Does anyone have experience to relay with Amazonka?
I am looking to do some S3 related work.
5 Likes
I use it for S3, SES and SNS. No complaints, does what it says.
This 7 post series I found to be what I needed to start: https://blog.rcook.org/blog/2017/aws-via-haskell/
3 Likes
I think it is excellent. The only thing to be aware of is that it is somewhat low-level, and you will need to write your own utility functions. However, this is intentional, as it tries to follow the AWS API closely rather than providing a Haskell layer on top.
3 Likes
We’re also using amazonka family of packages for S3, SQS, SES and SecretsManager. So far everything is great 
5 Likes
Can you state a bit more about your requirements? It depends on whether you need to access all the features of the protocol offered by AWS S3 (and also whether you need to target the original AWS S3 implementation or “compatible” S3 clones), whether you intend to use other services from the AWS family besides S3, and also whether the dependencies of Amazonka are acceptable by your project’s security auditing/policy requirements, etc… whether Amazonka is a good fit or maybe a different S3 client library might be more appropriate for your use-case.
2 Likes
Thanks everybody.
Not 100% clear right now. I plan a file server with an Rest API in front of it, as always the requirements will probably grow from there : )
I use it for work and love it. In addition to being very extensive, it offers some powerful abstractions (such as lenses for both values and exceptions), and uses established libraries (exceptions, unliftio, unordered-containers, etc.). It offers typeclasses and a transformer type to augment your effect type with the ability to interact with AWS, as well as a “Has” typeclass for configuring the environment in a flexible way. I also found it a great case study for some “intermediate Haskell” concepts.
2 Likes