Is the correct main bucket (and thus main bucket manifests) used if the user never invoked the scoop bucket add main command?

[Gitoffthelawn]

This question is related to issue #6171.

If the user never invokes a scoop bucket add main command, a main bucket is still used, but the URI it points to is different than after the user invokes the aforementioned command.

In this case (when the user has never invoked the scoop bucket add main command), is the correct main bucket (and thus the correct main bucket manifests) always used?

[HUMORCE]

The installer downloads ZIP file of Scoop/Main bucket repo. Then, when you run scoop update|install|etc., Scoop tries to turn itself and the main bucket into a git repo, if git is installed.

• The main bucket that comes with fresh installation of Scoop is the latest commit of main bucket repo.
• Scoop will not verify the remote url of bucket.
• In some cases, you may lose the main bucket on the first install. when that happens, you should fix it manually or just reinstall Scoop.

[Gitoffthelawn]

Great info there. Thank you.

Yes, for contents of bucket, the following 4 command lines are equivalent...

Initially, is the content also equivalent to the main bucket that comes with a fresh installation of Scoop?

For the next paragraph, let's refer to the main bucket that comes with a fresh installation of Scoop as "A", and the main bucket created with any of the 4 commands you mentioned as "B".

Over time, assuming the user does not invoke one of the commands you mentioned, will that content of "A" and "B" start to diverge because "A" is static, whereas "B" will receive updates? Or will they both receive updates?

IOW, is there typically a need for the user to invoke one of those 4 commands?

[HUMORCE]

is the content also equivalent to the main bucket that comes with a fresh installation of Scoop?

Sure

IOW, is there typically a need for the user to invoke one of those 4 commands?

No. After buckets added, Scoop will update them via git when appropriate.

Over time, assuming the user does not invoke one of the commands you mentioned, will that content of "A" and "B" start to diverge because "A" is static, whereas "B" will receive updates? Or will they both receive updates?

• If user does not have a git installation, Scoop will prompt about it.
• If bucket has not been converted to git repo, Scoop will always prompt and try to convert it. (main bucket only)

---

Scoop supports installing apps from the following sources:

• buckets
  • local bucket: ~\scoop\buckets\local
  • remote bucket
    • official buckets: you can list them via execute scoop bucket known
    • 3rd-party buckets: search on github or visit https://scoop.sh/#/buckets
• standalone manifest
  • local file: scoop install C:\app.json
  • remote file: scoop install https://example.org/app.json, scoop install https://github.com/ScoopInstaller/Main/raw/refs/heads/master/bucket/7zip.json

[Gitoffthelawn]

IOW, is there typically a need for the user to invoke one of those 4 commands?

No. After buckets added, Scoop will update them via git when appropriate.

When you mention "After buckets added", do you mean when Scoop automatically adds the main bucket when Scoop itself is installed?

[HUMORCE]

Any bucket that is git repo with remote url set up. you can check added buckets via scoop bucket list.

[Gitoffthelawn]

Any bucket that is git repo with remote url set up. you can check added buckets via scoop bucket list.

Got it. Thank you.

And because Scoop automatically converts the default main bucket (generated during Scoop installation) to a git repo when git is installed, the user never even needs to run the command scoop bucket add main. That's a nice touch.