this would be crazy good

This is a very interesting approach… Possibly abandon webContainers entirely in favor of an electron application that itself manages local docker instances... it would represent an overall architectural shift which may be best pursued as a separate fork rather than an optional configuration of a fork that otherwise supports webComtainers

This is a very interesting approach… Possibly abandon webContainers entirely in favor of an electron application that itself manages local docker instances... it would represent an overall architectural shift which may be best pursued as a separate fork rather than an optional configuration of a fork that otherwise supports webComtainers

webcontainers is what makes this project unique. there are numerous other projects that have full access to the file system, such as devika but one thing I like about the bolt approach is everything is containerized in the browser so people who are cautious about messing up their file system (or having to run a docker container just to run this) can use this without any worry. if you're already a developer you probably don't see this as a strength but personally I see this as the biggest strength of this project.

@cocktailpeanut are you arguing that an electron application packaged implementation of bolt should continue to use web containers or that bolt need not be packaged as an electron application?

Personally, I'd love to be running locally even if I did create a container to keep my OS safe. Downloading the code is much different than running locally. It would allow us to start with an existing project and that is #1 in my opinion.

@baberuth22 Then I think you are looking for a feature to "Allow Bolt to start from an existing local project". And that would be a non-trivial change.

I think that real path to this kind of feature and one the also addresses a number of limitations bolt is....is standing up a full Git integration.

This is applicable to your situation as. it could be addressed in the following way:

1. To start a mini project. and push it to your github account.
2. Then you clone the repo on your side.
3. then you completely change the project as needed, but in a way that bolts still can understand it.
4. Push it. then tell Bolt to sync with your repo.
5. Voila.

Ha! Now that I say all that, I think I realize where you're coming from. How many folks? solve your problem, including myself, is using the VS code ext named Cline. It's truly. a game changer. and it's free and it's open source.. And when you combine it with Super Maven. free code completion. You basically. have a. better vs code setup than both cursor and the new Github copilot. And I'm totally serious making that statement even compared to the new Github copilot

Here is a Youtube Demo by AICodeKing

@schaveyt I'd settled on https://continue.dev which seems closer to SuperMaven than Cline. I'll be checking it out today, Thanks for the pointer!

As you know, there are several open issues triangulating on a variety of Git support integrations. Maybe rather than hijacking this issue, maybe we open a placeholder for discussion of coding assistant integrations... @coleam00 it would be nice to have a Discord server for the Bolt Fork devs...

@hillct I'm releasing a Discourse community soon for us! Even better than Discord! More updates on that coming soon.

The problem is @coleam00 Discord is the current standard. You're asking us to open up yet another channel.

@nigelp For larger efforts like this a Discourse community is very common! And it's just another browser tab so you don't have to install another app if you don't want to!

The fact that you answered so quickly has convinced me. :)

I've been involved in OSS software efforts since 1993, when IRC and UUNet were the standard tools for community building. Times change. Fads come and go. I agree with @nigelp that discord is the current standard for software development and support communities, and has been for the past 5 years. That may change and Discourse may be the hot new thing but this is the first time in my professional career when someone has asked me to join a service that I haven't at least heard of prior. I'm open to being convinced though...

@hillct I appreciate you saying this! I should have clarified in my last comment that I'm certainly not saying Discord isn't a good standard!

So Discourse is actually something you have probably used before without realizing it. It's a "white labelable" product, so companies like OpenAI and Asana use it as their forum and it looks like their own thing.

I'm preferring it over Discord for this for a couple of reasons. First, it's more professional looking, especially when Discord does things like make everything Halloween themed like they did last week. Also, the plugin support for Discourse is phenomenal for moderation and integrating AI into the platform. I know Discord has a lot of bots you can bring in and things, but my experience with the Discourse integrations has been absolutely perfect for what we need.

Lol, I still need to get the app running but noticed this issue.

Congrats on taking the initiative and forking a project to take it somewhere else that they probably didn't plan to.

Before deciding what technology or strategy to use, maybe you should decide what the product is supposed to do.
Why build a motorcycle if you plan to have a family and a dog?

1. what is this supposed to do? generate some boiler plate for an app? Or is supposed to be able to help you code your app start to finish?
2. Some of the PRs will be people trying to gain traction for their own services.
3. You need to start gathering the most common needs of the people. (I like how you are letting the community direct where the project is going). Jira has a product discovery tool. I'm not saying to use that tool but you do need to track and organize all these incoming ideas and then give them weight.

(( personally I want an application using my hardware to help me read code in my project and be able to give some intelligent code to add a feature. Obviously, start to finish. ))
I have clearly stated my opinion so that you can understand my agenda.

1. I think that speed is one of the most important issues that needs to resolved.
   If you application is slow, it better be the best dam coding experience or it better have everything I need.
   remember you get 2 out 3 options.
   So, seeing that VScode has been awesome since day 1, electron would be a good base. Or better yet, follow up on cursor IDE - see what you can steal from them.

2. I want an app that I can use to code a full app.
   I don't want boiler plate. I thought I did. I thought it was cool to run a command and boom, I have some files built for me.
   AI can do this for me.
   But if AI can do it for me, then why can't it see all the files I'm working on that are in my directory.
   You'll need to have it index the code base but phpStorm does that too.
   I want to be be able to open up a project, wait for the app to do it's thing, and then I want to be able to add a feature to my code.

3. I need you to do due diligence for me. How can I keep up with every single fork of the LLMs out there? Benchmarks are needed. Yeah, yeah ... look I don't care. I steal read them and use them to cut out the crappy stuff so I can save time. So, yes, benchmarks are useful. Having a curated list of LLMS with pros and cons will also help draw people to your site.

4. Did I say speed?
   I want to be able to work with the AI in my code, doing most of the basic work for me and knowing what to do when I'm being lazy but I don't want to watch it spin for a minuet before it pops out an answer. A few seconds is fine but it needs to be quick.

5. I have not clue what people are talking about loading something on their machine and being worried about it getting dirty. Look, it's a fact of life that you will probably need to reinstall your OS because something went rogue. But an app like this? Lord no way, it will create some clutter but I can't see how it will mess up your machine. Just reinstall your main OS every year. It will clean out all that other garbase you have been letting pile up in the next to the muffler bearings.

Overall, I look forward to your hard efforts and wish you awesome luck to attract $$$ of venture funding!

@illuminate3 I appreciate your thoughts here! We are definitely doing a lot of planning as a community right now to determine the direction of oTToDev. Taking our time with that because I don't want it to just be me figuring that out and asking everybody to come along - I want it to be a collaborative effort! But it is certainly important that we figure that out fairly quickly!

I love your other thoughts too - yes speed, benchmarks, power to actually build a full app - all important!