Replies: 2 comments
-
What you are suggesting is no small effort. I'll speak for myself here, but the goal is not to become more popular than other frameworks like Phaser, and if it was, tooling is not the differentiator. These are all nice tools, I use some of these myself, but the benefits are negligible compared to the time and amount of changes needed.
This is purely a matter of preference and, to be frank, a baseless claim.
We already output esm/cjs/umd builds. Our build right now takes about 20 seconds on my macbook which is hardly an issue. If it were several minutes we'd probably look at using esbuild or vite or something.
This is subjective. I use starlight myself, I like it, but Docusaurus is perfectly fine too. |
Beta Was this translation helpful? Give feedback.
-
I appreciate the suggestions. We may take a look at some of these if they become major pain points, but I don't think we're there yet. We just completed a major migration to Docusaurus, so I'm not too interested in another docs framework unless it truly adds some new value (I briefly looked at Starlight and didn't see a major differentiator). Theming/branding is a separate issue, and we'll eventually tackle that. If people are truly interested in helping us grow the project, as @mattjennings suggested, it's not tooling but marketing we would love help with. Here are some concrete ways to help out as a community member:
There are things we want to do as the core team, but ultimately, we are community-driven, so we lean on community members right now to grow.
This is something I agree with, and we want to do (we're doing it... gradually). Docs are critical for our growth. We want to add more examples, we have the ability to embed live sandboxes (it's using Stackblitz), and we need to re-organize/fix the taxonomy of the docs to be more approachable. It's not easy because we have to maintain backward-compatibility with old docs permalinks, but we have some ideas on how to address that (either through Cloudflare rules or a Docusaurus redirect plugin). |
Beta Was this translation helpful? Give feedback.
-
This is a good project, with better ts DX. But its influence is still small. I didn't know it until last week. I think these ideas can make it move forward fast and be better.
now this project has many repos, it's too scattered. monorepo can gather more attention in one project and is easier to maintain and expand. pnpm make packages management more efficient. this monorepo includes
engine
,create-excalibur
,site
,plugins
four apps.plugins
app, then delete the origin repo, keep the only code entrancecreate-excalibur/templates
, then user typepnpm create excalibur
can choose the template they want. but I think we can only usevite
,vite
has more and more became the only choice to dev with frontend project. others like parcel, webpack seems to be outdated. I think can keep templates likevite
for web,vite+electron
for desktop,vite+react-native
for mobile. and remove other all templates repos.engine
to esm/cjs/umd bundleswebpack is so slow and tsup is faster, powerful and easy to config
site
using starlight to replace Docusaurusstarlight's UI is more modernization and it's easier to maintain docs, will make project look more professional.
At the same time, we can also reorganize the document structure to make it clearer and novice-friendly.
like we can update
examples
andshowcase
to show in codesanbox but not jump to github repos. it's more IntuitivelyI think maybe it's worth to discuss how to make this project grow up faster to compete with more popular frameworks like phaser or kaboom.
Beta Was this translation helpful? Give feedback.
All reactions