We're starting sieving on GFN23. This does not mean it's going to show up in BOINC this year and probably not next year. But we might need those candidates for DYFL depending on how large the next Mersenne prime is (DYFL is meant to keep above that).
Today (9 Sept 2019), when the uploaded factors were processed, it does not appear that the Manual Sieving Statistics were updated.
I'm well aware of that. In adding GFN23 to the stats, I needed to run all the GFN23 factors from 0-16000P through the program that does the stats generation. It took about nine hours to process the first 1P of GFN23 sieving and then the program died because it had lost its database connection during that time. That had never happened before, because we haven't recently sieved any GFN that we weren't already testing in BOINC. When testing and sieving happen at the same time, my program is always recording factors for currently loaded work and so the database connection doesn't time out.
Having added new code to make sure the database connection is still active, I'm now running the GFN23 sieving though a second time and it's been running for almost eight hours. The program is counting factors and sieve removals. When it's done, it will produce a new sieve file at the p=16000P level and then I'll never have to process that early sieving again. I need good data and I need that next sieve file, so I'm waiting.