To me this would be useful. A few things I'm wondering:

• Shouldn't there also be a check to ensure the Chrome container is up and running?
• Personally I would start workers and web in separate terminals. This way the logs remain distinct and easier to navigate. Would there be a way to somehow achieve the same here? OTOH I could see some value in them being intermingled, so probably not that important.
• Have you tried docker-compose.dev? If so, I briefly looked at Dockerfile.dev and it seems like it's missing the yt-dlp dependency. Could that be (part of) the problem you were having?

@irobot

Shouldn't there also be a check to ensure the Chrome container is up and running?

mmm I haven't had to test anything that required it so i haven't thought about it lol.
i think running the current 3 services represents the bare minimum, so maybe we can keep it as default, and add additional flags to run "everything else" or "full", which tentatively includes the chrome container. What do you think about that?

Personally I would start workers and web in separate terminals. This way the logs remain distinct and easier to navigate. Would there be a way to somehow achieve the same here? OTOH I could see some value in them being intermingled, so probably not that important.

If i need to dig deeper into the logs, I would run the separately, but usually I just want to start the dev environment to run a PR branch locally.

Have you tried docker-compose.dev? If so, I briefly looked at Dockerfile.dev and it seems like it's missing the yt-dlp dependency. Could that be (part of) the problem you were having?

I was trying to get the volumes setup "nicely" and also getting the relative paths to the .env aligned, but i run into some problem when trying to do the db migration. I can probably find out what my actual problem is with some time, but since there is an open PR to fix/improve it, I decided I'll just spend my time elsewhere instead

Shouldn't there also be a check to ensure the Chrome container is up and running?

mmm I haven't had to test anything that required it so i haven't thought about it lol. i think running the current 3 services represents the bare minimum, so maybe we can keep it as default, and add additional flags to run "everything else" or "full", which tentatively includes the chrome container. What do you think about that?

@xuatz Here's my understanding so far:

• Ingestion - chrome is essential in the process of saving hoarded pages, e.g. reader view, screenshot capture. meilisearch has no role in any of that.
• Viewing - chrome is not relevant for viewing already stored links, while meilisearch is.

In summary, depending on what you're working on, you might need only one or the other, or both. If we had to choose only one, it would be difficult for me to pick which one would be more useful to have as a default. Otherwise, for consistency it seems appropriate to have both.

Shouldn't there also be a check to ensure the Chrome container is up and running?

mmm I haven't had to test anything that required it so i haven't thought about it lol. i think running the current 3 services represents the bare minimum, so maybe we can keep it as default, and add additional flags to run "everything else" or "full", which tentatively includes the chrome container. What do you think about that?

@xuatz Here's my understanding so far:

• Ingestion - chrome is essential in the process of saving hoarded pages, e.g. reader view, screenshot capture. meilisearch has no role in any of that.
• Viewing - chrome is not relevant for viewing already stored links, while meilisearch is.

In summary, depending on what you're working on, you might need only one or the other, or both. If we had to choose only one, it would be difficult for me to pick which one would be more useful to have as a default. Otherwise, for consistency it seems appropriate to have both.

@irobot
let's have both by default then ^^

This is pretty cool, thanks @xuatz!