Components not an IDE?

David Bovill david at viral.academy
Tue Nov 24 07:58:09 EST 2015


I think we have an issue with the way we think of LiveCode and in
particular the IDE. We picture it as a CMS or an opinionated IDE. How about
thinking of it like a minimal core engine (or set of engines), and
installable components that can provide authoring and editing capabilities
if needed?

More like Node + Node Package Manager (npm) + Express? So we would have the
engines - JavaScript, and various platforms, we would have a package
manager, and maybe a good minimal reference app.

npm install livecode


That would be nice, and something I'd love others to get involved in.

The idea here is to break LiveCode down into core minimal components - and
help with the overall modularisation initiative.

I'm not clear how the current plans for the IDE are progressing in this
regard - it seems the plan is to take the existing IDE and make it easier
for people to add extensions? Again I would very much prefer the opposite
strategy - forget the IDE and look to integrating with other communities
first then adding component by component until we get something that
resembles the IDE.

On 24 November 2015 at 12:21, Peter TB Brett <peter.brett at livecode.com>
wrote:
>
>
> All the code and build tools for the HTML5 engine are public, so there's
> no reason why LiveCode Open Source Edition users who want these features
> can't have a go at implementing them themselves if they can't wait for me
> to get round to it.  After all, one of the main reasons for going Open
> Source in the first place was to enable & empower developers who have
> different priorities to the core dev team!


I'd like to pursue this at CCC and FOSDEM!

n terms of open source strategy is that it is difficult to build a
community around a large established code base that was proprietary.
Blender managed it, and I followed that initiative from it's launch, but in
many ways LiveCode is less compelling an open source proposition than let's
have an open source 3D and game engine. It's value proposition is very
subtle - to do with literate programming and democratising coding in my
opinion.

If we want to build an open source community faster we should start with
something minimal - and extend it. Not start with a big monolith and ask
for contributions. It seems that we now have the architecture to start
doing this work?

This is essentially what I find the core missing part of the existing open
source strategy. We should be looking to integrate with existing projects
by adding to them - not asking everyone to come to an entirely new way of
doing things. This means plugins and components and interoperability.



More information about the use-livecode mailing list