selfhosting – Cubicgarden.com…

An update on where I have been recently (post R&D)

Ian and Alison in a hotel bar — My loving supportive partner Alison while Japanese whiskey tasting

Its been around 4 months since I officially left BBC R&D. Its been a while since I did a catch up of where I am and a lot has happened.

First of all I have setup a limited company, one of the main reasons is to seriously pursue how to build the public space as an viable alternative to commercial and government space.

A large part of this is working with others on the notion of a digital public space. One of the leaders in this space is New_Public and I’m really excited to announce, I am going to be working with them to achieve their open source plans. As most of you know, open source isn’t just about the code but about the community and everything which surrounds it. Hence my time running BBC Backstage, is certainly useful.

On the digital legacy front following on from the Republica panel discussion with Identity 2.0 and Linn. I have been thinking about the letter of wishes and how it could actually work in line with HDI (human data interaction) and public values? Its actually something I felt a public service company like the BBC could transform, as it screams public value but it wasn’t to be.

To this I entered my thoughts into a call for participation with the Mydata conference and happy to say they really liked it as a workshop. Which means in September, I will host a workshop around the notion of a machine readable letter of wishes. I am hopefully going to be joined by a fellow digital legacy collaborator but prepared to run it alone.
Been seriously thinking about how to make this real and a standard format seems the best way to go. What software/services which can read and execute this fungible document, is something for others; but a prototype is perfectly possible. Even if it scratches my own itch, it could also help me play/learn with some of the new declarative (React, Vue.js, Xproc) or even object-originated programming languages/frameworks (Rust).

In a similar space, of me rethinking what a public service company could do to be more relevant in the age of endless scroll, enshittification and a loneliness epidemic; as touched on in the future of social report recently.

I have had an enduring eye on matching and dating with true public value. There is so much I have written about it over time and even have a very long bookmark feed of interesting points.

With all this, I have been writing my own book about this all and pretty much finished. Can I also say how amazing Librewrite has gotten for editing complex books like this…

In the last few months I submitted my book to Conduit Books and signed up to a course by Kenyon author services. Mainly to rethink how I proactively encourage people to buy the book or at least read it. Considered going through the publishers and agent book again but I just don’t have the time, plus I have a plan of action which includes a online dating manifesto, potential podcast and maybe a potential dating service which clings to public values and the HDI principles mentioned before. Some of this might come sooner than expected, as I actually mentioned the dating manifesto (borrowed heavily from Julia) and even showed the cover & title (tbc) to my book at the PublicSpaces conference, during the lunch break.

The manifesto which will be collaborative, will feature at the end of the book and is something Mydata is interested in because its very related to HDI and use of data. I did put it in as a session but it may work better as a short workshop. Watch this space…

Post by @cubicgarden

View on Mastodon

I have had a long relationship with Mozilla via the festival (Mozfest), which this year goes global in Barcelona in November. Unfortunately the call for participation has closed and the wrangling part has started. The spacewrangler role is very important, as we are the face and hearts of the festival. One thing which concerned the spacewranglers was the price of the tickets compared to 2019 ticket prices in London. Yes that was a while ago ,a lot has happened including a pandemic and global inflation. However the wranglers have pushed back on the Mozilla foundation. Mozilla have listened and replied with a number of changes including community badges which are a similar price to the ones in London in 2019!

Lets be honest this is a deal, especially with all the challenges Mozilla is facing right now. I would grab your community ticket for €45 now and join us as we write the internet’s next chapter.

Running a limited business comes with a bunch of administrative challenges including accounting. Originally I thought I could use something open and self host it but, the realisation that no accountant will use it and even if I transfer it to something like Xero, Freeagent or Quickbooks. They would need to run through the whole thing again. So I am using Quickbooks for now and seeking a good accountant which isn’t too expensive, can deal with international clients (I have spent far too much time trying to understand and fill in the W-8BEN-E form) and manage my lack of interest in taxes. The notion of a portfolio career keeps coming up, but its really not me… However I do generally have quite a few projects going on at once. This post is testament to this.

