This is the 101st post on my website. Humans are fascinated with milestones, especially when they line up with base 10 numerals. As a human myself, I enjoy these milestones, and it's not often I get a chance to do some meta content, so let's give in and see where this goes.

<nit>

Yes, I'm aware this is a look back on 100 posts. It'd be nice if this was post 100, but "Looking back on 99 posts" doesn't have the same ring to it.

</nit>

100 posts ago, post 1 on here, was published March 16th 2016, which feels like a completely different time. It's a post about a presentation I did at work about React Native - which nowadays I probably wouldn't consider a true blog post, but that's a different discussion. If we do ignore that, the first proper post is 28th June 2016, about ProtonMail. Since then, I've written 98 other blog posts of varying quality, length and topic. It's been an interesting ride, and I've learnt so much about myself, my interests, and the internet at large.

In times like this, a well-rehearsed author might ask their readers what they would be interested in hearing about, Q&A style. I however am not that well prepared. So instead, I'm going to ask the questions and tell the tales I think are most interesting, and hope that you find the answers interesting too.

#Why do you have a blog?

Honestly, I don't really know. By my recollection, I've had a website of some kind since around 2014 (2014-05-15 is apparently the day I registered theorangeone.net), as a personal online portfolio. At some point, that gained a blog area, and things spiralled out of control from there. The first post is from early 2016, but why I decided to write it is sadly lost to time.

I've always enjoyed teaching people, and generally found myself with knowledge others both don't have and want to know about, which is also why my theme is all over the place. I'm a huge believer in the idea that the best way to learn a topic more and cement your own understanding is to try to teach it to others. So writing things down for others to consume makes a lot of sense. Also, being able to write an express yourself is a very useful skill both professionally and personally - so it's helpful to have an outlet to practice.

The blog is also a bit of a driving force in itself. In order to write content, I need things to write about, so I need to be doing weird and interesting things which make good content. Many of the ideas for blog posts I come up with whilst writing others, or pop into my head at random in conversations with others.

#Covid

We're 3 years in at this point, I'm hoping I don't need to explain what covid is.

At the start of lockdown in March 2020, I found myself with a huge amount more time on my hands. Gone was the commute, needing to get in early to avoid traffic. In its place, lots of time and a desire to be productive and spend more time tinkering. Having gained a few hours a day back, I had time to experiment with things I'd been meaning to for a while, and share what I'd been learning. Deploying new applications to my server, experimenting with networking, or just writing up things I'd already done.

The start of lockdown was also the start of the selfhosted.show Discord server, which launched just a week later. The show, but more importantly the community around it, has been a huge driving force behind my blog's growth. Oddly enough, I didn't see a huge growth from my appearance on Episode 42, but that was only a 5-minute segment (Alex, Chris, you know where to find me 😉). Having a community to usurp was a valuable resource for collecting project ideas, discussing solutions, and generally driving each other forward. If you're interested in running your own cloud applications, I can't recommend the Discord (oh and the show, I guess) highly enough.

With all of that, It's no surprise that that's exactly the time my output drastically increased, sustaining 1 post a week for nearly 3 months, and rarely missing a month without posting something. I had so much time and energy and interest - I was unstoppable.

Why that momentum hasn't persisted is a long and yet not remotely surprising story, which starts in 2022:

#2022

If you look at the frequency at which I post, 2022 is quite an obvious decline. There were large gaps between posts and momentum decreased. I don't think the quality decreased, but even my analytics show a decline in views in that year too.

The cause of this: I'm not entirely sure. 2022 was quite a year for me for personal reasons, both good and bad, and as a result, the last 2 years started catching up, and taking a toll on my mental health. It's fairly well understood how to keep our bodies healthy, whether it be through exercise, a healthy diet or seeking medical attention when something doesn't "feel right". But mental health is equally as important, and if you ignore it for too long, and don't try and take care of yourself, it'll hit you hard.

