Twenty five years! A quarter of a century! That’s a significant portion of the lifetime of the World Wide Web itself.

Actually, this site did exist before September 30th, 2001 but it was very different. You can still look at the very first version of adactio.com if you want to wallow in the glory days of framesets and DHTML. But it was twenty five years ago that I started publishing my thoughts here, making adactio.com my online home.

The funny thing is, I distinctly remember when I started writing that I was getting into this whole blogging thing really late. So many others had been doing it for a while at that point. Now there aren’t many of them left.

It’s not just the journal though. At some point I added the links section. Later still, I added notes. Between those three sections, I usually end up posting something on this website at least once a day.

But I don’t put any pressure on myself. That might be the secret to the longevity of the site. I don’t worry about running out of things to write about. If I go for a while without writing anything, that’s okay. But that rarely happens. There’s almost always something I want to share. Like I said in that first ever post:

I’ll try and post fairly regularly but I don’t want to make any promises I can’t keep.

I also wrote:

Being a web developer, there’s bound to be lots of geeky, techy stuff but I also want a place where I can rant and rave about life in general.

I think the tech stuff has decreased over the past few years. I’m more likely to write about music or books. I expect that trend will continue.

I certainly have no plans to stop. It feels good to have my own little homestead on the web where I can do whatever I want.

After twenty five years, it feels good to occasionally peruse the archive and revisit the past.

Here’s to the next twenty five.

That’s why it pains me that I’m about to add more fecal matter onto the ever-growing pile of shit by writing yet another hot take related to large language models.

If you stop reading now, I don’t blame you. In fact, I applaud you.

Anyway…

The last time I made the mistake of writing about large language models, I set out my stall thusly:

Software is almost certainly the killer app for large language models.

I think the artists, writers, and musicians will be okay, or at least as okay as they ever were. It turns out that humans like things made by other humans.

And y’know what? If I had to choose which endeavour I’d rather see automated away—programming or art—it’s no competition.

I stand by that. But it doesn’t mean that software development gets a free pass to use exploitative extractive tools with a clear conscience.

You can use these tools in a thoughtless way or you can use them thoughtfully. I really like Robin Sloan’s approach:

The consensus vision seems to involve AI working for you day-to-day, a constant ambient presence, sparkles in everything; even cautious Apple has succumbed. I’m not interested in that.

My alternative: get your wish, then stuff the genie back into the lamp.

He’s got a set of rules:

1. You must actually and consistently use anything you build for two months before posting about it.
2. Ideally, never post about it. (I know I am breaking this rule, but my purpose is pedagogical, and anyway, I’ve got more apps I didn’t tell you about.)
3. Don’t let the AI agent name the app. You pick the name.
4. Don’t distribute the app, not even for free. Don’t post the code on GitHub. This is not software for glory; it is software for you.

For most of my career, I’ve wanted web development to become something that anyone could do. I really like the World Wide Web and I think it would be great if more people had their own websites. But I totally get that most people have no interest in learning HTML.

One of my biggest fears for the web was that it would become the domain of professionals only; a gatekeeping priesthood who get to make stuff while everyone else goes without.

The efficacy of large language models for coding would seem to lessen that possibility. Especially if you apply Robin’s approach.

The Session has an API. There are also weekly data dumps of everything on the site. The idea is that other people can use these to build other useful tools for themselves.

Lately, there’s been a big uptick of these kinds of tools, mostly made by someone with the help of large language models. Often they have no previous experience of making software.

This is good. But I get uneasy when I see people selling these tools to other people. Call me old-fashioned but I feel that if you’re going to ask people to pay for a thing, you should understand that thing. I think that’s why Robin’s list of suggestions resonates with me.

But…

To build software with large language models today, you pretty much have to pay some money to the absolute worst sort of people, the ones pedalling tokens to their “hyperscale” products. You can’t avoid being complicit because there isn’t really an alternative.

I think that might change.

I’m not going to make predictions. That’s a mug’s game. But I might attempt a gentle unspooling of one potential future…

First off, there’s a crash coming. That’s pretty much inevitable at this point. The only question is how much collatoral damage it’s going to do to the world.

Right now, tokens are subsidised. The hyperscalers lose money when people use their products. It’s possible that we are now living through a golden age of making software relatively cheaply with large language models. That will change when the hyperscalers start charging for the true cost of tokens.

Not for everyone though. I believe that enterprise software companies will happily pay the true price for tokens. It’ll be like Google Glass; that was an absolute disaster for the consumer market but went on to have a productive second life in factories.

You know when you’re in an airport and you see advertisements that are not for you, but for faceless corporations? You know the ones. They’re all “cloud” this and “sap” that. Now they’re all about “AI”. It’s not for us. It’s for them.