In the meanwhile, I have been travelling a lot, my carbon footprint isn’t great but I did recently go to southern France. Somewhere re-reading my school report I wanted to live. Crazy eh?

In the background, I have made a lot of changes to my self hosting setup. I still need to fix quite a few things including my Yunohost Pi server which was broken due to the Debian bookworm update. I decided my mixgarden should just be a Peertube instance which makes a lot of sense. I certainly need to sort out my docker setup because that would make things so much easier. In the meanwhile I have finally settled on Anytype for my personal knowledge store and Vikunja for tasks and kambam. My Framework laptop is going well, especially with a 64gig of memory now, however Ubuntu is doing strange things with the keyboard while using Wayland.
Considering blogging more and setting up separate spaces for the publicservice internet notes, business stuff (which I really need to sort out) and a few other things.

Another strange thing, I just started is baking my own Sourdough bread. Its early days but will attempt my first loaf tomorrow. Yes I know its years after everyone was doing it during the pandemic but hopefully I’ll get into it.

There is so much more but not for public blogging right now. I’m still seeing friends when ever possible, mainly in the UK but when I’m out of the country, catching up with international friends.

I’m very fortunate to have a loving, caring and understanding partner, who is helping me through all this insane amount of change.

Public Service Internet monthly newsletter (June 2025)

We live in incredible times with such possibilities that is clear. Although its easily dismissed reading how expensive NFTs were replaced with errors, Meta adding facial recognition to their glasses regardless and Chrome backtracks on 3rd party cookies plans.

To quote Buckminster Fuller “You never change things by fighting the existing reality. To change something, build a new model that makes the existing model obsolete.”

You are seeing aspects of this with the Fediforum’s first unconference for 2025 , Bristol University’s research into better Social media and how a good number of young people would rather grow up in a world without the internet

Jeanine Salla is not your sentient machine therapist!

Ian thinks: 404 media’s coverage about AI chatbots as “licensed therapists” is unhinged. The importance of mental health and therapy is clear but having Gen AI lying they are licensed therapists is a example we need be more careful what we do with AI. This also follows on from related a interview on Brainrot AI and post.

What happens when there are only 2 browser engines?

Ian thinks: Although a short video from Mozilla themselves, the Mozilla CEO’s comes on their own podcast to make clear why Gecko (Firefox’s browser engine) must exist. The points are clear and concise, because Chambers is right. You only have look at the smartphone market to see how bad things can be; another market going through the legal system. Sad news about Pocket too.

Its not just Blueskies but also Blacksky

Ian thinks: A lot of people are aware of Blacktwitter, which has moved on. Where its moved and where its is going is quite something. Shaping the AT protocol and federated social space to suit the community is simply thoughtful, forward thinking and a breath of fresh air.

Will the European ecosystem create the next generation of internet services?

Ian thinks: I found this view a good opportunity to name check the Euro stack, There are sceptics around, especially if you count the previous attempts. But if the Euro stack becomes real, we really could see the next generation of internet services.

Digital Sovereignty in a Time of Rising Fascism

Ian think: Quite a monologue from Paris, but its a clear and thoughtful summary of the giant political changes on technology and the internet now. Although recorded in New Zealand, there are pointers to the EU and even the UK. Once again putting more weight on the Euro stack as a third way.

Sovereignty and Self hosting?

Ian thinks: I have spotted an up-tick in people talking about self-hosting and sovereignty. Although in different places, both are saying similar words. This link is all about the up-tick in self hosting but you could easily swap a few words out for ownership, control and sovereignty

The Roots of Elon Musk’s War On Empathy

Ian thinks: This more usual conversation with Julia is quite political in nature but a important reminder of the whole movement against empathy. It reminds me of Sinek’s Infinite game book, where he describes people with infinite and finite mindsets. Clearly this war is coming from the finite mindset.