I think my main issue was trying to do too much at once, and sustain it for too long. Sure, during lockdown I was able to be very productive, but as my life changes, my output is allowed to change with it. Trying to keep up the same momentum on content, whilst trying to take on more personally is a recipe for a bad time - something my brain is starting to realise. There's only a finite number of hours in the day, and rather than forcing and pressuring myself into being productive, and feeling bad when I wasn't, I should have listened to my brain, done something it was in the mood for, and be sustainable. If I want to build something, then I should. If I want to write, then do. If I nerd-snipe myself into a rabbit hole of researching a random obscure topic and come out with nothing to show for it, then that's probably been quite a fun evening.

If you're reading this, and thinking you might be going through the same, I have 1 piece of advice for you: It's ok to take a break. It's ok to not be as productive as you may want to be, or be productive on the wrong things. Just because you're not spending all your evenings keeping up-to-date with the latest news, technologies, or tinkering with something, doesn't mean the time is wasted. In much the same way that exercise keeps us physically healthy, switching off your brain and going to do something not remotely taxing can keep us mentally healthy. Listening to my brain, and letting it either dive hard into researching some new tool or idea, or dive hard into the sofa - either is fine with me. If I let my brain switch off, then when it comes time to switch it on, it's in a healthier and more productive state.

To clarify, this isn't a cry for help, or a sob story, or me trying to seek pity. I'm ok, in fact in some senses I'm more than ok. Both personally and professionally I'm in a much better place than I was in 2022. Slowly but surely, I'm returning to writing, and I have lots planned (more on that later). Will I return to 4 posts a month? I doubt it. Will there be 4 months between posts, I doubt that too.

Posts per year

#Comments

Ah yes, back on to something more positive. In June 2022 I added comments to my website, based on Commento (Commento++, to be precise, since the original is pretty dead). From an engagement perspective, I've loved finding out if I'm completely wrong, what people think, or just hearing that I've helped someone out. Seeing people referencing my posts is 1 thing, but getting an email notification because someone found something you wrote helpful is another feeling entirely.

Just above the comments are icons for social media platforms. If you click on them, it'll create a post for you on your platform of choice, spreading the knowledge further. I have no idea how often the buttons get used, but I suspect not a lot - from what I've seen most people link to my posts as part of existing conversations, rather than starting new ones. Regardless, it's still heart-warming seeing people reference my posts in forums, Reddit threads and other places on the internet. If you're one of those people - thank you!

#Writing styles

Finding your voice in writing is incredibly difficult, and whilst I'm slowly finding mine, it's something that only comes with trial and error (more error, sadly).

Before 2018, I didn't write a huge amount online. Sure, there are blog posts from that time, but they were written more with the expectation of shouting into the void, and for my own personal reference than any expectation of an audience. Before 2018, even the idea of any kind of public speaking, written or otherwise, would have sent me running. In 2018, for reasons which are also a mystery to this day (mostly), I tried my hand at some public speaking (event commentary of all things), and my self-confidence has built from there. Finding my voice, building confidence, getting funnier (or at least I think so), can all trace their roots back to that moment.

I assume people read here for 2 reasons:

  • It's me, and they like my style of content
  • They're also interesting in the things I'm writing about, and enjoy hearing from like-minded people
<note>

Or both, both is good.

</note>

Having good content is critical. Sure, a skilled presenter can probably make any topic sound interesting, but I suspect that's not you, and it's definitely not me. There's no point focusing on writing style if you don't have something to write about, even if it takes longer. The same is true for getting started on YouTube: Focus on the quality of the content, not of your camera (unless you're a videographer, I guess).

Some of you reading this may have met me in person. Most of my posts are written how I talk, or at least how I talk on a semi-prepared subject. When reading, I "subvocalize", in that I read the words to myself in my head (not moving my lips) as I read them. To some, that may sound incredibly inefficient, and to others it's probably crazy to think of anything else. I'd assumed everyone read this way until hearing about it on Hello Internet. The fact I subvocalize is probably why my posts tend to read as more of a script - because as I'm writing it that's exactly how I read it. On the plus side, it makes it easier for me to get my thoughts down into words. The downside however being that some elements of tone and emphasis get lost in places I'd naturally insert them when speaking.

