-
Notifications
You must be signed in to change notification settings - Fork 6
Basic Concepts
Blitline takes an image, performs processing on it, and then outputs it somewhere (either to a temporary S3 location where you can go it, or to your own S3 bucket if you give Blitline permission to write to it)
A job represents a single source image that can have one or more operations performed on it.
A job can have a postback_url which will be called (by Blitline) when the job has completed processing.
You can submit multiple jobs to Blitline at once
A function represents an operation performed on an image such as blur, or crop, or watermark.
A function will always result in an output image, but you can decide whether or not you want that image persisted somewhere.
A function can have sub-functions, which use the parent function output as their input image.
You can continue to nest sub-functions within sub-functions.
A save instance represents the fact that we want to output the results of a function. A save instance MUST have a name of some sort. We call this an image_identifier. This image_identifier is what we will use to identify a specific image that we have persisted somewhere.
A save instance can also have a target location which would represent a custom S3 destination or your choosing (assuming you have given Blitline permission to write there).
If a save instance does NOT have a target location, a default Blitline location will be chosen. This location will store the image for 24 hours. This allows your application time to go get the image and move it somewhere.
When you submit a job to Blitline, you will get back some JSON that represents the images that you wanted saved.
These images will have a url defined for them. This output does not mean the images have completed processing, it only means that the location for where they will be stored has been defined.
You must either poll(bad) or define a postback_url(good) to identify when the job has completed processing.