Retropie table writeup

May 27, 2016 by · 2 Comments
Filed under: Development, Hardware, Linux, Personal 

A co-worker of mine asked me to do a writeup of my retropie table build and yes, I guess I should. I rarely write anything about what I do anymore 🙂 .

So, I got kids, and I thought it would be fun for them to play some of the old stuff I used to play and found the quite interesting and so set out to do something similar. I winded up going to Ikea with the kids and told them to choose colors, we winded up with a pink IKEA Lack table … ;).

I also got a bunch of old screens at home, I picked a very heavy 17″ monitor and removed all the casings etc greatly reducing the weight. My choice was slightly bad though for two reasons, the connectors are standing at a 90 degree angle to the screen so it winded up not fitting inside the table and it’s pointing out on the underside as it is now… secondly, the viewing angles are so so, I winded up rotating the screen 180 degrees as the viewing angles where much better from that direction and the /boot/config.txt in raspbian systems have options to hardware rotate the output. I also bought a set of 2 joysticks + 20 buttons with built in backlight and a xin mo based usb controller from ebay, a small speaker from a local shop and the power adapters I had at home, and a connector to fit for power input to the table. I decided to pick up a powered USB hub as well to fit so it was reachable from the outside. I also had a sheet of plexiglass lying around for many years which winded up useful. I also had a USB wifi dongle lying around so I reused that for connectivity.

We started out with measuring out and sawing up the holes needed and then removing the innards that are in the way. I used some paint masking tape to protect the table, more on this later. This work was very easy to do with a dremel with a circular saw addon. The underside wasn’t so important how it looked but I tried to make the cuts decent looking at least.

Once this was done, I test fitted and probed a bit on how to get the raspberry pi, monitor and power system fitted inside the the table before moving on to sawing the plexiglass sheet into the same size as the table, and then drilling and countersinking the screw holes, temporarily fit the plexiglass while drilling the holes for the joystick + buttons. Some sanding and fixing of edges followed. I removed the plexiglass sheet and drilled and countersunk holes for screwing the joysticks to  the topside of the table underneath the plexiglass.

Moving on, I screwed in the joysticks and fitted the power adapters for the monitor, USB hub and raspberry pi in the screen. I made 6 foam inserts to rest the monitor on and glued in place inside the table, wired up the monitor and put the monitor in. Removing the paint masking tape I realized that I used some shitty tape with much too hard tacking adhesive, meaning that I managed to pull away a bunch of the foil/tape (the “paint” on the table, it’s not painted, but rather foiled with a layer of colored plastic). When I realized this I started rethinking a paint scheme I had already planned and decided to do some modifications to hide the errors and to possibly heighten the feeling of the table.

I did the paintjob, made some simple paint masks etc and airbrushed the table with black borders, softish blue and red and green colors further softened with a few drops of white.

After this had dried for a few days I put the plexiglass on, screwed in all the buttons, joystick heads etc and installed the raspberry pi + other final electronics and tested the system. This is when I realized the problems with the screen viewing angle so I had to back everything up, remove the buttons, joysticks, plexiglass screen, and monitor. I winded up lifting some of the paint I had used to paint the table (the paint was really sticking badly to the surface). This lead me to question the surface of the plexiglass and figured I’d polish the plexiglass. I originally made the bad choice of trying to apply an old plastic modelling technique on the plexiglass, washing it with a layer of future floor polish. This looked absolutely horrible on this big surface so I winded up spending 1,5-2 hours removing the stuff again and then using some proper polishing compounds on both sides of the table making the plexiglass sheet incredibly nice looking (in my humble opinion). I Repainted the parts where the paint was removed, took out the screen, rotated it 180 degrees, re-fitted all the power adapters etc so it wouldn’t be in the way of the monitor and had to saw up a second hole for the DVI connector… I then reinstalled plexiglass, buttons, joysticks, etc… and now, a much nicer viewing angle of the monitor and a nicer looking plexiglass sheet, but paint job not as nice anymore. Shit happens. Oh, I also pulled a cable through from the screen to the front panel button so I can turn on/off the monitor from one of the buttons. The USB hub was glued into a hole made in the skirting so it sticks out underneath the table with two accessible USB ports.

