jwrallie

It’s mostly mission dependent as you mentioned, if the value of some images are different than others, compressing several pictures and selecting the best to get in raw is a good strategy, and you can also crop them to get interesting areas as mentioned.

It can also depends on the downlink bandwidth available and power/energy budget, so the optimal strategy will differ.

It can be misleading at times though, for example some interesting feature may happen to be just a compression artifact. A cloud looking like some strange human-made object for example, happened to us one time.

Do you have any resource to recommend for using AI with Emacs? As in, aware and able to interact with the REPL.

Or maybe a more fundamental question, how do you find out what are good packages to use?

AI companies did promise that, and for some time back in 2024 and early 2025 it did feel so to me. I just had a kid and suddenly could not work the extra hour every day I used to, but kept producing as usual.

Nowadays I’m receiving longer emails than usual and longer code to review. Less spelling errors, but long and pointless prose. Less superficial bugs but more sneaky, fundamental problems.

There is also the trap where you go from 0-80% immediately but end up spending time anyway fixing the last 20%, and it’s unclear if there was a benefit or loss in productivity. Sometimes the almost there rabbit hole of prompting can be misleading.

I have a feeling that the 39% served as a good framework for the LLM to work on the remaining part while keeping it consistent with the existing structure and style. So, in a sense, your 8 weeks are a big part of why it succeeded.

You can sometimes also support your podcast creators on Patreon and get ad free RSS feeds.

Reminds me of early test versions of Gemini Flash 2.0 (or was it 1.5?). It would argue something, try to prove itself right, realize it’s proof is agreeing with the user halfway through and, instead of admitting defeat, would backtrack to try another approach. So much fun.

With a global economy where products are imported from elsewhere, local pollution statistics that do not include that indirect pollution as their own are meaningless. Energy imports are also being abused for report purposes.

Very deep and well written. I think the most important thing I want to pass to my kids is that there are reasons rules are put in place, and there are reasons why we should respect the rules, but there are exceptions. There are times when rules can be bent or broken, and we need to know how to judge, when to take risks, and when we should be transparent about own our decisions and actions.

People need to be open to experiment and make their own judgement based on observation without relying purely on their environment and authority.

TOEFL will grade people on how good they are at writing, but it does not approve or reject candidates. Maybe the threshold for acceptance was set too low?

I share the same feelings of other comments.

The only case where I do find Flatpak useful is to install proprietary electron based applications that do not interact with the rest of the system so much, but I try my best to avoid using those in the first place.

For people interested in a minimalist setup, Tailscale + Samba can go very far. I personally use Foobar2000 and VLC on the client side.

It’s not pretty nor great with bad connections, but there is basically no maintenance.

It would be funny if they find out the television was made by LG and they happen to have another copy of the audio.

My family immigrated away from Germany 4 generations ago and I’m a Firefox user. Maybe you have a point :)

The other day someone asked me why I use Firefox. I realized the only other browser I use is Safari.