I'm not sure I understand : how can you product help for opening times or pictures of my local boulangerie ? What kind of data sources will help you automate the reviewing of its attributes ?
We are not providing opening times yet - we just check if place is permanently closed or not. But it is in the works under our experimental enrichment API (which is not yet open to public)
I started scraping restaurant websites in Zürich and extracted and hand-checked opening hours in the OpenStreetMap format. The goal is to build a corpus for evaluation purposes which maps website texts to correct opening hours strings for all restaurants in Switzerland. Maybe you can use that to benchmark your own hours extracting system... https://github.com/wipfli/opening-hours/
I don't know how well does Haiku handle OSM opening hours syntax, but with Kimi K2.5, I got better results when I asked it to provide opening hour ranges for every day of the week, and then constructed the opening hours manually.
Appreciate sharing this project - democratizing this data is indeed a very important step. Interesting that you settled on Haiku - did you have a chance to check flash-2.5-lite or gpt-5-nano performance?
This is obvious. Most people won't take the bus if it's too slow. Well in the US you can't bike without getting killed except in some rare cities, so cycling is not a competition, but it would replace the bus if it were allowed.
Let me present my take on why the federated alternatives struggle to replace X:
Twitter didn't succeed because it was a particularly good solution - it really isn't. It succeeded purely on the back of the network effect.
When every open-source alternative simply copies the existing restrictions without adding any unique value, why would users switch to an equally flawed version where none of the accounts they actually want to follow are?
My take is that Bluesky got the network effects right, with hyping and gradual release and careful curation of who seeded the network and that is - IMO - the important thing they got right.
I guess I don't get why all of these don't target better user experience goals - one login / trust plane and full distribution for decentralization.
The way I'd look at it is more like Bitcoin where everyone can decide how much compute to give to the pool to verify people or posts and maybe everyone shares chunks of the whole pie with copies like freenet.
Maybe I'm (likely) too dumb to get why this isn't what things in the fediverse is but they all are an awful experience and having so many hosts makes critical mass non existent so I can't be bothered to participate.
reply