Noel De Martin @noeldemartin@noeldemartin.social

@dajb @jamesravey @thoughtshrapnel I see, that's very weird indeed :/ I didn't notice it was working with curl, I thought it was outdated everywhere.

What I can say is that it must be some cache in the server side, because even clearing the cache completely or using a different browser altogether I still get the outdated version. But using curl I do get the new one.

Maybe it has something to do with that plugin? No idea.

@nolan Just the other day I tried to give this serverless thing a try, but I was discouraged to see that none of the solutions I found have a way to limit spending (DigitalOcean Functions, AWS Lambda, Netlify Functions, etc.).

Although I have to wonder, I suppose all of this can be configured through APIs, right? So why hasn't anybody build a 3rd party service/script to listen to these alerts and shut down the services automatically after crossing a certain threshold?

@nolan Also, tangentially related to this, I am a very big fan of using "conceptual compression" to deal with complexity :). It's somewhat of a cheat because it does create complexity, but it's encapsulated within simple concepts which is nice. https://www.youtube.com/watch?v=zKyv-IGvgGE&t=1047s

@nolan Something to mention is that nobody creates complexity for complexity's sake (although I agree with the "we love our own complexity" part). The thing is that we have problems to solve, and most of the time the "easy solution" (ironically) is adding complexity. To actually make something simple takes more time, not less. That's why complexity tends to grow over time. Even if essential complexity does grow with the scope of a project, I'd argue most is accidental complexity.

@weareopencoop @epilepticrabbit @dajbelshaw But at the same time, I do understand why most people don't want to self-host. Personally I think the problem is actually how software is architected, with data being coupled to apps. If something like Solid or other zero data protocols were more widespread, it would be a lot easier to use software that is aligned with our ideals.

PS: I think Doug's microphone is fine, but I've been looking into buying one recently so I'll be sure to avoid that one xD

@weareopencoop @epilepticrabbit @dajbelshaw I think there are a couple of reasons why it's been easy so far. The first one is that I'm the only one using these services, so I don't have to cater to anyone else's needs. And the second is that I rarely install updates, which I am aware is a security risk but to be honest I'm not too worried. I think software can be a lot more stable than people think.

@weareopencoop @epilepticrabbit @dajbelshaw I've been "self-hosting" (on DigitalOcean) a couple of sites and services for a few years and I have to say, as much as I dreaded the prospect of maintaining servers it's been pretty easy so far. I self-host my personal website, this Mastodon instance, and my personal Solid POD (Solid is a protocol for data storage).

Yesterday I reached 2000 movies in my Media Kraken collection, and I've watched 1208. It will probably continue growing at a quicker rate than I watch them, but I love having my own list. I never have to wonder what to watch next when I'm in the mood for a movie :D.

@nolan But wouldn't you need to have an html file with that name? For example, if I'm trying to have the url /user/123456, don't I need to have a /user/123456.html static file?

Maybe this can be worked around if the server allows you to customize the 404 html, and implement your thing in there. But the status code would still be a 404 I think, right?

Well in any case, regardless of it being possible or not I think we'd agree that by that point the complexity is not too far from doing an SPA :)

@nolan Yeah I also thought of something like that, but I guess you can't have dynamic urls, right? You have to resort to using query parameters or url fragments.

@nolan Another thing MPAs don't do (and probably never will): serve apps without the server-side hassle. There are many services and ways to host static apps easily and for free (it's arguable whether this is good, but it lowers the barrier of entry). But as soon as you need a server component, things get a lot more complicated. Building it this way also makes apps more portable.

Maybe I'm mistaken about the hassle of server-side computing though? That's been my experience at least.

@ccamara @humanetech @kytta @VincentTunru The only question that comes to mind is that all of this needs to happen in the server-side, right? At the moment, I am building my apps completely in the client-side, so I'm not sure how I'd manage to do anything with ActivityPub under that constraint.

In any case, I don't think I'll look into this anytime soon, because I want to focus on finishing the first version. But thanks for the ideas :).

@ccamara @humanetech @kytta @VincentTunru I see, I guess my notion of ActivityPub is indeed colored by the apps I've used or seen :).

I did know about Funkwhale, but I don't understand how it's using ActivityPub either so I guess I'll take a look. I do like the idea of enabling recipe variants, modifications, and such.

@VincentTunru @kytta Hey I'm curious about these use-cases. Why would you want a recipes app powered by Activity Pub? I've seen other people mentioning this type of thing before, but AFAIK Activity Pub is intended for social network-type apps, and a recipes manager doesn't fit that description. Is it in order to add comments to recipes and such? Or if you want to share the recipes, would posting a toot with a link to the recipe be enough?

I'll be in Amsterdam next week for #LaraconEU and I'm looking for a co-working space to go on Monday. Any suggestions?

The Laravel conference I was going to attend in Amsterdam was cancelled, but it's finally been re-scheduled for April 26th. This time it looks a lot more sure to happen 🤞 See you there #LaraconEU!

@VincentTunru I've been using it for a while and I like the idea, but the experience has been quite bad to make it work with Vite :/ I had it working, but now it's broken after I updated some dependencies. I'll have to give Vitebook a try, but it seems like it's still very experimental https://github.com/vitebook/vitebook

#NewProfilePic I've changed my avatar for the first in almost 8 years! And I've got @flicknelson@twitter.com to thank. For those who don't know him; he's designed the last couple of Laracon websites. So I'm very happy to both renew my image and strengthen my identity as a Laravel fanboy :).

@humanetech @baldur @benwerd Yeah, I'm "following closely" the Solid project and I don't know much about that project either. I just know they are working with the government and plan to give PODs to each citizen. Just what I've read on Inrupt's blog and some announcements in Solid World.

The community part definitely has a lot of room for improvement. But I know that's not Inrupt's main goal. The problem is that nobody is taking care of that.

@humanetech @baldur @benwerd I don't think Tim is acting as BDFL, and the spec is not driven by Inrupt. As far as I know, all the discussions regarding the spec are public and work like any other W3C standard. For example, this one recently happened and not everyone who participated is employed by Inrupt: https://github.com/solid/specification/issues/227

But yes, it's true that *most* people is involved with Inrupt. I think that's just because not enough people outside of Inrupt is interested to join.