Soemarko Ridwan

iOS & Web Developer ⟡ Coffee Addict ⟡ Scuba Diver

The Power of Flat File CMS
The Power of Flat File CMS

I am forced to take a break from Project Katara. Pictured above, this blog suddenly broken for whatever reason. As I finished writing the next build log, I want to post it, I just notice that it’s broken. Not DNS settings, I’ve tried other paths, DNS are fine, FTP is still up and running. Tried re-downloading Bludit and overwrite the core, still messed up.

I have backups both on server and behind Bludit admin, but they’re old. The ones on the server stopped backing up because my host disabled php’s zip during one of their maintenances. I never notice this until today. I’ll fix this after I fixed the blog.

I noticed that I can upload a clean Bludit and have a new installation up and running without issues. So I just copy everything from bl-content over. Broken again but in a different way, /admin led to 404. So something inside the contents are causing this. Let’s do over, go slower.

Clean install, move my custom plugin and theme, all good. Next just copy everything from /bl-contents/pages and /bl-contents/databases (except for the plugins database). It works. And since all of my pictures I hosted it on imgbb, I’m done. All good. The broken copy is still available on /blog-broken, but I don’t think I care enough to do a post-mortem, it’s there if and when I feel like doing something about it.

Everything is sorted within 15 minutes. Can’t even imagine what I need to do if something like this happen to a WordPress blog, which is an argument for paying to be hosted in, or you know, use Bludit.

I need (and want) to continue working on Project Katara.

Project Katara - Build Log 2: Enclosure Design

Not much to write. Now that the parts have arrived, I can find most of the exact models on GrabCAD, and the ones that aren’t available online are simple enough to design myself.

  • R10 800l/h Aquarium Pump
  • ESP-32
  • IEC320 C14
  • 0.96” OLED
  • KY-040 (don’t need the model for this, I just need to cut 7mm hole) — designed my own a knurled knob though.
  • Toggle switch (also no model, 16mm hole)
  • 1000 Watt submersible water heater (poorly designed, can’t find my exact measurement in GrabCAD, but I need to design this to align the water pump)
  • Chinese scavenged AC-DC power supply 5V 2A (just need to add mounting points)

I’ve split the enclosure into 4 parts,

  1. bottom for mounting the pump, sensor, and water heater. Screwed on to centre. 7.5 hours print.
  2. Centre the main enclosure. 12 hours print.
  3. Top is the lid, snap on to centre. 5 hours print.
  4. Clamp to clamp this to a container (the water bucket), and it also clamps centre and bottom together. 2.5 hours print.
  5. I’ve also added 0.2mm gap so I can add a transparency plastic between bottom and centre, you know the ones we use for 3D printed face shields. This is to protect the electronics from hot water vapours.

Total: ~27 hours of print time. I’m getting Volcano hotend + 1mm nozzles.

To be continued...

Project Katara - Build Log 1: DIY Sous Vide MK3 Prototype
Project Katara - Build Log 1: DIY Sous Vide MK3 Prototype

Ever since I saw the Anova tear down (as mentioned towards the end of previous post), I can’t get it out of my mind. I bet I can build that. So, as of a couple weeks ago, I started prototyping. I’m going whole hog. Got new parts, most of them anyway. I’m not pulling parts out of MK2, it is still actively used. The goal for Project Katara is $20 Anova, the whole thing, no rice cooker. And I aim to make this like a “commercial product”, you know, as far as DIY goes.

First the name, I love naming my projects. If it is not obvious, Katara is the water bender from Avatar The Last Airbender. The basics of this project is ESP-32 as the brain. Submersible water heater as the heat source. And an aquarium pump to circulate the water. A complete Sous Vide, aka the immersion circulator. The pump is key, if we don’t move the water around the heater, the water around the heating element can be boiling while the water an inch below it is still at 30ºC. Hence Katara.

Next, the prototype pic above. Since I already know the Sous Vide part, running PID on a submersible water heater based on the input from a thermal sensor, I need to test and prototype the things that are new to me. Mainly the rotary encoder and ESP’s Wifi libraries.

  • In the middle is the ESP-32, the brain. I missed the WiFi interface of MK1, kinda annoying going back and forth to monitor the temperature and making sure the thermocouple stay in the water.
  • KY-040 Rotary Encoder on the left. Another drawback of MK2, D-pad buttons are terrible interface to set temperature. Rotary encoder because that way I can have interface continuity between WiFi and the device
  • Spare 16x2 LCD, already have the 0.96” OLED on order. Seen there is the encoder running to set the “target temperature” and the local IP address of the ESP-32.

The basic of DIY Anova has been done multiple times and there many guides on Instructables and YouTube, a temperature controller, small aquarium pump, and a submersible water heater. However, for some weird reasons most (if not all) of them use a bang-bang controller that’s really meant for a cooler than heater. Seriously, like $5-10 more you’ll get a PID controller. See the graphs below, bang-bang is far from ideal for a Sous Vide.



To be continued...

Anchor Phone Stand: From Ad To Model In Minutes II

Another thing that’s “inspired by” Instagram ad.

So I pop open Fusion 360, model it from the picture, and it’s done within 30 mins. After about 30 min print, pop in the magnets, BAM saved like $10 or however much it was sold (I don’t remember exactly).

Instructions is in the thingiverse page.

Rice Cooker Sous Vide MK2
Rice Cooker Sous Vide MK2

Well, not necessarily an upgrade from MK1, but more of a remake on the same thing, minus WiFi and App, but plus LCD and buttons, and now that I have a 3D printer, a perfectly fitted enclosure. It is a simplification. I can’t use the same code because the Spark folks has drop support for the Core, took me like an hour to get “blink” running on it. It’ll be faster to remake it from scratch. And using a self-signed app is annoying for others in my family that could’ve use this as well.

