For folks using feature flags, how do you deal with updates of significant dependencies?

Anyone else started receiving user emails in

event.requestContext.authorizer.jwt.claims.username

in APIGWv2 with Cognito Auth? This has never worked for me in the past, kept on returning UserId (sub) but now, out of the blue it switched to email...

I’m after some advice from you excellent lot. We have lots of different business customers who have bought our products and sales are growing. Our main/flagship product has lots of different features and many many options for configuration of those features based on business requirements. Additionally, some of those features are part of the core product and some need a separate license. As we’re primarily using environment variables for configuration, what we’re starting to struggle with as that, when we introduce a new feature, or an additional configuration option for an existing feature, we need to update the environment variables for each individual deployment in Seed and redeploy. Some of our configuration is not env var-based and is done via JSON stored in parameter store and I’m starting to think that this is a better approach as it’s easier to write tools to auto update these values and they’d be picked up the next time a lambda cold-started (or we implement a TTL for values). This is all backend related but we have a similar challenge regarding our Vue frontends, the config for which is currently stored in individual config git repos and applied at build time to the core product code. I’m therefore curious as to what approaches others have taken to managing these kinds of changes.

Hi all, I have just built a free public API

"depguru"

(yes, it is serverless) to get all dependencies as tree (including transitive deps) of any NPM package without install:

<https://api.depguru.com/v1/npm/deep-deps/{packageName}/{version}>

For ex: https://api.depguru.com/v1/npm/deep-deps/express/latest

so…. Aurora Serverless v2 is kind of a disappointment, isn’t it? it’s expensive.. to be exact, our DB bill would be doubling if we switched. 🤔

https://twitter.com/cloudflare/status/1524374014839377920?s=21&t=Wl_3WjAR2HMscG3cWHhuYw thoughts?

Who knew so many of you found @thdxr beautiful?? https://twitter.com/thdxr/status/1524427171615363076

This is a little off the beaten track so this is an appropriate channel. I thought I’d share my recent pet (and first) Arduino project. I had a requirement to detect when my electric roller garage door had been left open as sometimes the button gets pressed accidentally and it opens without me knowing. Using a magnetic switch, an MKR1010 WiFi and AWS I’ve put together a first project that will send an SMS and email to defined numbers and email addresses when the door is left open for a defined period of time. It’ll then notify again when the door is closed, but only if it’s previously been left open. For anyone interested, the code for the Arduino and AWS resources is here. Any feedback, improvements, criticisms or even PRs to improve are welcomed There’s a high level architecture diagram in the readme but basically it uses an HTTPs request from the Arduino using the WifiSSLClient library and WiFiNina to an API Gateway endpoint in AWS which is secured using HTTP Basic Authentication. This invokes a lambda function which sends an email via Simple Email Service (SES) and posts a notification to a Simple Notification Service topic. The AWS SMS service is subscribed to that topic and so sends an SMS to each mobile phone subscriber when the topic is published to. It was a bit of a voyage of discovery, it’s a long time since I wrote any C, and never in any embedded project, but got something working in the real world with hardware through to cloud which I’m pretty chuffed about. Tomorrow out comes the soldering iron where I wire it up permanently and install in the garage. I posted similar in the Arduino forum too ( and gave a big old hat tip to SST )

are there any SST people in the nyc area?

Who are you supporting?

🔥 for Miami

🍀 for Celtics

Any Gitlab pipelines users here? Guess this question applies to GitHub jobs as well. With the support for docker containers in Lambda I got to wonder if I could drop my previous (unfulfilled) plan of ec2 on-demand runner fleet in favor of spawning Lambdas that could be assigned my team's CI/CD jobs. Anyone here tried going similar direction? Seems like all the building blocks are now available and this approach would allow for an instant runner spawning whenever one was needed without the overhead of managing ec2s.

man, I’m ready to rage-quit VS Code. anyone else have this issue where you save a simple file and it shows this for like a whole minute?oddly enough, it’s only in my SST project 🤔

saw this truck today. did a double take on their logo

How are people doing auth for their applications (minus cognitio)

@thdxr @Frank - there are some AWS limits on max lambda path length of 128 characters and 64 characters limit on lambda name. Would it be possible that you add some checks for this limits exceeded in sst before deployment? Some time ago we run into some problems with too long paths to the lambda handler and it fail during deployment but without any decent error from aws

If I have a Dynamodb Table declared in StackA, but I now want to refactor my code such that the same table lives no longer in StackA but StackB. Is there an ideal way to do that?

I'm struggling to remember the name of a JS framework for lambda. I think it implements the hexagonal pattern, I know some people in here use it

Will Mavericks get swept?

Looking at the Migration guide for v1, I see that many

arrays

has been converted into

Record<string, Object>

. Why is done this way? what’s the advantage of not having arrays? it’s easier to access an item this way?

Also, as part of the migration… why are you getting rid of enums, in favor of strings?

fields: type TableFieldType ⇒ string

Hey guys, Is there any recommendation on how to organize S3 buckets? for example, multiple buckets, one for each feature, or subfolders by service, .. etc? Following the micro-services model, we create a bucket for each micro-service (in our case, we have services running on k8s). But also, I wonder if CDK can define properties to “backup” those buckets?… Want to hear ideas from you guys 🙂

It looks like I’m going to need to build an API that allows a user to perform multi-field filtering and sorting against a MySQL (PlanetScale) table. Rather than cook up those queries myself I’m considering using an ORM or equivalent tool. I’m not familiar with what’s good in that space but I’ve read in here about people using Kysely, TypeORM, Prisma and probably others. What do you guys recommend I take a look at?

Hey guys. Dummy question here… I’m using a mono-repo, where I have a

data

folder, which is shared with all my lambdas. This folder contains all the Entities to connect to DB, etc. The thing is… every time I modify an entity there, all my lambdas gets updated (due most of them use almost 1 entity from this package). Is there a way to do tree-shaking, or whatever magic, to left my lambdas with only the classes that are imported and not the entire dependency package? If we have static references to each file/dependency, why this can not be done?

We’ve built a feature that provides an import from a web app of an calendar file (.ics) into a user’s calendar. It works fine on Android, Windows and Mac OS. However, iOS has restrictions that prevent it working. I’ve found an example from a local council that use a download with the unofficial Apple protocol webcal:// which I’d like to try out. I’m thinking I need to generate the file, write it to S3 and provide a link to download/subscribe to it to the user, however I can’t figure out how I’d do this with AWS - how do I serve a file using a different protocol?

What does the Cron stack use? Is it a lambda function?

Hey guys, not a complain but wouldn’t the quick response in slack messages negatively impact your productivity? I am just very impressed..

Probably last game for Dallas this season lol