Now one possible solution is to just build a lisp-like syntax on top of Agoric's Jessie. Then I'd probably be happy enough! And that may be less work than duplicating work in my Racket ecosystem, even if it seems silly. Plus it means that non-lisp users could easily contribute to the system.

But... it's a pivot. I'm hesitant about making another pivot. But it's probably the right one? I'm unsure.

Community feedback welcome.

I had a conversation with Mark Miller while at RWoT. Should Spritely be written on top of Agoric? From a duplicating work standpoint, probably.

Except I don't like working with Javascript! I'm a known lisp enthusiast. And I'm not very productive unless I'm happy in my hacking environment.

But MarkM pointed out that Agoric has worked on Jessie, a subset of Javascript that is closer to Scheme: https://github.com/Agoric/Jessie (Not entirely, but it is more beautiful)

(cotd ...)

Here is a big fight I am going to have to make with myself ahead. And actually, I can use community input.

The current plan for Spritely has been for me to use the tools I am writing in Racket. I've been very happy hacking in Racket personally and the tooling is coming along.

However, I am duplicating work that the Agoric folks are doing in making an ocap secure substrate rich enough to do this stuff: https://agoric.com/

Except... their stuff is in Javascript. (cotd ...)

@davidak I chose the latter, but I have guilt about it. It wasn't easy; MediaGoblin *was* my identity for many years. Ultimately I think I made the right choice, but I still feel guilty abou tit.

@davidak MediaGoblin's donors made ActivityPub happen. There wouldn't have been enough resources to be able to balance that work and allow us to initially justify it otherwise.

I feel guilty that I stepped back after the standards ended, but I needed to take stock of what the most important thing I could do next was: continue MediaGoblin (even now that Peertube and etc picked up AP) or try to pave a path for all the stuff the fediverse couldn't do yet but needs to?

@cjd It's also the issue/feature (bi-directional links / annotations) I'm most uncertain about in terms of whether to build it into new systems. Ups and downs both seem large.

@conservancy just published a blogpost: "Richard Stallman Does Not and Cannot Speak for the Free Software Movement" https://sfconservancy.org/news/2019/sep/16/rms-does-not-speak-for-us/

@cjd Another test: "Is the drama introduced less or greater than current behavior?"

In general, drama will usually exist and may be different; we can't assert that we can remove it. But we should do our best to try to reduce it, or at least if drama exists, try to maximize for only the cases where drama is "useful" (eg, enabling pro-democracy protests is one)

@cjd But you missed the biggest one that I've debated for a long time: Xanadu bi-directional links. Xanadu was very concerned about the possibility of what we now call "fake news" and added the feature that from your web of trust, you can see annotations of content. Links thus didn't go in just one direction, you can see what links *to* something.

But does that succeed? Could it help us discover false information, or just result in two islands of "truth"? Does it open up dogpiling attacks?

@cjd This is an interesting document but each section should include a "DRAMA POINT:" section, explaining how you expect drama to be raised.

But better would be "DRAMA MITIGATION:" and "DRAMA MAXIMIZATION:" on each section, explaining how each point occurs in either.

Currently it takes a lot of effort for me to try to extrapolate what *you* think the sources of drama are in each "feature".

@cjd No, you can't prevent people from being jerks, and you can't prevent people from being exposed to jerks. I have never, and will never, make that assertion.

I will assert that we can *mitigate* jerk exposure though.

@cjd Are you the "future of the fediverse red team" now? ;)

@ekaitz_zarraga @ricardojmendez I would also like to be able to pay other people to help, once we start building the virtual worlds stuff. This is a ton of work; I could really use help.

@ekaitz_zarraga @ricardojmendez It's hard to fund this work though. I am interested in figuring out how to find a more stable support structure for this work.

But it is the most important thing I can do with my life, I believe. I will not give up on it.

@ekaitz_zarraga Right now I am on a grant that Samsung Stack Zero is funding me on (thank you @ricardojmendez and friends!). That's given me the flexibility to flesh out these ideas in depth. And of course I have some income from Patreon, but that's not enough to cover my expenses independently.

Historically I've scraped together resources wherever I can find them, largely through contracting, though the first year I started doing this was from the MediaGoblin crowdfunding campaign

@cj not sure

I might seem all over the place, but I've actually been very singular in my focus towards distributed, rich virtual worlds for over a decade. It's just a lot of work to get there:

- MediaGoblin lead to ActivityPub
- ActivityPub is a distributed social system
- Liberated Pixel Cup gave us base graphical assets
- XUDD, 8sync, Goblins are all actor experiments so I can understand the technical foundation
- Datashards lets us securely distribute assets
- Ocaps bring us security