[azdle]

My meta side project for building other side projects: https://bodge.app/

I've always had a bunch of small side projects that I want to do that aren't worth the overhead required to actually put them together & keep them maintained. So, I built a small Lua-based FaaS platform to make each individual project less work whenever inspiration strikes. So far I've built:

* A current-time API for some hacked-together IoT devices: https://time.bodge.link/

* A script for my wife that checks her commute time and emails her before it's about to get bad.

* An email notification to myself if my Matrix server goes down.

* A 'randomly choose a thing' page. https://rand.bodge.link/choose?head&tails

* A work phone number voicemail, the script converts the webhook into an email to me.

* An email notification any time a new version is released for a few semi-public self-hosted services.

* Scrapers for a few companies' job listings that notify me whenever a new job is posted matching some filters.

* A WebPush server that I eventually want to use for custom notifications to myself.

* An SVG hit counter: https://hits.bodge.link/

Since I'm already maintaining it for myself, I figured I might as well open it up for others. It's free to play with, at least for now.

[alabhyajindal]

Very very cool! Just signed up. Reminds me of Val Town which I'm a big fan of. Did you choose Lua because you love using it, or for some other pragmatic reasons?

Do you think a service like yours with support for many variety of languages a good idea? Not in order to meet user demand but purely because I think it would "just" require running the program on the server using a different interpreter/compiler, assuming code sandboxing has been achieved to make the initial language work.

For example, I love the long list of languages supported by Code Golf: https://code.golf/wiki.

[azdle]

Thanks!

> Did you choose Lua because you love using it, or for some other pragmatic reasons?

A bit of both, though I'm literally drinking out of a coffee mug with the Lua logo on it that was given to me after playing a big part in making Lua a thing at a prevoius job. That might speak to my love of Lua.

> Do you think a service like yours with support for many variety of languages a good idea?

From a technical perspective, it would be relatively easy to add support for other languages, the biggest problem would be UI and documentation complexity. Each added language would either require a completely seperate set of documentaion or would require the docs to describe everything one layer of abstraction removed from the code people would actually be writing. Both of which would be less than ideal for my goal of extreme simplicity.

I think it can be a good idea, but to support something like that _well_ would require a pretty large team of people.

I do plan to support some level of 'other languages' for libraries, at a minimum some subset of native Lua libraries (ie. libs written in C). That means it would be possible to find a way to use pretty much any other language interpreter. However, I'm not sure that will ever be a top level feature, there'll probably always be some level of Lua glue code holding everything together.

[OccamsMirror]

Even better it could just support WASM and be language agnostic.

[azdle]

It's actually already using wasmtime as one layer in its sandbox. I just think that trying to support other languages, especially in a fully language agnostic way, would make things like documentation far more complex than I could handle and make the service complex enough that the only people who could understand it would be the type of person who don't really need a service like this in the first place.

[catlifeonmars]

This is great. One thing that is not immediately obvious to me is what kind of authentication it supports. Needing to authenticate the caller is the #1 piece of boilerplate that gets in the way of me completing mini-projects. It’s not hard, but it’s definitely a nonzero amount of effort. And this is coming from someone who has implemented many, many auth flows professionally.

[azdle]

It doesn't say because there's no special support for any auth protocols. Long-term I want to have out-of-the-box support for things like OAuth (for user-facing auth) or mutual TLS (for device/service auth). _Technically_ there's currently support for the cryptographic primitives required to do JWT (I added that because I wanted to support WebPush w/ payloads for myself), but those aren't documented because I intend to remove the current slightly-hacky custom APIs and replace them with some off the shelf libraries, but I'm still figuring out user-added libraries (and on top of that I'll also need to figure out support for native libraries).

Are there any auth protocols / flows you think would be important to support?

[catlifeonmars]

Ah makes sense!

> Are there any auth protocols / flows you think would be important to support?

- I think API key passed via basic HTTP auth would get you pretty far. This is ideal for serving machine-machine requests and just requires that both parties can securely store the secret.

- OIDC is great for interactions that happen in the browser or if the function is serving multiple users, but is more complicated to setup and/or use correctly.

OpenID connect is probably the best for contexts where you want something served by multiple users and those users are humans.

> _Technically_ there's currently support for the cryptographic primitives required to do JWT (I added that because I wanted to support WebPush w/ payloads for myself)

This is probably a good intermediate solution FWIW - expose signature verification and HMAC APIs and allow a user to bring in their own implementation.

[adityaathalye]

> My meta side project for building other side projects

Looks cool, congrats on putting it out there as priced service!

And, same!

Except, it's just a repo organisation system (structure, conventions, and tools) that lets me share common "parts" across multiple "projects". No monolithic frameworks here.

Libraries are functions. Apps are objects.

However, normally, we use these as distinct artefacts, eventually leading to the "diamond dependency" problem (and lots of other annoying development-time stuff caused by libs / code that is "over there" (elsewhere)).

My "meta side project" solves, essentially the Expression Problem as it manifests in source code management (particularly, cross library / service / project feature development).

[0] https://github.com/adityaathalye/clojure-multiproject-exampl...

[bisaacs]

Thanks for sharing! I signed up and tried it for something simple (storing a message via POST and displaying it in HTML via GET) and it was delightfully easy & approachable.

[azdle]

Great to hear! And thanks for saying so. I've definitely tried to make it as simple and straight forward as possible, but I really didn't know if it would be simple and straight forward to anyone but me.

[mattbettinson]

FWIW I think you should paraphrase this comment on the hero of the site! I was unclear what it did till i read your comment fully.

[azdle]

I know the homepage needs way more answer to "WTH is it?", I just don't really enjoy doing the 'marketing' side of things. I hadn't really considered just throwing something informal up there, but I guess I don't really know _why_, so, thanks for the suggestion.

[saadn92]

Love this, especially the commute idea

[azdle]

Thanks!