Show HN: Chunk – Code sandbox for back-end devs Chunk co-founder here. We spent the last 2 weeks building this to scratch our own itch: As developers, we often have problems that could be solved just by running a few lines of code. Sometimes, running this code on your local machine is fine. But other time, the code need to run automatically reacting to external events or to run continuously, which means, it needs to run on a server somewhere. So now, you have to find a cloud provider, to package or build the code and finally to deploy it. All of that for what could be literally be 4 lines of code. We couldn’t find an easier way to do this, so we built it. Chunk is an all in one web editor (think of the codesandbox experience) that allows you to write, deploy and run a piece of code in the cloud from a variety of triggers: HTTP, WebHook, manual or scheduled (cron). No setup, no build, no deploy. Chunk makes you go from idea to code running in the cloud in seconds. Let me know what you guys think! https://chunk.run July 28, 2022 at 11:19PM
Show HN: ColorAround – where you need to find a color around u using the camera Hi I made a game in an evening, I would like to share and get feedback. You get a random color palette and have to find it for yourself. ColorAround is an easy and fun color game. It is recommended to play on mobile devices https://ift.tt/bWn0T6m July 29, 2022 at 11:19PM
Show HN: TapeMeasure – Swift class for creating segmented, labeled lines in UI TapeMeasure is a utility class, written in Swift, for generating a visual tape measure, or any other graphical elements that need layout in a measured sequence. I created this class to help with a pet project I'm working on, where I needed to generate segmented, labeled lines to use as axes on charts and graphs. It occurred to me that this seems to be a pretty common task, involving a lot of boilerplate code and tedious calculations. Why not decouple the code and move it into a dedicated class? It's self-contained, nice and simple, and easy to unit test. It can also handle a lot of tweaks to the "tape measure", such as boundary clipping and shifting offsets. It's especially good for animation, where you can just change a parameter or two during animation, and call for the re-calculated data to re-render your measured line. That makes it great for animated sliding and scaling effects. Anyway, enjoy! https://ift.tt/6LPoFb7 July 29, 2022 at 11:09PM
Show HN: TensorDock Core GPU Cloud – GPU servers from $0.29/hr Hello HN! I’m Jonathan from TensorDock. After 7 months in beta, we’re finally launching Core Cloud, our platform to deploy GPU virtual machines in as little as 45 seconds! https://ift.tt/o7EtsVz Why? Training machine learning workloads at large clouds can be extremely expensive. This left us wondering, “how did cloud ever become more expensive than on-prem?” I’ve seen too many ML startups buy their own hardware. Cheaper dedicated servers with NVIDIA GPUs are not too hard to find, but they lack the functionality and scalability of the big clouds. We thought to ourselves, what if we built a platform that combines the functionality of the large clouds but made it priced somewhere between a dedicated server and the large clouds? That’s exactly what we’ve done. Built to make engineers more productive. We have 3 machine learning images so you can start training ML models in 2 minutes, not 2 hours. We provide a REST API, so you can integrate directly your code with ours. And, there’s a community CLI you can use to manage your servers directly via command line We provide a feature set only large clouds supersede. We have storage-only billing when the VM is stopped (for only $0.073/GB/month) so that you aren't paying for compute when you don't need it. We also provide the ability to edit virtual machines after they’re created to downsize costs. If you provision a NVIDIA A6000 and find out you’re only using 50% of it, stop the VM, modify it to a NVIDIA A5000, and you’ll be billed the lower hourly rate without needing to recreate your server and migrate data over! Our infrastructure is built on 3x-replicated NVMe-based network storage, 10 Gbps networking, and 3 locations (New York, Chicago, Las Vegas) with more coming soon! - CPU-only servers from $0.027/hour - NVIDIA Quadro RTX 4000s from $0.29/hour - NVIDIA Tesla V100s from $0.52/hour - and 8 other GPU types that let you truly right-size workloads so that you’re never paying for more than you actually need We're starting off with $1 in free credits! Yes, we sound cheap… but $1 is all you need to get started with us! That’s more than 3 hours of compute time on our cheapest configuration! Use code HACKERNEWS_1 on the billing page to redeem this free credit :) TensorDock: https://ift.tt/ke6vKaL Product page: https://ift.tt/o7EtsVz API: https://ift.tt/7Kk5dwt Community CLI: https://ift.tt/ChxJbIN Deploy a GPU: https://ift.tt/GUBzjPK I'm here to answer your questions, so post them below! Or, email me directly at jonathan@tensordock.com :) https://ift.tt/o7EtsVz July 29, 2022 at 01:58AM
