Git commits don't lie.

Reply to this note

Please Login to reply.

Discussion

They can have typos tho

๐Ÿ˜Ž

No dark band on top and bottom. Based.

#bullish

Few

Proof of work

๐Ÿค—

so they're liars. ๐Ÿคญ

git commit --amend

git commit -m โ€œboneless chicken wings are wingsโ€

Facts ๐Ÿ˜

Chicken is salad.

Hot dogs are sandwiches

FYI This is not true unless the hot dog was sliced in half and placed in sandwich bread. Itโ€™s been proven already..

Exactly ๐Ÿ™ƒ

straight to jail

No, they just identify as wings.

Is anyone actively working on git over nostr already?

Other than me I mean.

Dunno

Check out my kind 0 if you're on a decent client... looks pretty nice on gossip.

Currently embedding my utility library in the generator script so others can clone it too.

Yes, I working on it. Check out my long form nostr posts. I'll post a prototype CLI in Rust soon. Happy to discuss on a call if you like?

I prefer text, but we can discuss yes... That being said:

My main concern here is whether there is any point for me working

on my project publicly instead of privately.

Because if nostr devs are going in another direction anyway,

there is no point making the extra effort to put my work online...

I already put the clone script on nostr so it basically works.

There is a repo in my metadata right now, people can clone it.

Ultil relays start limiting how much stuff one can put in the

metadata fields I'm good.

I see no point working on anything related unless the nostr maintainers

specifically ask me to, because I always work alone.

I'm not a professional dev. So normally I have no need

for pull requests or issues, or whatever else people are familiar with.

Is your intention to build something that other people would use or for personal use?

At the moment I think the direction nostr:npub1xtscya34g58tk0z605fvr788k263gsu6cy9x0mhnm87echrgufzsevkk5s and seem to approve of is using the patch model. Everything else appears not to be fleshed out.

I think the patch model doesn't scale but is a good starting point because of it's simplicity.

I am not used to working in public either. There hasn't been any engagement with my long form posts on the issue. No one has commented on my draft NIP. My plan is to build a prototype and put it out there.

Eventually I want to publish some open source software. Not necessarily on nostr

but I really liked the protocol and I was like what the hell, let's see if it works.

That being said, I understand that there is one major problem with my approach

which would prevent frequently updated, big repos from being posted this way: bandwidth.

One would need to reupload the entire repo for the tiniest of changes.

Works for small / infrequently updated projects ( that's mostly what I work with ).

Fails spectacularly for big ones.

I'll take a look at your approach early next week if you'd like some feedback. I'll be a bit busy during the weekend.

Have fun coding.

Actually I just realized the bandwidth issue isn't that bad...

The repo could just mark certain patches as ACCEPTED, and it would be enough to write them periodically to the repo data.

Like every day or every week. No need to reupload everything for each patch.

I'm glad someone else is working on the problem. I'd love to take a look at your work. There is a catch 22.

As I understand it, you have put the tool to clone repositories from kind 0 events, in your kind 0 event. I can't clone the repo without the tool, which is only available in the repo.

No, I actually put the script that allows parsing that in a simple note.

There is no way to get around that sadly.

Turns out relays also accept twenty-something thousand chars in a kind-1.

I know it's an ugly solution, but it only had to be posted once.

but centralized time servers may! :)