Thread:Fandyllic/@comment-27075564-20170724155029/@comment-1077307-20170805001400

Well lucky for us if Wikipedia decides they like Lua, and this implementation is the one, then it will get driven back through MediaWiki and by sheer exhaustion and lack of will to fight it accepted by Wikia or keep divergent code.

To put a finer point on where I was going, there would never be enough infrastructure, staff, or developers to invent everything neat themselves and adopting a better alternate exposed server-side facility relieves that burden. If they want to evolve and grow alongside their peers, like FB, Google, etc... they need to expand their relative product similarly to allow this.

Toward the other, I've been very lucky. I was very lucky to be a part of many companies that happened to be early or first in many aspects of last 25 years of computing as we enjoy them now. In one case the first practical commercial "app server" and management products, at one time used by over half the fortune 500; the kinds tools and software servers that WM/Wikia could be built on. I've been lucky to get to be involved in projects creating 3D engines and supporting tool chains, and related actual custom Lua parsers, integrations and so forth for Lua IDEs.

So from my perspective, my guess and my hope is that this engine will stick. There are a ton of languages but they are very difficult to make, and there are fewer modern implementations than you would think to choose from, and are non-trivial to implement on your own like MW Code is for MW. The list is pretty short really.