This morning I tooted that our internet speed had jumped to about half a gig, something that is quite amazing in terms of infrastructure for rural Victoria. Admittedly it's only that fast on download, upload speeds are still comparatively slow
Wednesday, 1 October 2025
Of internet speeds past
Sunday, 14 September 2025
Using Acrobat's AI summaries with Trove
In my little bits of nineteenth century historical research I use digitised newspaper resources a lot.
The various digitised resources I use most often are nineteenth century Scottish newspapers via the SLV's subscription to Gale Newsvault for family history stuff, The Times of London's archives again through the SLV, Welsh Newspapers Online, Papers Past NZ, and above all, the NLA's Trove.
Trove is undoubtedly a great resource, but the quality of the digitised text, to put it politely, is variable.
Trove does provide OCR's summaries of the articles, but the quality of the digitised text can make the OCR'd text read as if it had been transcribed by a Martian - strange combinations of letters and punctuation followed by gobbets of reasonable text.
So, for years, what I have done is use the download option to generate a pdf, download the pdf to an ipad, and then sit and make notes on a 'proper' computer.
Latterly, if the pdf is too hard on the human eyeball, I've used J's old iMac, which now runs Linux, and Okular to give me a bigger image at a decent resolution to work with, and that's worked pretty well as a workflow.
Now, as I'm sure you're aware if you're an Acrobat user, Acrobat now behaves like an enthusiastic puppy, always asking if you want it to generate a summary of the document.
I've tended to ignore it, really because most of the PDF documents I look at on windows are boring things like credit card and electricity account statements, and there's usually only two important bits of information - how much we owe and when is payment due.
But instead of doing the majority of my work on a linux machine as I usually do, I researched the Panjdeh incident on my Windows machine, and typed my notes into Geany on the old Chromebook I installed Linux on, really as a way of assessing the usefulness of the converted Chromebook.
(Answer, very useful, and good battery life to boot).
Anyway, as I was working on Windows, Acrobat came along wagging its little tail, offering to generate a summary of every pdf document I opened.
So, for a number of longer documents, including some with poor quality OCR'd text, I did.
And they were surprisingly good, and the AI summary engine seemed to deal reasonably well with poorer quality scanned text, producing reasonable and good quality précis of the article texts.
Obviously you need to check the text yourself, but using AI text summaries turned out to be a useful way of assessing if the article was worth reading, it's not the first time I've slogged through a report of court proceedings to find that the report didn't add anything to what I already knew.
It's by no means a panacea, but it's certainly a valuable tool...
Thursday, 11 September 2025
What happens to our photographs when we die?
An interesting little question popped into my head - what happens to our digital photographs when we die?
Of course we've all wrung our hands about how letters and postcards have been replaced by email meaning that future generations have lost access to our correspondence, denying cultural historians access to sources that describe how people felt about things, but unless I'm very much mistaken, people's digital photographs have not really been thought about.
For example, and this shows the value of sometimes inconsequential seeming objects,I recently picked up a British World War One propaganda postcard from a postcard trading site. Transcribing it turned out to be interesting, with its hint of war weariness among the population as well as worries over the risk of German air raids.
Interesting, and something that one couldn't do about a contemporary conflict, such as that in Ukraine, because all the communication involved would be digital, and I don't see people collecting 100 year old WhatsApp messages they way they used to hang onto (and collect) old postcards.
Now obviously, one doesn't want to keep everything. Broadly speaking, there are two sorts of photographs in people's collections - the transitory and the significant.
The transitory are images like the cracked tail light on a rental car - you photograph it to show it was pre-existing damage, or the back of a wi-fi router to record the password.
Then there's the significant - examples being all my artefact photographs for the National Trust, photographs of old buildings, J's records of her artworks, and so on.
Once they would have been boxes of 35mm slides, and now they exist on a server somewhere.
And of course not everything physical survives - my geeky teenage photographs of closed railway stations in Scotland have gone to landfill in the course of various moves and relocations, along with pictures of former girlfriends, camping trips and the like.
Some of these may have had some value, some not.
And so with digital images, some have significance, for example some of my Trust photographs show the state of decay for some artefacts, and might be of value to future conservators, etc.
And obviously some work has been preserved - for example I know that some of my artefact photographs have been archived, but not all of them, and of course I don't know which ones.
And increasingly there is a problem.