Does this spell the end of making home-cooked apps with large language models?

I don’t think so.

The hyperscalers might become unaffordable to most people, but keep an eye on open models. And keep your other eye on local models.

Right now, open local models aren’t as good as the large language models that require data centres of computation and energy. But they aren’t massively far behind. And if the recent history of large language models has taught us anything, it’s that “good enough” is fine.

Take Google Search. They decided to fuck it up by forcing generated summaries into your eyeballs above the actual search results. These summaries aren’t 100% accurate. Let’s be generous and say they’re accurate more than 90% of the time. According to Google, that’s good enough.

So, sure, local open models may never reach the levels of data-centre-powered large language models. But they don’t need to. They just need to be good enough.

I’m really looking forward to that.

My issue with large language models has never been the actual technology, which is genuinely fascinating. My issue is with the power dynamics, of being robbed of agency, of the over-inflated hype and criti-hype being used to convince us of an inevitable future where a small group of very rich men are masters over our lives.

I’d love to see the technology decoupled from the current depressing narrative. I’d love to see a thousand homemade software flowers bloom from open local language models.

Summer is officially over.

I saw it out with my now-annual trip to Spain for the Cáceres Irish Fleadh. It was, as always, great fun. Nothing beats playing loads of tunes into the early hours while sitting outdoors in a beautiful old town.

I visited some lovely places this summer. A trip to Greece. A trip to Italy. Even my trip to Cork was blessed with bright sunshine.

But I think the highlight was the three weeks I spent in the north of Ireland. It helped that being there meant I escaped the worst of the heatwave here in England.

All in all, it was a thoroughly enjoyable summer filled with travel, music, good food, and sunshine. Now it’s time to hunker down at home.

There’s a bit of that in What We Can Know with its expository passages explaining what’s been happening in the 21st and 22nd centuries, but it mostly works.

As with all McEwan’s work, it’s extremely English. He just has to have some clashing of the classes in there.

The book has an unusual structure, switching between the present day and a climate-ravaged future one hundred years from now. A future scholar is obsessed with a poem from our time period that was never published. At times the book feels like a whodunnit, complete with all the suspects assembled together in a country house.

If the structure is unusual, the tonal swings are really something. We go from high-brow pondering of the utility of art in the face of catastrophe to downright schlocky soap-opera shenanigans. My eyebrows may have gotten whiplash.

But I guess it all worked somehow because I found myself turning the pages, eager to reach the final revelations.

Still, there was one thing that really bothered me. In order for the plot to work, our future scholar needs to have access to emails, chats, and other digital artificats from today. We get an explanation of how the cryptography has been cracked by quantum computers, but it’s simply taken as given that the files themselves would be available, having survived in “The Cloud”.

I can’t really blame Ian McEwan here; here’s just repeating the oft-retold advice that you should be careful what you put online because it’ll be there forever. But that warning simply isn’t true. The Cloud dissolves very, very quickly. Just ask the folks at the Internet Archive.

But if your credulity can stretch to accept the naïvely optimistic depiction of digital preservation in What We Can Know, it’s an enjoyable romp.

At the end of every Thursday afternoon at Clearleft, we wrap up the working week with an all-hands (video) meeting. Yes, I know that the week finishes on Friday, but Fridays have been declared the no-meetings day at Clearleft: a chance to concentrate on heads-down work without interruption. Besides, some of us don’t work on Fridays. So Thursday really is the new Friday.

That gathering usually finishes with a little quiz (like the “fonts or food?” quiz) . I was in charge of putting the quiz together last week. I came up with a movie mashup:

• Take two film titles.
• The end of one title is the beginning of the other title.
• Create one single title from both.

For each combined title, I smushed together the synopses from the Internet Movie Database.

For example, if you got this synposis:

Six unemployed steel workers form a male striptease act and King Arthur and his knights embark on a quest encountering many, very silly obstacles.

The answer would be:

The Full Monty Python And The Holy Grail

Got it?

Let’s see how you fare with these ten movie mashups I presented to the Clearleft crew (tap to reveal each answer)…

How’d you do?

