Family history and AI

 Every since Julie and Anna at the Library helped me track down some information on Fanny Elizabeth Bull, I've been going along to Julie's family history sessions, and very valuable the have been too.

Now, you might have an image of family history people as slightly weird older people wandering round graveyards, scratching down notes, and then annoying the hell out of people in registry offices and then spending the evening with a cup of tea drawing out part of a family tree with a cat and a scratch pad.

And certainly, if you watch cosy English detective shows on TV, that's very much how they are protrayed.

well, as you can see, the cat part is still true, but the rest of it is so very last century.

Digitisation, the rise of the family history behemoths, and expansion of various hobbyist sites such as Find a Grave (now owned by Ancestry) has turned what may once have been a fairly genteel hobby into something quite hi-tech.

There may still be people out there with nothing more than an A4 notebook and a box of HB pencils, but they are most definitely a minority.

So today we discussed such topics as the strengths and weaknesses of various online family tree builders, using local software such as Gramps, as opposed to the online solutions and the use of GEDCOM as a data interchange format.

We also touched on the 23andMe debacle, and on the use of genetic testing databases by various law enforcement authorities, all of which I was reasonably familiar with and could contribute to the discussion.

And then we moved on to the use of AI - and I found myself immediately out of my depth.

It's been two or three years since I did a significant amount of family history work, and that which I have done has been done using old school online techniques, such as when I was trying to trace Irene Hogg.

It turns out that most of the more committed family historians in the group have been making extensive use of AI to summarise newspaper reports, help transcribe records, and refine more complex queries.

The use of it in genealogical research is quite fascinating and obviously something I need to get up to speed on and to that end I've started a wiki page  as a place to dump links that I find relevant or useful ...

Finished!

I've completed my project to document the contents of Lake View House in Chiltern for the National Trust.

It's quite good to have finished roughly on time as I've avoided having to work into early winter when working in a building with no heating and no dampcourse to speak of can be a bit grim, not to mention the potential for moulds and other nasties.

Yes, of course you can rug up, and over the years I've posted various selfies of me working in thermals and a beanie, but it's more than a bit grim when you have to resort to fingerless gloves to type when working.

However, I enjoyed the experience and had quite a bit of fun along the way tracking down information about the various artefacts.

Not sure what my next project, if any, will be, but for the moment I'll concentrate on tidying up and systemizing my pandemic era family history notes, and perhaps finish off Katherine Scragg, I've still got to definitively trace where life took her after her assault ... 

Iceland, volcanoes and the fall of the Roman Empire

 Today brings news that there is evidence from Iceland and Greenland of increased volcanic activity around the time of the plague of Justinian, that might have helped push the successor polities in the west over the edge and also bring to an end Byzantine attempts to reconquer Italy and parts of Spain and North Africa. (The original Southampton University press release is available online if you want to follow up further.)

Basically the idea is that the increased dust in the atmosphere - a sort of quasi nuclear winter - would have caused crop failures as it did in 1816. This, combined with a population debilitated by disease would have left the various city states and other polities that emerged out of the end of the western empire dangerously weak.

Of course, this would not be the whole story - Roman civil administration had effectively collapsed over much of western Europe by the 450 as a result of the invasions of Germanic speaking peoples, but in a lot of cases there had been a melding in which the daughters of Roman administrators married the sons of Germanic warlords (and vice versa) to form a new governing class resulting in various kingdoms, such as in the case of the Merovingians in Gaul or the Gothic kingdoms in Italy.

However, as I blogged back in 2010, there's also a hypothesis that the global dimming event was not caused by volcanic activity alone but by meteorites hitting the Gulf of Carpentaria and an unidentified location to the north of Norway.

Of course, both theories may be true - the meteorite strike may have triggered volcanic activity, especially as both impact sites are not that distant from areas of high volcanic activity (Iceland and Indonesia).

 However, it's an intriguing idea, be it simply through volcanic activity, or something more dramatic like a large scale meteor strike...

Mrs Potts and garden archaeology

 A few years ago, when working in the garden, I turned up an old flat iron base

Heavily corroded and covered with clag, I put it to one side on a pile of bricks beside the shed and left it.

Yesterday, down at Lake View, I documented some Mrs Potts irons, which were a late nineteenth century innovation where you had a set of three flat irons with a detachable handle, the idea being that you left the bases on the kitchen range to heat up, fitted the handle to one, did your ironing and swapped the bases in turn as they grew cold, and reheated the irons until you were done.

Mrs Potts irons were quite common and turn up all over rural Australia, so it's no surprise that there's an example at Lake View

One of the bases looks like this (the picture's failed QA and will have to be rephotographed), but it has this quite sharp profile, three holes and "Mrs Potts Iron" embossed on it.

Now the flat iron I dug up a few years ago has a similar profile so I went out into the yard and had a second look

The last few years of being out in the rain and frost have freed some of the dirt and clag and quite clearly the profile is similar to a Mrs Potts base and importantly has three protrusions where the handle attachment fits on a Potts iron, and while you can't see this I could kid myself I can make out 'TS' in the correct place on the iron amid the corrosion.

So, tentatively, I'd say it was a Mrs Potts iron base...

