* Patronymics (& Matronymics/metronymics)

[elevator]

I am not 100% sure what you mean, but it is definitely clear that the GEDCOM standard falls short on many fronts. Unfortunately, within FH we are forced to work within these constraints, and so any method to record (and reference and search) this data inherits these constraints. I used to work with Gramps a lot (and TMG before that), and Gramps is actually one of the programs I liked the most in its ability to record names (and also places). Gramps specifically allows not only the researcher to differentiate between things like birth names, married names, nick names, a.k.a.'s, etc. But also allowed you to specify name element origin such as Matronym, Patronym, Toponym, etc. And this is even in addition to allowing you to separate out name elements like titles, suffixes, prefixes, and a host of other things. (Names in Gramps). Gramps is also similarly a project that has put an equal amount of thought into places.

Unfortunately, as mentioned, these are all things that fall outside the GEDCOM definition and so for our purposes we have to do with what we have. For me, recording every single name variance encountered in the sources has been the best way for me. For purposes of simpler and cleaner searching and reporting and such this has meant I have had to revert to "normalizing" certain names. No ideal, but it is what it is. Incidentally, I like FH much much better than Gramps overall, but I wish there was a way to better represent names and the various elements of them in a more robust manner.

Does my pondering whether filtering of names should be first/surname agnostic offer anything to these circumstances?

[davidf]

I am not 100% sure what you mean, but it is definitely clear that the GEDCOM standard falls short on many fronts. Unfortunately, within FH we are forced to work within these constraints, and so any method to record (and reference and search) this data inherits these constraints.

One of the issues that often occurs with Patronymics and "Farm Names" (Toponymics) is whether you put them inside or outside the "surname slashes" - this can impact on whether an FH filter (which offers Last Name and First Name input boxes) will find the name/individual (at best you have to try names in each of the offered input boxes)

My pondering is whether a single field filter would be more effective; in it you would enter any of the words that appear in any of the Name Fields or name parts and the filter/search would return any name which included all those words (or initial part of words) in any Name Field. (i.e. like many standard internet searches).

Such an enhancement would be GEDCOM independent.

[ColeValleyGirl]

My pondering is whether a single field filter would be more effective; in it you would enter any of the words that appear in any of the Name Fields or name parts and the filter/search would return any name which included all those words (or initial part of words) in any Name Field. (i.e. like many standard internet searches).

So, a lot like Edit > Find with the search restricted to Names?

[davidf]

So, a lot like Edit > Find with the search restricted to Names?

Well that searches across all records (and all name parts?) not just Individual records and the speed and output are different.

I am thinking of a change to the input of a Name to filter:

Filter by any part of name

Screenshot from 2022-11-10 10-45-15.png (6.04 KiB) Viewed 990 times

Hence putting in "Eric" will "return" all FirstNames including "Eric", but also "Ericsson" irrespective of whether that Patronymic was entered as a First Name or a Last Name. Also you could search for "Rev." which might be in the Name Prefix.

This filter box occurs in the Record Window (Individual and Family tabs), and the "select" in functions like Merge Records, Find in Diagram etc.

[ColeValleyGirl]

I'd find that single filter a lot less useful than what exists now.

[davidf]

Helen,

Can I push you as to why?

I've tried to suggest some of the benefits when I originally floated the idea - particularly in circumstances where you are not sure whether a name element has been entered "between the slashes".

As for disbenefits, I can understand that such a search would produce a longer "short list" (e.g. Surname = Ericsson when you only wanted Given Name = Eric) but most applications of the filter are either "exact search" where you know the full name and put it in to get a single return - which you then select by a click from a short list of one, or "part search" where you select the name you want by a click from a short list of many. It would become an issue if your search was so vague that the short list was unmanageably long and you could not refine it by entering more search terms - but isn't that pretty much the existing situation?

[ColeValleyGirl]

How would your proposal search for initials M. J. without returning a much longer set of results than necessary? Really useful when working with DNA and an online tree that somebody has helpful 'anonymised'.

[BillH]

With all the Norwegian and German names I have in my tree, I have long wished there was a filter on the records window that would find whatever you are searching for no matter where in the name it is rather than just the beginning.