I had fun putting this little quiz together but some potential movie mashups were left on the cutting-room floor. If there were a director’s cut of this quiz, maybe these would’ve stayed in the picture:

• Mad Max: Fury Road To Perdition
• Remains Of The Day Of The Dead
• For Your Eyes Only God Forgives
• The Land Before Time Bandits
• The Taking Of Pelham One Two Three Billboards Outside Ebbing, Missouri
• City Of Lost Children Of A Lesser God
• Run, Lola, Run Silent, Run Deep
• Carlito’s Way Of The Dragon
• The Man Who Would Be King Kong
• Man Of Steel Magnolias
• The Big Short Circuit
• A Matter Of Life And Death In Venice
• The Hills Have Eyes Wide Shut
• The Spy Who Loved Me And Earl And The Dying Girl

That sounds impressive, but it isn’t. Anyone can become an invited expert. The fact that I am now one proves it. You apply to be an invited expert and once that application is approved, you’re in. So you too could and probably should be an invited expert to a working group at the W3C.

I was strongly encouraged to become an invited expert in the web applications working group after weighing in on the matter of installable web apps (or progressive web apps or whatever).

Last week I had my first call but it wasn’t with the web apps working group, it was with the technical architecture group, a meta-group that helps the other groups if there’s a big-picture sticking point.

There’s a big-picture sticking point with installing web apps.

While most of the participants (read: browser makers) would like to spend their time deciding the details of what API to implement in order to support installable web apps, Webkit is rejecting the very premise of that work.

On last week’s call, Webkit outlined their position: there shouldn’t be a way for developers to allow users to install the current website. Leave it to the browser, they say. Also, is this even something that users want? Looking at the stats, it certainly doesn’t seem like it.

That would be a reasonable position if there were already a usable way to install web apps. But there isn’t. It’s technically possible to add a website to your iPhone’s home screen. In practical terms, it’s a convoluted usability nightmare.

(It’s hard to avoid veering into conspiracy theory territory and seeing this as some kind of malicious compliance. Especially when you compare it to how native apps are shoved in your face thanks to “smart app banners” better known as dickovers.)

So Webkit’s stance would make total sense if there were a reasonable way for users of Mobile Safari to install web apps already. But there isn’t.

Last week’s call was quite illuminating. It showed some incredible cognitive dissonance in the Webkit position. Let me explain…

On the one hand, installing web apps is kind of like bookmarking, they say. That’s true. We don’t have an API for bookmarking so why should we have an API for installing web apps?

That would be a fair point if the user interface for bookmarking and installing were in any way comparable. But bookmarking is literally front and centre of the user’s experience of a browser, backed up by decades of convention. Meanwhile the option to install a web app is buried five levels deep behind a “share” icon.

Also: we’re not talking about an API for installing web apps. We’re talking about an API for initialising the flow for installing web apps—the very same flow that’s triggered from that buried menu item. And that flow begins with a prominent option to cancel. So let’s not have any scaremongering about users somehow being tricked into adding web apps to their home screen.

Which brings me to the other point…

Webkit are concerned about allowing installed web apps getting access to more powerful APIs. They’re quite right not to want that! A web app launched from the home screen shouldn’t have any special privileges. If it wants access to say, geolocation, the user needs to grant permision just the same as if the page were in the browser.

Why, oh, why then did Apple limit push notifications to installed web apps?

It’s not like other browsers haven’t managed to implement permission-based APIs like push notifications. But apparently limiting notifications to installed web apps was the only way that Apple could think of implementing this API safely.

You see the contradiction, right?

On the one hand, Webkit is saying that installing web apps is like bookmarking. No big deal.

On the other hand, Webkit is saying that installing web apps grants special privileges. A huge deal!

Which is it?

It’s almost as if Webkit aren’t actually participating in good faith but rather have already made up their mind to drag their heels when it comes to any kind of progress on this topic.

Anyway…

On last week’s call, I was supposedly representing the interests of developers. No pressure!

I can’t make any claim to represent all developers, but I like to think I’m fairly representative of a typical developer who’s quite fond of the World Wide Web. So in my allotted ten minutes of speaking time, I said:

• Developers would love a way to initiate the installation flow. The alternative is to provide browser-specific instructions—always a bad sign.
• If browsers provided a prominent usable way to initiate the installation flow, this wouldn’t be such an urgent need.
• Users should have the option to install web apps just like they have the option to install native apps. Technically both options are available, but practically the scales are tipped very, very heavily towards native apps.
• Native apps don’t have the same security model as web apps, allowing them much greater access to user data. If security and privacy are values that the W3C thinks are important for users, they should do everything in their collective power to help tip the scales back in the direction of web apps.

Come to the excellent Prince Albert pub on Trafalger Street at 8pm on Saturday, September 5th. Head upstairs with your ticket. There you will be treated to an excellent triple bill.

First up is Figure In The Half Light AKA Charlie McConnochie:

Expect dreamy, fractured songs of dark and light.

Then it’s The Self Help Group:

Americana-influenced vocal harmonies, emotional tales and intertwining melodies.

The headliners are Salter Cane. Hey, that’s my band!

