Coronium ACE - Preview One

Solar2D General Questions/Discussion
#1

Hi,

I’m running a preview of a framework I’ve been tinkering with over the year called Coronium ACE. There is no client, as you can build one to suit your own needs very simply. Everything is based on regular HTTP(S) communications and JSON (or plain text).

The Coronium ACE toolbox is small, but used properly can build big things, or small things. There is also a compiled version that runs on the Raspberry PI, making it interesting for IoT applications.

I’ve decided on Docker as a deployment base, so you can run ACE anywhere Docker runs, which is almost everywhere now (even on the RPI). This also allows for many different deployment and scalability options.

More docs, screencasts, and tutorials are on the way. Coronium ACE is also the base for the upcoming OAuth2 service Coronium Arthur for Corona SDK, so I have a lot of faith in it.

Would love any feedback and testing experiences. As a preview, consider it pre-alpha. Have fun!

http://www.develephant.net/coronium-ace-preview-one/

Cheers.

#2

Hi,

A YouTube playlist has been started for Coronium ACE. Come learn more about using it for your Corona, IoT, and API projects.

https://www.youtube.com/playlist?list=PLlWSohAGdLDtC0ShNf9P2KyYqwEH79u09

New videos will be added frequently.

Cheers.

#3

Chris your work is amazing as always, if I was Corona I would snap you up in an instant!

#4

Thanks @Andrzej Zamoyski  :smiley:

There are 3 more videos up with lots of code use. Click here for Parts 3-5.

Cheers.

#5

Is the raspberry pi image close?

#6

Hi,

@scottrules44 - I’m still working out the deployment options. But it’s a high priority.

The main issue right now is trying to figure out the most simplified method of getting Docker and ACE on the PI. There really isn’t anyway to do that without bringing more external “work” into the mix.

I currently have a process that will get ACE up and running on a PI, but it’s not a one-liner (which I strive for). I’ve also considered a stand-alone engine, that can be dropped on a PI (or any box) without the need for Docker. But this would require either a serious refactor of the current code base, or just starting it as its own project (Coronium PI?). This idea is interesting to me since there are extended functionalities on the PI that are not needed on a server based version. It may be easier to maintain this separately.

I’ve even attempted to build my own Linux distro for it (buildroot). But I decided that was just “one more thing” I didn’t really want to add to my maintenance list. Plus I was never 100% successful as it is.

And, last but not least, Docker on PI will not “run everywhere” (something Docker touts) as all sources need to be re/cross-compiled to be compatible with ARM processors. This is a slow process, though I have been able to speed it up with a little internal distcc network.

So, where does this leave us decision wise?

  • Keep working on Docker PI integration with current code base.
  • Build a stand-alone app that can be dropped in (requires refactor).
  • Provide the full-blown build instructions and let the users figure it out.

_ All opinions, ideas, and use cases are very appreciated. _

Cheers.

#7

[quote=“Develephant,post:6,topic:336010”]

Hi,

@scottrules44 - I’m still working out the deployment options. But it’s a high priority.

The main issue right now is trying to figure out the most simplified method of getting Docker and ACE on the PI. There really isn’t anyway to do that without bringing more external “work” into the mix.

I currently have a process that will get ACE up and running on a PI, but it’s not a one-liner (which I strive for). I’ve also considered a stand-alone engine, that can be dropped on a PI (or any box) without the need for Docker. But this would require either a serious refactor of the current code base, or just starting it as its own project (Coronium PI?). This idea is interesting to me since there are extended functionalities on the PI that are not needed on a server based version. It may be easier to maintain this separately.

I’ve even attempted to build my own Linux distro for it (buildroot). But I decided that was just “one more thing” I didn’t really want to add to my maintenance list. Plus I was never 100% successful as it is.

And, last but not least, Docker on PI will not “run everywhere” (something Docker touts) as all sources need to be re/cross-compiled to be compatible with ARM processors. This is a slow process, though I have been able to speed it up with a little internal distcc network.

So, where does this leave us decision wise?