Show HN: Equities.fyi – Startup valuations based on public company data With the current down markets, we were wondering what private company valuations looked like when compared to similar public companies. We were inspired by the a16z article ( https://ift.tt/qRhaolT... ) for calculating valuations and figured out the sales-to-valuation ratio of public companies at the time of the last fundraising round and applied the change in the public market to get the current startup valuation. If things don’t look right, our defaults might be off - please feel free to change those if they’re too different from what you’d expect. https://ift.tt/cVmXv9f July 28, 2022 at 10:57PM
Show HN: Looria – A product (re)search engine About 1.5 years ago, I introduced my review aggregator BuyForLife on Hacker News, where it became the #8 most upvoted Show HN project of all time[1]. The idea of helping people to make better purchasing decisions continued to chase me over the last year. Here are some stats that illustrate how important online reviews are: • 90% check online reviews as part of their online buying journey • 43% visit 5-10 websites to research a product • 75% spend more than a day doing research before buying a product The top frustrations with the current process are: • Google full of SEO spam and Ads • Fake reviews • Fragmented trusted sources • Inconsistent information across sources Thanks to the recent advances in NLP (transformers, GPT-3, etc.) it became possible to solve these problems at scale, so I decided to team up with my co-founders Johnny and Tavis to build https://Looria.com . We aggregate and summarize the most trusted product reviews on the web like Reddit, Youtube, or Consumer Reports. Just like Rotten Tomatoes provides trustworthy ratings for movies, Looria offers ratings and reviews for all kinds of products. Our vision is to make Looria the go-to platform for making purchase decisions. Let me try to answer a few questions that come up frequently: # Why should you trust this meta review site over the 100's of other meta or search sites and 10,000's of fake review sites on the web? -> We are completely unbiased and transparent. We are not affiliated with any of the businesses or products that we review, so we have no incentive to skew our ratings in any way. Our team has a deep understanding of the product review space and the specific problems that exist within that market. We're also solving our own problem after spending countless hours finding and researching good products. # Why can't I just go to Wirecutter or Consumer Reports directly? -> Most people don't trust anyone blindly and look at multiple opinions (4-7 websites on average). One single place will never be "just enough" for a product you're planning to use for a long time and spend quite some money on it. There is often at least one deal-breaking issue that you will find after reading through different sources. Many people rely on Reddit for reviews. Redditors and other forum members are more interested in boosting their ego by showing their depth of knowledge on the topic (and correcting others on the topic), whereas corporate websites are more interested in generating clicks. # How do you avoid fake reviews? -> Fake reviews are a big challenge that even platforms like Amazon don't have under control. Good fake reviews are technically almost impossible to detect, even with sophisticated network analysis of the reviewer's profile. We focus on keeping track of and removing the special status of sources caught getting paid to write fake reviews. Some categories are more prone to astroturfing than others, and we account for that by restricting the sources. Actively curating the sources is part of our daily work. # What are Exit Reviews? -> Another big frustration with product research is the lack of long-term product information. We should reflect upon how a product performed over its duration of service instead of when it first arrived and people haven't spent much time with it to learn the quirks. That's why we're building exitreviews.com - a community where people can learn how products performed over their lifetime, where they failed, and how to fix them. Looria is still in beta and our data is far from perfect. We're working hard on improving the data quality, adding better filters, and scaling to many more categories. [1] https://ift.tt/HmeXFov https://looria.com July 26, 2022 at 11:22PM
Show HN: TypeScript query builder with full type inference Hey HN! Colin here - a TypeScripter, open sourcer, and engineer at EdgeDB. As the creator of Zod and tRPC, I'm interested in designing tools/APIs that use type inference and generics to make life easier for devs. This query builder represents another step in that direction. We set out to build an EdgeQL query builder that can express queries of arbitrary complexity (EdgeQL has feature parity with SQL, roughly) and infer the static type of the query result. We introspect the database and generate a schema-aware client that represent any query, including ones that use built-in functions, operators, string/array/tuple indexing, aggregations, conditionals, type casting, subqueries, computed properties, etc—things most ORMs can’t represent. This post mostly discusses the API design, which I think will be interesting regardless of familiarity with EdgeQL. I’d love to see some of these ideas bleed into future generations of TypeScript ORMs/query builders too. Best way to try it is to clone the sandbox repo and follow the instructions in the README[0]. Or jump into the docs[1]. [0] https://ift.tt/RSNka7t [1] https://ift.tt/3azK0ty https://ift.tt/XJz83KI July 26, 2022 at 04:24PM
