Kartik Luke Singh

@yeramianm If they enable the collaborator code it applies to everyone including installed apps.

@0xfluke They could always decide if they wanted a collaborator code but you’re saying now they can decide if they want installed apps to require one?

Shopify now requires a collaborator code to request access to a store even if they have your app installed?

@_swanand How do you like the new compiled version. I feel like I really enjoyed it in the SCP shorts format.

Ditch that crappy business book you just bought and buy some nice fiction. If you don't have any in your list, I will recommend "There is no antimemetics division" as your initiation.

@yongfook We ended up connecting it directly to the error logging and NewRelic and just have it run periodically and raise a PR.

If you haven't already, tail your server logs and filter by requests that take longer than 500ms. Copy paste it all into Claude and say "pls fix". Instantly faster app.

Step 1 in moving off of @heroku which we've been using for more than a decade. End of an era. I wish more PaaS tools copied some of Herokus UX because it's honestly pretty good.

@MichaelMartocci There's an SDK to remotely control Chrome that can connect and take the screenshots itself.

I constantly push screenshots back into Claude Code to give it context on how the interface is rendering etc Am I an idiot or that the only way?

@marckohlbrugge The tests after in general are not very good because AI can end up trying to fit the test to the code and get a pretty useless result. Telling it to write the failing test and then only make the changes required to pass it works well.

Is anyone doing true TDD with AI for Ruby on Rails projects? Do you see better results versus writing tests after?

opinion: this is a detrimental change for merchants. part of what makes building storefront apps easier than before is metafields. it has its issues (like inconsistent caching), but it also allowed us to scale up to serve merchants at insane scale. and merchants + their dev teams love it too. so much so that they explicitly look for apps that expose - and heavily rely on - metafields. this change effectively forces us to build our own infrastructure, stumble and learn how to do it right along the way. which is not bad, in isolation, but considering that merchants need apps to be as performant as the work they're putting into their storefront, this means merchants will struggle with building the right experience for their customers. infra cost to serve this - as a business - is probably something else to think about. i'm going to assume it's not significant for now. some will see this as a "skill issue". well maybe so, but i'd much rather use our time to expand what our products do. that's not to say i don't get why they did it. it seems like the decision wasn't easy. the way it's being done though doesn't feel right - more so for merchants (the actual end customer), than anyone else.

@FORSBERGtwo @kirplatonov Given I build an app that heavily uses metafields to provide functionality. This doesn't entirely make sense because either you: - Split it up but you're still loading the same amount of data - Load it from your own server through API and that will be even slower

@kirplatonov Ouch, guess it’s going to be more complicated to split it all out into many metafields instead. Hopefully they add a mutation which splits it out on write automatically, and aggregates it in on read, to keep the nice DX.

That's a painful change by Shopify out of the blue. I'm using JSON metafields since 2021 to store app settings and data for fast access by theme extension. And it's all going to break as not fitting new limit...

@stuartchaney That's interesting. Are you logging into the sessions in agent-browser manually or did you find a workaround like running it non-embedded?

automatically have agent-browser create a gif of the end-to-end test when creating a PR see all of the feature working in 10 seconds vs digging in to manually test 🤌 *saw this here a couple of weeks ago - can't remember whose gem this is or would attribute

@Shpigford Interested in your process here. I've been considering building a OpenClaw based executive assistant for my wife and I based on a shared note system. I tried Obsidian but it just mostly feels like a wrapper on having MD files with git.a

my wife and i have "docs" (text, images, PDFs, audio) spread across Notes, Evernote and Notion going back some 15+ years. really tempted to build something self-hosted that doesn't become bloated (evernote + notion) but still has powerful index/search + AI.

Shoutout to the STOQ team and @heysandy801 Very few apps with the level of support they have. Founder cares a ton and his team is +++responsive Great for restocks & back in stock alerts. We use it for an 8-fig client. He didn't pay me to say this - his team just helped me solve a fire for our client very fast though.

