blog.dataforce.org.uk

Slightly with which I make new posts, it has been my most recent post for a while. And I think it deserves an update.

When that post was made:

• Opus 4 was the best of the Claude models
• Claude-Code had only been GA for about a month.
• OpenAI were also in their 4 series, GPT-4o, GPT-4.1 and o3/o4-mini.
• Codex CLI had also only existed for a couple of months
• Google Gemini was on the 2.5 series

I think my initial feelings were fair and accurate for these models. In addition, Opus was more-expensive to use, so I was more likely to be using Sonnet for things unless I thought it really needed something.

As mentioned in that post, I had a new-and-fresh claude-code subscription. I had only intended to have it for a couple of months… Turns out I’ve never stopped paying for it, and have been using it for various levels of things since.

It’s crazy how fast the state of this has all changed, and how much my use and experience has changed…

The TL;DR here is that I am no longer sitting close to “Hype” and I now sit much closer to “Incredibly useful Tool”, but read on for a more fleshed out version of this!

AI in coding, Hype or genuinely helpful? I’ve always leaned towards ‘hype’ but a recent dive into using it for some real-world tasks started to shift my perspective… slightly.

I’ve long been generally skeptical of “AI” in all its forms. The idea behind it is fun, the technology is fun (if ethically problematic), but overall I’ve been rather underwhelmed.

I’ve pretty much had access to ChatGPT from the beginning. I’ve asked it questions and felt underwhelmed by the replies (I know it has gotten better). I’ve seen the hallucinations on things I know about, so I feel like can’t really trust it on things I don’t. It talks confidently, it sounds convincing, it’s probably entirely hallucinated. I have however also then seen friends using it to analyse games of online werewolf and summarise that day’s activities and chat to help them keep up when they’re too busy. It’s usually at least able to follow and accurately report on 50% of what is happening (it can’t read people though yet, so ymmv).

On the image side, I’ve used Stable Diffusion to generate images. I’ve also used image generation tools like Midjourney since the early days and seen it go from strength to strength. I’ve used it for quick throwaway joke images, but never really managed to get it to do quite what I had in mind (though usually good enough for a joke). I’ve also used and seen friends using Bing Image Creator to great success to generate meme-images for in-jokes among friend groups.

And while these tools are usually “good enough”, I’ve also then seen them generate 6-fingered people or obviously fake images. I’ve also laughed at the Will Smith eating spaghetti test (and then been impressed at the later improvements to this) or the fact LLMs can’t spell strawberry.

And I know other people have more success with it. I’m constantly being bombarded on social media by posts from people saying how AI is going to put people out of jobs, and that if you’re not spending 50% of your day in ChatGPT you’re wasting your time. But despite this I’ve just not succumbed to the hype, all my experience to date was very firmly at the level of “Meh, It’s ‘ok’. Not good. Certainly not a tool I expect to use all day every day”.

But earlier this month I finally dipped my first toe into the LLM/AI Coding pool…

What if I told you I’ve tracked every single night of sleep for the past 14 years using an app that’s basically been dead for over a decade?

If you actually know me, then it probably comes as no surprise at all. Anyone who has spent any considerable amount of time chatting with me will know a few things about me:

1. I don’t like change.
2. I like graphs and data.
3. I do not always have the best relationship with sleep.

So when I say “I’ve been tracking my sleep and graphing” the response is likely to be one, or both, of “of course you have…” or “but, why?”

The app in question is SleepBot which is one of the early arrivals on the sleep tracking scene and the initial premise was fairly simple. You press a button to say you went sleep, and another when you finished. And then you can rate the sleep and write some notes about it. This button can be in the app itself or a widget on your homescreen. Importantly - it also drew graphs.

Eventually you could also put the app on your bed next to you and it would record any unusual sounds you made, or motion you made. And it had a smart alarm feature, and it had web-sync capabilities. I never really used any of these features other than the basic tracking.

I have been a long-time user of proxmox for my virtualisation needs, starting back in the days of Proxmox 3 (prior to that I was running Ubuntu with OpenVZ rolled in by hand). Back then I only had a single server and resources were tight, so I deployed a lot of my workloads as OpenVZ containers.

As time went on, proxmox switched to LXC, and I dutifully converted all my containers to LXC and kept on going.

More time went on and I added more servers, and ended up clustering them for ease of management. Then eventually replacing them all with more powerful nodes so now resources were no longer a concern. I also eventually added Ceph into the mix (using proxmox’s built-in support for ceph) and 10G Networking so that I had shared storage for VMs and could the start doing VM migrations between nodes quickly.

But, LXC Conainers have flaws - live migration only works for full-VMs. LXC Containers have to do a reboot to actually move onto and start running on the new node. For some workloads this isn’t really noticable or a problem - but for others this is quite bad.

Also, as more and more of what I run involves Docker, it’s a lot easier/nicer/safer to run these workloads in actual VMs rather than LXC containers.

But installing and configuring full-VMs was a chore. With LXC Containers you could be up and running in minutes by just deploying a template. Full-VMs required a full installation to an empty disk. This could be automated using kickstart/preseeding etc (And I wrote a tool to help manage a pxe-boot environments for this purpose). But over time, this has now become trivial as well - cloud-init is now supported directly within proxmox and all the major OSes provide cloud-init compatible disk images, so getting a fresh VM is a matter of cloning a template, updating some cloud-init settings and starting the VM.

