Re: Pick One Platform And Stick with it! (assembling a team)

[Brent Picasso]

Hi Anfred,

Thanks for your passionate post. I've looked at the chip; it looks like it could simply run the standard Lua under Linux. From my understanding, eLua is best suited for OS-less microcontrollers where you don't have the luxury of a POSIX compliant OS. What might I be missing?

-Brent

On Fri, Jul 17, 2015 at 9:27 AM, Anfred [via eLua Development] <[hidden email]> wrote:

I have read most of the posts on in this forum.  I have seen many posts interested in doing an eLua port for a specific board.  Sadly, there has been little progress on most ports.  We must combine our efforts!

I am starting a team here.  I need people who are passionate about Lua.  I need people who are not willing to settle for the same old Arduino/RPI options.  People who believe in a world outside of linux and C code.  The future will be about code literacy.  About the average 12 year-old being able to read code like a it's a Harry Potter novel.  Lua is at the heart of that world.  Lua is the future!
 
OK maybe there will be many languages people use in the furture, but Python and Java have proven to be full of holes, in terms of how they can be used, especially by students and beginners.  I know there has been progress with Scratch and Java game programming etc.  At the end of the day, Lua has the best performance vs. simplicity of all beginner languages.  It is just a matter of not settling for less when we work towards the future of computing in general.  If you have ever read a few posts in the computercraft forum, you will realize this Lua shift is already happening.  Give beginners Lua, they will build tomorrow. 

There has been a shift in what hardware means: you can now get everything on one board for $9!  The best new board for eLua is C.H.I.P from next thing co.  The best hardware choice for anything is... well... CHIP.  For $9 you get , BTLE, WIFI, 512mb RAM, 4gb Storage, a 1ghz processor, and best of all, a composite video out.  Their github is here.

This means eLuaBrain would run very well on chip!  Why would you want eLua instead of a full linux distro?  Because the price could not be any better, no matter what you are trying to do, $9 for almost all hardware needs is as good as it gets, ever.  Want to make a bluetooth door lock? chip.  A simple 2WD robot? chip.  A board to learn computer programming with Lua? chip.

So I propose we do a port of eLuaBrain for chip.  eLuaBrain is for STM32 boards currently.  We need a port to use the Allwinner A13 SoC found on the C.H.I.P board.  If successful, the 512mb of ram would make Lua code run very well.  It would be like creating a Commodore 64. or a ZX Spectrum computer, except it would be modern.  Users could learn to code with wifi and bluetooth onboard.  They could make games, even whole, graphical operating systems!

I am not the best coder.  Join this team if you share the vision, regardless of experience.  Or if you are a seasoned programmer, looking for the next great hacking project.  Everything that goes up, must converge; the question is if you want to converge with linux and C code, or if you want to converge with something much more fun and accessible to the masses!

Thanks for Reading!  -Anfred  

---

If you reply to this email, your message will be added to the discussion below:

http://elua-development.2368040.n2.nabble.com/Pick-One-Platform-And-Stick-with-it-assembling-a-team-tp7578520.html

To unsubscribe from eLua Development, click here.
NAML

--

[ibisum]

>
> Thanks for your passionate post. I've looked at the chip; it looks like it could simply run the standard Lua under Linux. From my understanding, eLua is best suited for OS-less microcontrollers where you don't have the luxury of a POSIX compliant OS. What might I be missing?

A Lua-only system, as a standalone machine with its own complete environment - i.e. use that video output to give us an onboard editor/development environment.  A lot of current Lua-based systems are destinations for cross-compiler-based dev rigs; it’d be nice to use the power we’ve got to forge a Lua-only OS that can support its own development environment and be, indeed, pure 100% Lua, no host-OS.  That has been a goal of mine as a Lua neophyte for a few years now.

;
--
Jay Vaughan
[hidden email]
http://ibisum.ddns.net/

[Anfred]

Thanks for your reply!  I understand linux is an easy way to use the CHIP.  This may be off subject for eLua, when really we only need lua, instead of an embedded version.  I was hoping it could scale well, for boards that do have limited hardware, so then elua would be handy.  I am interested in an independent OS, though.  The idea is to not rely on linux or any other OS.  The eLuaBrain is a good example.  I know hardware constraints are the reason for making eLuaBrain, and the CHIP does not have limited hardware.  As you know, it is pretty powerful and capable of running linux.

Since hardware is becoming so cheap, I feel it is time to create an alternative to linux.  The idea here is to create a new standard for open source OS's.  Physical computing (embedded, robots, etc) has always used C, and more recently lightweight linux.  New hardware does not limit physical computing anymore.

The problem is that beginners and students are not inspired by C or learning linux commands, that has been proven over the past 30 years of them being used so widely.  After seeing what people have made in the computercraft forums, I realized beginners are inspired by lua, and scripting friendliness in general.  Linux is great for hackers, open source devs, and community.  The problem is 80% of the population just has no interest in linux.  When they try to care about it, they go back to their macbooks, with a sigh of relief.

If you want to build a wooden bench, must you buy a whole wood shop?  If you were to buy a whole wood shop, how would you know what to buy?  That is the kind learning curve C and linux provide.  The idea here is: when beginners are not intimidated, they thrive.  I feel that lua is the key to inspiring a whole new generation of coders.  I also feel that C plays a big part in discouraging beginners.  I know web developers that have been in web design for 10 years, yet they would never be interested in C syntax, and don't even bother with linux.    

[BogdanM]

Hi,

First, let me say that I realy like your passion. I agree that Lua is a great teaching tool, especially combined with platforms like eLuaBrain, and would like to see more of that happening. I still don't believe the answer is C.H.I.P or other similar platforms that were originally meant to run Linux. The reason is very pragmatic: drivers and OS services. Drivers that you'll very likely be unable to get up and running when you plug in an USB WiFi dongle in your C.H.I.P. board. Even if you get the drivers right, you'll still need the proper support around WiFi, plus a proper network stack. Linux provides high quality and high performance versions of those, and that is not to be taken lightly. The effort that was put into that is huge and replicating even a small part of it is pretty insane.
On a MCU board, things are a bit different. You can easily connect a "smart" WiFi module and be done with the WiFi part. People don't generally expect MCU boards to have a high performance networking stack, so simpler, less reliable solutions are generally deemed acceptable. And (quite important for me) MCU based systems are generally much easier to understand than Linux hardware. This is a very important advantage if your aim is education, since you can provide both a hardware and software education tool in a single nice package.

[Anfred]

You have some great points there.  I am still learning how to be a better coder.  I try to make up for that in big ideas full of passion.

If a microcontroller, instead of a full linux board, is the best way to get lua into classrooms, than that is what matters to me!  I wish nodeMCU 12E was not so limited in ram, otherwise it would be great for this purpose.  I would have to say the CC3200 is probably the best candidate, although another post in this forum has steered me away from broadcom chips.  That leaves only the Particle Photon board as an affordable wifi solution.  I need to get one for eLua now!

[BogdanM]

Dado found this:

http://www.wifimcu.com/

Looks pretty nice and already runs a Lua version which is based on eLua. It's also fairly cheap for what it has to offer.

[Anfred]

Hooray!  Thanks for finding the wifiMCU!  I was just on smartarduino.com, about 2 weeks ago, and that board was not available!  Very new, very interesting find!