Something I'm always trying to work on is making my posts longer. There's a fine line between giving people plenty of context and just boring them with noise. My posts are all fairly short, especially given the depth of subject they're talking about. I'd love to get better at expanding on topics more, and not running through them quite as quickly. I'm not quite after the same extreme as fasterthanlime has, but being able to talk a little more about the journey rather than merely the destination helps build the narrative. As someone who subvocalizes (a subvocalizer?), I tend to think posts are longer than they actually are, because they take a long time for me read. The reading time counter at the top of the page is based on Medium's reading time of 265 WPM. In contrast, presentations and conversations can be as slow as 150 WPM. I'm currently using the obsidian-reading-time plugin to show the reading time as I write.

#FAQ

<nit>

Because I'm not seeding these from you the reader, It's hard to call these "Asked", and those that have been asked haven't done so "Frequently". But hey, there's definitely Questions!

</nit>

#What's your favourite / least favourite post?

This must be how parents feel trying to choose their favourite child, if they had 100 children.

<fact>

Mum, Dad, come on - you know it's me!

</fact>

I don't really have a favourite. I often finish writing, immediately publish, then forget all about it and move on to the next thing that's peaking my interest. With that said, there are a few notable ones (in no particular order) I'm proud of and that I reference time and time again:

  • My First CVE isn't much of a post in itself, sure - But it marks a milestone I'm sure proud of.
  • Calming down Clickhouse is one that gets far more traffic than I would ever have expected. I wrote it whilst I was trying to deploy Plausible, and wondering why Clickhouse was writing 1GB an hour. Because Clickhouse is quite a niche tool, and those who deploy it are generally doing it at quite a scale (e.g. Yandex), it was hard to find resources for making it more lightweight - so I wrote my own. It's now referenced in the Plausible documentation, and gets a lot of views from corporate-sounding referrers.
  • Exposing your Homelab is probably the post I reference the most. For people getting started running servers in their home, it's not often clear the best ways to access those services from outside the home. There are plenty of different ways, and each has pros and cons, but it was hard to find a single resource discussing all the potentials - so again, I wrote my own. In the 3 years since I wrote it, a lot has changed. Tailscale's user base has absolutely exploded, Cloudflare Tunnels have proved far more popular than I could have thought, and I've come up with and seen yet more ideas for how to do this - so much so I might just start an in-depth series for each one.
  • One of the first conference-style talks I gave was an internal work one about optimising Django's ORM queries - after we were having a lot of performance issues as our applications scaled. Whilst the topic is niche, and it's a little outdated, the content is still sound, and I'd love to revisit the topic and expand on it, probably breaking it into a few separate talks (which sadly DjangoCon EU 2023 rejected my draft for).
  • My server CPU replacement was one of the more in-depth posts I've written. Not so much in length, but in that it followed a good narrative of what I was actually going through: Research, Implementation, Review. Power consumption is a big deal to me and lots of others (apart from people in the USA, for some reason), and I spent far, far too long looking into my options before settling, resulting in no noticeable performance difference (after some tinkering), and a ~25% reduction in power draw.

As for my least favourite, well that's a difficult question. My older posts are definitely a much lower quality than my more recent ones, not to mention significantly shorter, but I wouldn't necessarily call them "bad". They're a part of my history, and show a nice progression of where I've come from to where I am now. With all that said, there's a reason there's no Part 2 to the "Facebook Cleanup" post.

#What's the best / worst performing post?

Whilst I don't write for popularity, I love a good spreadsheet! I don't write for popularity, but it's still both helpful and interesting to know which posts perform best (and worst).