It could be an additional box so their would be a 'Last Name' box, a 'First Name' box, and an 'Anywhere' box (or something like that.

I wouldn't want the existing boxes removed as they are very useful as well.

Bill

[tatewise]

The Edit > Find... command will do something like that with just the Names box ticked.
However, it includes other Name fields than Individuals such as File Name.

The Search and Replace plugin with just Individual Names ticked in Search ONLY mode will do something very similar.

[BillH]

I have used both of those options in the past to try to accomplish what I want, but it would be a lot more convenient if it could be done on the records window. I never really understood why filtering was limited to just the beginning characters in the last name and first name.

Another option I've thought would be handy rather than an additional box would be a Contains radio button for the first and last name so the filtering would not be limited to just the beginning characters.

Bill

[davidf]

How would your proposal search for initials M. J. without returning a much longer set of results than necessary? Really useful when working with DNA and an online tree that somebody has helpful 'anonymised'.

Well "M._J." (underscore = a space) would match (at first thought):

• M./J./ - which is probably what you are anticipating - but also
• J./M./
• M._J.// - which might also be what you want (but would not get under present regime with M. in First Name and J. in Last Name)
• J._M.//
• /M._J./
• /J.M./

But if it was searching "all Name elements" we might have problems:

• M. as a prefix (e.g. French Monsieur), but only if there was also an initial J. elsewhere
• likewise for J. as a prefix and an Initial M.
• Similarly if single letters appear as suffixes (sure to be examples if qualifications and decorations are included e.g. B._A. (but not B.A. or BA or V._C. (but not V.C. or VC - which are single words)
• Possible other more obscure permutations (e.g. Surname Suffix M. and a nickname J.)

"M_J" (i.e. without the periods would match on any name with an element starting with M and an element starting with J) would give a long list - but it would still be pretty long under the current arrangement (any given name starting with M and a surname starting with J)

If you have a lot of names that are just initials, I guess you would attaching a period and if necessary a space between initials to avoid matching to full names already.

In practise would all the above give an excessively long list?

[davidf]

With all the Norwegian and German names I have in my tree, I have long wished there was a filter on the records window that would find whatever you are searching for no matter where in the name it is rather than just the beginning.

It could be an additional box so their would be a 'Last Name' box, a 'First Name' box, and an 'Anywhere' box (or something like that) ....

That is an option that would meet the requirements that I was thinking of - as well as miss-identified compound names (e.g. Ralph Vaughan Williams) which may have been entered incorrectly slashed.

The Edit > Find... command will do something like that with just the Names box ticked.
However, it includes other Name fields than Individuals such as File Name.

The Search and Replace plugin with just Individual Names ticked in Search ONLY mode will do something very similar.

Being able to do it on the Records Window or the Select dialogue would help identify variably formatted duplicates (e.g. Ralph /Vaughan Williams/ and Ralph Vaughan /Williams/) for merging or review in a single step. In fact it could identify duplicates that you did not realise you had.

... It could be an additional box so their would be a 'Last Name' box, a 'First Name' box, and an 'Anywhere' box (or something like that.

I wouldn't want the existing boxes removed as they are very useful as well.

In terms of "screen estate" I am not sure about adding an extra input box (the "Anywhere box"). Can we identify a moderately intuitive (or at least memorable) way to enable switching by means of a small button/icon to toggle (persistently) between the two alternatives?

[BillH]

In terms of "screen estate" I am not sure about adding an extra input box (the "Anywhere box"). Can we identify a moderately intuitive (or at least memorable) way to enable switching by means of a small button/icon to toggle (persistently) between the two alternatives?

See my last reply where I mention a Contains radio button. I actually thought it would be helpful to have three options:
begins with
ends with
contains

Bill

[AdrianBruce]

... But also allowed you to specify name element origin such as Matronym, Patronym, Toponym, etc. ...

Oh that's interesting. One thing that I have been guilty of is referring to patronyms as if they are always the equivalent of, or a substitute for, "family name". Of course(?), that's not true - in the name Pyotr Ilyich Tchaikovsky, "Ilyich" is a patronymic and "Tchaikovsky" is a family name.