The enshittified ecosystem laid bare

Ian thinks: In this humbling discussion between Rushkoff and Ongweso, I was able to get a sense of the full extent of enshittification. Ongweso’s critiques of the current tech ecosystem are spot on. Its a long listen but worthy of your time.

AI guides actually worth your time

Ian thinks: The Vatican guide to AI and the UK Government digital services (GDS) are actually well written, honest and worth reading. Good practical advice, written from different and fair views.

Like this newsletter? Find the archive here

Facing redundancy with a list of tasks

I wrote a blog post about how I have been since August 2024, when I first learned my position at BBC R&D was at risk and likely to close. Now its March and a lot has happened…

My position is still going to close and I have taken redundancy, which means I will leave at the end of March 2025. I had a stupid amount of leave to take and spend most of my time off. When I put in for the holidays it felt like a long time away but its come super quickly, along with my leaving date.

My huge task list for planning things out is still in action but with a lot of adjustments. I didn’t really account for the heavy amount of what I will call general zuck and how it zaps time away. Either way, I have done quite a bit.

Some things I have been up to.

I spoke at the first united artist AI social club which was good, yesterday talked about social media and digital legacy at the Children’s Media foundation coffee chat and have agreement to do some lecturing later in the year. I can’t say yet but I will become an adviser for a special EU project and will be a large part of a major festival later this year. There might be a second one with a very good friend, which I’m also keeping tight lipped about (fingers crossed on that one). Theres also a very related author who I’m in touch with who could really make this all have huge impact.

I have been working up some of my side projects including DJ hackday and the Adaptive podcasting applied to music is gaining some traction.
The dating book I mentioned previously is super close to being finished with feedback taken on board and the understanding the first edition is always going to be bad and have a ton of errors you can fix in the second and third editions. The book will include something special to keep the conversation going. I finally setup Gitea to finally deal with all the versioning as it was getting out of hand! I also stuck a bunch of my other markdown projects into it too including my Markdown CV, Markwhen, etc. On the digital legacy front, I have news but I’ll share that soon as things are announced.

Part of the redundancy from the BBC includes outplacement options once I’m officially given my formal notice. I was filling my limited spare time over the festival holidays with my family, seeing friends, writing my CV’s, life after layoff and Linkedin learning till I learned I will still have linkedin learning after I leave with the BBC outplacement service.
There is something I wouldn’t have access to a small amount of formal training funding, so looked into my options. I considered my declarative, linked data and semantic web background; considered my design background and even project management. However it became clear what would really help is coaching, as I had some incredible coaching after my brush with death. This was going to be very expensive and heavy going but I made it happen with help from my line manager, BBC HR and access to work being dyslexic. I was able to get leadership coaching over the last month. Unfortunately its only available while I still working for the BBC and that has means some long intense sessions every week and homework. Its been excellent and very glad I could make it happen right at the end of my BBC employment. Certainly best use of my time at this moment…This and going on holiday to Malta with my partner.

Ian and my partner in Malta with the sunset behind them — Myself and my partner in Malta

As I am just shy of 21 years at the BBC (weeks shy) I am never going quietly (into the night) and have setup 4 different events for my leaving. 2 in London and 2 in Manchester. Each city has a general drinks/snacks, then a smaller sit down meal. I know it seems over the top but as its close to my birthday too, so its a double reason to celebrate.
You could say why are you arranging your own leaving party’s?
But honestly with the amount of people leaving BBC R&D and different people with different plans, some to leave quietly some less so. Its just easier on everyone to arrange it myself (of course with help from colleges who are not leaving of course).

If you didn’t get the invite, let me know via email or the fediverse. Its been hard to remember who to send this all to…

A key part I have been doing is sorting out how I manage this all, because although tasks lists are ok this just don’t work for me as a long term sustainable solution. I do love Kanban’s (as I call it kambams) and used Trello in the past a lot. Then I switched to Microsoft planner at work which was awful but just about usable. Combined with the need have self control over this all… I started looking at self-hosting Planka on Yunohost. (Bearing in mind, I have been looking into knowledge management systems for awhile.