Show HN: PickCode – An educational coding environment for students after Scratch PickCode is designed for use on desktop and tablet, and supports creating chatbots, visual designs, and 2D games. There is plenty of functionality missing - you can't add media to games for instance, but the current version shows off the foundation of what I'm aiming at. I taught myself to code using MIT's App Inventor, so I have an enormous respect for block based languages like App Inventor, Scratch, Snap!, MakeCode, etc. PickCode is my attempt at adding options for students who want to learn more about programming without making the jump to text, or as an alternative to block coding for beginners coming to programming at an older age. The visual language is meant to lower the barrier to entry to coding but the far more important aspect for me is giving students the ability to make things they're proud of as quickly as possible. A JS/Python API for controlling the chatbot and game engine are in the works. As of now, there are sample programs to play with and an editor which saves your programs to local storage. Full user accounts, tutorials and administrator accounts for teachers to organize assignments are on their way soon. If you're interested in using PickCode in a classroom or want to discuss feedback, send me an email at charlie@pickcode.io https://ift.tt/tL2MZQX July 26, 2022 at 01:46AM
Show HN: Pipes puzzle (a.k.a. Net) on a hexagonal grid Hello, HN - I wanted to share this puzzle game I made during my vacation. I'm rather fond of the pipes puzzle where your goal is to restore a scrambled network of connections by rotating tiles. It's usually played on a grid of squares and this all started when I decided to make a programmatic solver for that kind of puzzle. Then I realized that with some minor changes the solver could generate new puzzle instances. I thought about what kind of puzzle to make and someone suggested a hexagonal grid. Adapting the generator wasn't too hard but then I had to create a way to play this variant. So I did just that =). I find hexagonal pipes a bit more difficult than the square variant because there's a larger variety of possible tile shapes. For an extra challenge I implemented wrap mode where the board can connect to itself (right to left and top to bottom), so there are no convenient outer walls to start from. The site is made with Svelte Kit, its code is available on github at < https://github.com/gereleth/hexapipes >. Hope you enjoy playing =). https://ift.tt/IYgyhbr July 25, 2022 at 12:20AM
Show HN: Redirect clicks using a formula Part of my day job requires me to review a report monthly. The URL for July's report looks something like this: fancyreport.com?from=2022-07-01&to=2022-07-31 Typing this URL manually every month was just enough pain for me to put up a simple script on somedomainiown.com/magic that redirected me to the report for the current month by automatically filling in the "from" and "to" values in the URL. This way, I could bookmark one link and always open the latest report with one click. At some point, I figured I should put this behind a UI so that I could share it with my colleagues and maybe HN too. A couple of weeks of hacking, and here we are with an MVP the little imposter inside me finally accepted to post online. P.S. the "formule" implementation gets the job done for me, but is pretty rudimentary. Would love to know if you have a specific use case that'd benefit from other parameters or a more advanced formula system. https://ift.tt/dtM7Pl9 July 25, 2022 at 02:35AM
Show HN: Zazow – Generative Art Toy (fractals, spirograph, satori, etc.) Im finally ready to share a fun side project that I've been working on. For 20 years, Ive been fascinated with Mandelbrot fractals. Back in the day, I made a MacOS9 app that rendered them, and I missed it. So, I decided to bring it up to the modern age and build a website to explore this and other similar fractals. In the process, I discovered some other types of generative art that looked fun, so I made those too. One is a spirograph, another is similar to an old AfterDark screensaver called Satori (anyone remember that?). Others draw paint splatters and random squiggly lines. The user is given a bunch of interactive settings to control how the artwork is generated. Im hoping to continue adding different styles of generative art to the website over time. Its a fun, relaxing project to work on and I hope other people enjoy using it to create pleasing images. https://www.zazow.com July 22, 2022 at 10:44PM
Show HN: Scribe, a Go library for writing, running, and generating CI pipelines Hi there. This is a passion project of mine I started after spending many painful hours working on CI pipelines. The general idea is that I wanted to stop making flimsy scripts in build & release pipelines and start writing software. I'm happy to answer any questions. Please let me know what you think of the idea and if there's anything you'd like to see. https://ift.tt/yMzVNkQ July 22, 2022 at 03:28AM
