Reply

That’s right. The sources are created on a “per tree” basis, so you can’t use the same one across multiple trees.

But the numbering is kept unique as without it the so-called “merge trees” feature can’t work.

The issue is one of language. (see http://kiwitrees.net/faqs/ for more details). There is no real ability to merge trees. The tool is actually and deliberately (in kiwitrees) called “Append trees” for that reason, but everyone insists on using the incorrect term “merge”. Even me it appears. I must avoid doing that.

So if you want to “merge” trees you use the “append trees” option which simply adds one tree onto the end of another (hence the need for all records having uniques IDs). Then you need to identify the duplicates, be they individuals, families, sources, repositories, media, or notes and use the “Merge records” tool to merge them one at a time.

For further clarification, the need for unique ID’s does apply to individuals in the same way as sources, So to answer your earlier comment “Assume the next record will be A2 (I changed)“, no it probably won’t. It’s more likely to be something like A123456, depending on the next lowest unused(not just next highest) ID on your site.
Ideally “John Doe” would be created in the same way, but for simplicity it is hard-coded to use only I1. I might look at the possibility of improving that one day, but as mentioned often, these IDs are machine-intelligible only so what they look like for humans should be irrelevant.

Nigel
My personal kiwitrees site is www.our-families.info