-
Notifications
You must be signed in to change notification settings - Fork 7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat(helm): --no-update
flag for helm repo add
#1142
Conversation
I'm sometimes opposed to making one subcommand do two different things. But... in this case, I think there is a strong reason why we should do this. The huge advantage to this is that when the Web UI comes out in a few months, installation instructions will be able to say:
That will be a lot simpler for our users than telling them to add a repo if it doesn't already exist, update it if it does exist, and then install the package. |
So I understand the existing behavior before throwing out an opinion... The current If that's correct, then I'd argue that
In which case it's probably more idiomatic to add resiliency to
As far as how this ties into the UI, I would suggest that UX/UI design effort includes a prominent, but distinct visual section whose purpose is to communicate the following:
As I think about the UI's role in communicating real-world usage of charts that they discover through search/browse, I think there are three distinct categories of info that we want to communicate:
The reason I think we should aim for distinguishing between those three categories is because for a large percentage of our users, they will already fulfill the prereq's, and they will not want to customize out of the box; and so by focusing only on "how to install assuming you know what you're doing" we're able to present the simplest, cleanest visual footprint. We don't want to be suggesting to anyone that helm is super complicated, hopefully we can provide the complicated info in a clear way, but in a way that connotes that it's only as complicated as you want it, in most cases it's a cinch! |
So the tl;dr of @jackfrancis 's post is: we should do the behavior @nicr9 describes as the default. |
Okay, I'll switch things around so that it'll update by default unless you specify I can see why someone would like to have either behaviour at their disposal. |
Update: I've not had a lot of time at work to devote to this lately. I've modified the default behaviour but haven't had a chance to update the tests and now I'm out of office on holidays. I'll finish this as soon as I'm back. |
Just ping me here or in the Kubernetes Helm channel when you're ready, and I'll review it then. |
Hey @technosophos, Sorry it took so long to get back around to this! Since a lot has happened in Just as a reminder: this PR changes the default behaviour for |
--update
flag for helm repo add
--no-update
flag for helm repo add
Code looks great. I'm going to make sure nobody objects to adding this in the Beta phase, then I will merge. |
When using
--update
new repositories will be added in the same way aswithout
--update
but if the repository has already been registeredthen the endpoint will be updated.
Closes #1141
This change is