Paul Kocialkowski's coding blog

Free software, programming and stuff

Fun

Impressions from FOSDEM 2014

Written by Paul Kocialkowski - 11 february 2014 - no comments

Well, this is mostly going to be about my experience at FOSDEM 2014. Thanks to being a student in a city that features an easy-to-access airport, I was able to attend this year's edition. While my travel schedule was real tight, I only found out the plane was landing not in Brussels, but in Charleroi a few days before departure. Thankfully, it wasn't too late to find another schedule that made it possible for me to arrive at the Friday Beer event around dinner time. The cafe, and the aisle that leads to it were incredibly crowded, to the point that it was barely even possible to make it to the entrance. And once there, despite the fact that the cafe had been reserved for FOSDEM attendees only, I sadly couldn't get in, since the interior of the cafe was apparently full as well. While waiting near the entrance, I was able to see Greg K.H., first of the numerous giants of the free software community I stumbled upon at FOSDEM.

The next day, I was really amazed to see so many people going to FOSDEM, on the way to the event. Public transportation was really filled up with free software hacktivists! Arriving at the ULB campus used for FOSDEM, it really felt spacious and seemed appropriate for an event that big. Lots of interesting discussions took place after the first talk I attended: it was really nice. There were also numerous stands, mostly divided in two buildings, with many interesting people to talk to as well. In the AW building, I enjoyed the Coreboot/Flashrom stand (free software BIOS), Hackable-Devices (apparently focusing on micro-controlers recently), OlinuXino (Allwinner single board computers) and OpenPandora (free software gaming device) with a prototype of the new OMAP5-based version of the device, running apparently really well on GNU/Linux without graphics acceleration blobs. However, the big slice of stands was in the K building, including popular GNU/Linux distros such as Debian or Fedora and desktop environments such as GNOME and KDE. The FSFE was also there, with real good-looking flyers about their Free Your Android campaign, promoting Replicant and F-Droid! I also spent some time at the CaCert booth, and frankly, I was amazed by the depth of the identity verification process. First off, having a single official document to prove of your identity is not enough for these guys, and things get worse when the signature on the ID card (kindly provided by mom at a time I couldn't sign it for myselef) doesn't match the one you produce. Not to mention you have to sign the paper before their eyes, else it's not valid. The ID card itself is also checked to be genuine, with the many UV lights at their disposal and descriptions of the expected results. So I was really surprised how strict the whole process is and I think it's really great that they are taking this very seriously.

The rest of the afternoon we spent in the legal devroom, where I could meet many great people such as John Sullivan, Bradley Kuhn, Karen Sandler and others. John gave a talk about Javascript, with all relevant infos regarding how to escape the Javascript trap, followed by Chris Webber about free network services and introducing the road map for the next MediaGoblin releases, with a really good-looking video explaining the goals. Chris was probably the most enthusiastic speaker I saw at FOSDEM: he was really passionate about his talk and created a wonderful ambiance in the room (summoning that feeling that urges you to get involved in something bigger than you), so kudos to him for being that good. After a while, we headed to dinner with the Coreboot crew, very nice and interesting people. It was a real pleasure spending time in their company.

On Sunday, we had to rush to get to FOSDEM in time for the F-Droid track by Daniel Martí, followed by an introduction to the linux-sunxi community by Olliver Schinagl, who kindly poked the Replicant project during the talk mentioning that I've been promising Replicant for Allwinner devices for the last six months. Olliver's talk really made me realize what an amazing platform Allwinner is, so I just went ahead and ordered a variety of Allwinner devices to port Replicant to, so we should get there in the near future! On the way to lunch, I quickly saw someone I believe to have been Harald Welte: while I would have loved to have thanked him for his great work, he went by faster than it took me to realize who he was. Time went by, and in the afternoon, I could only attend the Lima talk by Luc Verhaegen before leaving. The project is apparently steadily moving forward, however, without any mind-blowing demo this time. I really had to leave fast after the talk, to catch up with my transportation schedule.

I really have the best memories from FOSDEM, it was really nice and there is no doubt I'll attend next year's edition, hopefully presenting a talk about Replicant there. It was also really nice to see people grateful for the work I'm doing on Replicant. Such huge community gatherings are the best to gather the motivation to keep working on a free software project: actually meeting the community brings a whole different picture compared to what contributing to free software usually feels like individually, hacking alone in my dark room.

Connecting a TI-82 Stats and a computer, with free software

Written by Paul Kocialkowski - 11 may 2013 - no comments

TI-82 Stats.fr

As I am a student in High School, I was told to buy one of these TI-82 Stats.fr scientific calculators.
Despite the fact that it runs a proprietary operating system, I had to use it in class for nearly 3 years now and it's not even really convenient to use, especially the programming part. TI created the TI-Basic interpreted language, that the user can use directly by writing programs on the calculator, which is awful. As I was looking for a fast way to write large programs, I looked into the possibilities that existed to connect my calculator and my computer and then write the programs on my computer and transfer them to the calculator.

Making the cable

2.5 mm Jack cable

So I ended up making a cable using only a male DB25 port, two diodes and a 2.5 mm stereo jack: I lent the original calculator to calculator transfer cable at the time I wanted to do this, so I used a 2.5 mm audio jack I had instead. This jack had a microphone ring too, that I wired as ground. However, I guess it's better to not cut-off a wire and use a female 2.5 mm stereo jack connector instead.

