Access Violation First Thing in the Morning

I am now in what I’m calling parallel production on RM9, using RM9 first for any new work and then replicating the work in RM7 by hand. This is my third attempt to be in production on RM9. I abandoned the first two attempts but I think I might be able to make it work this time,

However, I left both RM7 and RM9 open over night last night. The first thing I did this morning in RM9 was to click the Home tab so that I could click on the RootsMagic Community to check on any overnight activity. There was an Access Violation immediately upon clicking the Home tab.

I shut down RM9, re-opened it, and ran all the database tools All seems well. However, it is disconcerting that leaving it sit overnight doing nothing seems to have generated an Access Violation.

It also raised a question. The official advice from RM on running the database tools in RM7 was to run them in order. But they are listed in a different order in RM9. Or at least they are listed in a different order under the Tools tab. They are still listed in the same order as RM7 under File => Tools tab. Does the order of running the database tools still matter and does the order that the tools are listed under that main Tools tab matter?

I too attempted to move to RM9 from FTM 2019 for real but ran into random access violations on mac after they seemed resolved months ago. Everything I did required far more clicks and screens than RM7 and especially FTM and seemed to end up in AVs or a locked up program. The great RM9 text size, find media and web search features do not overcome the very clumsy work process and many fatal bugs for me. FTM just works and displays my data better.

Jerry, did your OneDrive start to sync overnight while you were away from your computer?

That’s a really interesting ;question.

OneDrive can be paused for 2 hours, 8 hours, or 24 hours. It’s not like DropBox where you can pause it and it stays paused until you unpause. So you might wonder if maybe I paused OneDrive at 11:00 p.m. for two hours and went to bed. And then maybe it unpaused itself at 1:00 a.m. while I was sleeping and caused a problem.

Except that I basically never pause OneDrive because unlike Dropbox it doesn’t try to sync an SQLite database while the database is in use. So OneDrive didn’t start to sync overnight while I was sleeping. It was already syncing, except that it doesn’t sync SQLite databases which are in use.

Nevertheless and out of an abundance of caution, I’m going to try the following experiment. I’m going to leave RM9 up and running each night. For half the nights, I will not pause OneDrive when I leave the computer. For the other half of the nights, I will pause OneDrive for 24 hours when I leave the computer. Then each morning I will click RM9’s Home icon first thing in the morning to see what happens.

Then why keep using RM?

The Integrity and Compact tools are purely sqlite statements, respectively PRAGMA INTEGRITY_CHECK and VACUUM (iirc). And Rebuild Indexes is sqlite REINDEX with an extra app generated set of queries to update Birth Year and Death Year for everyone from their respective primary event date. Delete Phantoms is purely app queries.

The order in which the tools are executed doesn’t matter for a healthy database but might for one with problems. So I would suggest starting with the integrity test first followed by Rebuild Indexes. If the test failed, then you will want to rerun it after reindexing to confirm that it is now ok.

You might argue that one might as well rebuild indexes first and then test integrity. The counter is that then you may never know if there had been an integrity problem in a previous session.

Another question is whether to skip rebuilding indexes if the integrity test is ok because it takes a long time on large databases. If you’ve no concern about the Birth and Death years being out of sync with their respective facts, then there’s no need to REINDEX after an ok test.

Compacting the database rarely makes much difference. Maybe after very extensive deletions and merging in a large database. Or when the integrity test still fails after a Rebuild Indexes but that’s grasping at straws. I’d still advise doing the test before compacting but it is a waste of time to compact frequently.

Delete Phantoms is a sort of batch cleanup of detritus from the app because it did not clean up properly from some successful operation or it failed to complete an operation that would have otherwise cleaned up properly. (Note that RM hss incorporated no referential integrity controls available within sqlite.) Safest to do after an ok integrity check. However, I don’t think the app provides any reporting on the existence of phantoms nor does the tool address all possible types of phantoms. So you really have no way of knowing if you have phantoms before or after other than by observation. Viz the recent discussion about the ubiquitous Elizabeth Cutler.

