The efficiency of the colonial postal services

 Over on one of my other blogs, I mentioned that I'd acquired another example of a colonial period postcard.

It's one issued by the postal service in South Australia meaning it would have been posted there - before Federation the colonies ran their own postal services.

Strangely there's no South Australian post mark but there is one for Melbourne for the 31st of March 1898.

Flipping the card over

we see that the message is dated for the 30th of March, suggesting that the Melbourne post mark shows when it was received by the Victorian postal service, a day after it was written in South Australia and suggesting that the card was sent overnight by rail between Adelaide and Melbourne.

Nowadays, a letter or package takes two or three days to travel between the two, but it's worth pointing out that until about fifteen years ago, a letter to Sydney from Canberra would almost certainly arrive the following day, and a letter to Melbourne might just make next day delivery but would definitely be there the second day after posting.

Even more amazingly, an airmail letter posted on a Monday morning in Canberra to an address in rural Scotland would reliably arrive on the Friday.

Not only does this demonstrate how valuable a fast and efficient postal service was to people in the nineteenth century, when the post was the default means of communication, but also that the marked decline in delivery speeds has only happened once electronic communications became the default ...

Another damned camera ...

 When we went to Tasmania last year I made myself up a little retro photography kit consisting of a half frame camera and my big old Praktica MTL-3.

Well the Praktica didn’t have the happiest trip and I was going to swap the lens to another second hand Praktica body I have, but somehow I never quite got around to swapping the lens as planned, and my retro photography kit has languished on the top shelf of my book case.

And, it has to be said, that while my little dabble with film photography has been useful – particularly the discipline of trying to compose the shot – I really havn’t seriously got into the retro photography thing. Definitely a dilletantte.

So, today I swapped out my MTL-3 with a little second hand Canon Powershot digital camera.

When I brought an old Lumix out of retirement to use at Lake View I rediscovered the advantages of a lightweight digital SLR – while I’ve an old Fuji Finepix which I use for taking high resolution images for digitisation work, in practice it’s just a little too heavy and bulky to double up as a general use camera.

It’s just that little bit too heavy to carry round in a day pack, which is why I switched to a little point and shoot Ixus, which ideal for street scenes and the like, but has the problem that in the bright light of the Australian summer the display panel is drowned out by the sun and you often end up taking the picture blind.

The obvious solution would be to use a camera with a viewfinder, such as the revived Lumix, but being an old camera, the Lumix doesn’t produce particularly high resolution images, and while they're good enough for artefact documentation, they are not so good if you need to crop and fiddle with them a bit to extract part of the image.

I’m hoping that the Canon, being relatively lightweight, and with a higher resolution image collector will bridge the gap between the two, and let me do some reasonable photography ...

Dating UK railway ephemera

 What a confusing mess provenancing railway ephemera is!

Back in the middle of last year I had some fun with Victorian Railways luggage labels, and I've just been down a similar rabbit hole with British GWR labels.

Just as in Australia, there's no real way of dating railway luggage labels, except to say that ones with serifed fonts are generally older than those with sans serif fonts.

Now if the luggage label was an LNER label, we could say that it was post 1923 as railway companies were grouped into four entities after the first world war, and the LNER dates from the 1923 groupings.

Likewise, if the label said London and South Western Railway we could say the label was pre 1923 as the LSWR became part of the new Southern Railway in 1923.

All good, except in the case of the GWR, instead of merging it into a new entity to serve Wales and the west of England, they merged the smaller companies in the area into the existing GWR.

So, if you had a Cambrian Railways label for Builth Wells, it would be pre the 1923 grouping, as post 1923 Cambrian Railways had been merged into the GWR and any luggage label would have said GWR.

However, none of this is any use to us for dating purposes as the GWR reached Oxford in 1844 and continued to serve Oxford until it was nationalised in 1948.

So we're thrown back on hand waving stylistic arguments.

Just like the labels on medicine bottles, the first labels were almost exclusively letterpress.

