Strange Problem - 3rd column in ancestry sometimes missing

[Lynn McLaughlin]

I just started using TMG to GEDCOM last week, and it was looking great. Today I'm having the weirdest problem. I made a small version of my TMG project (32 people), and then in T2G I set the template to ancestry(private), generated a GEDCOM and brought it into ancestry.com. When I call up that tree, it comes up to one of my great-grandfathers on the far left, and when I click and ask for profile, I only get two columns, Facts and Sources. This happens with some of the people in the tree, but not all. I've spent the last couple hours trying to figure out what I'm doing differently, but to no avail.

To demonstrate, it comes up looking like this: 

If I click on Clarence and choose Profile, I see this and no matter how wide I make my screen, I don't see the third column:

If I use the browser back button, it looks the same as the first image, except now the Add Father/Mother boxes have Potential Father/Mother instead. If I click on the daughter and then ask for her profile, it looks normal:

But if I click on her father, Clarence here, it only has the two columns.

What could be causing this? Can it be something in the GEDCOM or does Ancestry have a weird bug? Seems like a combination of things, since I've never ever seen such a thing in an ancestry tree. I did one more test and used the Export function within TMG on the same project. It still comes up looking the same as the first screenshot, but Clarence's profile now has the 3 columns. 

thanks, Lynn

[John Cardinal]

Lynn,

 

That probably means there is invalid HTML in one of the person's events. Ancestry doesn't import a lot of what TMG exports, but it imports most/all of what T2G exports. Review the events for Clarence and look for mismatched printer codes or mismatched HTML tags you added yourself.

 

John

 

[Lynn McLaughlin]

I've never used an HTML tag, and the only printer code I used for Clarence is [:CR:] which doesn't need a match. I checked in the big file I loaded last week and thought had no problems, and he did have the same issue, so it is something about him - and he isn't the only one. If I turn off all sources in T2G, it works, so it has to do with a source. I've spend over 2 hours now trying to narrow things down, and I've found one source on the Name that is causing problems. It is for a book with a long title: Portrait and Biographical Record of Iroquois County, Illinois, Containing Biographical Sketches of Prominent and Representative Citizens, Together with Biographies of all the Governors of the State, and of the Presidents of the United States

Shortening the title worked. What is the max length? Is there anything in the ged file that warns me of this? I need a way to find all of these issues.

thanks, Lynn

--
You received this message because you are subscribed to the Google Groups "TMG to GEDCOM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to t2g-l+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/t2g-l/002601d6183b%24172535e0%24456fa1a0%24%40gmail.com.

[John Cardinal]

Lynn,

 

That source title is 241 characters. The maximum source title length is 248 characters, but Ancestry may or may not allow that length. Also, depending on how you have the T2G properties set, the source title may include HTML because most TMG source footnotes include the [ITAL:] and [:ITAL] codes to italicize titles, and that could push it over the official 248 character limit.

 

Can you copy the entire SOUR record from the GEDCOM file so I can review it?

 

The SOUR record will start with "0 @S1234 SOUR" record and continue until the next "0 @Snnnn SOUR" record. The "1234" part will be the TMG source number.

 

John

 

[Lee Hoffman]

At 4/22/2020 00:01 -0400, John Cardinal wrote

>That source title is 241 characters. The maximum source title length
>is 248 characters, but Ancestry may or may not allow that length.
>Also, depending on how you have the T2G properties set, the source
>title may include HTML because most TMG source footnotes include the
>[ITAL:] and [:ITAL] codes to italicize titles, and that could push
>it over the official 248 character limit.

John -

The GEDCOM specs I have (v5.5.1) says the max line length is
255. However, it notes that characters wider than 8 bit can require
larger buffer sizes. I do not find where Source title length is any
less, but I might have missed it.. Further, it may be that Ancestry
has effectively set that limit by limiting the buffer size to 255 and
thus truncating anything wider.

