It makes perfect sense that the indexer is a lower priority part. A patch in the upstream isn’t a priority for me either as we needed a fork anyway. But when you do start working on it, whenever in the future, could you please check out this boost::mapped_file patch? I’ve proofread it countless times so it should be ready for you whenever you’d like. github com/CGDogan/orthanc-indexer/commit/eb3b5fa8a40a6b592e9573aaa69c12f263bdc193 (to get the .patch file, add “.patch” to the end of the URL) I’ve seen an enormous boost in efficiency of the plugin with this patch, it reduces the number of redundant copies by the plugin by 50%. My scope of work is changing and I won’t be submitting more patches (my local fork isn’t too relevant) so keep up the good work and regards.
If you like the patch, Orthanc users will certainly see a visible benefit from a similar patch you could prepare for FilesystemStorage.cpp . Noteworthy drop in memory & electricity usage in orthanc. In fact, I would like to even propose a RegisterStorageArea3 since this is what would make Orthanc really efficient whether or not someone already uses the PostgreSQL plugin but the order of priorities is up to you as always.