I have about 200,000 MCC entries and I understand dc55232 has about 500,000. As you all know, MCC has great many variations in frosting and mirrored field, and there are also many specimen with no images. So I thought to encourage data entry, we should set aside a "staging area" for all unsorted varieties and entries with no images. So I thought CCT330 may serve as the staging area for the yr2000 silver 1oz frosted ring Panda. Perhaps there is a way to designate these specimen as "waiting to be sorted"?
What you suggest is the best way to handle unsorted specimens. If one "coin type" branches out into multiple "variety types", then the unsorted or unknown variety specimens should be left in the "root coin type". I think we're all still learning how to handle all this data we're collecting, including me. Eventually we will figure out the best practices for interacting with the data in the Coin Compendium. For now, we don't even have standardized words and phrases to describe a thing like a "root coin type" and "variety type", so I guess we'll have to figure that out as we go too.
Right now, the mintage estimates code assumes everything has been sorted, so it is wrong in those cases where it's not all sorted yet. That is something for me to fix. Continue using the "root coin type" as a staging area. I don't think there's anything else we should do to designate those specimens as "waiting to be sorted", since the CC should do that for us automatically.
I have reached the point where I have to accept that I can do no more than I'm already doing (most of my time is spent ensuring the CC remains funded), so these bugs and new features are probably going to have to be done by someone else. I am already in the process of formulating a plan to achieve some of our goals. One thing worth studying is this (also Google for Pywikipediabot to find much more info):
https://www.mediawiki.org/wiki/Manual:PywikipediabotThe CC is growing fast, and it's already so large that it's becoming difficult to fund efficiently, so we need more automation. We already use Pywikipediabot to maintain the CC's infrastructure I built, so if you're interested in finding ways to do more with less effort, Pywikipediabot is probably the most advanced way to begin.
I also use Jitbit Macro Recorder, which is very simple and very powerful. It's much easier to use than Pywikipediabot, and can achieve a lot in the same ways. In fact, it might be possible to use its image recognition features to automatically identify and sort varieties! That would be an interesting experiment. I think maybe I'll try testing it on something easy, like 1998 large dates, just to see if it can do the sorting with a low error rate. If so, then at the very least, in might be capable of finding and flagging coins that could be incorrectly sorted, for a human to look at later.
Another tool to look at is Actionaz. It is free, and also quite powerful, but not as easy to use as Jitbit.
http://www.jitbit.com/macro-recorder/http://www.jmgr.net/