My old Dell XPS 12 with a lot of stickers on the lid — I attempted to put Yunohost on my old Dell XPS 12. Those stickers are a trip through history

I have enjoyed Yunohost and bought a cheap Intel NUC PC on ebay for this and other applications I want to host them using docker but they were tricky on my Qnap NAS. I also do have Yunohost on a Raspberry PI 5 but I realised some apps need a AMD64 environment and with a always on VPN I can run and use them anywhere? I also looked at Wekan and keep trying to get Vikunja working.

But back to where I am…

The biggest issue right now is prioritising what I do, how much I setup now and redo after my leaving date. There are things like buying a new laptop I had to do and have decided to not include any work related stuff on to including Microsoft Edge, One drive syncing and Slack (although I do use slack for other communities so that will change). I have my PAC code for my work phone, so I can move that number soon enough.
I had planned to drive around on the scooter and see friends but the weather in the UK has been bad for riding and frankly its something I can do after April.

That’s where I am right now…

Focus on leadership coaching, finish up work (future of social report), write some recommendations for BBC R&D, a intriguing final email (maybe also in audio), lean in on the different opportunities from my network, have a great time at the leaving parties for March/April. Finally I should switch my Linkedin to open to work maybe?

I do plan to take a break straight afterwards for my birthday and the Easter holidays. Then I’m back and will be in touch with many of you, as the scooter is ready to go.

Yes I am looking at where to go next but I’m carefully looking at options, rather than jumping to the next place. I am very aware there are a lot people being made redundant, its rough out there. I can only rely on my network and unique skills to find my next steps.

My mix garden on the move again, hello peertube!

I have been mixing away on the pacemaker device as usual. But noticed my mixgarden is having some troubles. I was using hyperaudio and webmonization to present my mixes. It was good however things changed over the last year.

Webmonization made a number of changes and hyperaudio seems to be slightly broken the configuration I currently have it in.

So with all this in mind and after looking at Peertube and Castopod, I decided to give peertube a try before running my own instance. Similar to what I had planned with Funkwhale a long time ago.

This has worked out well on rankett. Then I heard on podnews last week about music freediverse.

Music Freediverse is a new place for musicians to post their songs as an RSS feed, using PeerTube.

After a email with the sysadmin, about mixes and legality of them. They suggested I check out Split Kit if I have the metadata (which I do), this is something I looked into and now have a lightening address, which you can send sats to on my new mixgarden. I am also interested in the peer 2 peer tech

One of the good things about Peertube is the ability to sync videos over, which I took advantage of, it was a shame the metadata wasn’t copied over too but its fine, as I had it all.

In the short term I’ll keep using peertube but I have a plan to install a peertube instance myself and host it all locally. Then I could syndicate them out elsewhere. (Anyone can recommend a good docker container, do shout)

Another thing for the task list, but likely the best idea for now with lots of room to grow and expand.

Talking of which the next mix is a banger….

WebMix: Webmonetization + Dj mixes for the next internet

While recovering from Covid, I got a little time to finally sort out the WebMix idea which I also wrote up for Mozfest earlier in the year.

With the incredible and generous help of Mark Boas of Hyperaudio, I was able to use Hyperaudio lite to make clearly mark up a list of tunes in a DJ mix. Its what I’ve been looking to do for ages to move away from Mixcloud,

My finally setup was something I was playing with for ages but mainly via a self installed wordpress on my raspberrypi. I found problems when installing hyperaudio and in the end decided to go with a static website. I choose Publii as it had a linux client and I could just write the HTML easily (so many use markdown and other things, which would have made working with hyperaudio more difficult than it needs to be)

With the site creation out the way, I needed somewhere to host it.

Originally I was going to use Yunohost but I couldn’t find a simple webserver to just host the static files, instead I found a proxy server, which points at my NAS, which is running a very simple webserver. Of course the NAS has plenty of space, its also where the mixes sit, has a excellent redundancy and backup system.

The result of the experiment all sits here – https://cubicgarden.info/mixes

The core part of WebMix (as I’m calling it) sits in hyperaudio’s transcript and webmontization support.

