cal morris

@aandreug @framer @Strava Only 1.4km per step

@CalMorris @framer @Strava Lol that’s fast

standard level of activity for our team right now @framer, we ship🛳️ gotta keep the @Strava streak alive 🤣

This week I am OOO but that doesn’t mean the shipping stops 😅 Look out for what we’re shipping soon @framer 👀

@superset_sh @pascalordanoo ngl the first time i saw this I genuinely thought i was going insane and was like “that’s enough for tonight” 😭 I’ve found it most commonly occurs on Claude, but might be irrelevant 👀

W̸̢̛͉̟̮̱̃̈́̾̈́͠e̷͎̳͔̩͊̈́̾͝ ą̶̜̯̈́̃͝p̷̛͙̪̾̈́͝o̷̢̜̯̾̈́͠l̸̛͉̟̾̈́͝ơ̶̢̜̾̈́͠g̷̛͉̾̈́͝i̸̢̛̾̈́͝z̶̛͉̾̈́͝e̷̛̾̈́͝ f̸̛̾̈́͝ơ̶̾̈́͝r̷̛̾̈́͝ t̸̛̾̈́͝h̶̛̾̈́͝e̷̛̾̈́͝ ḭ̷̛̾̈́͝s̶̛̾̈́͝s̷̛̾̈́͝ư̸̾̈́͝e̶̛̾̈́͝, w̷̛̾̈́͝e̸̛̾̈́͝'l̶̛̾̈́͝l̷̛̾̈́͝ l̸̛̾̈́͝ơ̶̾̈́͝ơ̷̾̈́͝k̸̛̾̈́͝ i̶̛̾̈́͝n̷̛̾̈́͝t̸̛̾̈́͝ơ̶̾̈́͝ i̷̛̾̈́͝t̸̛̾̈́͝.

been learning @superset_sh 's language lately

happy customer after migrating to Forward and switching to TypeScript SDK!

Domain registrars may be one of the most overlooked single points of failure in our stack. We spend serious time and resources securing our cloud infrastructure and workloads, but our domain registrars rarely get the same level of scrutiny. Even though our domains underpin everything we operate. 🧵

To all @framer creators, I wanted to share a quick update on the current state of template reviews on the Framer Marketplace. As you may have noticed, we’re experiencing a higher volume of submissions, which is impacting how fast we get back to you. The Marketplace’s natural growth, combined with the second season of the Framer Challenge, has pushed submission numbers to levels we’ve never seen before. In April alone, over 1,200 templates were submitted. That’s a record for us. The Challenge in particular has been incredible to watch, with so many of you stepping up and building at a pace we genuinely didn’t anticipate. We’re on it, and we want to be transparent about it. Here’s what we’ve done to address it: → We’ve pulled in reviewers from another team to focus exclusively on templates for the next few weeks, until the queue is back to normal. → We’ve implemented internal AI workflows to help us catch details faster. → We’ve improved our internal triage process to prioritize submissions that have been waiting the longest. → We’re actively monitoring the queue daily to make sure no submission slips through the cracks. These changes are already in motion, and our goal is to recover our usual review timelines within the next two weeks. We’re not where we want to be yet, but we’re moving in the right direction. If your template hasn’t been reviewed yet: please bear with us. You haven’t been forgotten. Every single submission will get a response. We know how much work goes into building and submitting a template, and we don’t take that lightly. You will hear from us soon. Thanks for your patience, and thanks for continuing to build 🙏

We're looking for a new batch of beta testers for Framer with Claude / Codex. If we missed you the first time, feel free to enter again. x.com/koenbok/status…

@TomasJansson @mdo @vercel Would he interested to know how you’ve got it setup, as Vercel support confirmed that there is no isolation to us recently. We thought moving to a custom environment (my assumption) would solve this, but have seen the issue since.

@CalMorris @mdo @vercel What? On Vercel I’m quite sure that is not the case. We have custom environments and have been using it for two years or so and have not seen any cache bleed. But it depends which cache, with data cache I think it is kept between deploys, which make sense for a data cache.

