Louis Kessler’s Behold Blog

Yesterday, I logged into my account at my webhost Netfirms and I was met with a somewhat alarming message:

That was not pleasing to me. I knew what that meant. Likely I’d need to make major revisions to my website to get my Behold blog and my GenSoftReviews site to work under the new version of PHP.

My Behold blog and GenSoftReviews sites are 12 years old. I developed them both myself with WordPress. GenSoftReviews uses a WordPress plugin called WP Review Site that I purchased and then customized to my liking. My Behold Forum uses bbPress version 0.8 that was able to integrate with WordPress.

I spent many months customizing my blogs and forum to my liking, starting with the Behold style that I created to make my blog and the forum completely match the rest of my site. I added a user database for my Behold and DMT trials and purchases and automated the sending out of trial keys and recording of purchases. I created an integrated login system so people could post comments on my blog and messages in my forum. I added sophisticated spam filters to prevent the multitude of spam from getting onto my page. I added my newsletter system into the framework. Almost every single thing is tweaked and customized exactly to my liking.

The programming language for this is PHP and the database is MySQL. I had never used either of them prior to this endeavor, so it was a trial by fire. I’m proud of what I created and it has worked almost without a hitch for the past 12 years. That is of course without upgrading the underlying versions of WordPress and bbPress that I was using. I couldn’t upgrade them, really. The customizations I had done were extensive and some of the plugins that I was using were no longer available and were not being upgraded to work with new versions of WordPress.

Flipping the PHP Switch

I knew what would happen when I selected a PHP version 7 or greater: My blog would stop working. I tested it out and sure enough, only an error message appeared where my blog should be.  I changed it back, and it worked again.

I spent the next couple of hours adding PHP 7.4 to my computer. I went back to my live blog and tried a few things. I flipped the PHP switch on my live site again and got the error again. I flipped it back to 5.6 and … oh oh, I still had the error.

This was no ordinary error. This was the dreaded Error 500 – Internal Server Error, that told you absolutely zip, zero, zilch about what was going on:

So how do you figure out what’s causing an error when no information is given? Into my Wordpress PHP code I went. For the next 3 hours, I was debugging it live online, line by line, putting in “here I am” statements and tracing to find what line is causing the error. I found out it was the line that was trying to initialize the MySQL database.

    $wpdb = new wpdb(DB_USER, DB_PASSWORD, DB_NAME, DB_HOST);

I spent two hours trying to get WordPress to initialize the database and tried everything including setting up test programs, and scanning the web and StackOverflow for this type of problem and solutions. I almost went as far as changing the password on the database. The funny thing that I noticed was GenSoftReviews was still working, but what that meant didn’t yet register on me.

It was now 1 a.m. I used Netfirms Support chat and got help from one of their support people. I was trying to figure out from the support person why the PHP change and then changing back now resulted in my blog not working. We tried a number of things and finally I was given a ticket where a Technical Specialist would contact me in 24 to 48 hours.  It was 2:30 am and I went to bed.

The next morning I was right back at it with some new ideas. I tried various things and continued debugging. Overnight and for much of the day, I had a sad little message posted on my blog and forum:

After a few hours working through it all, I checked my email and I had got this message:

Umm. What!?. This is an automated message from WordPress to me. Sure enough, lots of WordPress files were missing on the server.  And there were extra files as well. What I had on my computer which was supposed to be a working copy was different than what was online.

So I used BeyondCompare to mirror the tens of thousands of files on my computer in my blog directory back onto my website at Netfirms. When that completed a half an hour later, my blog appeared and worked fine!

An earlier email from the morning said this:

What had happened earlier that I didn’t realize was that WordPress on my website updated itself to its latest version. That I knew would crash my blog just as would the PHP upgrade. It should have twigged on me that because GenSoftReviews still worked. It couldn’t have been the PHP upgrade and downgrade that caused the problem since that would have affected GenSoftReviews as well.

Phew. Problem solved. But no images were being displayed in my blog. Another whoops. The images were uploaded from my blogging program Open Live Writer. Open Live Writer updates the blog posts into my blog’s MySQL database at Netfirms, but the images are put into the wp-content/upload folder with the WordPress code. I had never thought of syncing those images back to my computer.  So I inadvertently deleted them when I mirrored up my files.

Another support chat with Netfirms and they were able to restore that folder for me from their backup.

By the way, I was very pleased with the Netfirm support chats. There was no waiting and the support person at the other end was very courteous and knowledgeable and helpful!  It was not like this 5 years ago at Netfirms. They have really upped their game impressively.