DB25 connector

Schematics for the connector are easy to find on the internet but they do not mention whether the connector is seen from its front or back. As a result, I soldered the wires in the wrong order and ended up with a non-working cable. However I though this might happen so I inverted the cables layout and it worked!

Schematics

The software part

TILP2

As the cable was ready, I had to figure out the software part. First, I had to find a computer with a female parallel DB25 port (my recent computer didn't have one since printers, that used that port, are now essentially wired with USB). Then, I looked up what software could handle this. I was amazed to see that there is a whole set of free software libraries to handle the TI calculators, so hopefully, not much was to be done here. I found tilp2 to be a nice front-end for all these libs and decided to go with it. At first, it couldn't use the parallel port because lp was already using it, so I had to unload the lp module so that it could work. After that, the automatic detect function detected my calculator just fine, as a TI-83 (while it is in fact a TI-82 Stats).

Downloading and uploading programs with the calculator worked just as well. Though, I discovered that the programs are not in plain text but have headers, a checksum and a special encoding of what's not plain ASCII text (such as the functions names, the operators, etc). Then, what I needed was a compiler for the TI-Basic language, and I could hardly find one. What came closest to it was the tibasic program, that is free software for Windows, but was ported to GNU/Linux (a matter of writing a Makefile and minor modifications). Though, the files produced by that compiler turned out to make tilp segfault when sending them. After looking closer at it, it turned out that there were differences in the produced header compared to the one of files that came out of the calculator.

The ti83c compiler

So I decided to quickly write such a compiler, matching the expected format for the TI-83, to compile my programs. It was very easy and fast and ended up as the ti83c compiler. Since it is based on the tibasic compiler, I kept its same license and added the original author as a copyright holder of ti83c too, so it's not released under the GPL, but another free software license described in the copyright header block. I didn't test ti83c extensively, but for the programs I wrote, it went just fine.

Contributing to OpenStreetMap part 2: Gathering data on the field

Written by Paul Kocialkowski - 30 december 2012 - no comments

I could keep writing a lot about the ideas behind the project, my personal motivation and such but well, OpenStreetMap is one of the rare projects I'm contributing to that actually require people to get out and see things for themselves ! So that's very good for us hackers that are used to do our work behind a screen: for once, we're required to get some fresh air. That has been a good opportunity for me to discover more of the city where I live, do some sports by the way and discover many relaxing places in a natural environment.

Bike

From the moment I started mapping, I always used a bike to move around town. That's probably the best way to catch every detail surrounding your ride, making it easy to stop at any moment and any place. I've done mapping on my feet a couple time too, it's good when there is a high-density of POI (Points Of Interest), like in the town centre. If this is not the case where you live, you are most likely going to waste a lot of valuable time. On the other hand, this can be a fun way to spend some hours to kill in the middle of the day.

The first device I used to do mapping was the Neo FreeRunner and its embedded GPS. I also got an external antenna to improve the reliability of the traces. On the software side, I was most of the time using Hackable:1 and TangoGPS but at some point switched to SHR with FoxtrotGPS. That was pretty nice to use, except that the keyboard with very small and required me to carry a pen. I attached the FreeRunner to my bike using cable ties though I had to drive very carefully to avoid damaging the phone.

Walking paper

Carrying a sheet of paper and a pen can also turn out to be very useful to draw a quick sketch of the ways and their names. When possible, printing walking papers (black and white is fine) with the already mapped OSM informations helps a lot too. I'm not a regular user of these methods but from time to time it helps, especially when there is not a lot of informations already available on OSM. Another kind of complementary mapping technique that I used from time to time is voice recording: this permits to be very precise in the description. These techniques are used best along with regular GPS tracking.

Tablet

As soon as all the streets were properly tagged, I focused on adding particular POI such as stores, bus stops, public buildings, etc. Thanks to the Cadastre, we have the detail of every building available in OSM, so we can precisely place POI without the need of GPS traces. The FreeRunner remains relevant for this task but just as well as other devices. At some point, I decided to switch to Android devices to map (with the OsmAnd app, that is free software). Since most of these devices come with a camera, it also permits to take pictures of the place quickly while mapping. A phone is fine for that, but the best I've found is a tablet with a large screen: you can place the POI precisely that way and enjoy the large keyboard. OSM mapping is one of the tasks you'll really enjoy doing with a tablet more than any other device.

Arduindoor

Written by Paul Kocialkowski - 02 september 2012 - no comments

Quite a while ago, I bought some device with a laser pointer. I did lost it over time and eventually ran into it just a couple days ago. In the meantime, I bought an Arduino and started experiencing with it, especially with photo-sensitive resistors (photoresistors).

Installation So I wondered what kind of fun could come out of a laser + a photoresistor + an Arduino. Actually, it didn't take long to figure out what to do. Remember all the Mission: Impossible-like movies and shows, where people are to move into laser-secured rooms, without interrupting any of the laser beams? Well, that's pretty-much the idea I came up with: "securing" my room door with such things.

Setup The setup is pretty easy: the Arduino is on one side of the door, with the photoresistor and another 10k resistor that we need to have a changing voltage. The laser is on the other side of the door, directed right to the photoresistor and connected to the Arduino via a couple wires. So the general idea here is to make the Arduino power the laser and get the tension from the photoresistor (or the 10k resistor) a couple times every second. Here is a scheme:

Scheme

Here is a demo video of the Arduindoor system running (first time making an alarm ring, second time playing/pausing music) :


If you're interested in the idea, you can setup your own with the additional help of: