Hello folks,

In my app, the client can upload some amount of files from his machine (let's say, with a 100 MB upper bound). I have a doubt on how to architecture this. In the client, I represent files with the following data structure:

{file: ArrayBuffer, type: string}

I can think of two options:

1) In the client, I convert the files to a base 64 string representation and I call a server action to transfer those strings to the server side in multiple steps. In the server action, I inject those files in an S3 bucket and later, in my payment webhook, I inject the S3 URLs for the files in my database.

2) In the client, I inject those files in an S3 bucket and later, in my payment webhook, I inject the S3 URLs for the files in my database.

I know (2) is much more efficient in terms of bandwidth but (1) is very safe (no risk to expose my S3 bucket information to the client).

Two questions:

- Is (1) really bad ?

- Is there some way to implement (2) safely and preferably free ?

Thanks