And we could say that if they were for an obscure country station such as Abbeydore, we could guess that they would print a batch of labels and keep on using them until they ran out, possibly forty or fifty years after the station opened.

It would be possible to get a mixture of styles, because when Abbeydore opened, they would have sent labels for Abbeydore to all their stations where you could check luggage.

Some stations would have got through their labels faster than others, and others could conceivably still be using ones from Queen Victoria's time in the 1930s.

In the case of Oxford, that would not have been the case.

Due to the university it would always have been a busy station handling lots of luggage and consequently they would need to reorder luggage labels for Oxford every two or three years, meaning that while you might get a mix of styles they are all likely to have been printed within ten years of each other.

If we look at the label, we see it's printed using a rather stylish sans serif font and looks as if it might have been printed using lithography.

J W Tyler, the original owner of the hat box died in 1909.

To me, at least, the fonts used do not look to be like those used in the Victorian or Edwardian periods. Sans Serif fonts, especially stylish ones with thickened strokes look much more like you would expect in the 1920's or 1930's, so I'm guessing the hat box was used by someone travelling to Oxford some time between the first and second world wars ...

Microsoft's AI price hike

 If, like me, you have a personal Microsoft 365 subscription you'll have received an email telling you that the cost of your annual subscription was going from A$109 to A$159.

(Actually I didn't get the email, not right away, it was only when J told me about her price increase email I went looking and found that Outlook - bless its little cotton socks - had classified the email as junk - go figure.)

Well, I hissed and swore, but I had expected an increase in part to our dollar doing one of its periodic nosedives in value. I did think an A$50 increase a bit steep, but I guessed that Microsoft had been listening to some of these catastrophising exchange rate analysts - the same ones that told us last year that our dollar would go up in value - and decided to hedge its bets, and I must admit I didn't really read the email properly.

Now, I need Microsoft 365.

I use One Note extensively - while I used to use Evernote to manage my research material I've moved over to One Note as a note management tool, and the 1TB of storage comes in handy given I've a vast number of photographs of artefacts.

And I've found that while Libre Office will happily do 95% of everything you can reasonably do with Office, it doesn't cope that well with weird macro dominated data recording spreadsheets and grant application forms using equally weird templates and mandating strange fonts.

So, Microsoft 365 it is - and that's why both my desk and work laptop use windows despite my liking for both linux and parsimonious solutions to recording data in the field

But when I read the email carefully, I saw that part of the increase was to pay for the AI features Microsoft was building into its applications.

Now, I'm a cynic about AI, and a lot of it seems quite immature and singularly useless.

It certainly doesn't need to be everywhere, and being offered the option to generate a summary of my credit card statement or an email about a nineteenth century coronial inquest is not helpful.

However, I will admit that while the little search summaries sometimes generated by Google's Gemini a few months ago were useless, recently, they've improved, and when researching nineteenth century medical instrument manufacturers they've been reasonably useful, providing you read critically and are alert for the odd howler.

I havn't used Microsoft's equivalent product, but I have no reason to believe their products have not followed a similar trajectory, and may prove useful to me in what I do, which does involve quite a bit of web based research.

So, while I'll hiss and groan, I'll put up with it.

I suspect that most users of Office don't really need the brave new world of AI, but then it's always been the case that 90% of Office users only use 10% or the features - the problem has been guessing which features are the key 10% - so no change there.

What does worry me are the  adverts appearing on TV (it'll soon be the start of the school year here in Australia) about how CoPilot and CoPilot based PC's will help students power through assignments, and in one advert, the actor playing a uni student is shown asking CoPilot to cite her sources.

I'm sorry, if you want to do research oriented work, you need to learn to do research, learn to critically read your sources, and weigh up their validity, in short you need to learn to think for yourself.

AI is a tool, not a panacea.

 And as for the students and their clever PCs, I keep on thinking about when I learned to sail yachts and they made us learn to do some very basic astronomical and solar navigation, because, as the course tutor said, when you're out on the ocean and the electrical system has died taking out the GPS, how are you going to find your way home?

The same goes for outsourcing your thinking to a machine ...