Closed tokee closed 1 year ago
I will close this. To put it short, noone will benefit from it.
It will not improve playback by using PyWb instead of SolrWayback by exposing a primitive CDX-server that PyWB can call. As long prefix search and POST requests can not be handled with the currect version of WARC-Indexer+Solr, there will be no playback (or very little) playback improvement. Also old WARC-files harvested with Heritrix does save the post-data in the request.
Use Outback-CDX + PyWb to have playback work for the most complated SOME sites such as Facebook. This also require the harvest was done with WebRecorder technology.
We have all the information, why not let SolrWayback act as a CDX-server, så that there is not a need for a separate index? (suggested by Yves Maurer at a webinar)