Fork me on GitHub

anybody doing robotics with clojure? i recall seeing a library for embedded devices that had a lot of clojure.core implemented... i wonder if it's feasible for low-level stuff (yet)


Not a Clojure implementation (nor have I read it in detail), but @UCFG3SDFV explains how to use Clojure on Raspberry Pi's:


That's the closest thing to Clojure+robotics I can recall ever seeing


But yeah, not really robotics/embedded 😞

Adam Helins06:06:42

It is more about targeting lInux ARM devices (like the RPI) that genuine embedded work (in the microcontroller sense)


I used to be active in RoboCup (about 5 years ago), and there was a team who built their server in clojure, and an derivative language to run on their robots. Its called ferret;


Wow that's super cool. Compile it down to C++11 executables! genius

Drew Verlee14:06:45

Can anyone recommend a good resource on learning the rete algorithm or forward chaining rules systems?

πŸ‘€ 3
Dane Filipczak15:06:27

Production Matching for Large Learning Systems, Robert B. Doorenbos

Drew Verlee16:06:24

Would you or different my someone you know be interested in doing a group reading of this?

Dane Filipczak17:06:14

Sure! Additionally, a literate /learning implementation might be fun to approach as a group

Jimmy Miller19:06:35

I'd potentially join in on that

Drew Verlee19:06:22

I'll see if I can find some others to join us. I'm excited your interested. :)

Drew Verlee21:06:00

@U5K8NTHEZ great. I'm away from my computer tell Monday, but hopefully we can get 2 more and try to break the paper into bite size parts.

πŸ‘ 3
Drew Verlee12:06:25

I'm going to write my thoughts in a google doc so it's easier to share.

Drew Verlee13:06:42

@U5K8NTHEZ @UU1GF5MGQ I wrote an overview of how we can match segments of the paper to a meeting time frequency so that we can finish it fast enough that we dont lose interest but slow enough to learn the material.

Drew Verlee13:06:05

Apologies for any mistakes in that doc, the next thing on my list is to fix my keyboard.

Jimmy Miller13:06:40

Two questions. Kind of confused on the timeline. It lists 10 different reading goals, that are biweekly and then says we would finish in 6 weeks. Something isn't adding up in my brain. Maybe I'm misunderstanding? The other thing I'm unsure about is the desire for 6 people. I've found that this is often the hardest part. I have no idea what time zones you all are in, but finding that many people who want to read a thesis and can meet at the same times might be a bit tricky. I've seen lots of groups not exist because of goals like that. I guess I'm just curious why you want that minimum number?

Jimmy Miller13:06:58

I will add that I've been quickly reading through and it is a super readable paper.

Drew Verlee15:06:48

err i said bi weekly, what i meant was twice per week.

Drew Verlee15:06:38

no hard reason on the number of people, it means if 2 dont show we still have 4 which is comfortable.

Drew Verlee15:06:03

agreeing on a time is a big deal

Jimmy Miller16:06:00

Makes sense on both counts. But would definitely recommend making the number of people a loose goal. I'm in us east time zone and can do this during work hours. Not sure I could do twice a week outside work hours.

Drew Verlee17:06:09

It's your call, no point in meeting if no one can meet. Given the density of the material, I suspect we will have questions and discussions points long before the meeting then. Once a week should be fine, but then I would up the amount of reading. Unless we want to make it a two month thing.

Drew Verlee17:07:30

@U5K8NTHEZ @UU1GF5MGQ I made a discord sever. I suppose we could try to make a short lived channel on this slack to, idc. The messages will disappear really fast though. The goal for today is to set a first meeting date and reading. It seems like we should do about 20 pages or so a week and meet at the end to discuss it in person for maybe half an hour to an hour. Meanwhile we can ask questions or discuss it in the discord at anytime. Feel free to invite more people. These things are never perfect and if your not finding value in it then its easy to back out πŸ™‚


I once saw a (I think strangeloop) talk about a programming language where the code is granularly stored in a sort of an immutable registry. So if Alice and Bob write/use the same function independent of each other, it actually is the same function regardless of the name. Roughly like that. But I don’t remember the name of the language.


That's a pretty cool idea