  • Keep working on Docker PI integration with current code base.
  • Build a stand-alone app that can be dropped in (requires refactor).
  • Provide the full-blown build instructions and let the users figure it out.
    _ All opinions, ideas, and use cases are very appreciated. _

Cheers. [/quote]
I am happy with whatever, as long as there is a tutorial out there on how to get this on my pi.

Thanks for all your hard work

#8

Hi,

Are you running RPi 1b or 2?

Cheers.

#9

I have 2 2b

#10

Hi,

A YouTube playlist has been started for Coronium ACE. Come learn more about using it for your Corona, IoT, and API projects.

https://www.youtube.com/playlist?list=PLlWSohAGdLDtC0ShNf9P2KyYqwEH79u09

New videos will be added frequently.

Cheers.

#11

Chris your work is amazing as always, if I was Corona I would snap you up in an instant!

#12

Thanks @Andrzej Zamoyski  :smiley:

There are 3 more videos up with lots of code use. Click here for Parts 3-5.

Cheers.

#13

Is the raspberry pi image close?

#14

Hi,

@scottrules44 - I’m still working out the deployment options. But it’s a high priority.

The main issue right now is trying to figure out the most simplified method of getting Docker and ACE on the PI. There really isn’t anyway to do that without bringing more external “work” into the mix.

I currently have a process that will get ACE up and running on a PI, but it’s not a one-liner (which I strive for). I’ve also considered a stand-alone engine, that can be dropped on a PI (or any box) without the need for Docker. But this would require either a serious refactor of the current code base, or just starting it as its own project (Coronium PI?). This idea is interesting to me since there are extended functionalities on the PI that are not needed on a server based version. It may be easier to maintain this separately.

I’ve even attempted to build my own Linux distro for it (buildroot). But I decided that was just “one more thing” I didn’t really want to add to my maintenance list. Plus I was never 100% successful as it is.

And, last but not least, Docker on PI will not “run everywhere” (something Docker touts) as all sources need to be re/cross-compiled to be compatible with ARM processors. This is a slow process, though I have been able to speed it up with a little internal distcc network.

So, where does this leave us decision wise?

  • Keep working on Docker PI integration with current code base.
  • Build a stand-alone app that can be dropped in (requires refactor).
  • Provide the full-blown build instructions and let the users figure it out.

_ All opinions, ideas, and use cases are very appreciated. _

Cheers.

#15

[quote=“Develephant,post:14,topic:336010”]

Hi,

@scottrules44 - I’m still working out the deployment options. But it’s a high priority.

The main issue right now is trying to figure out the most simplified method of getting Docker and ACE on the PI. There really isn’t anyway to do that without bringing more external “work” into the mix.

I currently have a process that will get ACE up and running on a PI, but it’s not a one-liner (which I strive for). I’ve also considered a stand-alone engine, that can be dropped on a PI (or any box) without the need for Docker. But this would require either a serious refactor of the current code base, or just starting it as its own project (Coronium PI?). This idea is interesting to me since there are extended functionalities on the PI that are not needed on a server based version. It may be easier to maintain this separately.

I’ve even attempted to build my own Linux distro for it (buildroot). But I decided that was just “one more thing” I didn’t really want to add to my maintenance list. Plus I was never 100% successful as it is.

And, last but not least, Docker on PI will not “run everywhere” (something Docker touts) as all sources need to be re/cross-compiled to be compatible with ARM processors. This is a slow process, though I have been able to speed it up with a little internal distcc network.

So, where does this leave us decision wise?

  • Keep working on Docker PI integration with current code base.
  • Build a stand-alone app that can be dropped in (requires refactor).
  • Provide the full-blown build instructions and let the users figure it out.
    _ All opinions, ideas, and use cases are very appreciated. _

Cheers. [/quote]
I am happy with whatever, as long as there is a tutorial out there on how to get this on my pi.

Thanks for all your hard work

#16

Hi,

Are you running RPi 1b or 2?

Cheers.

#17

I have 2 2b