Plausible helpfully provides an export of all page views for all time, but secretly paginates it to 100 pages, and there's more on my website than these 101 posts. To get around that, I had to write my own export script to grab the raw data from their "breakdown" API - which is surprisingly fast for an analytics API. Pulling it into Excel LibreOffice, and a quick sort later, I have the top and bottom posts of all time, as of May 4th:

Top 5 (by views, descending):

  1. HomeAssistant Temperature & Humidity Sensor with ESPHome
  2. Mount NFS inside LXC containers
  3. Server build 2020 - Proxmox setup
  4. Nvidia GPU passthrough in LXC
  5. Exposing your Homelab

Bottom 5 (by views, ascending):

  1. Hacktoberfest 2019
  2. Hacktoberfest 2018
  3. Cyber Security Month 2016
  4. Fix steam under linux with an iGPU
  5. React Native intro dev meeting

However, there's a problem. "Automatic builder discovery with BeautifulSoup" was only published a few weeks ago - Calling it the "11th worst post I've ever written" is a little misleading. It's not the worst, it's just not had as much opportunity to be great as others. Instead, the statistics should to take into account how long the post has been published, rather than looking simply at absolute view numbers. After a little bit of VLOOKUP magic, some manual data massaging, and the aforementioned filters, I had a spreadsheet with some additional useful columns:

Path Visitors Page Views Publish Date Publish Year Days published Views per day

/

=VLOOKUP(A2, $PageDates.A$2:B$101, 2, 0)

=YEAR(D2)

=DAYS(TODAY(), D2)

=C2/F2

Top 5 considering publish date (descending):

  1. Server build 2020 - Proxmox setup
  2. HomeAssistant Temperature & Humidity Sensor with ESPHome
  3. Docker in LXC
  4. Mount NFS inside LXC containers
  5. Nvidia GPU passthrough in LXC

Bottom 5 considering publish date (ascending):

  1. Cyber Security Month 2016
  2. Hacktoberfest 2018
  3. Hacktoberfest 2019
  4. Fix steam under linux with an iGPU
  5. React Native intro dev meeting

There's no change in the top posts, which was surprising. Cyber Security Month 2016 rockets to the bottom slot though, probably propelled by its age, and the fact "react-native", "hacktoberfest" and "steam on linux" are fairly well-searched topics.

You'll notice at no point here have I mentioned absolute viewer numbers. Plausible has the ability to make the analytics dashboard public, allowing anyone to view traffic numbers. I personally don't have that enabled (sorry), and don't really plan to. The numbers are useful for me to determine which posts are performing well, or where my posts might be linked from, but they're not especially useful to anyone else. And shrouding just how many views I get in a little mystery is also quite entertaining.

#Where are your readers?

I, and the server which runs my website, are in the UK, but according to the analytics, most of you lot aren't.

As a surprise to literally no one, the top country reading my content in the last 12 months is the USA, at 27%. All of my content is in English, because sadly it's the only language I speak, and the USA makes up a huge portion of the english-speaking internet world, especially those interested in technology. For a time I considered moving my website there to better serve the audience, as my monitoring notes that response times are almost 6 times higher in New York than they are in London, and California is almost 10x. I elected against it, because I don't want to slow things down for anyone else, not to mention building a CDN yourself is hard, and I'd rather not route yet more of the internet through Cloudflare. After the USA is Germany at 12%, followed by the UK at a lowly 6%.

Looking at were readings come from is more interesting, for me at least. Plausible tracks "Top Sources", which uses the Referer header to determine where the person was before they hit my site. I take SEO quite seriously to try and improve reach, which shows as 41% of traffic comes from Google search, and 2% from my search engine of choice, DuckDuckGo. The source currently only uses the Referer header - some when I should really start adding UTM parameters for the links I post around.

From a security perspective, and because I'm a nosy person, I quite enjoy seeing people's self-hosted applications appearing, and seeing what people are running and why they might be taking note of my site. A few internal Jira instances, random internal IP addresses, podcast players, chat applications and email clients. Referer headers can sure be a privacy problem!

#Do any of your posts have an interesting story behind them?

