How I made Huge Mistake with Interpretation of Laravel Downloads
The PHP FW Trends project was introduced 2 years ago to compare download statistics over "what they say on the internet". The methodology is based on Packagist data, which worked but had some flaws.
Also, monorepo and split packages downloads lead to a problem when one monorepo download has hidden 10-30 package downloads. There is no way to detect what packages are downloaded or used exactly, so this project should be stopped to avoid showing miss-leading data far from reality.