Lee

[Lynn McLaughlin]

Here it is:

0 @S2@ SOUR
1 REFN 2
1 TITL unknown author, <i>Portrait and Biographical Record of Iroquois County, Illinois, Containing Biographical Sketches of Prominent and Representative Citizens, Together with Biographies of all the Governors of the State, and of the President
2 CONC s of the United States</i> (Chicago, IL: Lake City Publishing Co., 1893). Hereinafter cited as <i>Iroquois Co. Bio.</i>

I've figured out what the deal was.  I read at 

http://wiki-en.genealogy.net/GEDCOM/Field_Sizes 

that the max line size is 255 and I assume you got 248 by subtracting the level number, space after it, 4-charater tag name and space after that. My TITL line is a total of 245 characters, including the <i> and <\i>. Because you used the CONC, the lines are short enough.

But also, ancestry has a limit of 256 characters, including the control codes, for what is puts as the source, and if you start with "unknown author" and go up to 256 characters, it will be cut off before the </i>. And that, as you originally suspected results in them being mismatched and driving ancestry's display crazy.

So I think a change in T2G (sorry, once a software engineer, always one) is to place the </i> so that the > would be the 256th character printed. So it will work if the CONC line starts with

2 CONC s of the Unite</i>

Nothing after that will show up.

But it would probably be nicer if T2G could issue a warning that the text for source 2 (or whatever) will be truncated.

Thanks, Lynn

--

You received this message because you are subscribed to the Google Groups "TMG to GEDCOM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to t2g-l+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/t2g-l/006101d6185a%24b5a15220%2420e3f660%24%40gmail.com.

[Lynn McLaughlin]

The line length is 255, but each line starts with something like "1 NOTE " which is 7 characters, leaving 248.

Lynn

--
You received this message because you are subscribed to the Google Groups "TMG to GEDCOM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to t2g-l+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/t2g-l/5ea07a85.1c69fb81.17a2a.0b65SMTPIN_ADDED_MISSING%40gmr-mx.google.com.

[John Cardinal]

Lee Hoffman wrote:
> The GEDCOM specs I have (v5.5.1) says the max line length
> is 255. However, it notes that characters wider than 8 bit
> can require larger buffer sizes. I do not find where Source
> title length is any less, but I might have missed it.. Further,
> it may be that Ancestry has effectively set that limit by limiting
> the buffer size to 255 and thus truncating anything wider.

Lee,

The SOUR record is described on page 27 of the GEDCOM 5.5.1 specification:

n @<XREF:SOUR>@ SOUR
+1 TITL <SOURCE_DESCRIPTIVE_TITLE>

SOURCE_DESCRIPTIVE_TITLE is defined on page 62 with "{SIZE:1:248}". That defines the allowable length range as between 1 and 248 characters.

The maximum line length is 255 (though that is ambiguous due to some loose wording), but the line length only affects values indirectly. For example, if a value is allowed to exceed 255 characters, GEDCOM allows CONC/CONT records so that no single record exceeds the maximum line length, but the resulting value does.

John

[John Cardinal]

Lynn,

 

The limit is 248. See my message to Lee where I described the SOUR.TITL definition.

 

It's impractical for T2G to fix this.

 

T2G can truncate the result, but that would not be appropriate for most target applications because they will accept more than 248 characters despite the GEDCOM limit for the value.

 

Truncating a few characters early might avoid the Ancestry issue, but source values might include various HTML tags and keeping track of which ones need to be inserted is tedious. Tedium aside, shoving end tag(s) into the value and truncating the value wherever that avoids the Ancestry issue would result in loss of data on the T2G side.

 

I will do some testing, but I am in the middle of something else and so I won't change anything soon.

 

John

[Lee Hoffman]

At 4/22/2020 14:26 -0400, John Cardinal wrote

>The SOUR record is described on page 27 of the GEDCOM 5.5.1 specification:
>
>n @<XREF:SOUR>@ SOUR
> +1 TITL <SOURCE_DESCRIPTIVE_TITLE>
>
>SOURCE_DESCRIPTIVE_TITLE is defined on page 62 with "{SIZE:1:248}".
>That defines the allowable length range as between 1 and 248 characters.
>
>The maximum line length is 255 (though that is ambiguous due to some
>loose wording), but the line length only affects values indirectly.
>For example, if a value is allowed to exceed 255 characters, GEDCOM
>allows CONC/CONT records so that no single record exceeds the
>maximum line length, but the resulting value does.

OK, I missed p. 62. Duh! Should've looked there, but wasn't thinking.

Thanks -

Lee