thomas536

Looking for short toy impls

For $10-20 per month (the rough cost of middle of the road web hosting), what would you pay to be searchable? Your favorite niche vertical like your twitter network, full fidelity search of the glory days of Usenet posts, your bookshelf?

Switching costs to new systems are extremely high. Apache Airflow has a large ecosystem/Medium posts, cottage industries to help make Airflow more palatable, batteries included for integrations.

But it has many warts and gotchas. Are there any bad enough that would be worth your company's time and money to switch to something else?

Have you ever found yourself wanting to jump start your professional knowledge on a niche topic and wanted to pay X hundred USD to pick a professional's brain for an hour? Has anyone found a community that facilitates this?

An example might be wanting to understand the deep nuanced differences between practical applications of two distributed SQL engines (which wouldn't be answerable on Quora, Stack Overflow, etc.).

Common references might be Wikipedia and Stack Overflow, but I'd imagine everyone has their own personal niche ones for your field that aren't widely known.

There are a handful of examples of open source, real-time, turn-by-turn navigation software out there. For example, https://github.com/valhalla/valhalla gives great turn-by-turn directions. I'm looking for best practices around navigation that will tell you to turn right in 300 feet as you're driving.

I'm curious about the high level, non-obvious gotchas in making them really useful to match current best offerings. Things like smoothing of gps (specific to the use case), better estimation techniques for current position/speed between gps updates, best practices around time estimation to next turn given current (estimated) position/speed, how best to determine when to reroute (how far is too far off route, how long to wait until rerouting), cheats people use to paper over bad data, etc.

Thanks!