1. FTM is unable to represent any leap day date in the Old Style calendar. For example, 29 Feb 1604 (o.s.) should be represented as 29 Feb 1604/05, but FTM considers this to be an invalid date.
2. FTM will not accept dates prior to 100 B.C.E.
3. Often it is impossible to determine whether a date in an original source is New Style or Old Style. FTM assumes all dates entered are New Style, leading to documentation errors.
4. When working with a source wherein all the dates are Old Style, one must convert them to New Style before entry into FTM. It is easy to slip up here, leading to incorrect documentation. FTM needs an option to change the assumption to Old Style.
5. The date forms, "child", "infant", and "stillborn", intended only for death dates, are nevertheless permitted in any date field and are undetected as "unlikely" by the date error checker. 

6. Partial dates, such as "Jan", or "4 Jul", or "Feb 1700/01", or "1700" may be entered. Year-only dates in the double-date form, e.g. "1700/01", are not accepted.

1. Ranges: If using a M/D/Y format, the comma will disappear if one end of the range includes a month & day while the other end of the range includes only a year.

 
 

Enter:	Bet. June 1, 1975-1980
Display:	Bet. June 1 1975-1980


There is no problem if both ends of the range include a month & day. 

2. In Ancestor and Descendant Trees, choosing the "birth-death" date display option may incorrectly show these dates if they were entered as a range. The first date of the range is used, which is incorrect if that is a death date. The "birth-death" dates should always give the broadest span by picking the earliest birth date and the latest death date. This wart seems to have been introduced with version 5.

version 5
version 4.4
 

3. GenRpts: One can enter "stillborn", "infant" or "child" in place of a death date, indicating an early demise. The program shows no finesse when plugging these into reports. The GenRpts, which are intended to show a verisimilitude of "natural writing", end up awkwardly worded: Thisshould be "Jack Spratt died as a child." or, "... died as an infant.", or "... was stillborn."

4. In trees, when the output date format chosen is "born-died", a  birth date entered without a year – "September 16" – will display with an absent death date as "September - " . This output form is supposed to display just the years of birth and death, and has done so until version 5.

1. Inconsistent and incorrect handling of Old Style dates involving March 25^th: The manual specifies that a double date "falls between December 31 and March 25". [Vers 4, Chap.3, p.94, italics theirs]
1. If March 25 stands alone or begins a date range, FTM will not convert it to a double date. (No problem there.) However, if a date range begins in Jan or Feb and ends with March 25, then FTM converts March 25 to a double date. (March 26 through 31 will not become converted if used as the ending date of a range beginning in Jan or Feb.)
2. If a date range begins in March and ends with any day in March, the ending date will be converted to a double date including days 25 through 31

1. Date range: FTM does not sort date ranges properly; the month & day are unrecognized. Example--Mary is married twice. You know the first marriage occurred June 1, 1980. For whatever reason, you know the 2nd marriage occurred ‘before 1-1-1985’. Enter a marriage date for 2nd marriage as "bet. June 1, 1980 - January 1, 1985". Mary will sort as the 1st wife. This problem pertains to marriage ordering, but not children lists. [McE]

2. Dates with "Bef." and "Aft." modifiers sort in peculiar ways. "Aft." dates without a specific day sort before dates with specific days. E.g. "Aft. 1676" is ordered before "1 SEP 1676".

1. It is not possible to distinguish different "qualities" of dates in FTM. A date range which is little more than an educated guess is indistinguishable from a range where the bounds are grounded in other hard facts. Users should be able to add a prefix of their choice to individual dates.