Personally, my solution is to avoid M$. I find that other technologies almost always work better with less hassle. But I recognize that that doesn't work for many people.I have two comments.
1 - Microsoft does not guarantee that old files will work with future versions. I don't know what the solution is.
No argument. I thought about doing that, but decided to stick to my usual software engineering dictum, and not pre-optimize.Maybe avoiding a spreadsheet is best. I think that a small database with a web front end would be better. That way when someone does find a single cross, they can add it without any sending of emails (where is that email address? What is that guys name?), etc.
It may be that you're right, and we'll end up with enough traffic and updates that managing a spreadsheet will become unwieldy. In that case it will certainly be worth stuffing it all in a DB. That would actually be quite a good problem to have
Yes, no argument to that either. Again, I figured something was better than nothing. If there are updates and/or refinements to what's there now, great, let's capture them.2 - Some of data in the cross ref is questionable. Bearing cross references can be a minefield..
It should go without saying that anybody using this data is expected to use some common sense and skepticism. Check those specs before ordering!