Hi, Pulsar community, As you may notice, the Client Feature Matrix [0] has been linked on the Pulsar doc site [1] for quite a while, providing an overview of feature supportability on language-specific clients. As the outcome of PIP-108, it has addressed the initial community request [2] for technology evaluation and selection.
However, it has the following limitations to continually serving the purpose over time: 1. Visibility: Not prominent for users/maintainers to notice it. 2. Process: No required review/approval or version control. 3. Accuracy: A bit out-of-dated with limited chances to get it updated (possibly caused by 1&2). To make it more robust and prominent to better serve the adoption purpose, I think we can make the following improvements: 1. Deliver a more robust Client Feature Matrix and required documentation through a thorough review and update. 2. Move the matrix to the Pulsar repo and display it inline on a web page. Refer to this reference [3]. A quick question is about the granularity of the feature sets presented in the new matrix. Generally, we have two options: 1. Display a full version as we do in the existing feature matrix to provide more detailed tech capabilities. 2. Display a compact version with high-level features to provide better readability. Feel free to share your thoughts. [0] https://docs.google.com/spreadsheets/d/1YHYTkIXR8-Ql103u-IMI18TXLlGStK8uJjDsOOA0T20/edit?usp=sharing [1] https://pulsar.apache.org/docs/next/client-libraries#feature-matrix [2] https://github.com/apache/pulsar/issues/9723 [3] https://beam.apache.org/documentation/runners/capability-matrix/when-in-processing-time/ Cheers, momo-jun
