Pages are called Given_name Surname (Year_of_Birth-Year_of_Death). In case YoB or YoD is unknown, it is replaced with a question mark (?). Question marks are not very informative and they cause problems with forms (as they are a reserved character in the user interface).
It would be good to replace the question marks with something else. rtol 07:25, April 13, 2010 (UTC)
Robin's proposal[]
- Birth year unknown: do an estimate (using "c"). Make it the same as spouse birth year if any unless there's contrary indication such as parent or child birth year; if none, make a stab at it. If we later find we have a duplicate, we can fix one or other to distinguish. Better than a question-mark, which would be more likely to create a duplicate and has its own problems. Even a fairly wrong estimate doesn't actually matter, because we don't take date from pagenames as a rule.
- Death year unknown: omit hyphen and question-mark entirely, just as we do for people known to be still living. Do an estimate if you think you're within a few years but otherwise nothing.
- Both unknown: do just the birth year estimate as above.
Do it gradually as pages come up for edit. Move existing pages; change redlinks on sight.
— Robin Patterson (Talk) 07:57, April 12, 2010 (UTC)
Rtol's proposal[]
- Birth year unknown: do an estimate (using "c") -- Birth years can be estimated from the age of the parents, siblings, children and spouse, and from the time of death. Record the reasoning in Birth_date_notes.
- Death year unknown: do an estimate (using "c") -- Death years can be estimated from birth year, wedding year, and birth years of children. Record the reasoning in Death_date_notes.
It does not make sense to me to come up with a rigid procedure on estimating birth and death years, as circumstances vary and we have little experience with what works best.
We should, however, provide some guidance. SMW allows us to calculate such things as the average age of the mother at birth (in case the birth year of the mother is known), the average age at the first marriage (in case the wedding year is known), and the average life span (in case the death year is known). I've made a start here. rtol 07:25, April 13, 2010 (UTC)
- I agree with the second and third paragraphs, but I doubt if we should encourage estimates of death dates, because:
- There's a wider range even if we do have another life fact to go on
- If someone finds a better estimate, there will be page moves: not a big problem, but can get messy, especially if (because of the wider range) a third estimate is made
- — Robin Patterson (Talk) 07:06, April 14, 2010 (UTC)
Comments[]
I don't know if I am invited to this conversation, since I'm not an admin...but...I think estimating is a really bad idea. I read the c#### as a 'close' estimate. Say, within a year or two..five at the outside. Say I was looking for a person whom I knew was born in 1854 but knew nothing else about them. The circa could be off by decades and I might not follow up on a 1873 death record because information here could be entered as c1900 just to fill the box.
I guess what I'm saying is that people might use information from here as 'fact' and build their plan of research off of it. If people start estimating without sources (based on average life spans and whatnot) we could send people off in the wrong direction. I wouldn't want to be reading microfilm for hours based on an average or uneducated estimate.
I would prefer that some way be found to note that the date is unavailable. I like the bef. and aft. abbreviations, but a common replacement for the question mark would be useful for places where even that is lacking. Perhaps a 'U' or a different symbol? Ideally, the question mark would be 'made' to work, somehow. (I know - you've probably already tried to make it work, but it's just so perfect!)
Anyhow, that's my two cents, Lanica 20:15, April 15, 2010 (UTC)
- You are very welcome to this conversation.
- The question mark is a reserved character in the graphical user interface software of this wiki. We cannot change that.
- We could use NA instead of ?, or est1970 instead of c1970. rtol 20:37, April 15, 2010 (UTC)
- Yes, we cannot use question-marks with the current software unless we are willing to invite trouble. One or two of our most prolific contributors seem to think they can get away with it, but I fear that they are setting a bad example and probably storing up trouble for other editors.
- I appreciate Lanica's concern about "c". I doubt if "est" is very much broader in apparent range. Some expressions have distintly different meanings for different people, even within a single country. (Example - do you know that "Next Monday" can have two totally different meanings, a week apart?) Anyway, if I see a birth year shown as "NA" (or "est....") I will not change it unless I later find a close estimate to replace it with.
