My coffee mug

Hello world, and welcome to my corner of the web. This is where I write words about what I'm working on, and post photographs of things I've seen.

I'm a Software Engineer at the Wikimedia Foundation, and so of course my personal website is a wiki (running on MediaWiki). In my spare time I volunteer with WikiClubWest to work on Wikimedia projects, mostly around my family's genealogy and local Western Australian history (especially to do with Fremantle). I try to keep up with issues on all the things I maintain (but usually fail).

I also try to find time to work in my workshop on various woodworking projects. Recently, that's been focused on restoring a chest of drawers and building a metalworking bench.

Travel features in my life, not because I really hugely want to go elsewhere but because just do — and also because then I can do some more interesting mapping on OpenStreetMap. Sometimes I ride my bike to get there.

I'm currently reading the following: A Puritan Bohemia (Margaret Sherwood, 1896), and Arrowsmith (Anon), and Doctor Thorne (Anthony Trollop), and Perth (David Whish-Wilson, 2013), and The Railway Adventures (Geoff Marshall; Vicki Pipe, 2018).

To contact me, you can email me, find me on Matrix as '@samwilson:matrix.org', or the fediverse as @samwilson@wikis.world. If you want to leave a comment on this site (by creating an account), you need to know the secret code Tuart (it's not very secret, but seems to be confusing enough for most spammers).

Spideroak not for backups


· Spideroak · backups ·

I don't know what Spideroak's deal is, but they seem to no longer even advertise the fact that they provide a backup service. Their website at https://spideroak.com doesn't mention anything about the 'Spideroak One' product (I thought maybe it'd be under 'Securing Terrestrial Communications and Collaboration' but no). Seems you have to go to 'Spideroak for business' which takes you to https://crossclave.com where it's listed under Products. But I'm not a business!

I guess they're more interested in space these days. Probably time for me to finally finish moving all my stuff to rsync.net (that was supposed to be a 2022 goal…).

The web is already decentralised


· fediverse · websites · RSS ·

On the current decentralisation movement by Manuel Moreale, 2022 December 16:

The more I look at this "issue" the more I'm convinced the solution is already right there and it's called the web. Want to have an unblockable, unbannable user profile? Buy yourself a domain and get a personal website. Want to have a space where you can say and do whatever the fuck you want? Get a webspace and put up a blog. Do you want to keep up with what other people are doing and saying online? Start using RSS or, and this is gonna sound like a very radical idea, bookmark their websites and every once in a while open them in your browser and see what they're up to. Want to also have discussions? Add comments to your website. Don't care about other people's opinions? Don't add comments to your site. I honestly don't get why people are searching for some new technical solution to a problem that in my opinion doesn't exists.

Incidentally, that site has nice typography and formatting. Maybe I should revisit SimpleText one of these days (that's the skin I'm using here) and fix it up a bit.

Webmentions as likes


From 100 Days of IndieWeb Challenge by James, 2 January 2023:

The IndieWeb advocates for a web-centric approach, where we leverage the building blocks of the existing web as well as a few new ones to build tools that make communication and publishing on the web easier.

Fundamentally, the internet does not have to be controlled by walled gardens. You can publish content on your own website and share that content with friends. You can even send likes to people's websites and receive likes on your site, too, using a tool called Webmention. For me, participating in the IndieWeb is partly my belief in the importance of owning your own data but also the opportunities that affords.

I've been meaning to add Webmentions to my site, but haven't yet. I do like the idea of each page displaying any (or perhaps just a count of) in-bound links. Webmentions (and old-fashioned automated crawling to find links) are a good way to find which pages out there have linked here, but they don't necessarily indicate why they've been linked. So I don't like to think of them as equivalent to 'likes', but anyway this whole thing of Fake Internet Points™ is pretty ridiculous and probably best avoided.

Pattypan to upload scanned letters


· Wikimedia · scanning · H.M. Wilson Archives · Pattypan ·

Pattypan is a great tool for uploading lots of files to Commons. I'm using it today for a set of scans of airgraphs. I scanned them all yesterday, and gave them hopefully-useful filenames that contain a prefix, date, and bit of info about each. These documents all look pretty much the same when viewed as thumbnails, so it's important to be able to determine at least something by glancing at the filename. That said, it's never good to rely on metadata being in the filename nor on all files having the same structure to their names.