After everything was fitted and tested to work I started to look at the backside what could be done about it… I had to make a raised area to increase the depth of the table as the buttons I got wont fit properly otherwise. I used a 1 cm floor skirting around the hole and then took the sheet from the monitor hole and sawed into two pieces which fits over the hole, drilled a lot of holes in it to at least create some air ventilation into the table.

At this point I’ve installed some games, used it for a bit, let the kids play around and I’m absolutely happy with it. The software side I didn’t need to do anything about, it worked more or less out of the box. I had to make a usb quirks hack to split the controller into two halves, I had to rotate the screen in the config.txt and that’s it, then just follow the installation howtos. Retropie was a really happy surprise, I wasn’t expecting things to be that smooth to install. I do wish the Amiga emulator was better integrate, it would be nice to be able to do the same thing as with the NES images, just drop them in and they work… but I understand that each game needs its own “fixes” to get up and running… I will have a look and see if it’s possible to improve the situation somehow, at least so I can start games with just the joysticks and buttons.

FOSDEM 2016 is over

I went to FOSDEM 2016 this year with 8 other colleagues of mine and had a really really good time. A lot of good speeches and stuff to talk about and I feel very motivated for some new projects. Some of the stuff going on right now is incredibly exciting, especially with regards to containerization etc which is something I have a lot of personal and work related interest in. I will be looking into more details in that for the future…

What I did miss was a more “general networking” track with low level stuff like iptables, netfilter, iproute, wireshark, snort, etc. I’m just not sure if this is the right conference for that though. Gathering my thoughts and working out some of the project details in the upcoming week if I get time.

Using AWS EC2 instances for large builds

I experimented a few years ago with using EC2 spot instances (virtual server on the internet, but using unused server capacity). It was fairly successful, being able to run large calculations that should have taken weeks in a matter of days.

Since I started at my current job I’ve been running into building increasingly complex yocto images which keeps growing in size, at this point most images I build can take up to 6-7 hours to build on my laptop. This is an i7-4558U 2.8GHz cpu and 8 gigs of RAM so it’s not bad really, just not enough for these types of builds.

Again I started experimenting and I am really happy and impressed. So far all experiments are for open source projects etc, so nothing that has any non-disclosure agreements or corporate etc etc, I’d like to but this isn’t up to me really. I’ve setup an AMI on EC2 which I can instantiate and have up and running in 2-3 minutes, and then mount a 100 gig EBS volume where I store the sources and build data.

The same build that generally takes up to 6 hours on my laptop takes approximately 30-40 minutes on an EC2 c4.4xlarge machine (16 cores and 32 gigs ram).

My key findings so far is:

  1. Keep an AMI with all the build tools necessary/or script the setup.
  2. Keep an EBS volume with the long term stored data, gits etc for building and mount somewhere suitable in the AMI.
  3. Keep a micro instance (these are free/very cheap) around for mounting the EBS volume when you just want to check stuff out, mess around etc but not make builds.

Qt5.5 QtPositioning PositionSource unable to load due to missing symbols

December 31, 2015 by · Leave a Comment
Filed under: Development, Linux, Ubuntu 

I’ve slowly been working on a “Magic Mirror” for a long time now, trying to pick up some new technologies and making something fun. I’ve been working with Qt on and off for the last 10 years or so, but only peripherally, looking and awing over what they do. QML is absolutely awesome in my humble opinion.

A few weeks ago I started using Qt5.5 and ran into some issues executing the Magic mirror in a qmlscene now that I continued the project. It was fairly easy to reproduce but it seems to only affect the Qt binaries I’ve installed from the installer downloaded from I’ve had verification that the code works with packages built from source, and trying to verify this on my own as well right now (building as we speak).