- The real solution to Lanica's concern about being led astray is that our SMW has now reached the stage where she and I and other ordinary genealogists can start using inline queries or semantic search to collect all potential individuals in a defined date range (and other combinations of facts) irrespective of what the article name says. If we don't already have a forum or help page setting out exactly how that works for Familypedia, with clear examples, it's time we did have. Remind me "tomorrow" if nobody else has supplied links to any such page.
- — Robin Patterson (Talk) 15:11, May 30, 2010 (UTC)
Can we make a decision to change the standard? The easiest is to use "NA" instead of "?" rtol 19:02, May 30, 2010 (UTC)
"Question marks ...cause problems with forms (as they are a reserved character in the user interface"
Is the problem with the Media Wiki software, or with the forms routine? The former possibility is the more serious, but if its embedded in the MediaWiki software I'm surprised that it has taken so long for this to be flagged as a problem. More likely its in the forms add on.
The purpose of the DOB and DOD in the title is not so much to provide information, as to make the article unique---e..g., so that you can distinguish between two John Smiths, one born 1607 and the other born in 1797. It doesn't really matter what you use for cases where the DOB or DOD is not known, as long as you can create a unique identifier. Given that, the mechanics of what the system needs to work efficiently should dictate the answer.
As to the need to eventually change a DOB or DOD in the title, that's not really relevant to the problem at hand. You have exactly the same problem even when both of these dates are known when the article is first written---after all, new information often leads to new understanding, and in this case, having a need to change the DOB or DOD should be fairly routine---might come up slightly more frequently when the initial date is unknown, but I suspect that's still a marginal problem.
In anycase, the critical point on the DOBS and DODs is that there's a place in the forms to record them---and that affords the opportunity to change them if the need arises. That won't change the title, of course. But that's something you can deal with if needed (as noted elsewhere) by renaming the article.
As to which of several different approaches could be used for the title in place of a "?" does it matter whether there's a uniform site policy about this? If "?" interfers with the system performance, that should probably be excluded. But whether you use "c1798" vs "unknown" vs. "something else" probably doesn't make much difference. You might have a preference, and think that should be the rule, but the truth is, unless there's an overwhealming need to specify something (like "no '?' permitted in the title"), it probably matters little. I would guess that most people are just going to create cards for their ancestors using whatever criteria seems appropriate to them for titles. If you make a rule up for this, then you just have to police the system that much more to be certain the rule is applied. As someone noted, that's apparently not the case now, and so probably won't happen no matter how you change the rule---other than getting rid of the "?".
For some, the more rules you have, the fewer problems are created. For others, the more rules you have, the more problems are created. The trick is in knowing when rules are important. Create rules for those instances, and not when having the rule is not important. In this case, getting rid of the "?" in the title is apparently important. How that is accomplished is probably not.
Bill Willis 23:36, May 30, 2010 (UTC)
- I have just changed the template {{Showfacts formbuttons}} I believe that fixes the problem with question-marks, at least for the "tree" pages and "pictures" pages. Please test it, everybody. Thurstan 04:12, May 31, 2010 (UTC)
- I think we should make a decision rather than reinvestigate whether this really is a problem. It really is. The question mark is used as a wild card in the user interface, and it cannot be used in a page title.
- If you really do need more convincing, use the little search box on the top left to find "Fronilde of Castile (?-1014)".
- Autocompletion works, but a click takes you to "Fronilde of Castile (" - I just created that page using the button "Create tab: Ancestor tree" on her sensor page.
- There is no use in trying to solve this problem, because it cannot be solved by users of this software.
- Let's therefore agree on a change in standard as everyday pages are added using the current standard. rtol 05:04, May 31, 2010 (UTC)
- I understand that a change needs to be made. I have no preference what is used in place of the question mark. I will be happy to follow whatever convention is decided upon. Bill Hunsicker 23:53, May 31, 2010 (UTC)
- Thanks, Bill. — Robin Patterson (Talk) 14:42, June 1, 2010 (UTC)
Suggested abandonment of this forum for an existing overlapping one[]
This forum is about dates alone, so I suggest that (rather than changing its title again) we now work on the more specifically-named Forum:Standards on dateless individuals. — Robin Patterson (Talk) 14:42, June 1, 2010 (UTC)