Matthew Lang

I need to find a new forge now that Sourcehut is prohibiting the use of LLMs. There’s an interesting discussion about this on Lobsters.

I’m still trying to assess whether I am in breach of this with the two paid accounts I have at Sourcehut. The change to the terms is minimal, but I am not entirely convinced that I am breaking the terms and conditions of their service. With these two accounts, I have used LLM tools to generate code that, for the most part, I have amended before committing it.

I also have a repo for a web application I used to publish the content for this blog, which calls on an external LLM to generate alt text for images.

But better to be safe than sorry, so I think I will end up moving these accounts elsewhere. GitHub is out, so I am looking at GitLab again and perhaps Codefloe, which has the benefit of being hosted in Europe, but is only supported by donations. Self-hosting isn’t an option for me. I don’t have the time to maintain my own forge.

A shame that I am making this move, because I like the Sourcehut product. It’s minimalist and does everything I need it to.

I’ve been looking at several PostgreSQL clients as I migrate my local development databases away from Docker containers and back to a native PostgreSQL installation.

As I evaluated several desktop apps to manage these databases, I was disappointed to find that many require a monthly subscription.

The subscription model makes sense for server-dependent software. The server is part of how the product works and is often a key component. The product might use the server to sync information between devices or to process information in the background. The subscription model covers the ongoing costs of running and maintaining the servers on which the product depends.

Why a subscription model for a desktop app though? In most cases, the desktop app runs on your own hardware. In the case of a PostgreSQL client, it can open databases that exist on my laptop or even on another remote server, but the product itself doesn’t need a server to do this.

The justification is revenue predictability. One-time purchases tend to spike at launch and taper off, making it hard to forecast earnings or plan for the future. A subscription smooths that out.

I get why developers go down this route. Predictable revenue is easier to build a business around than the feast-and-famine cycle of one-off sales. But predictability is a benefit to the developer, not a justification rooted in the product itself. A $20/month subscription for a desktop app will cost $240 a year, every year, for software whose core functionality was largely complete on day one.

There’s a better model from back in the day before subscriptions. A one-time purchase with an optional paid upgrade for major versions, the approach taken by apps like Postico and TablePlus, gives the developer ongoing revenue tied to actual new work and gives the consumer a cost that reflects what they’re actually getting. If the subscription model is justified by ongoing server costs or continuous development, fine. But if it’s just a pricing decision dressed up as a product decision, it’s not a product that I think I will give my money too.

Over the last year, a line has been forming between my laptop and my phone. And it became really clear last week when I tried out Obsidian Sync for a few days.

I thought it would be good to run Obsidian on my phone as well as my laptop, so I set up Obsidian Sync, got the remote vault working with my local vault, and set it up on my phone. I must admit, it was tricky to get my local vault up and running with it, but that wasn’t the reason why I cancelled it.

After a few days of use, I found I didn’t like using my phone to interact with Obsidian. I preferred a bigger screen, and so, I found myself reaching for my laptop more during this time. The sync feature wasn’t the problem. Trying out Obsidian Sync solved a different question. Which device do I want to use? The answer was obvious: the laptop, almost every time.

It was from this point that the divide between phone and laptop was clear.

My laptop is for head-down work. Coding, writing, researching. Anything that requires me to spend more than a few minutes on a single task. Anything like that and I use the laptop. There are two benefits to this.

The first is that the laptop is a dedicated workhorse. There are very few social media applications on it, so notifications are infrequent.

The second is that I no longer require as many app subscriptions. Many subscriptions are based on the premise that they can sync your information between devices. With just one device, that cost disappears. And given that just about everything is a subscription these days, that’s not a bad thing.

For everything else I do, and there’s not much left thankfully, there’s my phone.

I use it mostly for messaging family and friends as well as for golf club-related comms. I use it to arrange teams matches, organise competitions for the club’s junior section and organise other junior events.

I do use social media on my phone as well, but I keep my interaction light there. Reddit is still a good source for many things. I’m more of a lurker on Instagram. I prefer to share most of my photos on my blog rather than on Instagram. Then there’s BlueSky and Mastodon for news.

There’s one exception to the laptop rule. I have iA Writer installed on my phone and my laptop. Why? Just for when the need arises for putting some words down and I’m away from my laptop. For everything else, though, the divide between them is clear: each has its own dedicated use.

It’s taken me until I am almost 50 years old to get to this point and recognise that each device has its purpose. It’s taken a while to get here, but at least I have arrived. Better late than never.