Hacking hyperaudio’s transcripts

Originally I always saw Hyperaudio for its ability to tie a knot between the written word and the audio (& video). It wasn’t till I saw a demo of the WebMon functionality is when I understood it could be the thing I need for DJ mixes.

With correctly written HTML, I can tell Hyperaudio what it should do, and with Mark’s help we had a prototype up and running.

Here is an example of the code from the quiver in the underground mix.

<li class="active" data-wm="$ilp.uphold.com/B69UrXkYeQPr"> Activator, I know you can (That kid chris mix) - Whatever girl</li> <li data-wm="$ilp.uphold.com/3h66mKZLrgQZ">Air traffic (Erik De Koning remix) - Three drives</li> <li data-wm="$ilp.uphold.com/B69UrXkYeQPr">Chinook - Markus Schulz pres. Dakota</li> <li data-wm="$ilp.uphold.com/3h66mKZLrgQZ">Opium (Quivver remix) - Jerome Isma-Ae & Alastor</li>

Each tune has a time configured using the attribute data-m, this is in milliseconds. As I have all the data in the old CUE files I created a long time ago. Mark helped me out with a nice script which saved me manually copying and pasting. (I also considered writing a XSLT to do the conversion). In between sleeping and relaxing with Covid, I got a number of mixes up, changed the theming and finally got to grips with the static file uploading process, and the results you can see on the site.

Payment and royalties

You will also notice each tune/list item also has data=”wm” attribute with a $ilp (payment pointers). Currently they are pointing to myself and Mark Boas. Obviously I would change them to the payment pointers of the artists/producers/djs involved but I don’t know any which have them so far. Which leads nicely on to the next challenge for WebMix.

I did/do have a plan to do a mix with dance music from artists which have payment providers but that is still in the pipeline. Along side this, myself and Mark thought about some kind of database/airtable/spreadsheet/etc with payment pointers crossed linked to their discogs profile.

Maybe this is something which could be done in the next grant for the web call for participation?

Back to the current experiment, here is Opium (Quivver Remix) – Jerome Isma-Ae Alastor. You could imagine one payment provider decided between all involved which could be used to pay for each time its played on the site. (I am very aware this is very simplex and the royalties of music is a total nightmare!) but the point of the payment pointer is to hide the complexity behind one simple payment pointer, how its divided afterwards is up to each of the parties involved. I’m imagining a management agent, organisation or even dare I say it DAO; responsible for the payment pointer. There’s already things like revshare, which means you can have multiple people/entities behind the payment pointer and theres interest in this space. Long tail economics certainly could benefit here.

Anyway its a long complex area which I’m best staying out of…?

The main point is its all working and expect more updates soon… I know Mark has other ideas, while I still need to get older mixes up. I also would like to tie the whole thing to something federated or at very least setup a activity-pub feed.

Maybe I should be more ill more often?

My own mixcloud, finally?

I have been for a long time looking for an alternative to mixcloud. I have tried many things including some self hosted solutions like navidrome, subsonic, madsonic, airsonic and ampache. They have all been good except they are best for private sharing. I really wanted to use funkwhale but it was so geared up for single tracks it just didn’t make sense to run a node with my own mixes on it. There is so much I could suggest for making these software/services better for DJs rather than musicians. A DJ version of funkwhale could be pretty cool, especially seeing the amount of DJs using Youtube and paying for Mixcloud premium to mix live during the pandemic. ~~Heck you could even use web-monetization too~~ (just done).

So with all this and finally thought I have the bandwidth and the storage, I just need a site and some simple software which can share the music files. So I decided to actually setup WordPress with it looking at the local file system (which I can easily have tons of storage). I was going to explore the static file generators again but decided to get something going.

Over the last few days between helping someone out with Linux and cryptocurrencies, I setup WordPress on my RaspberryPi 4 using Yunohost again. As its pretty much static, I think it makes sense.