This is the sample code not working with qmlscene:

import QtPositioning 5.5
import QtQuick 2.0

Rectangle {
id: root

PositionSource {
id: positionSource

Bug is reported to here:

Google photos refused to back up via mobile Hotspot

December 25, 2015 by · Leave a Comment
Filed under: Android, Phone 

I ran into a problem making my father’s phone sync pictures to Google photos, it simply refused to run, just saying “uploading photo 1 out of 489”.

After messing around on my own I started searching the net and finally found this rather obscure explanation, which fixed it for me. Apparently the phone realized it was running over a mobile Hotspot and refused to run because of it. See this blog post to fix it.

Systemd oneshot, ExecStop and RemainAfterExit

July 14, 2015 by · 1 Comment
Filed under: Development, Linux 

I wanted to create a systemd service file that just ran 3 small commands today in a sequence on ExecStart and another set of reverse commands on ExecStop. My initial idea was to use bash syntax with ; between the commands (I keep forgetting that Systemd is not bash…), and the service file was set to being a oneshot file, which meant the ; was actually interpreted correctly, but all the ExecStop commands where also run directly when running systemctl start service.

So I read up a little on Systemd and ExecStart, this is what the page has to say:

When Type is not oneshot, only one command may and must be given. When Type=oneshot is used, zero or more commands may be specified. This can be specified by providing multiple command lines in the same directive, or alternatively, this directive may be specified more than once with the same effect.

So… that means the ; syntax will only work with oneshot apparently. Also, oneshot means that ExecStop runs directly after ExecStart. Further reading the documentation seems to indicate that RemainAfterExit=yes will make the service stay around according to systemd, so it will only try to execute the first time you run systemctl start, but not the second one. I don’t think this fixes that it runs ExecStop on start however, but I’m not sure.

ABS plastic with plastic modelling accessories

I’ve been building plastic models on and off since I was a kid, and very much so as of late, meaning that I have a lot of modelling accessories available at home.

I accidentally broke off one of the screw holes for the BabyNES Raspberry PI case ( almost instantly due to the bad print quality. Reading around I didn’t find any obvious glues to use except acetone. In fact, I found a whole heap of warnings that the fat plastic would probably be hard to glue etc. Hence, I decided to test Revell Contacta Professional glue ( and it worked like a charm from what it looks like!

Since this worked so well, I tried some Italeri putty for plastic models ( to cover up the sides which shows a lot of layers and some defects. Except the putty smelling really bad, it tacked on really hard and dried enough in just a few minutes to be sanded down.

Since the above went so well, I went ahead and did some more tests by printing a set of battered barrels ( which turned out fairly crappy due to the print quality (mind you, much better than the first BabyNES case). Either way, perfect chance to test what could be done using those modelling tools/supplies. Here’s some before through after shots:

2015-03-14 19.43.32 2015-03-19 21.01.34 2015-03-30 22.52.22 2015-04-11 20.48.44

I think a lot of the plastic modeller materials could be used with ABS and possibly PLA plastics as well, especially where you want to make a structure, and then want to improve the details. There are much better printers out there with much better calibration of course, but these materials combined shows some great promise IMHO.

3D printing

February 11, 2015 by · Leave a Comment
Filed under: 3D printing, Development, Hardware, Personal 

wpid-img_20150209_132426.jpgI started working at a new company a while back if I failed to mention it called Pelagicore, doing automotive in vehicle infotainment systems the right way with open source, working in the community etc. They have a 3D printer (Makerbot thing-o-matic) that’s been standing around for a while really grabbing my attention and I’ve wanted to find the time to use it since I started there. I finally took the time and am happy as a clam.

