We still depend on it as long as we set it as image processor but now we
can switch to another image processor without changing the code around
error handling.
We now rescue from unknown errors during image processing which should
make the app more robust.
We only reference MiniMagick when rescuing errors but when it's not
loaded, that code fails to find the error class itself to apply the
rescue block.
The rescue block is covered by a spec but the code passes there as
MiniMagick is loaded.
We can see this error only in development, staging and production.
I could have split this into several commits:
* DRY direct linking to images.
* Check S3 config before direct linking.
* Just check if service is public instead of relying on name.
Developers may copy a staging or production database or experiment with
S3 storage. But when the S3 config is missing then calling `service`
raises an ArgumentError due to a missing name.
Now we only try to call `service` if the S3 config is present.
Defining an alternate s3 configuration set to `public: true` means we can use it selectively. It sets the objects to `acl: "public-read"` by default (read-only) and means any image tags for those assets can use direct public links in the src attribute (without hitting the ActiveStorage::Representation endpoint). The default non-public service will still be used by default for any other files on instances using s3.
We introduced a list of formats we support and forgot to add webp. Now I
added that as allowed format again and modified the error message.
I removed the first sentence from the error message because it's very
similar to the default error which is shown as well.
I chose `has_one_migrating` as method name for two reasons:
1. It reflects Active Storage's method `has_one_attached`.
2. And it has the same length as Paperclip's `has_attached_file`.
Therefore the commits don't need any whitespace changes.
When we change it to `has_one_attached`, we will also remove the
Paperclip options which then don't need whitespace changes either.
While we migrate from Paperclip to Active Storage, we need to use both
at the same time to avoid any downtime or lost images.
Once the migration is complete, we want to use the same name for
attachment as before. Using Paperclip and Active Storage at the same
time creates a name conflict on a couple of methods.
I'm using alias_method as a temporary solution to access Active Storage
methods. We will remove that after the migration. I declare Paperclip
afterwards so that we have those methods declarations for backwards
compatibility now.
The changed code obtains the image dimensions when you first upload an
image. Unfortunately it's also triggered when thumbnails are refreshed.
That doesn't change the size of the original image though.