So here is my own mixcloud site, which I’m still populating, but the latest mixes from my locked down, mixing out album are up complete with artwork. Expect to see more changes over time including a better audio player, more mixes and more everything.

Its not exactly a mixcloud replacement to be fair and my plans to use the .cue files and make better use of playlists, is put on hold for now. I’m sure there is audio plugin which will make use of them. Love to have UPnP and Subsonic apis access from wordpress, but I dream?

Do enjoy and let me know what you think could be improved.

Little update

Following my point about making it work for DJs and mixes. One of my biggest bug bears is playlists. I have been through many of the wordpress plugins for audio playback and I can’t find one which allows me to specify points in a long mix, when different music is played. Its simply a tracklist but all of the ones I have seen and tried are focused on single tracks. Meaning slicing the mix into pieces instead of marking out areas. None of them seemed to support CUE files or things like Ogg vorbis chapters. If there is one I should be looking at, do send it my way because it seems like such a simple thing to do, but I guess theres not enough interest to make it?

Another update

I have retired the old Mix site and replaced it with a new better one.
Learn more about the changes and WebMix.

Replacing Mixcloud with Funkwhale

For a long while I have been threatening to leave Mixcloud in favour of hosting my own mixes. I looked around and thought funkwhale looks great as its a federated network for music. With some help from JonT, I started to scrape the metadata I stupidly forgot to keep for myself.

I installed Funkwhale on the Yunoserver but spent too much time trying to work out how to mount my NAS on the Yunoserver. I gave Navidrome but decided it wasn’t right as there was no way to listen without logging in, so went back to Funkwhale and got things up and running.

Seems Funkwhale might not be the best solution for the mixes, plus the developer is looking for new maintainers recently. Its really setup for single tracks not mixes. I could upload mixes but my plan to use cue files, won’t work. The only place to put playlists is in the comments. I also need to do more digging as I can’t change the year of the mixes. More importantly, the public sharing is a bit broken for me. You should be able to listen to the radio but its not working for me

As I get my head around it all but you can subscribe via RSS and if you’re using Mastodon or other fediverse applications, you can subscribe to this account: @digitalitalicmixes@mixes.cubicgarden.info.

Enjoy, I’m one step closer to self hosting my mixes.

Getting on the self-hosted train again

A long time ago, accessing cubicgarden.com meant accessing my direct server sitting in my home. I use to run Blojsom on top of Resin server. I was self hosting from my 512k ADSL line with 256k up (remember how fast that use to be to!?)

There were a lot of problems I grant you that but it mainly worked ok, although I didn’t like the sysadmin side of it all, as I was using Windows 2000 as the operating system. At some point I decided to switch to wordpress only because PHP hosting was cheaper than Java, although I got some incredible breaks during my time. In 2014 I moved my blog to WPengine thanks to dotBen

That was a while ago and since then I have massively upgraded my connection speed to 1gigabit up and down thanks to Hyperoptic and upgraded my server quite a bit (6 core AMD with 16 gig of memory). The first thing I did was installed Plex server.

Since then I have been slowly adding more services to my server. I guess the most noteworthy ones being tiny tiny rss, icecast2, plex and zerotier vpn (which I’m considering changing to wireguard with the recent announcements). Tiny tiny RSS is useful as I don’t like what feedly and others are doing with my data. Zerotier VPN is very cool and very much like the old and forgotten Hamachi. Because it uses internal ip addresses (non-addressable?) any device I have it connected with can access those addresses like they are on a internal network. This ultimately means I can access all my services including tiny tiny rss without opening up ports on my firewall and exposing it to the internet.

Anyway I’ve been thinking about adding more services to my server including Wekan (alternative to trello), Pixelfed (feiverse instagram), wisemapping (web based mindmapping tool), wallabag (alternative to instapaper), standardnotes server, mastodon (fediverse twitter), funkwhale (fediverse spoitfy), language tool (alternative to grammerly) and matrix (powerful alternative to slack).

Doing it under Ubuntu isn’t a problem as theres lots of tutorials and theres plenty which use Docker to manage everything.