People's collections of potentially archivable material are changing - emails have replaced paper, digital photographs have replaced analogue film, etc etc.
And of course, there's also the problem of obsolete media - recordings on cassette tape, video tapes and the rest, plus if they were digitised, where did the digitised version end up, and how is it preserved?
Answers on a postcard?
Saturday, 6 September 2025
Multi factor authentication and the outback
Australia is a big, really big, sprawling country, and as a consequence there's a lot of places you don't get mobile coverage.
Sometimes you can get a wifi connection because the local pub has satellite wifi.
If it's Starlink, it's usually not too bad, and wifi calling and text messages can get through.
If however, it's the NBN's aging SkyMuster, or some other solution it can be too slow for wifi calling, and guess what, text messages sometimes don't arrive.
I'm talking seriously slow, the sort of speeds that make you long for character mode email and text based web browsing.
Really frustrating.
And of course you can't then complete the authentication process.
And Google's 'check your other device' solution can be just as bad, especially when you don't actually have your other device to hand, like it's a couple of hundred kilometres away.
The solution, of course, is to do all your set up somewhere with white lines and traffic lights before you go bush and making sure you click the 'remember me' box if there is one.
Of course, you don't always remember...
Bunsen labs ditched
I said I'd try Bunsen Labs Linux in a real world situation to do real work.
So I did.
Using Libre Office to review a document I started to get an annoying intermittent flicker - it could have been a latent hardware fault or it could be that the Radeon screen driver shipped with Bunsen Labs wasn't optimal for my hardware.
Well, only my pride was affected, I had very little work on the machine, so I wiped it and installed Ubuntu, remembering to click the third party drivers box.
I deliberately chose Ubuntu as they have particularly good support for Lenovo machines.
Well, changing operating systems seems to have cured the flicker problem (maybe).
It's certainly better but it does come back occasionally. The only thing to do is try it for some time and see if it is just as bad with Ubuntu as it was with Bunsen Labs.
(AMD also provide Radeon drivers for Ubuntu, and if the flicker comes back, I might well give these a go. Unfortunately, they don’t provide generic Debian drivers, and Bunsen Labs is based on generic Debian but there is a wiki page on AMD Radeon on Linux).
Bit of a pity, because I quite liked Bunsen Labs, but to be fair they did warn you on install it was a hobbyist supported distro, and that there might be problems ahead.
While I'm obviously disappointed, it won't stop me from trying Bunsen Labs again on other hardware...
[update 09/09/2025]
Well, I was still getting an intermittent flicker with Ubuntu 24, so I did a little digging.
lspci was correctly showing the graphics card to be an AMD Radeon, but as I still occasionally got a flicker, so nothing ventured I downloaded the latest AMD driver for the hell of it
Monday, 18 August 2025
Bunsen Labs in use
I was impressed by by Bunsen Labs Linux running on a VM, so much so I decided to try it on a real machine.
The only machine I had to hand was my old desk laptop that had been gathering dust for about nine months since I'd upgraded to Windows 11 - not by choice but sometimes you have to stay compatible with the world.
Anyway, long story short, it's an AMD Ryzen based Lenovo laptop, and even now after some six years of use, a pretty meaty machine.
Installation was easy, it just flew and it gave me a working system in around 45 minutes. I probably spent longer trying to get it to boot from install USB. (Just to be different AMD based laptop didn't use a magic function key combination to get into the boot menu. instead you needed to use a sim ejector (or a bent paperclip) to poke a special recessed magic button on the case when the machine was powered off. This caused the machine to boot into the boot menu - and yes I did have to track down the manual online to find this out)
Sunday, 17 August 2025
Bunsen Labs Linux
For the last twenty or so years I've been reusing old computer hardware for various of my projects, something that has invariably involved installing Linux as often software bloat on both Windows and OS X has reduced the usefulness of the hardware (and which is why I've been able to pick up some pretty good machines for not a lot from hardware recyclers and refurbishers.)
I've played with quite a few distributions over the years, but these days the two I feel most comfortable with are Crunchbang++ and Ubuntu.
Crunchbang ++ I tend to install on resource limited hardware - which is why I used it when installing Linux on a Chromebook, and Ubuntu on anything else.
Crunchbang started out as a custom Linux distribution designed to use fewer resources than most mainstream distributions.
Development of the original project halted in 2015, but it spawned two successor projects, Crunchbang++ and BunsenLabs linux.
