grima - whispering into alma's ear with APIs
This project is maintained by zemkat
This document leads you to the right documentation to get started.
Great, go to the Grima sandbox hosted by Kathryn Lybarger and using (read-only) fake data from Ex Libris.
You’ll need to get an Alma API Key first.
Then go to the public Grima and follow instructions. This one is hosted by Kathryn Lybarger, but it will have access to your institution’s data by using your Alma API key.
You’ll be sharing your API key with Kathryn, which is probably fine for testing, but eventually you’ll want to stop doing that. We don’t store the key, so once you stop using the public grima, you’ll also stop sharing that key.
Institutional concerns and policies will probably mean you need to install grima at your institution. (Handing out API keys willy nilly to employees who then hand them to Kathryn is unlikely to be a sustainable model.)
If each person who wants to use grima can install software on their own desktop computer, see Grima on the Desktop. This uses the nice fact that computers are ridiculously powerful, and that your desktop computer can pretend to have another computer and another internet inside it running your own website! Possibly your IT people will think that is too weird and want to be more involved.
If you (or your IT people) can install web-apps onto a web-server, see Grima on the Server. This includes both 20th century servers and cloud native deployments.
Grimas help simplify your workflow, but to do so they need to match your workflow. Write the perfect grima yourself, by following these instructions for New Authors.
Starting with 2019-12 release, we will provide stable images at
docker.io/zemkat/grima-cloud:2019-12 etc. You may want to use
the instructions for deploying Grima on the Cloud.
In order to build your own images, alter the recipes in the Makefile (on GitHub) to point to your own image name. You’ll also need to update any of your cloud deployment configuations to use your images as well.
The first build of cloud images requires around 15 minutes per architecture, but using the build cache further builds (altering only the PHP) should only take a few seconds.