Using the @newrelic MCP is such a huge unlock for debugging performance issues.

@stevekaliski We've been using @DevinAI for pretty much this exact thing. We have Slack messages, support threads, etc. trigger agents that run in their own environments and return PRs. Really interested in Toolshed. Getting context is key and its the main thing our developers are needed for.

At Stripe we have a tool called "minions" -- it lets us kick off async agents built right in our dev environment to one-shot bugs, features, and more e2e. I have team, project, and personal channels dedicated just to working with minions. I like to think of it as a new type of pair programming -- "pair prompting." Read more --> stripe.dev/blog/minions-s…

@nateberkopec @yongfook @jessethanley Where are you migrating to? I'm in the process of migrating our Heroku databases to Planetscale using Bucardo (because Heroku doesn't allow replication). Leaves us fairly open to trying alternatives for our Rails app.

@yongfook If your spend is over 1k a month these PaaS alternatives are not really aimed at you. I’m migrating @jessethanley next month, let’s go Jon

I’ve got over a billion rows and more data than would fit on my laptop. I can’t just “download the database”. I looked into Heroku alternatives a year ago and it felt like they weren’t ready. If you’re starting out maybe, if you’re porting a production setup, not so much.

@yongfook @excid3 Had the same experience. Called me a ton of times to save a very small amount. Luckily our contract expires in a few months and we are already moving our databases to Planetscale after which shifting shouldn't be too hard.

@excid3 I literally just signed an Enterprise contract like 7 days ago... no wonder they were quite uh, insistent I sign. It's mostly just a different billing agreement though.

It's crazy that we live in a time where AI can help build any app you want... ... but to deploy it live there's still only like 3 choices Heroku, DO, or roll your own infra.

@_m27e Though Herokus internal recommendation was some sort of read from the S3 backups of your DB kind of process that Planetscale told me was much worse than Bucardo. Currently I think it's either Bucardo or pg_dump / pg_restore. Neither are zero downtime.

@_m27e Yeah it's a bit scary but possible. Was able to switch over one of our smaller apps with about 30 minutes of maintenance mode. But there's a bunch of pitfalls that delayed the migration like it doesn't automatically pick up schema changes.

Everybody wants Heroku customers to migrate to them, but nobody wants to put any work into a production-grade migration plan. Everybody's migration docs are a joke for anything that isn't a toy app.

@ShopifySupport I will definitely do it next time. We kept unsetting and setting the metafield and it eventually started working. I'll add this to the community post when I get more examples.

Thank you for your response! To help us investigate further, we’ll need a specific identifier called the X-Request-ID from your browser when you encounter the issue. Here’s how you can find it: For Google Chrome: 1. Right-click the page and select 'Inspect' (or press `Cmd+Alt+I`). 2. Go to the 'Network' tab. 3. Reload the page or trigger the issue with this tab open. 4. Look for a request in red (status code 400 or higher). 5. Click that request, select the 'Headers' tab, and find the 'X-Request-ID' in the response headers. 6. Copy the X-Request-ID. For Firefox: 1. Open the menu and select 'More Options > Web Developer Tools'. 2. Go to the 'Network' tab. 3. Reload the page or reproduce the issue. 4. Find a request with an error status (like 404). 5. Click it and scroll to locate the 'X-Request-ID' in the headers. 6. Copy the X-Request-ID. Please copy/paste the X-Request-ID directly in your reply (rather than sending a screenshot). Once you have the X-Request-ID, please DM it to us along with the best email to reach you so we can look into this further. If you run into any issues collecting this information or have questions, feel free to DM us as well and we’ll help you out.

@ShopifyDevs @ShopifyEng Seeing concerning unreliability with the Shopify metafield system. We're seeing cached metafields showing up on locale domains very frequently. Any other Shopify developers seeing this as well. community.shopify.dev/t/metafields-o…