More descriptive inputs and images key names

[book000]

From the user-side, we feel that the current names and behaviour of inputs and images are confusing.

Regarding inputs and images, the users of this action think that "passing multiple images to docker-manifest-action as input" will result in "one combined manifest image output".
Therefore, each with-keys was considered as follows:

• inputs = "images to be merged as input"
• images = "merged manifest image(s) to be created as output"

This idea seems to differ from current implementations. It is plausible to judge by the documentation, but we believe that it is better to be able to uniquely judge by the key name if possible.

What is your opinion?

Discussion Comments before moving on (issue#134)

Hello, thank you for developing this great action.

v0.3.0 has been released, but applying this release without updating with will result in an error.
The job log where this is occurring is following: https://github.com/tomacheese/my-pixiv/actions/runs/4021878692/jobs/6911501009

Deprecations rather than Removed, so I assumed these would work with the update applied without any changes.
Is this behavior expected?

P.S. I have read the action.yml and I don't know why this error occurs even though required: false is specified.
Well, I am concerned that required: true is specified in src/input.ts.

---

Also, if possible, when configured deprecated inputs, please mention this in the deprecationMessage, so that appropriate warnings are displayed and easily noticed.

https://docs.github.com/en/actions/creating-actions/metadata-syntax-for-github-actions#inputsinput_iddeprecationmessage

I look forward to your reply.

[book000]

Also, it appears that images and output / outputs are mixed in the input key to which extra-images is changed.

These can be found at:

base-image and extra-images are deprecated in favour of inputs and outputs.

Release v0.3.0

docker-manifest-action/README.md

Line 25 in acf2524

> **Warning** -- As of v0.3, `base-image` has been renamed to `inputs` and `extra-images` has been renamed to `output`. The former inputs will work but will be deprecated into v0.4!

docker-manifest-action/README.md

Line 52 in acf2524

output: namespace/image:latest

---

I'm not sure, I haven't read the code, just the documentation, but aren't inputs and output (images?) reversed?
If I just recognise the words, I think inputs is the list of images contained in the image to be created (same as previous extra-images), and outputs is the name of the image to create (same as previous base-image), but I'm not sure after reading the warning in README.md#L25 and Release page.

Sorry if some parts are difficult to understand due to the use of translation tools.

[auguwu]

Deprecations rather than Removed, so I assumed these would work with the update applied without any changes. Is this behavior expected?

The behaviour is not expected. I'll create a new release making it not required.

Also, it appears that images and output / outputs are mixed in the input key to which extra-images is changed.

This is also a mistake, I'll be fixing that also. :3

I think inputs is the list of images contained in the image to be created (same as previous extra-images), and outputs is the name of the image to create (same as previous base-image), but I'm not sure after reading the warning in README.md#L25 and Release page.

inputs is the same premise as base-image, where it is the inputs for the merged manifests to be created and images (output/outputs) are the outputs of the merged manifest, i.e:

inputs: namespace/image:latest
images: namespace/image:latest-amd64,namespace/image:latest-arm64

namespace/image:latest-amd64 and namespace/image:latest-arm64 will be merged into namespace/image:latest.

Sorry for any misunderstanding with the documentation on translation tools as they are not accurately the best and I am also not good at speaking English even though it is my native language. I'll be updating the documentation to make it more understandable and have it peer-edited with the rest of the Noelware team.

Thanks again for using this action~! 🐻‍❄️

[book000]

Thank you for your reply.

Regarding inputs and images, the user-side of this action think that "passing multiple images to docker-manifest-action as input" will result in "one combined manifest image output".
Therefore, each with-keys was considered as follows:

• inputs = "images to be merged as input"
• images = "merged manifest image(s) to be created as output"

It is plausible that the decision should be based on the documentation, but we feel that it would be better if it could be uniquely determined by the key name if possible.
However, this is my personal opinion and I am not sure what others think, so I would like to hear from other users.

[auguwu]

However, this is my personal opinion and I am not sure what others think, so I would like to hear from other users.

I would like that also, so I'll make this as a discussion rather than a issue (so that other people viewing/using this action can comment on it also!)

[book000]

Thank you for moving this to discussion. We have rewritten the title and description for people viewing this discussion.
I would appreciate it if you could edit them if they are inappropriate.