Is there anyway to get additional environments on @vercel without making the jump from Pro to Enterprise? We have a single app we split publish to multiple sites and that's our key to adding another site (it's a good one, I promise!). Wondering if I'm missing something.

after upgrading either @superset_sh (Version 1.8.4), i cannot seem to paste screenshots that are on my clipboard into codex or claude anymore but it does work if open codex via Ghostty, anyone else ran into this?

@saddle_paddle @framer let’s go! can’t wait to test it out 👀

@CalMorris @framer Ah yes that makes sense, more of a reason for us to focus on multi-repo workspaces soon!

To go a little bit deeper into how we've managed to ship so many changes, I thought I'd share a bit about my workflow at @framer since joining, which I’ve been tweaking! 1. Before I joined, I made a promise to myself that my ~/.dotfiles would be clean, tidy, and repeatable across systems! The heart of this is Chezmoi, and it's been a joy to work with. 2. Making my workflow in an "AI-native" fashion, meaning that I can easily hand off a task to an AI agent (agnostic of the underlying model *cough* rate limits *cough*) through the use of custom CLI tools without breaking my current train of thought. No matter the tool I am in, I can open up a terminal and hand off a task. 3. Plan. This part is critical with any system you are building, but having a solid plan means that your team can give feedback early, and you can agree on what needs to be built and a high-level "how" before effort is wasted on implementation. Make the plans "live" where they feel most natural: GitHub issues, Notion, Slack, or a local markdown file. This has come in so handy, as now once I've made a plan and reviewed it (and I make one for everything, even something small), I can point multiple AI agents at it and have a great starting point. Plan → Review → Iterate → Repeat until complete → PR → SHIP 🚀 4. @superset_sh has really saved the day and made my agent orchestration so much smoother. @NckLcs recommended this in our Slack a few months ago, and it has made it a breeze to set up worktrees and span context across projects. It just doesn't get in the way and works exactly the way I want, excellent! 5. Using AI throughout the entire lifecycle. What I mean by this is when I have something in a PR, instead of manually opening the (normally broken these days) @github UI, I just ask: "Fetch the latest feedback and validate it against my local changes," closing that feedback loop quicker and resulting in a much faster time to merge. 6. Tests. I know, nothing new here, but with the rise of AI throughout the entire cycle, it's crucial to have a solid test suite from day 0. These are just some of the ways that I've boosted my own productivity and work at @framer, and that's the beauty of working here; nobody is forced to work in a certain way, we are all working towards the goal of shipping 🙌🏼

Yeah, that’s neat! I’ve actually used the teardown/startup to manage a custom setup so that I can quickly hand off to a worktree (and know which ones are/are not managed via superset) Also, I think it’s difficult to orchestrate the testing of worktrees in repos that expect things in parent directories and/or use docker to spawn test scripts where they are all competing for port allocation etc. This is not a superset problem however, it’s a problem of worktrees and not having the necessary setup for it in the repo(s)

Ah that is a common pain point! The way we've been suggesting to handle it in the short-term is to add a step to the setup script to add another worktree for your second repo: docs.superset.sh/setup-teardown… That way it's always available if necessary for an agent. We definitely need multi-repo support soon though lol. Another variant is using the cli and having the agent spawn a workspace for the second repo: docs.superset.sh/cli/getting-st…, then it'll have context and you can still directly work/look at the second project too Both aren't ideal for sure though

@saddle_paddle @framer Thank you! I think it would be awesome to group a set of trees by purpose instead of repo, I find myself working on features that span multiple repos at the same time, and it’s handy to say “check the other repo, implement from there”. maybe there is a better approach 😅

@CalMorris @framer Glad we could help! Let us know if there are any more features you want :)

@val_savd @framer Welcome!! 👋

Okayyy I can say it: I’m a designer at @framer now! 😮‍💨