Upgrade Necessary

I was still being presented with this message::

This is a window I was now getting when I try to go into Admin mode for my blog. Prior to last night, I had never seen this message before. I don’t know what triggered this message to start happening, but I did notice it at some point last night and dismissed it as something I can’t do and not to worry about. 

Maybe I accidentally hit that “Upgrade WordPress link”, or maybe WordPress itself may have detected an error in the plugin when I switched to PHP 7 – I’m not sure which. But something caused Wordpress to merrily start upgrading itself in the background. That’s why the database wouldn’t open. That’s why all the files were different. That might have initiated those emails.

That “Database Upgrade Required” message prevents me from getting into the Admin mode in Wordpress. I tried using the:
      define(‘WP_AUTO_UPDATE_CORE’, false);
directive that is supposed to turn the display of the message off, but it didn’t for me. So instead I just hacked the WordPress code and commented out the calls to the routine:
      wp-admin/includes/upgrade.php

Netfirms is forcing its users to upgrade to PHP 7. As I result I will also have to upgrade WordPress and bbPress. I guess after 12 years of smooth sailing, it’s come to this and I’ll finally have to bite the bullet and update everything.

Sigh! That’s not what I wanted to have to do now. I’ve got updates to both DMT and Behold that I’m working on. But neither of those will be of use if I don’t have a working website to present them.

I’ve got an adventure ahead of me. It will be a lot of work, and a lot of learning, but it should be interesting and fun as well.

 
Followup, Jan 6, 2022: I finally completed the upgrade to PHP 7.4 on my sites, not without a few challenges. See: A Bit of Downtime on my Sites.

In my last post, I discussed a methodology that I could quickly put together an ancestors-only tree for my niece at MyHeritage.

I was able to get back to about 3rd great-grandparents on most of her lines. But it was her mother’s father’s mother’s side that started to get interesting.

My niece’s mother’s father’s mother was Emma Blanche (Smith) Graham (1883-1976). Now you can instantly spot that I’m in for a challenge with a maiden name of Smith. Smith of course is one of the most common surnames there are. So how can I ensure I get the correct John Smith out of two million John Smith’s?

Mayflower?

Following one of Emma’s ancestral lines I assembled at MyHeritage, it led me back through Smiths of Niagara Peninsula (Upper Canada) in the 1800’s to a Wilcox line in the 1700’s that led to Elizabeth Cooke (1641-1715) who was born in Plymouth, Massachusetts. Hmm. Plymouth was where the Mayflower arrived in 1520.

Her father was Jean John Cooke. One of the Record Hints that MyHeritage gave me was this one from WikiTree:

Jean John Cooke was born in Leiden in The Netherlands. Instantly, I recognized that as the city where the passengers on the Mayflower lived before their voyage in 1520. This year is the 500th anniversary of the Mayflower’s arrival! Might the picture WikiTree has for Jean John Cooke be the Mayflower? Could my niece be one of the 35 million Mayflower descendants?

I visited Leiden in 2014 for the Gaenovium Conference. What a beautiful city! And I had the pleasure of meeting and spending time with Tamura Jones, who just happens to be an expert with regards to Mayflower descendants. 

I sent off an email to Tamura asking him if this Jean John Cooke might have been on the Mayflower. Tamura confirmed for me that Francis Cooke was on the Mayflower along with his eldest son John who was a boy a the time. His wife Hester and other children came later.

This Jean John Cooke was the son who was on the Mayflower. Eureka! I can say now that it’s a fact that my niece is a Mayflower descendant, right?

Not so fast. Tamura then told me that he could not find the Wilcox line I supplied him in the lists of descendants he had. He said I should check that line.

So I went to our friend Google and came up with this: I2742: Daniel WILCOX (1631 - 2 Jul 1702) (ksu.edu). It’s from an obviously well researched and sourced genealogy of the Needham Family.

It indicates that Daniel Wilcox (1656 – bef 1730) was the son of Daniel Wilcox (1631 – 1702) and NOT Jean John Cooke’s daughter Elizabeth Cooke, but a previous wife, possibly: Susanna Thompson.

So Daniel Wilcox and his full brother Samuel Wilcox, are not descendants of Elizabeth Cooke and thus not descendants of John Cooke or Frances Cooke.

The extensive references at the bottom of the page talk about this and indicate that “there is no evidence that Elizabeth was the mother of his sons Daniel and Samuel”. 

I immediately scratched out the fiction of Elizabeth Cooke being an ancestor and replaced her with the fact of it being possibly Susanna Thompson.