Technology and the road trip

 As I've written elsewhere, we're just back from a road trip to Port Lincoln in South Australia.

Technologywise, we took more or less the same as we took to Tasmania  last year - I took the HP ProBook which continues to function well as a travel computer, supplemented by the little Lenovo M8 tablet to read the news in the morning, and my second hand Canon Powershot.

I did take a film camera with me as well but it ended up staying in the bag for the whole trip.

Everywhere we stayed provided internet, but of variable quality - for example in Glenelg the internet in the short stay apartment was quite slow, and we ended up using my 4G travel modem, especially as J had to make a zoom call one morning.

In other places, for example in our rental cottage in Goolwa it was blisteringly fast.

In a couple of places, motels used these damn captive portal solutions where you have to jump through hoops to log on, and we simply used the 4G modem in preference as it provided reasonable speed for minimal hassle ...

Landfill and Windows 10

 Back in February I was fairly scathing about a news report that suggested that the end of support for Windows 10 would result in a tide of older machines incapable of running Windows 11 being dumped in landfill.

I still don't think that will be the case as regards home users and most corporates have made the jump already - most corporates replace machines every three or four years - but it is interesting that my favourite supplier of cheap refurbished laptops is running a promotion on cheap sub $300 laptops and almost all the machines are Windows 10 machines - most of their more expensive offerings are Windows 11 based suggesting that some corporate are already into their first Windows 11 refresh.

It also suggests that most of the buyers of refurbished machines are already on the lookout for a Windows 11 machine as a second machine, which does tend to suggest they are not installing linux on them.

And yet linux does have an advantage in that it is not the product of an American megacorp, and in these days of uncertainty about the USA and its role in the world, there may increasingly be people who would prefer their computer operating system to be independent of the megacorps.

And it's not just the paranoid - there are cases of various governments moving to linux, or at least experimenting with it as an option.

Perhaps and just perhaps, and given some people's dislike for AI in everything, the end of Windows 10 might nudge the uptake of linux on the desktop along a bit ...

Documenting artefacts - a tweak to the mechanicals

 For the last eight or nine years I've been volunteering as a cataloguer for the National Trust of Victoria, first documenting the contents of Dow's pharmacy in Chiltern, and then Lake View House, also in Chiltern.

And the methodology I use has not changed much from the method I first described back in 2017.

There were some changes along the way, such as the use of One Drive directly rather than uploading the data from a USB stick at the end of the day but the procedure at the end of the project was still essentially the same, as is the procedure used at Lake View, except that the data is entered into a spreadsheet designed to ease the ingest of the data into the Trust's digital asset management system - so while the way I structure the data may have changed, the procedure is basically the same starting with a descriptive entry written in longhand in an A4 notebook which is transcribed into the appropriate columns of the data entry spreadsheet.

As a procedure it's robust and it works well - don't knock it, and longhand entries in notebooks have a certain permanence.

But yesterday, I tried a slightly different procedure.

I was working on the contents of the kitchen, which is a detached brick building separate to the main house. The kitchen lacks power sockets, and lacks any free surface on which to write, the available surfaces being covered in exhibits, and I didn't want to have to carry artefacts from the kitchen to the main house to document them one by one.

However, we have plastic chairs in store for outside presentations etc and a large brick floor area suitable for photographing artefacts on against a white sheet (and equally if I needed to use my lightbox, it can be run off a powerbank).

So, I thought, why not take a second laptop in, sit on one of the plastic chairs, and create a basic spreadsheet not unlike the original ones I made up at Dow's and then cut and paste the data as required into the bigger and more formal spreadsheet.

To do this I used my linux based IdeaPad 1 and created a base spreadsheet using Libre Office Calc

There's nothing magical about using Linux here, you could equally well use a Mac or a Chromebook, or even a second windows device.

I settled on the IdeaPad as it has decent battery life and can drive my endoscope if required to capture a makers name or serial number on an artefact. I did think about using my lightweight research machine but the lack of power sockets in the kitchen worked against it - it only manages two to three hours on a full charge, and can be slow to recharge. The IdeaPad usually manages a little over four hours on a charge and recharges fairly quickly meaning it can be charged over lunch for an afternoon session.

Likewise there's nothing special about using Libre Office Calc, Numbers or a Mac or Google sheets on a ChromeBook, the only special thing I did was save the spreadsheet in a recent Excel 365 format.

After documenting about ten or twelve artefacts - which roughly takes an hour I would upload the spreadsheet to a scratch work area on One Drive via the web interface.

Documentation sessions are separated on the worksheet by blank lines to make sure that we are copying the correct data - it would be a bit of a problem if objects were accidentally duplicated.

I would then go across to the main house and my desk, and on my main work computer - a Thinkpad running Windows 10 cut and paste data from the scratch spreadsheet to the main sheet, doing any editing and restructuring required along the way, and add in the photo data.

While it sounds a bit of a pfaff, I found that not having to retype the data entries from my longhand notes did speed the process up a bit.

Obviously I could have simply uploaded the data and taken it home to work on, but I did want to check the quality of the photographs as I went, and perhaps do a little basic photo editing to make a makers' name or feature clearer.

All in all, I think using a second machine like this was a win, even if the procedure sounds a little clumsy...