MindfulMaverick

joined 4 months ago
 

I'm trying to find a place where you can ask broader development questions, not just specific error messages.

StackOverflow and Codidact are way too restrictive, if your question isn't a precise technical issue with a reproducible example, it gets shut down immediately. Reddit and Lemmy seem more focused on news and memes; actual questions and discussions tend to just sink without engagement. And honestly, the kind of specific error-driven questions StackOverflow excels at are things AI can solve instantly now.

What I'm really looking for is a community (forum, Discord, whatever) where you can get help on broader topics related to software engineering.

Does anything like this still exist? Somewhere with actual humans willing to discuss the process of building software, not just fix syntax?

[–] MindfulMaverick@piefed.zip 2 points 2 days ago

I'm trying to download all fics from that specific forum. Sorry I wasn't clear.

[–] MindfulMaverick@piefed.zip 1 points 2 days ago

If I used sqlite or any other SQL database I don't think users could collaborate on building the database, so I was thinking of json files committed to a git repository online.

 

I'm currently using a pagination, link extraction, and Python filtering process before feeding links to fichub-cli to download all stories from a specific forum. The workflow is detailed in this post: https://piefed.zip/post/1151173 . Looking for a more streamlined, possibly one-command solution that could crawl the forum, extract thread links, and download them automatically. Any suggestions?

 

Tried Connect since I used it for Lemmy, but no luck, keeps giving me "The page your browser tried to load could not be found" error when I try to log in.

Has anyone had success with other apps?

 

I'm looking for advice on building a collaborative caching system for APIs with strict rate limits that automatically commits updates to Git, allowing multiple users to share the scraping load and reduce server strain. The idea is to maintain a local dataset where each piece of data has a timestamp, and when anyone runs the script, it only fetches records older than a configurable threshold from the API, while serving everything else from the local cache. After fetching new data, the script would automatically commit changes to a shared Git repository, so subsequent users benefit from the updated cache without hitting the server. This way, the same task that would take days for one person could be completed in seconds by the next. Has anyone built something like this or know of existing tools/frameworks that support automated Git commits for collaborative data collection with timestamp-based incremental updates?

 

A massive NIH study of nearly 400,000 adults over 20 years found that daily multivitamin users had a 4% higher mortality risk compared to non-users. The research showed no mortality benefit whatsoever—contradicting the belief that multivitamins serve as health "insurance". Interestingly, multivitamin users typically had healthier lifestyles overall, yet still showed increased risk.

For healthy adults without diagnosed deficiencies, the healthiest nutrients come from food sources, not processed pills. Some specific concerns include potential buildup of excess iron or niacin from daily use. This reinforces that supplementation should be targeted and evidence-based, not indiscriminate.

[–] MindfulMaverick@piefed.zip 2 points 2 months ago

Imagine Stack Overflow, but you have some probability of not seeing the top answer. That's pretty much what happens with federated instances blocking each other. I wouldn't use that.

[–] MindfulMaverick@piefed.zip 3 points 4 months ago

And I'm saying you can encounter the same issue of not finding a post whether you save it or not. If you suggest saving content to reference later, there also needs to be a way to search within those saved posts. Otherwise, it’s essentially the same problem as not being able to find it without having saved it.

[–] MindfulMaverick@piefed.zip 1 points 4 months ago* (last edited 4 months ago) (2 children)

You would need to be able to search saved posts which would also be useful, but not nearly as much.

 

I think that's some important missing functionality.