Show HN: I built an all-in-one virtual production platform for indies After being frustrated with the budget and learning curve requirements of the big names in the Virtual Production field, I decided to create my own all-in-one platform designed with beginners and less-technical folk in mind. It takes complicated tasks in other engines (tracking, compositing, floor plans, etc.) and makes them two-click solutions. And, with the free companion mobile app, you can track cameras, scan your sets, and (coming soon) do full body motion capture! There's a ton more features that I'm super proud of, and I'm open to any and all feedback! Thanks! Https://SceneForge.app July 21, 2022 at 03:43AM
Show HN: I built a handheld CHIP-8 game console to teach myself embedded systems A while back I wrote a CHIP-8 emulator (which is considered the Hello, World! of emulators and is more accurately a virtual machine since historically CHIP-8 was an interpreted language running on top of the COSMAC VIP to make game programming easier). But a few months ago I got really interested in embedded software, so decided it would be neat to port my emulator to a STM32 MCU and design a console around it as a learning experience, since CHIP-8 never existed as a physical system. I didn't know much about embedded software when I began, and even less about electronics, but I managed to write all the firmware from scratch and even designed my first PCB, resulting in a finished (though not very polished) handheld CHIP-8 console. For those curious, the GitHub repo also has links to my dev blog about the project as well as a build guide. Thanks for looking! https://ift.tt/8APjd54 July 20, 2022 at 03:12AM
Show HN: Ex Platform for SWEs Hey guys, just sharing this project that I've been working on. Let me know what you think. This idea came primarily from my own experience in the software industry. After a while, I was feeling discouraged by my job, mostly because I felt expendable (like a code monkey), burned out, and had unchallenging work. It felt like the employers I was working for didn’t have an idea of what was going on or how their employees were feeling. They would talk a lot about the technical challenges, but never did we talk about any mental/emotional challenges. So I started thinking about what could be done to mitigate this problem and came up with this solution. An employee experience platform that tracks multiple psychological dimensions (specifically customized to track problems that can be faced in a software engineering environment) and gives the managers some insights into their teams/employees. I'm still currently working on the MVP but would love to hear what you guys think about the idea and landing page. The MVP should be out in 2-3 weeks from now. https://smartteams.ca/ July 17, 2022 at 01:24AM
Show HN: Hacker News Mods - A collection of tools/mods for HN Hey HN! I built Hacker News Mods as a place to collect any tools or sites related to HN. We just started building one mod/tool for HN per week, and I thought it’d be a good idea to showcase all of the tools we’ve created, as well as tools that others have created as well. The site is pretty scrappy, so any feedback is appreciated! Also, please submit any projects that we don’t already have listed! Thanks, Jarren https://ift.tt/DjkOYmM July 16, 2022 at 02:09AM
Show HN: Eesel – Federated search without API integrations Hey there! Amogh here from eesel ( https://eesel.app ). eesel filters your browser history to show the docs you need for work, right in your new tab. You can see recent docs, filter by app or search by title or content. We're trying to solve a pretty universal problem. Everyone's work is spread across apps - there's a project brief in Google Docs, issues in Jira, a mockup in Figma, PRs in GitHub - and with this kind of sprawl, it can be a game of trial and error to find the links we need to do our job. Trying keywords in the address bar only works if we remember the title and it's specific enough, search in apps can be slow and noisy, company "knowledge hubs" in Confluence or Google Drive are usually not up to date, and we ultimately just ping each other on Slack to find things. I was struggling with this acutely as a PM at Intercom, and it felt ridiculous that I could search the web faster than my company's docs. It was around this time that I also discovered an Effective Altruism blog post on Operations ( https://ift.tt/DZvefq1 ) and how "maximising the productivity of others in the organisation" can have this multiplier effect for your own impact. That's when it clicked - here's an "operations" problem that felt tractable for my skills and I could potentially multiply my impact by solving it. This is what gave the conviction to prototype something on the weekends, and things spun off from there. Let's talk about the solution more. The magical thing about eesel is that we don't use APIs. When it comes to "search across apps", integrating with different APIs is a pretty default way to approach things. That's how we started, but things felt uneasy - could we really build API integrations with _everything_? There's so much out there, and this list is pretty much always changing. If we really did want a search across all work apps, we'd have to play catch up with old and new APIs. You could argue that these were just the schleps ( https://ift.tt/Ud79qkD ) we had to overcome, but it was amidst this we realised that uh, the browser exists. We mostly work in the browser, and the great thing about