Anyway, after creating the spreadsheet for these files, I did some tweaking to the data columns:

  1. Filled the source column based on the name column:
    • Formula: =CONCAT("{{HMW|wiki=",B2,"}}")
    • Result: {{HMW|wiki=Airgraph 1943-12-30 Adele to Murray}}
  2. Extracted the date:
    • Formula: =REGEX(B3,"[0-9]{4}-[0-9]{2}-[0-9]{2}")
    • Result: 1943-12-30
  3. Crucially, any column that uses a formula needs to be copied to a new column with values only (i.e. "Edit / Paste special / Values only"). Otherwise Pattypan can't see the formula result.



I've been reading a bit about Airgraph letters (aka V-Mail, in the US). Some articles say it was developed by the US military, and the Wikipedia article could do with some improvement about the detailed history of it. This 2021 video from Mark Felton is interesting; the clips from the US military in that come from this film it looks like, and there's also this Pathé News film showing the Airgraph system from the UK point of view.

I'm uploading about 65 scans of airgraphs to this category on Wikimedia Commons.

Airgraphs to Murray Wilson, pile.jpg

What to archive


· archiving ·

Interesting article in the New York Times. Makes me even more keen to keep working on software that makes it easier to keep personal archives — but not keep everything.

Your Memories. Their Cloud. (archive), by Kashmir Hill, 2022 December 31:

I noticed a philosophical divide among the archivists I spoke with. Digital archivists were committed to keeping everything with the mentality that you never know what you might want one day, while professional archivists who worked with family and institutional collections said it was important to pare down to make an archive manageable for people who look at it in the future.

“It’s often very surprising what turns out to matter,” said Jeff Ubois, who is in the first camp and has organized conferences dedicated to personal archiving.

[…] Mr. Ubois said it’s hard to predict the future uses of what we save. Am I socking this away just for me, to reflect on my life as I age? Is it for my descendants? Is it for an artificial intelligence that will act as a memory prosthetic when I’m 90? And if so, does that A.I. really need to remember that I Googled “starbucks ice cream calorie count” one morning in January 2011?

Pre-internet, we pared down our collections to make them manageable. But now, we have metadata and advanced search techniques to sort through our lives: timestamps, geotags, object recognition. When I recently lost a close relative, I used the facial recognition feature in Apple Photos to unearth photos of him I’d forgotten I’d taken. I was glad to have them, but should I keep all the photos, even the unflattering ones?

Bob Clark, the director of archives at the Rockefeller Archive Center, said that the general rule of thumb in his profession is that less than 5 percent of the material in a collection is worth saving. He faulted the technology companies for offering too much storage space, eliminating the need for deliberating over what we keep.



· new years · archiving · Fremantle ·

It's the first morning of the year, and I'm at the new cafe on High Street, Milkmaid Coffee Bar. Seems like a very nice place. I've added it to OpenStreetMap, although I'm not quite sure I've put it in the right place… I'll go outside in a bit and check which bit of this building I'm actually in. The facades of these places can be misleading, especially from the aerial imagery.

It's probably the done thing, on January first, to figure out What One Should Do in the new year — but I don't think I'll bother. I'm mainly working on genealogy stuff, and am hoping to spend the rest of today scanning a few dozen airgraph letters from 1943. The were mostly written by my great-grandbother Edith Wilson to her children who were away from home serving in the army and air force.

(Bah! It looks like Every Door created two nodes when I edited it just now: 10297567092 and 10297572193. That's weird, but I guess it's because it didn't finish refreshing the data after saving, and I was editing a node that I'd only just saved.)

Automattic is a great model


· Automattic · WordPress · hosting ·

Very interesting thread this morning from Chris Trottier about why Automattic (the company that makes WordPress) is such an interesting model to look at when thinking about financial sustainability of social media.

Most questions about the Fediverse's long term sustainability and survivability can be answered by looking at Automattic.

1. Does it scale? Yes. 2. Can it be extended and customized? Yes. 3. Can the non-technical average Joe be comfortable enough to self-host? Yes. 4. Can institutions embrace it? Yes. 5. Can an ecosystem be built around it? Yes.

The operating model has already been built and been proven long ago.

Read the rest of the thread…
Retrieved from ‘https://wiki.samwilson.id.au/index.php?title=Welcome&oldid=3677