I'm afraid to say the answer for this one is quite boring: No. Every post has a story behind it, of course, but in most cases said story is just that I found it interesting enough at the time to write about, or that it played enough of a part in my technological life to talk about. No story stands above the others. But of course, that's with my definition of interesting - not yours. I'm not going to sit here and talk about why I wrote every one of my posts (that'd be quite tedious for us both), so I guess we'll never know if any posts have a story you find interesting behind it.

With that said, "How I lost some data" does speak for itself when it comes to interesting stories.

#What's your flava theme?

This is both very easy and very difficult to answer: I write about things which interest me. What interests me? You know about as well as I do.

I'm the kind of person who is more of a "generalist" - I don't really specialize in a single subject area, and like to know as much as I can about everything, because it means when I need to research or repair something, I have a wide gamut of knowledge to pull from (which is massively helpful during incident response!). A "Jake of all trades", if you will.

There's definitely a strong focus towards technology, specifically self-hosting, security, privacy and programming - all domains which interest me anyway. And if you're reading this, they probably interest you too. Self-hosting has been quite the numbers kick over the past 8 years, helped along by the community around the selfhosted.show podcast, which is definitely where the most engagement comes from. Professionally, I'm also part developer, part sysadmin, part infosec, which aligns quite nicely with my interests, and is why there's the odd in-depth programming post kicking around.

#Do you plan to keep going past 100?

All the time there are ideas in my brain and time on the clock, I'll keep producing content so long as you keep reading it (or consuming it, in whatever format it may be). It's not been the smoothest of roads, but I can't imagine I'd be the person I am, professionally or otherwise, without having experimented with all these technologies and written about them in some form.

#What are you looking forward to for the next 100 posts?

At this rate, the next 100 posts won't arrive until 2030 - which is a long way away. I'll be 7 years older, 7 years wiser, and probably have a blog post backlog about 70 times longer. The state of my server, website, profession, basically everything could completely change in that time, or it could be identical (well, I'm hoping I'd do a server upgrade within 7 years).

Something I'm at least hoping for with the next 100 posts is a reduced barrier. My drafts and ideas folder has 47 items in at the time of writing, and I'd quite like to bring that number down. Either by just accepting that a topic isn't enough to write about, or by actually starting to write about it and see where it heads.

#Can you tease anything you're working on?

Well, you've got this far, so I should probably give you some kind of reward - even if the reward is a bit of a tease.

In the short term, there are a few things I'm vaguely working on. Who knows when they'll come out.

  • A better, more resilient and more secure networking setup to expose my home server to the internet - without the need for a reverse proxy on my VPS (WireGuard HAProxy Gateway V2)
  • Moving away from Proxmox, towards "something else" (not telling 🤫) that's a bit easier to work with under the hood
  • More long-form posts, derived from conference talks (Yes, I'm a reason person who exists in the physical world, honest)
  • Home Automation, and solving some laziness problems with technology

Something I've started doing recently, in an attempt to ride my motivation a little more is taking more notes about the things I'm working on. It's far easier to convert notes into a post than it is to try and remember everything that happened, especially if I'm trying to build some form of narrative. There's 16 items in there already, that I'm slowly working through. Some might get completed this month, this quarter, this year, maybe even this week - but they form a slightly more refined backlog.

And of course, there's always those 47 draft items.

#What's coming next?

No one knows - I'll go wherever the wind (i.e my own motivation) takes me.

I work on things in fairly random orders, jumping around quite a bit. I'll write down some notes on a topic, maybe I'll complete it next, maybe I won't. I tend to get interested in something, write about it, and hope nothing distracts me in the meantime before posting. Even on a day-to-day basis, exactly what I spend my evenings doing changes massively.

I'm looking forward to getting back to more bleeding-edge content, trying things that other people aren't or don't want to commit to - that stuff seems to perform quite well. With that as a driving force, I'm hoping to keep up some kind of momentum, but we'll see how that goes. What post is coming next? I don't even know.

Now enough of this long-form style content, and back to "regularly scheduled programming", which is neither regular, scheduled, nor necessarily involves programming.

Share this page