That said, I made this with 100% spare parts, so personally I spent $0 on this. I bought an Uno clone and LCD shield for “prototyping” purposes, but it never left the anti static bag. It is always faster to jump directly prototyping on the board I actually plan on using, zero code adjustment.

The longest part of this build is waiting for the enclosure finish printing. If you need a guide for this, just follow the Sous Viduino by Adafruit.

Links: enclosure and code. A quick breakdown on my specific setup:

  1. Arduino as the brain, and Uno because I want to use the LCD Shield.
  2. LCD shield to simplify things. I’ll have buttons for interface. Up / down to control the temp by 0.5ºC increments, left / right control by 5ºC. The set button to start and stop the PID. 1
  3. Temperature sensor as PID input, I use the type K with MAX6675, because I bought quite a few of it right before people jump to MAX31855. Even the DS18B20 will work as well.
  4. Fotek 40A SSR is fine, it’s good enough to control the rice cooker. Utter shit as a product, I wouldn’t trust it as mission critical projects, but I already have 4 of them because they’re like less than $2 each.
  5. And I can’t remember the exact module code for the Aliexpress AC to 5VDC module to power the Arduino from the same plug. 12VDC into Vin pin would be better, but regulated 5V into the 5V pin is fine.

Downside with this rice cooker setup that you’ll need to make sure that the temp sensor is in the bath when the PID is working or you’ll boil whatever you’re cooking in the bath. I have one occasion where the probe pop out for some reason and the water reached 70ºC when I wanted 60º.

I plan on doing something like a “thermal runaway” protection in the code like in 3D printers. See if the temperature isn’t rising after a certain time while the heater is “on”, kill everything.

Also, I saw this Anova teardown, it is so simple. That thing is literally the same as this project, but instead of rice cooker, just add a $2 submersible heating-element, and a motor to “circulate” the water. I see MK3 in the horizon, another perfect project for #CoronaCraft.


1: This LCD Shield is terrible idea, it is super annoying to use. I'd 100% use a potentiometer to set the temperature for MK3.

Goodbye Dropbox, Hello MEGA
Goodbye Dropbox, Hello MEGA

I recently hacked my Nintend Switch, there I tested and found out that MEGA is the fastest way I can grab large files down to my computer. And it is a totally viable replacement to Dropbox. Apparently last year Dropbox has limit the basic account to 3 devices , which I never noticed until recently.

Because for a long time, I only need 3 devices. iPhone, iPad, and my 2015 MacBook Pro. A few months ago I replaced my (very old) iPad mini to an Android device. Dropbox limitation has been annoying but I didn’t find a need to look for replacement. Now, I just got the new 2019 16” MacBook Pro, the limitation has become a deal breaker. I don’t need a large space. Only used ~5 GB out of the 11 GB I’ve obtain out of various challenges / easter egg many many moons ago.

Moving into MEGA has been painless, I just create a folder called dropbox, and drop everything there. ez pz. On most devices I just sync this particular folder. It took awhile because I have like thousands of files, after everything is synced, it is seamless. Just launch a red M app instead of the blue box, everything else remain the same, and to some extend, MEGA is better as it doesn’t try to do anything else other than a synced folder, the way Dropbox used to be.

An Audience of One

Tom is totally on point in the video. Web 2.0 days is totally the Internet’s utopia. “It was going to be the Age of Mashups: take data and do interesting things with it.” If we’re still in the Web 2.0 days, we won’t have these many messaging services, and even if we do, there will be an app that does them all like Trillian (member Trillian? I member…).

A few weeks ago, I built and released WhatsDP. A simple tool that’ll update my contacts with the display pictures in WhatsApp, I want to see pictures instead of initials when people calls or in iMessage. This would be something trivial in the Web 2.0 days.

It started as something I wanted to do, so I look for ways / existing apps that’ll do it, closest one is Vignette. Not only that it doesn’t support WhatsApp, I have to go through every single one of my contacts and add their Instagram or Facebook account. Nope, that’s a non-starter.

And then, as a developer myself, it turns into an app idea. However, after further research, I’ve found that there’s no kosher way of pulling that info, even with WhatsApp Business API. WhatsApp Business API is only good for sending OTP and phone tree. And turns out, there are very little services that have both phone number and display pictures, think Gravatar but with phone number instead of email. We no longer live in the utopia, it'll never be widely accepted in todays world.

So here we are, I cobbled together some JavaScript to scrape some images off WhatsApp web. It works, for now, and at the very least, for me. Even if I’ll be the only user that it ever has, my contacts look so nice now. I got what I wanted at the beginning of this process.

Just because something is going to break in the end, doesn’t mean that it can’t have an effect that lasts into the future. Joy. Wonder. Laughter. Hope. The world can be better because of what you built in the past. And while I do think that the long term goal of humanity should be to find a way to defeat entropy, I’m pretty sure no-one knows where to start on that problem just yet.

So until then: try and make sure the things you’re working on push us in the right direction. They don’t have to be big projects, they might just have an audience of one. And even if they don’t last, try to make sure they leave something positive behind.

WhatsDP: Pulling display pictures from WhatsApp to your contacts

Firefox extension.

The second brain child of Appify. It started as an app idea, turns out it was impossible to do it right (as in without some weird hacks — how do you even scan a QR code on the same phone?!). The cleanest way was with a browser extension.

  1. You need to export all your contacts to vCards.
  2. Run the extension (like the video)
  3. Delete all your current contacts, and add the one you just downloaded from this extension — this is the scariest steps, but don’t worry… you already have a backup of your existing contacts. If you don’t delete first, you’ll end up with duplicates.

And you’re done.