I noticed this, when it was only read a few files from my project, and I had to ask it to read ALL the files.
I then had it make a mistakes file and write every mistake, so it would learn, it kinda worked but it would still make the mistakes. It clearly wasn't reading all of it.
So I made a checklist, and it had verify every item on the checklist, that was my work around to both lazy and short mindedness of the agents. Turn mistakes into items to check for. Traded processing time for better results, ok for me on smaller projects. My run times went from 5-10 minutes from 3 per task, need to start logging tasks effectiveness/efficiency to reduce processing time.
I keep seeing people saying loop engineering is the way to get around these issues, I guess I'm kinda doing that in an adhoc way. Since I'm already looking at adding cost and goals(kinda).
I think things are moving fast, tested that new vibethink-3B, works on many small tasks/fast, and playing with ornith-35B with a draft vibethinker-3b as a draft gave me some good speed/results.
Was just trying to see how small I could go and get acceptable results, but yeah, larger Qwen 3.6 with MTP is going to be better. Cant wait to see how AI model (unsloth/local-llm/heretic/reaper/etc communities) are tweaking/engineering quality down into smaller models. Lots of new things coming out.
Somewhat true, the finances might fall apart, might, for some of the biggest companies, but AI is exploding in services industries of all sizes. Also, ai hardware is lasting way longer than 1 year... So a bubble? maybe, too many assumptions yet.
Having issues with coding a render for good looking realistic smoke coming off burning incense, opus 4.8 & gpt-5.5 both have code issues, glm-5.2 did it. Amazing.
The real time 3d fluid dynamics appear to be the tricky part, I wish I still had opus access, would love to see if it can do it.
I really liked when XP came out, and you could have both the clean 2000 look with advanced font rendering like cleartype. That was the perfect combo for desktop use.
I had my xp running a blackbox 4win and coLinux/cygwin,
then moved onto vmware/virtualbox with windows vm and linux desktop,
now I'm doing win11+wsl2 and loving it for cuda/ai/building projects.
I wonder. I spin up tons of projects, but I do it differently.
I'm also on windows. But I'm using a WSL linux vm to do the work. Then I have claude ./sync-to-windows, so the src is copied to a windows share via /mnt/drive/work and I just run a windows built.bat.
I have claude take care of the sync/build.bat so it can do things like clear cache, add options, look at logs, etc. This is also how I build multi platform, build on linux, src works on linux/osx/win
I'm a big fan of streaming radio, and been using apps for decades, so decided to make a modern version. Also, wanted to try AI coding out, and it worked out really well. Been having a blast.
I made my first vibe coded game, made casino style slots with lines, bets, free plays, auto play, in a linux shell, with color graphics, amazed how well it came out. Amazed how well it came out, with fixing graphics and adding features.
Might be nice on older games that dont have 4k HD texture packs, or even games like Skyrim or replaying older Witcher 1&2. Its upto the user right? Let people play older games with some better looking gfx. Seems like an easy win, use it or dont use it.
Newsgroups took a dive when people moved to web forums, scattered all around the web.
Now people are moving back to a centralized website, reddit, that's interestingly almost like reading newsgroup threads.
Always thought the biggest problem with forums, not feeding RSS feeds back into newsgroups to keep it alive. But then, you couldn't reply, and was only for archival purposes.
At least with Reddit, I can use ifttt to push content into reddit, and make the discussion happen on reddit, in dedicated subs.