So much for my niece being a Mayflower descendant, at least on that line.

Churchill?

We did get a not-so-small consolation prize out of it though. If you take a look at that Daniel Wilcox link I have above, at the bottom of the page in the references it states:

The Churchill Centre, "Mayflower Ancestry: For and Against"
http://www.winstonchurchill.org/i4a/pages/index.cfm?pageid=50
"No genealogies have been more carefully prepared, or reach a higher standard than, the Mayflower Society genealogies. There is solid evidence that Daniel Wilcox married a first wife prior to his marriage to Elizabeth Cooke, granddaughter of Francis Cooke. There is no evidence that Elizabeth was the mother of his sons Daniel (Churchill’s ancestor) and Samuel. There is circumstantial evidence that she was not. In genealogy, absence of evidence means absence of conclusions."

Checking out Sir Winston Churchill’s ancestry, he does in-fact connect to Emma (Smith) Graham’s line.

Sir Winston was in fact a 5th cousin of my niece’s great-grandmother, making my niece a 5C3R (5th cousin, 3 times removed) to the British Prime Minister.

Just the Facts

The Needham Family site is a fantastic resource. You can see the numerous references at the bottom of each individual. It would take years to redo that work.

So I decided to go through his site and cross reference the ancestors I had collected and change any information I had to what he had. As I did that, Needham pointed me to another excellent study that was of Benjamin Wilcox by John Blythe Dobson, and I cross referenced and changed my information for my people from that study as well.

Of the 129 ancestors I had found for Emma (Smith) Graham, Needham had information on 84 of them, and Dobson had 32 of them.

I put the information in my spreadsheet so that I could quickly visualize and access the information at Needham and Dobson’s sites:

Notice the people in orange. They were fiction I obtained from other people’s genealogy.

Dobson stated:

"We know of no basis for the recent claim that she was a Sarah Hart or Hort, b 16 Apr 1684 at Dartmouth, daughter of Thoas Hart or Hort and Margaret. Not only is any such person absent from the town’s vital records, but …". Needham states "Some claim she was Sarah Hort, daughter of Thomas Hort. I have seen no definitive proof of this claim."

which negated the Hort name and Sarah’s parents and grandparents.

And Needham only gives Susanna Swift as “Susanna” with a 1612 birth date, not the 1622 that I had. So the Susanna that married Ralph Allen, likely wasn’t Susanna Swift. So scratch her parents. Needham also didn’t give a surname for Rachel Sherman, so I removed that as well.

There could, of course, be later scholarly research that updates what Dobson or Needham have found, but I’d like to see it with extensive references that can be followed before I’ll believe it.

Prime Minister?

Notice the Borden ancestors in the spreadsheet above. Needham pointed me a site with the Descendants of Richard Borden. That site pointed me to information about Sir Robert Borden (1854-1937), who happened to be the 8th Prime Minister of Canada.

So now I can assuredly add this Prime Minister as well to my niece’s cousin list. He would have been her 7C5R (7th cousin, 5 times removed).

Seaver?

One other connection I managed to make. While searching to verify the fiction or fact of a “Thomas Bloomfield” ancestor, I came across Amanuensis Monday - Post #286: 1684 Will of Thomas Bloomfield (1615-1686) of Woodbridge, N.J. by the incredible genealogist Randy Seaver on his Genea-Musings blog.

Randy’s genealogical work is also of the gold standard that I would 100% trust.

Searching his site for more information, I found his page Genea-Musings: Surname Saturday - BLOOMFIELD (England > colonial Massachusetts > New Jersey) and from that page I was able to tell that Thomas Bloomfield was Randy’s 10th great-grandfather.

He’s also my niece’s 10th great-grandfather. So that makes Randy and my niece 11th cousins.

Others?

I’m sure there will be more connections that will come up for my niece. Once a genealogy gets back this far to Colonial America and England, there’s much more to be found.

These first discoveries are exciting for me. My own genealogy by comparison heads back to Romania and Ukraine in the early 1900’s, so I’ve never really got to experience these sorts of family connections the way so many other genealogists do.

And I feel much better knowing that these connections are not fiction, but fact.

A few months ago, I decided it would be interesting to investigate my niece’s genealogy.

I entered information about her parents and grandparents into a new tree at MyHeritage and let MyHeritage’s Record Hints and Smart Matches go to work. It wasn’t too long before I had a couple of hundred people representing a lot of her relatives out to about 2nd cousins.

