Thanks for the feedback.
"Tabular" text template gave a horrible looking huge table with lots of blank entries and wasted space, while "None" didn't create any TfS at all, meaning I had to re-enter all the details. Presumably I could modify the coding to save in my existing simple format, but the vast majority of FH users (including me) would not have the knowledge or inclination to do that.
None is intended for people who either don't want TfS, or actively
want to create it themselves (perhaps copy and paste from the Internet).
'Tabular' is intended to record what data you did and didn't enter -- I don't expect it to be used a lot, but it's a better Default than None, perhaps.
Did you try with one of the other Autotext templates I provided, e.g. England and Wales Death certificate, -- or creating your own (Tools > Manage Autotext > Autotext for plugins > Select the plugin you're creating a template for.) -- for example, for the csv transcription you currently use. I'm trying to make it possible for people to use whatever Autotext suits them -- rather than dictating what they should use. No coding would be involved -- just rich text editing. If you use the Help button, you'll find details of what tokens you can include in the Autotext.
(All the above is covered in the Help).
Even though I identified the informant as the principal's father, this did not carry across and create a witness fact.
Did you identify the record for the father? Or leave the parents as 'Name only/not mentioned'? When I test with an identified father, the witness fact is created... If you can replicate the problem and provide screenshots, I'd be grateful.
by creating the source first as a self-contained entity
This is what FH does -- you would not run the DEA until you were happy with the Source record you've created; and then use the DEA to enter the information derived from the Source that's needed to create the associated facts. (The DEA tries to extract information from Templated Sources, such as the principal's name or place of death, but allows you to adjust it or provide it if it can't find it in the Source. You can check the data as often as you like before you hit the OK or Cancel buttons, and review it and tweak it again (including the quality assessment) in the Results set that is generated.
Nothing is going to suit everybody, and I already know that you don't like AS or DEAs, so it was good of you to test it. I'm interested that you say a death certificate is simple -- all the fields presented for you to edit are included in a Death certificate I've seen (English/Welsh, Scottish, Canadian, America, Australian, New Zealand). For example a Scottish birth certificate will name:
- the deceased, with name, possiblly name and occupation of spouse and whether the spouse is deceased or not
- date time address and place of death
- sex and age of deceased
- name occupation and whether they're deceased for both parents, plus maiden name for mother (Canadian death certificates also sometimes provide place of birth of the parents)
- cause of death and who certified it
- Informant with relationship to deceased and residence
- Where and when registered plus registrar
Not quite so simple...
I do try to keep distinct in my mind what is definitely true beyond doubt (e.g. the fact that the source says the principal died on a particular date), and my assessment of whether that statement is true and relates to the person I assign it to
So do I -- once the DEA has run, I will use the result set to assigned individual QUAY (Assessments) to each fact -- and maybe even delete a 'fact that I disbelieve, or add a Fact Flag to mark it Rejected.
as I design the new version of Sourcery.