[davidf]

...
Another option I've thought would be handy rather than an additional box would be a Contains radio button for the first and last name so the filtering would not be limited to just the beginning characters.

Bill

That (a "contains option" or similar - wildcards?) would be a help when you are not sure if the name begins Mc or Mac etc.

But a "contains" option would not handle miss-slashed compound names - for that I think the "Anywhere" box is probably the easiest option.

[ColeValleyGirl]

If you have a lot of names that are just initials, I guess you would attaching a period and if necessary a space between initials to avoid matching to full names already.

In practise would all the above give an excessively long list?

They're not names in my tree -- they're names in online trees of DNA matches. Searching my tree for the corresponding initials via firstname and surname (with a year range) usually gives a manageable list -- importantly, with no false matches.

I agree there needs to be a better mechanism for filtering to handle more complex naming patterns, but I don't thing a blunter search is the answer.

[tatewise]

The Edit > Find... command will do something like that with just the Names box ticked.
However, it includes other Name fields than Individuals such as File Name.

The Search and Replace plugin with just Individual Names ticked in Search ONLY mode will do something very similar.

Being able to do it on the Records Window or the Select dialogue would help identify variably formatted duplicates (e.g. Ralph /Vaughan Williams/ and Ralph Vaughan /Williams/) for merging or review in a single step. In fact it could identify duplicates that you did not realise you had.

Both the options I suggested show enough detailed results to identify such 'duplicates' and allow merging.
i.e. They display a Result Set window like a Query, and that is similar to the Records Window.
Yes, it needs an extra step or two to setup the search but is available now.

[ColeValleyGirl]

Mike, davidf explained why he didn't think Edit > FInd was a solution when I suggested it: https://fhug.org.uk/forum/viewtopic.php ... 17#p130066

[BillH]

It's obvious that everyone works differently and has different needs. What works for one may not work for another. My suggestions were what would work best for me, but I understand others would have different needs. Maybe someday CP can incorporate something which will allow more options for filtering on the records window other than just "begins with" filtering. How (or if) CP do that is of course up to them.

Bill

[BillH]

Mike, davidf explained why he didn't think Edit > FInd was a solution when I suggested it: https://fhug.org.uk/forum/viewtopic.php ... 17#p130066

Helen, how are you getting that link to a specific post within a thread? That would be very useful.

Bill

[tatewise]

I was actually responding to Bill's point, who said he did use those options.
I also mentioned the problem of Edit > Find > Names looking outside Individual records.
The Search and Replace plugin had never been mentioned before and its search is focused on just Individual Name fields.

[tatewise]

Bill, you display the specific post and copy the browser URL in the address bar at the top.

[ColeValleyGirl]

Helen, how are you getting that link to a specific post within a thread? That would be very useful.

Bill

Click on the heading for the reply and copy the url from your browser address bar. Past it into your post and use the link icon to convert it to a url.

No doubt somebody else has a quicker way to do it.

[BillH]

Mike and Helen,

Duh... I really should have known that.

Bill

[davidf]

Are we getting to a stage where a Wish List Request is beginning to become relevant (and be exposed to people who are not interested in Patronymics etc.)?

The Requirement might be something general along the lines of:

Name Filtering

The Name Filter Function in the Records Window and Selection dialogues should be able to handle situations where Names have been miss-slashed or inconsistently slashed (e.g. Compound Surnames, Patronymics and Toponymics).

This would make it easier to filter and select names (particularly non Anglo-Saxon "Given /Surname/" patterns) as well as catching accidental duplicates due to inconsistent input.

Notes:

1. A user switchable (but persistent) option to swap between the current Last Name and First Name input boxes and an Anywhere input box (where "Anywhere" could include all Name Parts - suffix, prefix, surname prefix, nickname, used etc.) might achieve this.
2. At the same time adding the functionality to handle part matches other than the current "begins with" - either by "contains" or "begins/ends with" or full implementation of wildcards - would enable matching irrespective of "Mc" or "Mac" and similar.
3. If Soundex could be incorporated into the filter it would allow searching on name variations/miss-spellings (Taylor Tailor, Tailer) and avoid searching on created columns using the Soundex function.