Monday, January 04, 2010

READERS' RESPONSE: Census "view maps" at

Last week's blog entry titled Census "view maps" links no good, solicited several reader responses.

FIRST: Ruth Himan (via FaceBook) said: "I know I should post the errors I find in software. But sometimes I do not know if it is just the user. But I need to slow down a bit and correct and make aware when I can---worse case someone can correct my thinking and all is right again. My findings will always go in as an inquiry with samples. Thanks for sharing and especially sharing with a "fix". Your Genealogy Cousin (I know somewhere we will hook up)."

SECOND: Randy Seaver gave a shout out citing Ol' Myrt's blog entry in his Best of the Genea-Blogs - 27 December 2009 to 2 January 2010.

THIRD: GEOLOVER posted a response to my blog on Randy's site just a little bit ago this morning:
Geolover said...

Thanks for this: [Randy wrote] "Census "view maps" links no good, by Pat Richley on the DearMYRTLE's Genealogy Blog. Ol' MYRT finds a problem with the map function - it points to the county as if it's a city or town. Well done, MYRT - this slipped past everyone else!"

However, when the maps were rolled out there were many such critical comments in the blog announcement.

The core of the problem is the gedcom-format failure to distinguish between villages and Township-type municipalities by the same name. Even googlemaps does not handle this well: try to get it to come up with a Madison Twp. in Ohio and it comes up with Madison, Wisconsin, without telling you, and incorrectly linking the tree place/event. Similar problem with Bowling Green Twp, Licking County, Ohio; 9 times out of 10 you get Bowling Green, Kentucky.

It is not entirely the mappers' programs' fault. It is the gedcom-format failure to designate what unit of government is referred to.

What *is* the mappers' fault is failure to map Township-type subdivisions in order to distinguish between Counties and towns by the same name. The mapper-programmers are fundamentally city-based, and unable to provide maps that account for the fact that most of our ancestors did not live in cities.

I post here because there is no way to comment on the post on Dear Myrtle's site.

Monday, January 4, 2010 10:16:00 AM PST

ITEM: Regarding no comments at Myrt's blog: Ol' Myrt here doesn't permit even moderated comments on my blog, due to an overabundance of spam comments -- time consuming to reject, and not the way I'd like to spend my days. Readers merely click on my email at the bottom of a posting to send an email. Lately, responses have also come via FaceBook email or wall postings.

ITEM: GEDCOM in my view has nothing to do with the problems Ol' Myrt reported. GEDCOM is a file format that facilitates sharing of compiled genealogy data between genealogy software programs. Neither Bing nor Google maps are GEDCOM file format compatible.

Ol' Myrt here totally gets that the problem isn't the map site's problem, but it's in the character string sent by the user (be that you, me or a site like

The case I cited concerned sending a character string (from the label of their census page, NOT through a GEDCOM file) to maps at where:

The three fields were broken down by Bing as:


But users know to be:
state can easily overcome this for the "view map" option on census index pages -- without regard to anything GEDCOM. Their coders merely need to strip out county names when requesting a search at a linked map site (be it Bing or Google), though I still prefer Google Maps at this point.

ITEM: Township versus city. You are entirely correct that modern day maps don't handle historic places well (such as townships, mentioned in your comment). Ghost towns will also be problematic. We agree it is difficult at best to plot an ancestor's place of residence on 21st century maps without benefit of GPS.

ITEM: Blog - Indeed the blog got quite a few upset responses (from a total of 167!) to the announcement back on 20 Nov 2008 titled See People In Your Tree on a Map. THAT data being sent to an external map website from IS GEDCOM based, since the data came from users' compiled family trees which may have been uploaded as a GEDCOM file, instead of typed in letter by letter at the website. But this isn't entirely a GEDCOM file format issue, either.

It's also a user issue.

GEDCOM file format doesn't require users to type the town, county, state and country when specifying an ancestors' event location. People frequently type the state or country in the first field, because that is all they know at that point in their research. So even GEDCOM isn't at fault. Its just plain difficult to track how users will employ a piece of software.

SUMMARY: The problem Ol' Myrt described was an ' census index page link to an external map site' error -- something is quite different from an 'user input on trees linking to external map site' error.

Try saying that three times fast, after eating four Saltine crackers without a drink to wash 'em down first.

Happy family tree climbing!
Myrt :)
Your friend in genealogy.

No comments:

Post a Comment