Discussion on changes to RFC

[Patrick Magee]

Hello Everyone,

I have submitted this issue: https://github.com/openwdl/wdl/issues/338

In hopes to spark a discussion around the RFC process, specifically encouraging engine implementation, and how an engine actually knows whether or not they have implemented a new feature request.

It has become increasingly apparent that if we want engines to implement features in a timely manner, we need to provide guidance for how those features are supposed to work and when to consider a feature actually "implemented".

 I would appreciate your comments and suggestions!

Cheers, 

--

Patrick Magee Software Developer at DNAstack T: +1-613-545-5315 dnastack.com @DNAstack

CONFIDENTIALITY NOTICE: This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message. 

[Mike Lin]

Pat, thanks for this; I'll add some specific comments about test cases on the issues.

I think miniwdl has just, in the past month or two, gotten to the stage of feature-completeness where it can provide a sandbox to prototype PRs through the process. We're now on a documentation sprint with a new Getting Started tutorial for users and also planning to add a 'worked example' of forking the code to prototype a new WDL language feature (nominations welcome for the tutorial subject). It's been a slog to this point so we're really excited to start inviting more community contributions to the project!

Mike

--
To unsubscribe from this group and stop receiving emails from it, send an email to community+...@openwdl.org.

[Christopher Llanwarne]

Mike - that's awesome news! Congratulations!

Pat - I'm also definitely in favor of this. It gives engines something to develop against, and makes sure the intent of the original spec change author is what actually appears in the engines. I don't know whether *every* small PR needs test changes, but for new features we already ask for examples as well as text - writing them up as test cases too seems like not too big of an ask.

Chris