it is that it's built on web standards. From HTTP and URLs to HTML and CSS, all apps in the browser follow the same predictable patterns: documents are accessed via URLs, content lives inside the HTML, there's a page title, there's a favicon, and so on. It's not a perfect replacement for APIs, but it felt good enough. We didn't need to manually integrate with each app, and could instead rely on existing web standards. And that's what we did. eesel works with any app in the browser, including apps without APIs (like that internal company tool), or apps that don't exist yet (the new Product Hunt hit). Not using APIs also meant that we could go an extra step with privacy - eesel works fully locally by default and you don't need to login to _anything_ (even eesel!). Simply install and it works. We want to keep building on this approach and improve how we work in the browser. For instance, eesel uses keywords to automatically organise pages into Folders, and there's Commands to take actions (spoiler: you can customise a JavaScript to inject on a page, like this script that goes to a Notion backlog and clicks the "New" button - https://ift.tt/8Ceq3ZI... ). Alright, that's a lot of writing from us. We have a bunch of ideas, and would love to hear about where you think we should take this next. https://www.eesel.app/ July 15, 2022 at 05:44PM
Show HN: Mapedia.org – A Crowdsourced Learning Map Hi HN! We're happy to announce the launch of Mapedia.org, an open source crowdsourced learning map! Mapedia is a new kind of learning platform at the crossroad between Wikipedia, Google Maps and Khan Academy: a learning map built collaboratively to support online learners to learn any topic seamlessly. We built an interactive learning map of topics to be able to visualize the different fields of knowledge, what concepts are included in them and how they relate to each other. This allows for curiosity based exploration, identifying knowledge gaps (unknown unknowns) and figuring out what to learn next (and in which order). For each topic you can then find community and expert curated resources, learning advices and smart recommendations in order to learn as efficiently as possible. We want people to spend time learning rather than figuring out how to learn, and in particular to empower self-directed learners. The idea came out of the frustration and inefficiency of learning online, and I've been working on it for 2 years now. The vision in itself for it is not so new, Mapedia is rather a different take on it that particularly believes in the potential of crowdsourcing and online communities. Our roadmap includes implementing learning groups based on shared goals rather than shared course/learning material, customizable "constructive" feeds of learning materials and adaptive learning paths. The topic map is obviously far from complete and we are still in the early product iterations, but you can checkout a few examples here: https://ift.tt/U9eVA32 -> The explore map from the top level topics https://ift.tt/uphciXS... -> the map focused on functional programming, showing how concepts relate to each other https://ift.tt/36FQTPy... -> the page for the functional programming topic, with curated resources https://ift.tt/IOdBnNs... -> an example of a learning path (this feature is in a very alpha version) Let us know what you think! We're very open to feedback and suggestions https://mapedia.org/ July 15, 2022 at 11:57PM
Show HN: App to help with planning, producing, reviewing, and publishing video Hey everyone, I've started working on an app to help companies, freelancers, and even individual content creators plan, produce, review, and publish their videos and streams in a more organised way. We're just getting started so not trying to sell anything at this point, but I do need feedback on where we're headed. I recorded a video walkthrough of some mock-ups at https://www.youtube.com/watch?v=mTtB_EkUA-M, or you can take a look at them yourself at https://ift.tt/za751Vr (it is somewhat self-explanatory, but in the video I add more context). You can think of it as a potentially better, more organised alternative to having a bunch of stuff in random Google Docs and other places that you have to pass back and forth. I also eventually expect to add a lot of automatic post-processing type features (captioning, lighting/sound adjustments, "umm" removal, etc.) for people who don't know how to do that on their own. Any kind of feedback at all is welcome, positive or negative, on whether there's an idea here that you think I could develop into something people would pay for, or if there's something you'd like to see in an app like this. Thanks in advance! July 14, 2022 at 05:15PM
Show HN: Typewaiter, the typewriter that doesn't wait Howdy HN, just recently made this silly typewriter variant and think some of you might enjoy it so thought I'd share. Here the cursor takes no heed of what you're typing and just advances at a steady pace, with the effect that typing something reasonable-looking requires you to type at a very steady rhythm. There's also the bonus that the space character is no longer needed—you can (and kind of have to) just wait—so the keyboard minimalists among us can shave even more space off their devices. Only desktop for now. It's an interesting contrast to the last one I made—which requires you to do all the heavy lifting in moving the cursor position yourself— https://ift.tt/VxK42dZ https://ift.tt/tRUF0Z1 July 11, 2022 at 03:57AM