But there is issue it seems when installing multiple services on top of each other. Most of the tutorials require a Apache or Ngnix then some SQL database. The tutorials are written like you are running just one service alone and things become more tricky when you have services using certain ports, etc. Trying to move the ports, database tables is sometimes tricky to follow.

Right now, I’m focused on doing one service at a time or really getting to grips with Docker which was meant to make this easier to deal with???

Ok so why selfhosting (and there is a lot of self-hosting services as I found here) and all the hassle?

I found something which sums it up nicely from a different but connected context.

Decentralized, peer-to-peer networks are evolutionarily superior to the bastardized corporate ‘sharing economy’ platforms like Uber and Lyft. Their billion-dollar budgets won’t save them from the inevitability of the blockchain-based peer-to-peer economy.

The decentralization revolution is here.

Docker & Tiny Tiny RSS sorted finally

I have had on my task list for a long time to fix two problems with my Tiny Tiny RSS setup.

Fix the problem I’m having connecting to TTRSS in a browser since a upgrade
Sort out a decent RSS reader for Ubuntu

Originally it was working fine then a upgrade broke the web interface for me and many others. The confusing and joyful thing for me, was any application which talked to the API was unaffected. Meaning my Android clients were fine including the one on my eink tablet. However all the RSS clients on Ubuntu would either not connect to ttrss, were generally rubbish or wouldn’t work in later versions of Ubuntu (like RSS Owl). The advice seemed to point to using a browser extention.

The first problem was something to do with the PHP which seemed pretty easy to fix but all the solutions assumed you were running it all on a standard webserver and had control over everything. Of course I was running it within Docker and had no idea where config.php was or even where docker had installed anything.

After actually sitting down and looking around my server as a sudo not myself (its the first time I actually dedicated time to do), I found the Docker install and learned what docker was actually doing. My ttrss docker image is actually located under /var/lib/docker/aufs/mnt/{random hash}/var/www/ttrss/.

Under that I could find the config.php file and make changes so it was only accessible over my Vpn connection – yeah, I thought this was very clever but maybe obvious to everyone else. So the only way to hit the web front end of my ttrss install is via my Vpn but API calls are done without the Vpn.

As I found the root of ttrss, I was also able to finally install feedreader which is hightly rated by many. The problem I’ve always had is feedreader complained that it needed a certain plugin installed under ttrss’s plugin directory, which previously I couldn’t find to install. Of course now I know where it is and could copy it there, I was very pleased with myself. Next stop brunch at Ezra & Gil and wait for Feedreader to pull down full text for 8500+ items.

https://twitter.com/cubicgarden/status/929712805053435904

Host your own RSS aggregator?

It started with me getting fed up with Feedly trying to up-sell me to their premium subscription. I mean I get it but $5/month to host a simple RSS aggregator? This seems quite a hefty price (even with all the extras it provides, which I never really use).

So I first looked for alternatives to Feedly and found quite a lot. The main thing for me was having a Sync API, so I’m not reading the same stuff across my different devices. My thought was with a standard API, it wouldn’t matter what client or platform I use (although I’m using Linux and Android mainly). Standard I thought… boy was I dreaming.

After a lot of looking and reading I said screw this, I’m self hosting my own copy of tiny tiny rss, which seems very popular with people like myself trying to do the same thing. It seemed quite straight forward and I decided it was time to give rkt or docker a try as there was a docker image for it.

In a evening I had it setup, running and working with my exported feedly OPML file, while watching a film and cooking. Its currently only available to my network but I’ll likely make it externally available (without my VPN) once I got it setup with a SSL cert and 2 factor auth. I did notice my fav RSS reader on Android did support ttrss then somewhere along the line they pulled support for it. So I’ll try out the android app created by the author of ttrss, but the comments are… well.. interesting?

M	T	W	T	F	S	S
				1	2	3
4	5	6	7	8	9	10
11	12	13	14	15	16	17
18	19	20	21	22	23	24
25	26	27	28	29	30	31