chazeon

It seems US and China are trying to reach an agreement to use AI to pick military targets these days.

I have thought about using tessaract, using it to OCR the TOC and generate something like this. But there are just so many edge cases that make the whole process fail. For example, how do you handle it if the title breaks into two lines? What if the page number is not recognized correctly? For example, 10 can be 1o What if there are dots? Maybe you can use GPT to clean the extracted text.

In the end, I found ChatGPT-4's multimodal capability can recognize text + page number pairs well if I feed screenshots of TOC into it, and I have settled on that.

I have been thinking about this, but for a while now, I have settled on using ChatGPT's GPT-4v's multimodal capability to generate a text file containing the titles and pages based on screenshots of the TOC. After that, I used a pikepdf-based Python script to bake the TOC into the PDF I had.

The upside, compared to Krasjet's approach, is that this works not only for text-based PDFs but also for scanned PDFs, even old scanned journal papers.

The downside is that, before baking the TOCs, you need to make adjustments to the PDF as sometimes the empty pages are not included. You also need to calculate the offset for the prologs, cover, etc. I have a script for this kind of adjustment, but there always is manual intervention involved.

TBH Hertz having EV is a good thing for me as a customer because they are cheap. It’s difficult to rent a gas car for $50/day near NYC but they are renting out EV at this rate. Although recharging is a hustle, I tried several times and never get a Tesla, and there’s no enough fast charger for other brands.

Several multi-protocol proxy clients support this functionality, some notable open-source examples include:

  - [sing-box](https://github.com/SagerNet/sing-box)
  - [clash-meta](https://github.com/muink/Clash.Meta) and other clash-based clients
  - [xray](https://github.com/xtls/xray-core)

Close-sourced client include [Surge Mac/iOS](https://nssurge.com/).

I remember when I designed a list of checkboxes this case was specifically considered. The last one needs to be updated when all of the checkboxes above are checked or one of them is unchecked.

Can I use a regular m.2 PC Wi-Fi card on these ARM SBCs? Suppose the SBC's CPU is mainline supported / armbian supported, and the PC Wi-Fi card works on x86 machines under Linux.

But how do you sync on Obsidian? Obsidian Sync?

With iCloud Apple indeed handles well update conflicts in Apple Notes. I have tried to set up Obsidian or any other Markdown-based notetaking system, the sync is so often and I had to give up. Apple Notes does handle this pretty well. So I finally moved to Apple Notes.

Of course, a lot of this is about coding on paper. I would wonder if it is possible to run the projection on screen instead. Say I have three iPads; their placement can likely be some input. The projection will not get a resolution comparable to a screen soon.

One issue with iPad is the pencil does not automatically reconnect to another one after touching anther screen.

The CPU yes, but not for peripherals. The Wi-Fi on my Orange Pi 4B dropped connections after 5 mins and cannot connect afterwards using Armbian image.

You won’t need to really run server all the time, you can run locally and use it as a lightweight CMS, and run only when you need to edit and generating anything. You can choose to only upload the generated HTML files to GitHub.

See my explanation on my reply to the original post.

Yes.

Some static site generators have good support for generating pages from a dynamic source, say API, database or anything you can access using programming language.

See eleventy, for instance, has the “Javascript data files” [1], it run some JS code to generate a list of posts, here we can fetch from pocketbase, then, we can generate pages dynamically [2].

Pocketbase in this case just act as a lightweight CMS.

Is this what you’re talking about?

[1]: https://www.11ty.dev/docs/data-js/

[2]: https://www.11ty.dev/docs/pages-from-data/

I feel like this will be super interesting when pairing with static site generator such as eleventy. Using static site generator is sometimes difficult because of lacking a web backend and ghost is too heavy for me.

Is ESP32 enough for me if I want to SSH onto some machine?

Wayland isn’t going to save linux desktop, but stream deck would.

Feels like very dangerous, if I ask it to delete my home folder I guess it will delete it right away

Lens and matrix size is everything when you cannot control the environment, when you can change the lighting and design the scene, then you can make up for the smaller sensor.

While building physical simulation workflows, the input for calculation steps was intentionally not autogenerated from some config file but retained a template that require users to check. This way, it is easier for scientists to have a good idea so that they know they are doing the "right" calculations.

Pandoc works pretty well these days. I writes in LaTeX and my Professor works in Word. The conversion works just fine.

Battery life, RAM size, and size and weight. I used to use an XR, it is okay in terms of battery life, but it is very unpleasant to hold in hand. Then I switched to the 12 mini, it is nice to hold, but the battery life it much worse than several later generations, it is okay for COVID life, but not much for a regular life. The battery life won't be an issue if you are using a Pro Max phone, and Pro series will feel better than using a non-Pro one (iPhone Xs vs Xr), but I believe if you are using a later generation, you won't want to go back to those two.

iCloud has become a real nightmare when I try to build something in Swift Playground and open it also on Xcode on a Mac. On iPad I was constantly reminded of conflicts, saying that the file is edited elsewhere, while in fact it was just I leave the Xcode window open. Sometimes it just revert to an older version. It is really annoying.

I expect the old behavior is to see other people's stars, but when I filter only the stars, there is no entries there.

Now the best hope you get is Ladybird

Back in the day, VLC or LaTeX used to use fc-cache to rebuild its font catalog when you installed some new font, and it could take quite some time if you have a lot of fonts installed. I wonder if Quicken does something similar.

Back in the day, VLC or LaTeX used to use fc-cache to rebuild its font catalog when you installed some new font, and it could take quite some time if you have a lot of fonts installed. I wonder if Quicken uses something similar.

It's really painful to look at all those screws. A few days ago I opened up a ThinkCentre m900. There is only one screw and you can get access to RAM, and SSD.

I also had a problem with Sandisk 2TB portable SSD. But it dies for a few days and comes back. It's wired. I will not keep anything important in that drive after this incident.

I do computational physics and I use Snakemake. On HPCs, we only have user-level access. We are not allowed to perform long-running processes on login nodes, could be killed at any time due to a violation of rules. That said, anything that depends on Docker is a no-go; anything that uses a server-client structure we will try to avoid (although it might be possible for us to host a daemon elsewhere that we as students pay out of our own pocket). We also deal with a lot of tools that are not well-written in Python or modern languages, you wouldn't want to build any CFFI onto it.

So Snakemake, and similarly, Nextflow, suits our needs well. It is a user-space CLI tool that does not require any privileges, it optimizes for running bash command / any CLI-based tools. A bonus for Snakemake is that it uses Python and our other scripts use Python too.

So I guess DevOps tooling, which heavily bias towards docker or whatever container-based execution, is really a different space.

A fleet of three ThinkCentre PCs, m920x2 and m720x1, they use 8th Gen Intel CPUs. These three run Windows 11, Arch Linux, and macOS. Homelab is running mostly on the Arch Linux one with Docker supplemented by the Windows 11 one running some longer time job because Windows only upgrades every few months.

I also have a FriendlyELEC, an arm box that is running OpenWRT as a router and also serves Ubiquiti's controller.

There is two part of software/hardware I should be retiring/moving; both are Homebridge, one on Raspberry Pi Zero W and one on another Windows PC Hyper-V Ubuntu, but I just don't have time to move them yet. The Raspberry Pi Zero is especially annoying because it is so underpowered that it feels painful to wait for anything to upgrade.