Jump to content


Member Since 06 Aug 2006
Offline Last Active Yesterday, 06:37 PM

Topics I've Started

Searching Family Events

02 February 2017 - 11:17 AM

I have discovered what appears to be a useful feature of searching family events such as marriage and divorce. I considered posting this message in Tips, but I thought that I would first confirm that my observations are correct.


Suppose we have person A who has children with each of person X, person Y, and person Z. Suppose I have a marriage record for person A married to person X and a marriage record for person A married to person Z, but I don't have a marriage record for person A married to person Y. That doesn't necessarily mean that person A and person Y were never married nor that their children together were born out of wedlock. It may simply mean that I haven't yet found the marriage record for person A and person Y.


In any case, if I set up a search for Number of spouses -> not equal to -> 0 AND Marriage -> exists -> is False, then who is found by the search? Well, the search certainly finds person Y and the search certainly does not find person X or person Z. These results are expected. But the search does also find person A. This is a little unexpected, but nevertheless it can be a very useful result. The result is unexpected because person A manifestly does have a marriage fact - indeed person A has two marriage facts. But the result is also very useful because person A does not have a marriage fact for person Y. Hence, I was thinking of making this into a Tip, assuming my understanding of the search results is correct.



GEDCOM Export crashing RM

06 December 2016 - 04:32 PM

The following scenario repeatably crashes RM on my machine. (RM, 64 bit Windows 7 Home Premium). I don't know if will do so on anybody else's machine or not. It's not worth spending any time to fix in RM7, but perhaps my scenario could be tested against RM8 at an appropriate time.

  1. Begin a GEDCOM export. Forget to restrict the export to a particular group, and RM begins exporting everybody in my database.
  2. Realize my mistake and cancel the export long before it finishes.
  3. Begin the export a second time, except this time remember to restrict the export to a particular group.
  4. RM immediately crashes.

I have to shut down RM to get around the problem. Upon restarting RM, I can accomplish the GEDCOM export without incident provided I remember to restrict the export to a particular group before I start. I think it's the cancel and then starting again that causes the problem. I'm dubious that using a group has anything to do with it, but the scenario above is the one that's reliably repeatable on my machine.



100% of narrative report text needs to be under control of sentence templates

29 November 2016 - 11:33 AM

I have written about this problem before. In some ways, it's more of a "wish list item" than an "issues" item, and in some ways it's more of an RM8 item than an RM7 item because RM8 is going to be the first real opportunity for this problem to be addressed. But I think that maybe this is the best place to post for now.


For many years, all the way back to Family Origins days, I have used carriage returns in fact notes to control white space in narrative reports. I use carriage returns at the end of one fact note to force the next fact note to a new line or to a new paragraph. RM has added support to deal with such vertical white space, but the support that was added is not quite flexible enough for my needs. In any case, I have taken a decision to move vertical spacing control out of the fact notes and into RM's sentence templates. That means I will remove carriage returns from the end of fact notes and instead I will insert carriage returns into the beginning of sentence templates.


There are both advantages and disadvantages in making this change. The exact advantages and disadvantages don't matter so much as the fact that  I can't make the change completely because there is text in RM's narrative reports that is not under control of sentence templates. In particular, the spouse sentence in descendant narrative reports is not under control of a sentence template (or at least, it's not under complete control of a sentence template). And because of that, I can't use a sentence template to force the spouse sentence to begin with a new line or a new paragraph.


I think of the data printed for direct descendants in RM's narrative reports as consisting of four parts. If an individual has multiple spouses, then items #2, #3, and #4 are repeated for each spouse. This is for individuals with a spouse and/or children, and it's for the NGSQ and NEHGS formats. The outline formats work a little differently.

  1. Data about the individual
  2. Data about the couple
  3. Data about the spouse
  4. A list of the children

My problem occurs between items #2 and #3. The data about the spouse does not begin on a new line or a new paragraph even though the sentence template for the birth fact (the first fact) does begin with a carriage return. Therefore, I have to retain the carriage returns in the last couple fact such as marriage or divorce in order to get the data about the spouse to begin with a new line or a new paragraph. It's the first line of the spouse data that I'm calling the spouse sentence, and it's the spouse sentence that is not honoring the sentence template.


The list of children is also problematic when it comes to sentence templates. A child in the list can also be carried forward to subsequent generations as an adult depending on whether or not they have a spouse or children. If they are not carried forward to the next generation, then their sentences in the list of children are based on sentence templates. But if they are carried forward to the next generation, then their sentences in the list of children are not based on sentence templates. Suppose for example that you replaced all the standard sentence templates with sentence templates in French. You would discover that the list of children was sometimes still in English because the list of children was sometimes not using the sentence templates.


Therefore, I would like to reiterate my wish that RM's narrative reports would be 100% under control of user controllable sentence templates. And furthermore, RM needs to support multiple sets of sentence templates, so for example you could have a set in French and a set in English and readily switch back and forth between them. Or you could have a set for verbose reports and a set for terse reports and readily switch back and forth between them.