Due to all of this, almost all of the VMs I create these days are full VMs. Anything new - it’s a VM, which gives me all-the-good-stuff™.

But I still have a lot of legacy LXC containers. These all end up suffering any time I do hardware/software maintenance on the host nodes, or if I have any problems that require putting a host into maintenance mode.

It’s time to fix this.

I have recently added some iscsi-backed storage to my proxmox-based server environment, primarily as an off-server location to store backup data.

For a multitude of reasons, such as the sensitive nature of the data, the fact that the physical storage lies outside of my control, and just good security hygiene - I wanted to ensure that the data is all encrypted at rest.

I wanted to be able to use this iscsi as a storage target for proxmox allowing me to just add the volumes to VMs allowing HA, and I didn’t want to have to do encryption inside every VM incase I accidentally forgot to enable it for one of the VMs (remember, the storage is hosted external to me so I have no control over the physical access to it) so to do this I have made use of LUKS encryption on the iscsi block device that I am presented with and then I run LVM over the top of this. (LVM-on-LUKS as-opposed to LUKS-on-LVM)

Once again, I have replaced the theme of this blog.

Unlike the previous theme, this one is actually one I mostly ended up designing myself rather than just finding one that I mostly liked and running with it, and given that, I figured I’d talk a little bit about the thoughts behind it and how it came to be and what else I’ve done behind the scenes.

This post is part of a series.

1. DNS Hosting - Part 1: History
2. DNS Hosting - Part 2: The rewrite
3. DNS Hosting - Part 3: Putting it all together (This Post)

In my previous posts I discussed the history leading up to, and the eventual rewrite of my DNS hosting solution. So this post will (finally) talk briefly about how it all runs in production on MyDNSHost.

Shortly before the whole rewrite I’d found myself playing around a bit with Docker for another project, so I decided early on that I was going to make use of Docker for the main bulk of the setup to allow me to not need to worry about incompatibilities between different parts of the stack that needed different versions of things, and to update different bits at different times.

The system is split up into a number of containers (and could probably be split up into more).

This post is part of a series.

1. DNS Hosting - Part 1: History
2. DNS Hosting - Part 2: The rewrite (This Post)
3. DNS Hosting - Part 3: Putting it all together

In my previous post about DNS Hosting I discussed the history leading up to when I decided I needed a better personal DNS hosting solution. I decided to code one myself to replace what I had been using previously.

I decided there was a few things that were needed:

• Fully-Featured API
  • I wanted full control over the zone data programmatically, everything should be possible via the API.
  • The API should be fully documented.
• Fully-Featured default web interface.
  • There should be a web interface that fully implements the API. Just because there is an API shouldn’t mean it has to be used to get full functionality.
  • There should exist nothing that only the default web ui can do that can’t be done via the API as well.
• Multi-User support
  • I also host some DNS for people who aren’t me, they should be able to manage their own DNS.
• Domains should be shareable between from users
  • Every user should be able to have their own account
  • User accounts should be able to be granted access to domains that they need to be able to access
    • Different users should have different access levels:
      • Some just need to see the zone data
      • Some need to be able to edit it
      • Some need to be able to grant other users access
• Backend Agnostic
  • The authoritative data for the zones should be stored independently from the software used to serve it to allow changing it easily in future

These were the basic criteria and what I started off with when I designed MyDNSHost.

I run ubuntu on my servers, and since moving to Hugo, I wanted to make sure I was using the latest version available.

The ubuntu repos currently contain hugo version 0.15 in Xenial, and 0.25.1 in artful (And the next version, bionic only contains 0.26). The latest version of hugo (as of today) is currently 0.32.2 - so the main repos are quite a bit out of date.

So to work around this, I’ve setup an apt repo that tracks the latest release for hugo, which can be installed and used like so:

sudo wget http://packages.dataforce.org.uk/packages.dataforce.org.uk_hugo.list -O /etc/apt/sources.list.d/packages.dataforce.org.uk_hugo.list
wget -qO- http://packages.dataforce.org.uk/pubkey.gpg | sudo apt-key add -
sudo apt-get update
sudo apt-get install hugo

This repo tracks the latest hugo debs in all 4 of the architectures supported: amd64, i386, armhf and arm64 and should stay automatically up to date with the latest version.

This post is part of a series.

1. DNS Hosting - Part 1: History (This Post)
2. DNS Hosting - Part 2: The rewrite
3. DNS Hosting - Part 3: Putting it all together

For as long as I can remember I’ve hosted my own DNS. Originally this was via cpanel on a single server that I owned and then after a while I moved to a new server away from cpanel and moved to doing everything myself (web, email, dns) - hand-editing zone files and serving them via BIND.

This was great for learning, and this worked well for a while but eventually I ended up with more servers, and a lot more domains. Manually editing DNS Zone files and reloading BIND was somewhat of a chore any time I was developing things or spinning up new services or moving things between servers - I wanted something web-based.

There wasn’t many free/cheap providers that did what I wanted (This was long before Cloudflare or Route 53), so around 2007 I did what any (in)sane person would do - I wrote a custom control panel for managing domains. and email. and it was multi-user. and it had billing support. and a half-baked ticket-system… ok, so I went a bit overboard with the plans for it. But mainly it controlled DNS and throughout it’s lifetime that was the only bit that was “completed” and fully functional.