For a long time both projects were very similar as regards installation and the user experience and I did run BunsenLabs linux on an old netbook for a number of years, but for the last few years Crunchbang++ has been my go to lightweight distribution.
However, when I was working out what I could use in my Linux on Chromebook project I came across quite a few reviews that mentioned Boron, the latest Bunsenlabs distribution as being quite slick and resource efficient, though not quite as minimal in its disk usage as Crunchbang++.
So, I thought I'd take a look, and this morning I built a BunsenLabs VM using VirtualBox on my Dell Latitude.
Like CrunchBang++ installation was via the standard Debian installer and once booted and logged in you are presented with a customised OpenBox desktop not that different from the standard Crunchbang desktop, albeit in a nice blue green shade and with the time and connection status on the bottom left rather than on the top right
Like Crunchbang there is an option to install additional software
Wednesday, 6 August 2025
Linux on an old Chromebook
Some time ago, for what seemed entirely sensible reasons, I bought myself an old Chromebook.
In practice, it has turned out not to be quite as useful as it might be. However, the screen’s in good condition and the keyboard is nice to type on, so I wondered, could I install Linux on it?
When I say install, I mean replace ChromeOS (which is a barebones quasi-linux under the hood) with another version of Linux entirely.
The machine is an Asus C202SA, which means it comes with an Intel Celeron N3060 processor, 16GB of eMMC storage and 4GB of RAM. Not the fastest device on the planet, but by no means the slowest.
The original Linux based EEEPc 701SD had an even slower Celeron processor, only half as much in the way of storage, and far less in the way of RAM – a measly 512MB – but I successfully installed Crunchbang Linux on it way back in 2014.
Using my previous experience with the Eee, I reckoned that it should be able to run a current distribution of Crunchbang++ successfully.
The Crunchbang install image size is typically a little less than 6GB, and when idle it only uses around 512MB RAM so it should work. The install process has a breakpoint to avoid installing large applications such as LibreOffice, so with a bit of luck minimal install should be smaller than the typical 6GB.
A web browser, a lighter weight word processor such as AbiWord, a text editor and a lightweight spreadsheet such as Gnumeric should give me most of the functionality I’d need.
So, how to install?
Chromebooks are designed to run ChromeOS and have a number of features to prevent people installing alternative operating systems.
However, for a few years there was a project, GalliumOS, to develop an alternative to ChromeOS for Chromebooks.
The project’s now been discontinued, but the project wiki has a wealth of information about installing alternative operating systems on older ChromeBooks.
In the case of my Asus, you need to replace the startup firmware (the BIOS if you are old school), with an alternative firmware image.
Chromebooks typically have a write protect setting on the firmware and this needs to be disabled.
Mr ChromeBook Tech supply replacement firmware for Chromebooks and have a pretty comprehensive list of models and how to disable write protection.
In the case of ‘my’ Chromebook it comes up with
meaning that you need to crack the case and remove a screw from the motherboard.
Fortunately the C202 and variants are designed for easy repair, and opening up the machine is straightforward with no nasty glue or anything like that involved, and there are a number of videos on YouTube, mostly featuring intense young men explaining exactly how to take one apart.
So, first things first.
It’s a 64-bit machine so I downloaded the latest 64bit ISO image of Crunchbang ++ (aka Cbpp), and using Rufus, made a dd style boot volume. The latest image is only available as a torrent, meaning I needed to install µtorrent to download the image.
µtorrent is a paid for application these days, but there is still a basic free version, but you need to be resolute and ensure you select the free version, which comes with some mildly annoying ads.
Then the first slightly scary bit – cracking the case and then using a prying tool (a standard mobile phone and case separator to separate the two halves of the case. Mine came from ebay for less than five bucks.)
Wednesday, 23 July 2025
Fanny Stepniak and the 1907 RSLDP congress
I've written before that Fanny Stepniak, Sergei Stepniak's wife, was a bit of a mystery to me.
There's evidence that she worked with Constance Garnett on Russian translations after Sergei's death, but very little to suggest she continued to be politically active.
Well, that's possibly not the case.
I've just finished Robert Henderson's book on the pre 1917 Russian exile community in London, and in his description of the 1907 party congress of the Russian Social Democratic Labour Party - the pre revolutionary underground political party that eventually became the CPSU - he mentions that Fanny Stepniak was personally included by Lenin in the vote of thanks to the organisers at the end of the party congress.
