Michael Dowling

We now have a formal proposal to add let expression to JMESPath. Feedback welcome! github.com/jmespath/jmesp…

@jsaryer I started using Helix and really took to it. I’m super jealous of Lua scripting in neovim though. I hope Helix uses Lua for their plugin model.

Took the evening to migrate my ancient vim setup over to neovim, and so far I'm pretty impressed. The fact that I can now use lua and never have to deal with vimscript is already worth the switch.

The number one missing feature in JMESPath, and usually the answer to "what's the JMESPath expression to select ...", is scoped variables. Here's a refreshed proposal, let me know what you all think: #issuecomment-1464113092" target="_blank" rel="nofollow noopener">github.com/jmespath/jmesp…

@jeremeamia @seaofstarsgame While I’m waiting for this one to come out, I’ve been playing Chained Echoes. It’s pretty good, retro JRPG fun

I played through the @seaofstarsgame demo yesterday morning. Great gameplay, artwork, and battle mechanics. Captures best parts of the great, turn-based RPGs of the past, but with modern flair. Snappy menus. Excited for the release later this year and to experience the story.

@samgoodwin89 Yeah good question. Sometimes if clients know a more direct way to access a resource, changing the endpoint to be resource specific can improve performance. It shouldn’t come at the expense of client performance though. We’re looking into it.

@mtdowling Why can’t the server perform that work instead of pushing it on to the clients?

The ruleset.js file in the @aws-sdk/client-s3 is a 1.4MB pure JSON object that is almost entirely whitespace. 😲

@samgoodwin89 It’s a set of rules used at runtime to change endpoints for requests. S3 has the largest set of rules because there’s a lot of features that affect the endpoint of a request in various ways. We’ll work on reducing the size of it.

What is this leaky thing??? This should not be in a SDK I don't think.

@joshbarrnz Service teams that have this kind of docs split don't write docs in their models (placeholder or otherwise). The docs team runs a simple script that creates a placeholder for missing docs for every shape that they fill in.

Hey @mtdowling, the Smithy docs mention a use-case where an AWS docs team extends a service definition. Do they apply custom traits? Using the default smithy documentation trait fails with a conflict if a service team has provided a placeholder value. Love to know how it’s done!

Do browsers typically parse then evaluate CSS selectors with a kind of tree-based interpreter, or do any of them do more advanced things like emitting and interpreting bytecode?

@elfidomx These docs cover creating traits, though they don’t cover how it works in Java though: #trait-trait" target="_blank" rel="nofollow noopener">awslabs.github.io/smithy/2.0/spe…. This page lists implementations which shows how others have built generators: awslabs.github.io/smithy/2.0/imp…

@mtdowling is there any documentation about creating custom traits and/or generators for smithy? Any hint will be appreciated

@cameronc I won’t be there this year. I don’t think there’s a talk solely about smithy. I’d guess talks about AWS SDKs would touch on it though.

@mtdowling Curious if you or anyone else working heavily on Smithy will be at re:Invent? Any sessions planned?

@kubukoz The yolo of Smithy’s type system

Nulls are a thing in smithy because of the Document type (which is basically "arbitrary JSON"). I'm making a query language for Smithy, so null documents need to be handled (and writeable by users) - hence, the null literal. Only typechecks as a document.

that sad moment when you have to add nulls to your language...

@eigenseries Yes. They haven’t updated to use this new version yet.

@mtdowling Is this the same Smithy that the Rust AWS is generated from?

The Smithy team released Smithy IDL 2.0 (a tool for generating SDKs and services from models). Code generators can now generate fewer nullable types thanks to the addition of default values. Also has DX improvements like mixins to reduce copy/paste: aws.amazon.com/blogs/develope…

@lsegal 😭

@mtdowling HBO (Warner) has been canceling and removing content left and right to make room for $$$ in the Discovery merger, with the goal of focusing more on unscripted content. It seems like the precedent is clear that the HBO brand is secondary.

The idea that Discovery thinks HBO Max is the thing that needs a rebrand and not the other way around tells you everything you need to know about how out of touch those execs are.

@velvetbaldmime Great series of posts!

Building a fullstack Scala application with Smithy4s, Weaver, Skunk, Scala.js and Laminar, and deploying the whole thing to Heroku and Platform.sh! Jobby is here! Part 1: blog.indoorvivants.com/2022-06-10-smi…

@GaneshNJ Well deserved!

Personal update: officially a Senior Software Engineer.

@russellrcohen This makes me glad that UUID isn’t a built-in data type in Smithy

"New UUID Formats" is not a phrase I ever expected to hear, but it definitely makes sense: ietf.org/id/draft-peabo… TLDR: Everyone uses UUID for database primary keys but UUIDs are bad primary keys, actually.

"Delete and don't notify" in Apple Calendar sure did just notify a big ol' list of people that I'm not attending a meeting... I guess I've been aggressively declining optional meetings on accident for a while now 😬 Back to Outlook.