Johnny Cash meets Joy Division: that’s one attempt to describe the sound of the legendary post-punk folk-noir band from Brighton who are back with their new album, Deep Black Water.

Entry will be £10 on the door, but advance tickets are £8 so get yours now and I’ll see you on Saturday night!

We sometimes make it work though. Like crossing the Atlantic together.

Dan and Sue were in Italy last week. Their son Rex was attending a mandolin academy in Acqui Terme as part of the terrific Mando for Kids programme. As a mandolin player myself, this very much aligned with my interests.

Jessica and I flew out to meet up with them and spend a lovely few days enjoying the Italian food, landscape, and culture. And I got to hear the magnificent sound of massed mandolins, tremeloing together to create lush waves of beautiful music.

Somehow I got roped into a little demo of Irish tunes on the mandolin. It was a bit nerve-wracking, and I totally flubbed some of the tunes, but the lovely audience was very forgiving and receptive.

During the Fleadh, I went to quite a few sessions at Cumann Cultúrtha Mhic Reachtain. I ended up chatting to a fiddler named Rus from New Mexico. He told me about a book he had written twenty years previously about his time in Kerry coaching a basketball team and learning Irish music.

I had heard of this book! I can’t remember where exactly, but there aren’t that many books with a connection to traditional Irish music, so it’s no surprise that it was on my radar.

Anyway, Rus very kindly sent me a copy to read. It’s called Paddy On The Hardwood.

While it’s ostensibly a mix of basketball and trad, it’s actually about 90% basketball and 10% trad. But that 10% is great. Rus was learning from the late great fiddler Paddy Jones.

While I wish that there was more of that, I understand that is not really a book aimed at me. It’s a book aimed at Americans who might not know anything about Ireland but already understand basketball. As someone who doesn’t know anything about basketball, I was left to flounder in jargon and acronyms.

But even as someone who doesn’t have much interest in sport, I felt kind of uncomfortable about the way different sports were presented. There’s exasperation that basketball isn’t more popular in Ireland. The popularity of Gaelic Football is only described as something that gets in the way of basketball. The interest in Irish culture extends as far as music, but not indigenous sports.

I was also really uncomfortable with the level of fat-shaming. This is from the very first page of the book:

A chunky girl with a navel that shoudn’t have been exposed leaned against the wall behind him.

I rolled my eyes and hoped that would be the end of it, but it only got worse. There’s a barrage of “humourous” remarks from this white coach about the weight of a Black player on his team. The power dynamics alone made that a poor decision.

Whenever there’s a fish-out-of-water story, the humour can come from two sources; the fish or the water. You can tell a self-deprecating story about yourself, or you can point and laugh at everyone different to you. There was a bit too much of the latter for my taste. The punching down grew tiresome.

The book was written two decades ago. Were it written today, I would hope it would be less mean-spirited and colonial in its outlook.

While this mostly isn’t a book for me (apart from the musical excursions), you might like it if you’re into basketball and you’re okay with some outdated attitudes.

I ended up sitting next a lovely local fiddler named John. I saw him again the next day when I was at Ulster University for my workshop. He had a stand there, selling a new book of tunes that he had published. It’s called Pass It On. I couldn’t resist buying a copy. Then he mentioned that he also had a few copies of his brand new debut novel.

Now I was really intrigued. I told him to give me the elevator pitch and he said it was a semi-autobiographical work about growing up in Belfast druing The Troubles and finding solace in traditional Irish music.

There aren’t that many books about traditional Irish music, and certainly not many novels. “I’ll take one!” I said. He handed me a fresh copy of When The Boys Came Out To Play.

I started reading it when I got back to Brighton. To be honest, my expectations weren’t high. I shouldn’t expect too much from a self-published novel by a first-time author, right?

But I was completely captivated. I know it’s a cliché to talk about a book being un-put-downable, but I was seeking out opportunities during the day to crack it open again and read just one more chapter.

There’s nothing particularly clever about the prose, and that its strength. Some of the best writers understand how to let a story unfold through clear, direct writing. It sounds simple but it’s extremely hard to do it well.

This is a story told from a young boy’s point of view, and it genuinely feels like you’re seeing the world through his eyes. More than that, it feels like you’re right there experiencing events, some terrible and some transcendent.

This isn’t an epic story. It’s just the story of one child and his family. But I was completely invested in it.

Now look, just because I really like a book doesn’t mean that other people will. I always try to keep that in mind when I’m writing about any book I’ve read. But in this case, I really think John Mackel has written something excellent and I’d love for more people to discover it.

Do yourself a favour and get hold of When The Boys Came Out To Play. You won’t regret it.