The party congress was held in exile in London and is notable for being where the Bolshevik faction under Lenin gained control of the political direction of the party.
Strangely, the congress was held in the Brotherhood Church, the christian anarchist group that gave rise to various Tolstoyan communes in England, including Purleigh, from which Tom Ferris and Bertie Rowe travelled to meet with Tolstoy in the winter of 1902-3, and the communes at Whiteway and Stapleton just outside of Leeds.
I'm guessing, and it is only a guess, that as well as being in contact with Constance Garnett, she was also in contact with the Maudes, and used her contacts to help arrange the use of the Brotherhood Church by the RSDLP...
Saturday, 19 July 2025
Field postcards
I've written both about the use of indelible pencil and more generally about the use of ordinary pencil on postcards in the nineteenth and early twentieth century.
I've recently acquired two examples of first world war British field postcards, one written in pencil, the other in indelible pencil
The pencil example is addressed to a Miss N Abel in Clifton Road in Aberdeen and dates from 1917
A most impressive array of bottles...
Ever since I documented the contents of Dow's Pharmacy in Chiltern for the National Trust, I've been fascinated by nineteeth century medicine bottles.
When we were in Melbourne recently, I noticed an impressive display of nineteenth century medicine bottles in a pharmacy on Wellington Parade in East Melbourne.
To try and give an impression of the size and range of the collection I've used an online jpeg stitcher to combine my images.
It's not ideal, but it does give an impression of the collection.
I'm happy for you to right click and download the images, but for a more detailed view here are the individual images in left to right order
If you do want to take a look yourself, remember it's a working pharmacy, not a museum, the staff have a job to do and customers to serve, so it's best to go at a quiet time and it would probably be appreciated if you bought a couple of things when you visited.
Sunday, 13 July 2025
LibraryThing and the Athenaeum
Maybe I have a new project.
Up at the Athenaeum, we have a heritage book collection –
basically all the books they bought between 1862 and the early 1970s when it
ceased to function as a library and reading room.
As far as we can tell they never threw anything out, only
replacing books if they remained popular, which gives us a
picture of reading tastes, and how they changed from the Goldrush era
onwards.
Now we have an excel spreadsheet listing the roughly three
thousand or so books but data quality is not great.
Publishers and authors names mis-spelled, different
abbreviations for the same publisher,
etc etc.
I’ve spent the last few weeks checking publishers’ names to
see if it was fixable, but really it’s not, the simplest solution is to
recatalogue the entire collection.
This of course is a problem in itself, we’re not a library,
we don’t have a catalogue system as such, all we want to do is catalogue the
collection as accurately as possible, and then load the data into Victorian
Collections for long term preservation.
Well, I think we have a solution - LibraryThing.
Originally designed to help people catalogue their personal
libraries, it’s been used successfully to catalogue
small research libraries.
While is does have add on modules to give it the functionality
of a larger scale OPAC, we only need its cataloguing capabilities. After all we
are not going to add, lend, or deaccession material.
Data can be exported in both excel and MARC format, which
means that not only could be load the data into Victorian Collections, we could
potentially load it into a Library Management system, such as our local library’s
Sirsi Dynix system, if desired.
But the real killer is that we can validate entries against
both the British Library and the National Library of Australia, meaning we don’t
need to create every entry from scratch.
The British Library link is especially valuable as in the nineteenth
century and early twentieth century about 90% of the books in the collection
were imported from the UK, with the remainder coming from the US, especially
after the first world war when there seems to have been an interest in Westerns
and crime novels – the original pulp fiction.
Legal
deposit rules mean that the British Library should have a copy, and hence a
catalogue record of every book published in the UK in the nineteenth century.
(It’s not quite perfect sometimes when I check entries manually I find there are minor inconsistencies
in entries between the BL’s catalogue and the National Library of Scotland’s
catalogue entries for the same book, but they are probably not significant enough to cause a problem).
Nineteenth century Australian publications might be more of
a problem.
Not all are in the NLA’s catalogue, but the State Libraries
of NSW and Victoria respectively are fairly comprehensive.
The only problem is that LibraryThing does not link to them,
meaning that in these cases we would have to create a manual entry.
The other problems that I’ve come across are Book Club
editions, and books published in Australia in the second world war.
