Hi Karl,

I have been exploring the API, RAM Concept API and have found that the current API only allows one request at a time making it impossible to construct a tendon data extractor since thousands of API requests creates a performance bottleneck. I am not sure, but it seems like RAM Concept is a legacy application wrapped by a modern programming language or something like that. The way forward for you guys is what CSI did, bulk extraction through database tables. ETABS had the same issue of being single threaded, but they solved it by exposing the database tables making it damn fast. You should store the analysis or design or any report to the tables, as you do now, but the problem is it is stored in memory during runtime and it can't be found once the model is closed. Then expose it through API to bulk extract, maybe something like batch queries or direct table access. Once you do that, you would decrease the hassle since we know the tables, we can directly extract them, we are familiar with the structure, and we can develop programs we can sell or use in house since performance wouldn't be a bottleneck anymore.

The issue here is basically the synchronous API design limiting throughput - exposing the underlying data store (SQL, SQLite, or whatever you're using) would solve this.