As I said, this is already a wish list item and maybe I should call it one of my chief "hope list" items for RM8. But in the meantime, it's an issue for me with RM7 as I try to move spacing controls out of fact notes and into sentence templates. That's why I posting this message under issues. I have a workaround, but it's very dissatisfying.



Quick and Easy Reports

04 October 2016 - 11:25 AM

There was an interesting message on RM's Facebook page recently where a user was requesting help making a report of everybody in their database who had died in Indiana. The answer is that there are certainly ways to create such a report from within RM (SQLite is not necessary), but such ways are neither quick nor easy. To that end, I offer the following suggestion that might address reporting needs for a lot of quick and easy reports.

  • First of all, I would suggest that a feature be added to People View where the contents of the View (including the lines not presently on the screen) could be printed. In effect, this report would become a new kind of Custom report. Due to a variety of limitations in People View, such a report would not be quite as powerful as a full blown Custom report, but it would certainly be much quicker and easier to use than a full blown Custom report. To this end, such a report should be initiated by a new option on the People VIew screen itself. People View already has options for Customize this view, Show Everyone or show a group, and Show Alternate Names. A fourth option should be added called Print this view. And the Print this view option itself should not have any options and should not pop up any windows that ask any silly questions. It should just create the report with one click. Having created the report, it should then be possible to print the report or to save it to a file.
  • Second of all, I would suggest that a feature be added to People View where each column could be filtered by right clicking the title at the top of the column. You can already sort the column by left clicking the title at the top of the column, so we would just be adding filtering to the existing sorting. The filtering feature would be most beneficial if it could be in effect for multiple columns at the same time. The filtering would not be intended to replace the ability to display a Group in People View, but rather would be intended to enhance the ability to display a Group in People View. Of course, the filtering could be used even if there were not a group in People View.

If this new feature were in place, the Facebook request could be satisfied by going into People VIew, making sure that Death Place is one of the columns being displayed, and filtering the Death Place column with contains Indiana.


It strikes me that this suggested feature would be quite an easy thing to program. And most of the time, I don't even think you would need to print the report or save it to a file. Getting the "report" displayed in People View would usually be sufficient.



Perils of Geocoding

09 September 2016 - 11:51 AM

I certainly think that RM is far and away the best genealogy program on the market. But regular readers of this forum will also recognize that I think RM includes a number of features that I think are not very satisfactory. For the most part, I choose not to use such features or else I use them and I have workarounds for their inadequacies. One such feature is geocoding. Or a better way to say it might be that RM's geocoding feature does not meet my needs. Geocoding may really be quite adequate, and it may meet the needs of many RM users quite well. But it doesn't meet mine.
When RM4 first came out with geocoding, I was very excited because I expected, for example, to be able to enter GPS coordinates for burials down to the individual grave site and for RM4 to record and map such coordinates. But geocoding doesn't work that way. Rather, geocoding is automatic and geocodes only down to the county level. It doesn't seem to allow me to enter my own coordinates even when I know the exact coordinates of an event - the church for a marriage or the hospital for a birth or death, for example, in addition to the grave site for a burial.
Here is a real example where RM's geocoding is of little help. In his will of 1879, David Cross of Sullivan County, Tennessee bequeathed to his son Elijah Cross some land in Anderson County, Tennessee. David Cross's will also stated that David had inherited the land from his father, but did not give his father's name. I need to prove David's father by finding Cross men of an age to have been David's father who owned land in Anderson County. A man named Micajah Cross who might have been old enough to have been David's father purchased land in Anderson County in 1806. 
Anderson County was formed in 1801 and subsequently was split multiple times to form many additional counties. Much additional research has determined that Micajah's land was in modern day Scott County, Tennessee which is several counties away from modern day Anderson County. I can even locate Micajah's land down to within 50 to 75 feet of its exact location in Scott County. Never mind that Micajah's land became part of Campbell County upon its creation in 1806, and never mind that Micajah's land became part of Scott County upon its formation in 1849. I can conclude from its actual location that Micajah's land was not the land that David had inherited because the land was in Scott County in 1879. But if I enter a deed event into RM for Micajah Cross in Anderson County in 1806, RM geocodes it in the geographic middle of modern day Anderson County.I think it should let me geocode the land where it really was, even if that location is in modern day Scott County.
RM seems to geocode counties by their geographic center in modern times. RM seems to geocode states by their geographic center in modern times. So I was expecting it to geocode countries by their geographic center, and I was expecting it to geocode United States as a point somewhere in Kansas. But it seems to geocode United States to Washington, D.C. So I don't what RM's geocoding rules really are.
To tell you the truth, I'm not sure how I think RM's geocoding should be changed to make it better. It's a very difficult problem. The one thing that I do think is that it should allow me to enter precise coordinates for events when I know them. In the meantime I put coordinates for such events into notes for the events and RM can't map from the notes.
Here is a URL for a cautionary tale about the use and abuse of geocoding when geocoding is based on the middle of geographic areas. For example, how would you like to live in Kansas and have every geo-reference to the United States end up in your front yard?