Due to the shortages of materials in Britain book exports to
Australia almost stopped, but a few UK publishers entered into licensing
agreements to have local editions produced here in Australia by Australian publishers,
and not all of them seem to have made it into the NLA catalogue.
Likewise, there are some post world war II pulp fiction
reprints that were produced locally but don’t seem to be in the NLA catalogue.
Again these would have to be investigated on a case by case basis.
However I’m confident that we can use LibraryThing to
automagically ingest in excess of 90% of our holdings.
Probably the next step would be a trial run of a couple of
shelves worth of books and see how it goes.
That should allow us to refine and document our methodology
and perhaps come up with a more realistic estimate of the number of person
hours involved.
Friday, 4 July 2025
Smartwatches - what's the point?
A long time ago I bought myself a cheap no name fitness tracker, which worked quite well, allowing me to track my bike rides and to beep when I had a new email or text message - a feature that turned out to be quite useful, especially when I was wearing nitrile gloves and documenting an artefact - I could glance at the tracker and decide if the message was worth degloving for.
In time I replaced it with a brand name device, an Inspire HR, which actually did a little less, but came with nicer management software.
However, it did everything I cared about, was light and comfortable to wear, and didn't need to be charged too often.
Then, three or four months ago, Telstra, our phone and internet provider, emailed me to say I had a pile of loyalty points that were about to expire.
Unfortunately, the points were not enough to make a serious difference to the cost of a new phone, but they did have the Ryze wave smartwatch available, and I had enough expiring points to cover it, making it effectively free.
Now, I've always been curious about smart watches, so this seemed to be an ideal way of finding out if there was a use for one in may life.
Out of the box it did everything that was expected of it, had a nice legible display, and was a perfectly competent device - a bit bulky on my wrist but comfortable enough.
But...
I realised after a couple of months that I was only using the same functionality that I was from the Inspire HR, or indeed the cheap fitness tracker,
Basically, it really wasn't adding anything to my life.
So I stopped wearing it and went back to my battered and scratched Inspire HR.
Now, it's entirely possible that if I had brought another brand of smart watch, it might have had some attribute that really helped make life easier or better, but reading through the specs for those made by Google, Garmin and Apple, I don't really see anything stand out as regards to capability.
So, are they just expensive doodahs, or am I missing the point?
Wednesday, 2 July 2025
Linux on an old imac
I appear to have accidentally done something useful.
J's old 2017 iMac had been sitting in a corner in the study, waiting to be wiped and taken to the recycler.
Too old, too slow.
So, today I decided to wipe it.
Rather than do something straightforward I thought I'd try installing the latest version of Ubuntu out of curiosity - I'd heard good things about the current version of Ubuntu and older intel based imacs, so it seemed worth a go.
Certainly, when I tried installing Ubuntu on an older imac five or so years ago I found it really didn't work that well, so it seemed like a fun idea to try installing a newer version of Ubuntu on more recent hardware,
Obviously, bluetooth mice and keyboards don't work until you've installed the new operating system so you need to use an old wired keyboard and mouse. The other thing I found was that you needed to plug the install USB into the back of the mac and not into a USB socket on the keyboard, otherwise I just used a standard installation USB I'd made a few months ago using Rufus and writing the volume as a dd image.
Well, surprise surprise, it recognised the image and booted cleanly
Sunday, 29 June 2025
Indelible pencils
Friday, 13 June 2025
Data recovery at home
J was looking through some documents to do with her mother's death that she had got some years ago from her sister, deciding what needed to be kept and what didn't. In among the papers was a CD labelled Graham family photos.
We had been living in England when J's mother died, (this had been round about the turn of the millennium) and it had been left to her sister to get rid of her mother's effects, and it looked like an album of family photos had disappeared.
So, did the CD have the missing photos?
Well, J's sister's husband been a photographer with The Age in Melbourne, and later on a TAFE photography lecturer, as well as being someone who was both interested in the history of photography - he built himself a copy of a nineteenth century glass plate camera at one point - and had dabbled a bit in family history, so it was just possible he had copied them, perhaps for a project of his own.
Unfortunately we couldn't ask him exactly what had happened as he died a few years ago.
And of course computers don't have CD drives these days.
So, nothing ventured, nothing gained, I bought a $20 external usb cd drive from ebay and connected it up to an old Windows 10 laptop.