTML page with all the code that satisfies the above criteria?

On the phone I don't.

In Vim, I use multiple shortcuts that I made for this use case.

Sure. But even when you add noise to the described experiment, you get useful data.

That is the point I am making: Experiments of a single person can be useful.

The critics of single person experiments usually come up with examples vastly different than the Theanine experiment described here. With long term experiments which are only conducted once. But the Theanine experiment was looking for a short term effect and can be conducted many times. The hammer experiment I made up would be an extreme example of this type of experiment which leads itself well to be conducted by a single person.

What I am trying to point out is that if you are a skeptic, it would be better to try and find weaknesses in the experiment at hand. Not making up completely different experiments.

    an effect large enough to be
    apparent just from plotting

And how large is that? Without putting a number on it, how do we come to the conclusion that the effect is not large enough? That it didn't show in their sample of data points could have been just random chance.

But before we take the measured effects at face value, I think it's important think about them more. They report significant p values of their success in predicting if the capsule holds Theanine and also for the effect of the capsule when it holds Theanine. Both negative correlations. My first thought reading this is that the Placebo tasted more like Theanine and thinking they took Theanine had a positive effect on the outcome.

An experiment of 1 person can very well produce useful data.

It depends on the setup of the experiment.

Imagine an experiment where a person's thumb gets randomly hit with either a hammer or a feather once per day. And they then subjectively rate the experience. After 1000 days of collecting data, I doubt that we would wrongly come to the conclusion that the hammer treatment leads to the nicer outcome.

The setup of the Theanine experiment which is the basis of this thread looks good on first sight. I have the feeling that the interpretation could use more thought though.

I write into a text file on my phone. And later move those lines into the file on my computer.

It's a bit cumbersome, but not too much. But yeah, I look forward to when I finally get around to writing an app for it.

This is great. The author defines their own metrics, is doing their own A/B tests and publishes their interpretation plus the raw data. Imagine a world where all health blogging was like that.

Personally, I have not published any results yet, but I have been doing this type of experiments for 4 years now. And collected 48874 data points so far. I built a simple system to do it in Vim:

https://www.gibney.org/a_syntax_for_self-tracking

I also built a bunch of tooling to analyze the data.

I think that mankind could greatly benefit from more people doing randomized studies on their own. Especially if we find a way to collectively interpret the data.

So I really applaud the author for conducting this and especially for providing the raw data.

Reading through the article and the comments here on HN, I wish there was more focus on the interpretation of the experiment. Pretty much all comments here seem to be anecdotal.

Let's look at the author's interpretation. Personally, I find that part a bit short.

They calculated 4 p-values and write:

    Technically, I did find two significant results.

I wonder what "Technically" means here. Are there "significant results" that are "better" than just "technically significant results"?

Then they continue:

    Of course, I don’t think this
    means I’ve proven theanine is harmful.

So what does it mean? What was the goal of collecting the data? What would the interpretation have been if the data would show a significant positive effect of Theanine?

It's great that they offer the raw data. I look forward to taking a look at it later today.

Isn't the price of a KW of solar panels similar to the price of a KW of nuclear power these days?

I wonder what hinders us to replace the roofs of all houses with solar panels and put batteries in all cellars?

It might still be useful to build out nuclear power plants. But the solar+battery approach seems like an easier first step to increase the available power, doesn't it?

Is this solution different from just executing this:

    docker run -it --rm -e DISPLAY --net=host -v $XAUTHORITY:/root/.Xauthority -v /tmp/.X11-unix:/tmp/.X11-unix debian:12-slim

And then run the GUI app in the container? For example for Firefox:

    apt update
    apt install firefox-esr
    firefox

That is an approach I sometimes use to try GUI applications in a Docker container.

Why would there be an uppper limit for the number of buses?

I would think that as long as there are cars on the road, each new bus is a net positive, as it replaces 100 cars.

So you expect the number of buses in the metropolises of the world to stagnate?

Segregation is not neccessary to have more seats.

If we double the number of buses, we can increase the number of seats per bus and still have a higher overall passenger capacity in the city.

In my city, I see buses that have only seated places all the time:

https://www.moia.io

It is an interesting mix of a private and public service. Mostly it is used by people who prefer a better ride experience than on a public bus. And are fine with paying a much higher price per ride. They pay out of their own pocket to the service provider, Volkswagen. But for people with disabilities, using the Moia buses is free and paid by the city.