There were a few enticing hints on her mother’s father’s side that attracted me. I started following them and they’d each add another 30 relatives and take me back another generation. In very little time, MyHeritage had 80 source suggestions for me with over 1000 matches and every time I analyzed and confirmed a suggestion, the numbers would continue to grow.

I’m the type of person who likes a clean email in-box, as well as a clean MyHeritage hint list. I saw that I’d soon have thousands or even tens of thousands of people in my niece’s tree, and I really didn’t have the time to build that and resolve the multitude of hints that would result.

So I took a step back and thought about what I was doing. My niece had a few lines that were going back to Southern Ontario in the early 1800’s, and they got there from New Jersey where they were in the 1700’s and previously England. Now I’m in the realms of early America, and I realized these are genealogies that many genealogists connect to. A lot of research has already been done on these individuals and the same people are in the trees of many genealogists.

I had never had this problem with my own genealogy, since all my ancestors arrived in North America between 1900 and 1930 and I’ve never had to deal with early America and English roots. That really looks so interesting.

But I concluded it would be a waste of my time to redo all the research others have done. It’s all out there already. I just needed some way to connect to those people.

Pedigree-Only

I decided what was best was, after allowing inclusion of my niece’s relatives out to her 2nd cousins, would be to only add her direct ancestors to the tree.

At MyHeritage, that ended up being fairly simple to do. I checked the Smart Matches by person and found those that gave additional spouse or parent information for any of the direct ancestors.

Matthew Borden is one of my niece’s lines that traced back to England. I currently have that his wife’s name is Joan, but I don’t have her maiden name. Getting her maiden name, birth date and death date and place might lead to finding her parents.

MyHeritage has 78 Smart Matches for Matthew, and the Smart Match summary indicates that there is new spouse information.

Clicking on the "Review 78 matches” button reveals the 78 matches with other people’s trees at MyHeritage ordered first by the ones that provide the most additional information:

I check to make sure that my birth year and death year and places match, that the wife is listed as a partner, and that the son is listed as among the children. Since I am doing just the pedigree, I only have one child included.

Then I take a look at the partners, and see if the partners listed for the 78 matches all agree.  This is what I found:

• Joan Reeder (51 times)
• Joan Glover Reeder (9 times)
• Joan Mary Reeder (1 time)
• Joan (6 times)
• no partners listed (11 times), but 5 of these listed Joan Reeder as Matthew’s mother (presumably incorrect)

So they all seem to agree that the maiden name is likely Reeder. The Glover was sometimes in parenthesis, so I’m thinking it might the surname of her spouse of a previous or later marriage. With only 1 entry of a middle name of Mary, I’m not going to believe that until I get further evidence.

If I would have had conflicts here, then I would have gone off to our good friend Mr. Google and see if there’s something on her. I could look up “Matthew Borden” “Joan Reeder” and see what pops up. I’d be looking, not for a genealogy containing the two people, because I’ve already got plenty of those from MyHeritage’s Smart Matches, but for some scholarly work detailing the ancestry WITH SOURCES!!  The sources will show that it was detailed first-person research. Another family tree is NOT considered a source.

Here’s what Google gives:

The third link has nice information, specifically that source at the bottom:

So what I’ll now do now is Review the first Smart Match of Matthew Borden, confirm it, and mark the data in my tree I want to update. 

I may want to also update Matthew’s information if it’s better (e.g. birth and death date and place, occupation), and add his parent’s information if I don’t have it.

I was originally going to use Matthew’s parents as the example in this post, showing how conflicting information can be resolved, but this case is more complicated and his mother is still an unknown with several different people possible, so I chose to use Matthew’s wife instead.

The only people I will accept information for is Matthew, his wife, his parents and possibly his son Richard who is the direct line ancestor we are interested in. I will not select information about Matthew’s other children or siblings. By not selecting that information for inclusion, those other people will not be added to my niece’s MyHeritage tree and it will remain pedigree-only at this generational level.

I’ll then go down to the bottom of the Smart Match. I’ll be sure NOT to press the “Extract all info” button, but just click “Save to tree”.

Now I’ll go back to Matthew Borden’s Smart Matches. I’ll check the next few to see if they might have useful additional information. Once I’ve got most of the information I want, I’ll go up to the “More actions” dropdown and tell MyHeritage to confirm all the remaining matches.

However, I will not save any of the confirmed information to my tree. The confirmation simply makes them available in my profile of Matthew Borden, and removes them from my list of Smart Matches so that I can now concentrate on the next person.

I then simply continue this same procedure for each of the Smart Matches until I’ve exhausted them.

I’ll report on the results of this in an upcoming post.