You are reading content from Scuttlebutt
Re: %UJwRnvFGb

What is a gesture-based composition?

In smalltalk 80, if you want to pipe data & events between two running applications, you draw an arrow between the windows with your mouse.

When you have drawn that arrow, an editor pops up that allows you to configure the message being sent, what event triggers it, & the order of the args -- as well as whatever arbitrary smalltalk code you want to add.

(In other words, you use a mouse gesture to compose together arbitrary pieces of running code.)

The window that's popping up is literally an editor for live-editing the source of the application on the rear end of the arrow, so you can jump in and modify the application to provide an appropriate event if it doesn't already.

Re: %UJwRnvFGb


Given the current unhappiness out there with Chrome / Chromium, and it being effectively the only browser (== UI) in town, a part of me is wondering how hard it would be to implement the UI part of say #ssb in a smalltalk such as #pharo.

We know it is capable, cross-platform, supported, and can be made quite small, so deployment can be pretty easy, as it can mean just shipping a smalltalk image.

I don't have a lot of familiarity with the particulars of SSB, but if you're just connecting to a running sbot it can't be substantially harder to do it in smalltalk than javascript. Unlike mastodon, there isn't a whole lot of heavy webtech you're tied into here -- you've got json & markdown, but no html (so there's no particular reason to use html or css or whatever).

I was considering modifying fern to do ssb (basically, forking it & making a version that hooks to sbot instead of, & I don't see any particular difficulties in doing that, so I don't expect any in a graphical client either.

Voted this
Join Scuttlebutt now