αιмѕσмηια

its a gooood model

Now that's an ordering that pleases me 😄 Just discovered Tool Questor, a nice platform for discovering open source alternatives for your workflows. Thanks @kmhaneem for featuring @junobuild 👍 toolquestor.com/category/backe…

If you feel adventurous, custom functions in TypeScript for @junobuild are released. ✅ Docs and announcement coming tomorrow (or so), after I review an awesome contribution first (👍).

Knight Lautrec of Carim #darksouls

CS50x, learn basic.

Privacy isn’t a feature — it’s a right.

Step 3: Ask it's AI about what you are interested in: (I typed "Where can I trade NFTs")

I have been developing on ICP since genesis. Cycle costs have been a concern since day one. When you develop on web2, you try to move as much as possible to the server. This is because it is safer, you have more control, plus you can allocate processing power to ensure things are fast even for users with slow computers. On ICP, you move as much as you can to the client, getting the browser to do the work whenever possible. The feasibility of most app ideas hinge on if you can come up with some clever tricks to offload most of the app's processing to the browser. As an example, the reason @swaprunner was at all feasible to build is that the whole thing runs in the browser! That is, the ICPSwap and Kong code that we call runs in their backends of course, but the swaprunner page calls those backends directly from the browser, not costing swaprunner a cent in cycles! Swaprunner is, I think, one of the more popular apps on ICP, and yet it would not be able to sustain itself with the high user loads but infamously low liquidity/volumes traded on ICP if it had to do any interesting work for users in its own backend. As it is, the swaprunner backend keeps track of the registered tokens in user wallets and some trading aggregated trading stats, and that is already enough that it is strictly speaking unsustainable, and with a 2.5x gas cost increase, definitely so. Ads, you might think? Have you seen ads on ICP apps? There are reasons why you haven't (that this post is already too long to get into, but they exist). When I wrote my first ICRC1 token indexer, SneedScan (OGs will remember), it was very different from all other indexers I have ever seen or written on other platforms, in one prominent respect: it had no timer that kept the indexer running to ensure it was up to date. The obvious choice (you wouldn't even reflect over it on other platforms) would have been prohibitively expensive on ICP. So, instead, the indexer was demand driven - every time someone visited the frontend website, the browser would send a little nudge to the indexer to catch up and index any new transactions. A most ridiculous solution anywhere else, a most ingenious solution on ICP. Every app I have written on ICP has been like this. Sneed Hub is very full of clever tricks - perhaps you would be amazed to learn that to show you all the members of an SNS DAO, we download *all* the neurons from the SNS into the user's browser and makes the browser index them! The SNS canisters don't even have an index over the SNS members, it would be too costly to maintain. And it's not like it is just me. All devs I know that have survived on ICP work like this, furiously counting every cycle and trying to cut costs in ever more innovative ways. When I worked with Tirex on TACO DAO, I was thoroughly impressed by someone who counted their cycles even harder than I, and it shows in the performance - all of TACO DAOs dapp canisters together use up fewer cycles than their SNS indexer canister alone (Dfinity supposedly does not have the same incentive to write efficient canister code for the SNSes, since it is not their cycles being burned). And it is not like Dfinity doesn't know this. Despite the outlandishly terrible performance of the SNS indexers, Dfinity has cut corners everywhere to deal with the expensive cycle costs. Did you know for example (and this comes as a big surprise to many) that Dfinity's DAOs, both the NNS and the SNSes, that we all make a big deal about how the votes go on the blockchain...actually throw away all your votes after a little while to save money. No permanent voting record on the blockchain, for any of the DAOs. Too expensive. Why is the Dfinity forum not on-chain? Why not the Internet Computer dashboard? Take a wild guess. Too expensive. Why doesn't Caffeine run on-chain? When not even Dfinity deploys on ICP because it is too expensive, and the only ones who do are those who spend years learning every trick in the book to make it affordable, then the proposal to raise cycle costs by several hundred percent comes as very little short of a slap in the face to us devs who actually care about deploying on this chain. The attitude of investors who have never even tried to ship an economically viable product on ICP but understandably listen to Dfinity on this and now are there to tell devs that doubling or tripling or 5x-ing the costs (or whatever Dfinity has landed on this Tuesday) will be fine, and that greedy devs (pretty unanimously deep in the red) have practically been stealing these cheap cycles from their pockets, is also not comforting. It is time for ICP investors to understand the truth about their investment. No, it is not cheaper than AWS. Yes, it is amazing, alien tech, but that won't matter if it is not affordable. There are no users, no liquidity, and there is an architectural issue with serving ads. There is, simply put, no money here, and very soon there will also be no builders left if you continue down the track of making the platform even less attractive. The platform is amazing, but before it has been very thoroughly battle tested, it is a tech bet, not a business. Until then, don't expect revenue, what you can hope for is attracting enough highly competent tech enthusiasts, perhaps even some of them scouts for bigger business evaluating new tech (I invite you to seriously consider this aspect), and let them cook. That is the only, I repeat only, way to prove your platform - your investment - to the world. Without devs on an application platform, you truly have nothing. And you're about to drive the remaining devs away. Vote no on mission 70 before it is too late. nns.ic0.app/proposal/?u=qo…

From static to motion, @canva's professional design suite just got even bigger. Created by animators, for animators. Welcome to the family, @cavalry__app 💚 Read more on the latest news: canva.com/newsroom/news/…

☕ blogged a full stack walkthrough (TypeScript + Rust) about replacing GitHub Actions secrets with short-lived OIDC tokens daviddalbusco.com/blog/building-…

Sunday funday ☀️ The Console UI/UX got updated with a bit of polishing and the recent deployments were added to the Satellite overview 🍻

test NeoBr-UI

☕ blogged about a neat Svelte snippet pattern I discovered recently to improve the type safety of my guard components. daviddalbusco.com/blog/better-ty…

I'm getting rid of third-party providers as default for local dev on Juno. 👉 Faster DX, better narrative alignment, easy identity switching ✅ Another long-time todo resolved 😃

Tired of the auth flows while developing? Introducing the **one-click** sign-in ☝️

I love the Neo-Brutalism UI style, so I made a library for myself to use in future projects to keep things consistent. of course @junobuild UI are the reff💙

The pokedbots screening

the generate are more faster now. dpl4s-kqaaa-aaaal-asg3a-cai.icp0.io

🚀 Sign-in to Juno Console with GitHub is here!

@SnassyIcp DO IT!!!!

@aimsomnia It is a great idea, I will definitely see what I can come up with! 😃

imagine a marketplace with neuron and canister auctions and private token OTC deals