I made my first 3D print at work two days ago, first of three pieces for a BabyNES Raspberry Pi 2 case (, a small hatch. When I went on and tried building the two larger pieces I ran into troubles. Running out of time that day, I got around to try and fix it today. Disassembled the extruder stepper motor and realized a screw had dislocated from the vibrations. Re-screw it and assemble everything again, and this time, it worked! Got top and bottom parts done before leaving for home today and am fairly happy. They will do.wpid-img_20150211_163900.jpg

The print quality is pretty shoddy, a lot of settings needs to be tuned I guess, and the print speed was probably too high for the unit. Judging by the software installed, I should probably install newer software and redo the calibrations using those instead. Still, awesome to work with this 🙂


Project build speed importance

I began writing this several years ago but never published it for various reasons. I think some of the thoughts are still really interesting, I joined a company named Pelagicore some months ago and we do some rather large builds at this company, similar in build times in the project I refer to in this text, but with some major differences. This build is based on yocto and it actually works really well once you get through the first build (currently a scratch dev image build is up to 4 hours in time, but after that it’s able to rebuild only changed software and recipes as necessary. The scratch build is heavy admittedly, but we are working on some ways to improve that as well, such as sstate caches, icecc distributed build systems, and so forth. I think with modifications we could get those 4 hours down to under an hour, which would be really sweet. Anyways, I thought I’d post this text as is, even though it’s not finished, just because and since I re-read it right now 😉 .

Original text

Again, I’m struck by the importance of a proper build process in projects. No matter how small or big the project is, the build must be kept fast and working from get go to the end — which means its end of life cycle. I am currently working in a large project with a huge code base where the build system is, in my opinion, next to completely collapsed. Build time is in excess of 1-4 hours, and the automated test-suites take from 3 hours to 7 hours to complete, depending on how thorough a test-suite you choose. A gigantic heap of time is spent just…. waiting… waiting…. waiting. Just to make a simple one line edit and then compile to see if it compiles can take up to 4 hours. I’ve previously been a bit spoiled with good and simple build systems, only occasionally running into really crappy build systems — for some reason, a lot of scientific open source software seem to be stuck in this category.

This time, I think I hit pay-dirt on “how not to do it”. Instead of focusing on the bad parts, I will try to focus on the things to do and to keep in mind. Nobody seems to like a grumpy person anyways, which I really am sometimes.

1. Keep the build system simple and manageable. Try to maintain the build system in a logical fashion and in a single language/system (scons/python, Makefiles, etc).
2. Expandable (new directories, files, etc)
3. Scalable (multiple CPU’s, machines, threads, etc)
4. Try to use as few frontends as possible (a single top makefile for example, with targets depending on what you want to do), and keep people informed of changes. Wiki with history via RSS is _perfect_ for this. This doesn’t mean you can create a script to build something, then 2 months down the road delete it because it is no longer needed, it causes a lot of stress for developers who just barely had a chance to find the script.
5. The programmers are (most likely) your customers.

DIY OSD update

December 11, 2014 by · Leave a Comment
Filed under: Development, FPV, Hardware, Robots, Video 

Late update, but video is working as expected, GPS is working to some extent, but because of local weather I havent really tested it outdoors… I’m considering changing video transmitter and power system however. I’m currently running on 1S 160mAh batteries with a stepup regulator to 5v to a 10mw transmitter, which gives a really lousy battery time for the video system, also I’ve got too many batteries flying around at the moment I think.. Limiting to a bit fewer might be a good idea. Currently, there’s a battery for the video link/osd, one built into the camera, one for motor/rc receiver/servos, and one in the rc transmitter, 4 batteries in total of various sizes etc… a pain to charge them all. Hooking up the videolink and OSD to ESC/BEC is a trivial change and would remove a lot of cruft (stepup module etc), but will also add the EMC from motor/ESC to handle for the videolink etc as it’s a completely separate system today.

I’ve also upgraded so I’m using a mobius actioncam. Everything all in all means the system will draw more power and weigh more as well. Removing some batteries and running from the mains system would be really nice.

Next Page »