  1. It’s the only software I’m aware of that will produce the reports I need for family reunions. This includes NGSQ and NEHGS formats. This includes an excellent index of names and an excellent index of places, with the indexes being separate. This includes excellent support for sentence templates. This includes being able to jury rig the database on a case by case basis to force selected people who don’t have a spouse or children to be carried forward into the next generation. This includes being able to post process reports with Microsoft Word before printing.
  2. It’s one of few pieces of software that I’m aware of that supports hints both for Ancestry and FamilySearch, with a direct interface into both. I could be wrong, but I think there is only one other piece of software that supports both Ancestry and FamilySearch in this fashion.

Note that @Rwcrooks asked that question of Rooty, not of you. Rooty has done nothing but claim from the beginning because RM just doesn’t work on his Mac, even when in one instance it wasn’t a RM problem, it was his internet connection. He also is singing the competitor’s praised for be useable.

RW: As my comment stated I am no longer attempting to use RM9.

Jerry: only RM and FTM get hints and sync with ancestry at least on mac. Heredis and MFT lack this crucial feature and also add all media to the database file which then becomes huge. Right now FTM is in month 3 of no sync to ancestry.

RM does provide nicely formatted NGSQ and NEHGS reports, but I find that the people I supply my reports to couldn’t care less about which standards they adhere to. They aren’t professional genealogists and are happy to get any sort of readable report. Plus it’s been a while since my last submission to the NGSQ and I don’t see me submitting again.

And the indices are nice, and don’t seem like they’d be too hard for competing products to add. I wonder why they haven’t already done so.

That’s probably true, but I care - and for a couple of reasons.

I have never actually been able to find any document that describes the standards for formatting NGSQ and NEHGS reports. By formatting, I mean the indentation of each line of the report and when the report goes to a new line or to a new paragraph. All I have been able to find is the standards for numbering people in such reports. In any case, one key thing I like about the NGSQ and NEHGS numbering standard for such reports is that they require that most individuals be listed twice - once as a child in a family grouping with their parents and again in a family grouping with their spouse and children.

By contrast, RM’s implementation of the outline type formats separates the children in the furthest back generations in the report from their parents and siblings because each person is listed only once. I have seen old compiled families histories that obviously were prepared on a manual typewriter that used an outline type of numbering, and nevertheless the compiler included a list of children with each set of parents. I would find this kind of formatting and numbering to be completely acceptable. It would probably even be preferable to the NGSQ and NEHGS reports when the audience is a family reunion. That’s because the general public will understand outline type of numbering much better than the NGSQ or NEHGS type of numbering.

I don’t think RM’s current implementation of outline type reports is acceptable for family reunions because of the lack of lists of children for each family. I think this could and should be added as an option to RM’s outline type reports. If so, I would probably use the outline format instead of the NGSQ or NEHGS formats. So I’m not hung up on the NGSQ and NEHGS reports, per se. I just want each set of parents to have a list of children, and the NGSQ and NHGS reports are the only ones supported by RM that support such lists of children.

The other reason is that the NGSQ and NEHGS numbering does not carry children forward to the next generation unless the child has a spouse or children of its own. For the most part, this is desirable behavior for reports for a family reunion. That’s because a family reunion report includes living people and everybody wants to be sure their newest grandbabies are in the report. Carrying everybody forward to the next generation including the newest grandbabies would result in many more pages of paper, making the reports thicker and more costly for me to produce. I probably wouldn’t care as much if I were distributing reports electronically, like giving everybody at the reunion as USB memory stick containing a PDF of the report. But I’m persuaded that 50 years from now, my printed reports will survive in a way that electronic reports will not. And people always thank me for producing the reports on paper.

I do have a trick I use to force a few selected individuals forward to the next generation in RM’s NGSQ and NEHGS reports despite not having spouses or children when the selected individuals have a lot of facts or notes. I wish RM had a feature to support this behavior on a case by case basis so I wouldn’t have to use a trick. But it remains the case that for most individuals who do not have spouses or children, I do not want them carried forward into the next generation in printed reports for a family reunion.

That’s a very good question. I think RM’s indices in printed reports are a very important and perhaps unappreciated feature of the RM product line. All the way back to Family Origins, RM has always done the report indices very well and usually much better than their competition.

…and yet we keep hearing about your difficulties.