Show HN: Evryca – fractal thinking tool to brainstorm and organize thoughts Made a tool to organize thoughts. Actually it is a mind tree, but in a more web-friendly form. It has pivoted from what I originally started building at evryca.com. Some years ago I got the idea of fractal conversation, instead of old-school tree/ladder-like comments. I wanted to see only comments related to the current level. I started making "something" with fractal comments. This "something" was a project discussion platform. But it turned out that even I myself don't use it, and the idea of fractal comments stuck there unused. And recently it dawned on me that it may be a conversation with yourself — thinking, brainstorming, taking notes, writing. So made this kind of cork/whiteboard, where one can dive into the subject and, being in the flow, write and see only related notes and rearrange them later. I'm trying to make it flawless and add keyboard shortcuts where it's possible (Ctr+Enter to submit idea, drag-and-drop to rearrange, Esc to jump level up). So finally I've made a tool that I use myself and will update it gradually (sorting, touch devices, ex/import (json, text), boards, more navigation with keyboard and other stuff, and login). https://www.evryca.com/ July 9, 2022 at 03:03PM
Show HN: I made a small Space Shooter Thank you https://opengameart.org for your service. People like me who can't draw or compose music are in your debt. The game, "Glitch and Rush" has been made in a few hours here and there. I've been tinkering in GameDev for years, if not decades, but I never released anything. The goal here was simple: - come up with a core mechanic - implement a single level - add some polishing - RELEASE I hope you'll enjoy it. BTW, my best time is 2min23s (yes, I'm bad at my own game, it's ok, I'm ok). https://ift.tt/SwrLDBF July 9, 2022 at 02:28AM
Show HN: I made an app to help insomniacs learn how to sleep again Hi HN! I suffered from chronic insomnia for over a year and tried everything from cutting coffee, blocking blue light, to taking melatonin and antihistamine, but couldn’t find anything that worked. I even bought a $500 research-grade EEG device to track my sleep, which was honestly kind of depressing because it showed that I was sleeping less than 4 hours per night for weeks straight. In the day, it took an immense amount of energy for me to perform even the most mundane of tasks, such as doing my laundry or ordering groceries. At night, I felt an overwhelming sense of loneliness and resentment as I lay in bed wide-awake, reading and re-reading Sleep by Murakami or mindlessly scrolling through reddit/ HN. My performance at work suffered, my personal relationships suffered, and my happiness suffered. When I finally decided to see a sleep specialist, I was put on a 3-month long waiting list. Eventually, I was able to get my insomnia treated, but I realized that there is no reason why anyone should wait 3 months to get treatment when the same therapy that I received can be delivered online. My co-founder and I both have experience in digital health, so we decided to partner with sleep experts to create a mobile app to help people with insomnia get better sleep using psychology. We launched in February this year, and have already helped over 500 patients improve their sleep permanently. Our data shows that our program is just as effective as group, in-person sleep therapy, and we’re doing a clinical study with Brigham and Women’s hospital and Harvard Medical School to prove the efficacy of our product. On average, our users sleep 74 minutes longer than before and spend 52% less time awake in the middle of the night. If you have trouble with sleep, try our app and let us know what you think! https://slumber.one July 8, 2022 at 07:53AM
Show HN: PostgreSQL proxy for integration testing things you don't control Hi hackers! I recently tried to get a stack of microservices running in a docker-compose sandbox for integration testing using their original (e.g. prod) configuration. This means I didn't know what credentials they'd be using or what database name. So I hacked together a postgresql proxy using an existing project ( https://ift.tt/TqBX8ve ) and simply allow any credentials to connect and make sure the desired database exists before returning control to the caller. Sharing this in the hopes it is useful for someone else. Lucas https://ift.tt/ms1TBEA July 8, 2022 at 12:12AM
Show HN: Barfi – Python flow based programming with GUI What? Another flow based programming library for Python? Yes. All the FBP libraries out there for Python need to be run as a self contained application. They are not components that could be integrated into your existing data workflows. Barfi, on the other hand can be integrated. At the moment it has a Streamlit component that you can use in your Streamlit apps. Currently, I am working on a Jupyter notebook widget. https://ift.tt/nZWbd0C July 7, 2022 at 02:30AM
Show HN: C3 – A C alternative that looks like C Compiler link: https://ift.tt/czU4tGx Docs: http://www.c3-lang.org This is my follow-up "Show HN" from roughly a year ago ( https://ift.tt/SMC6yEY ). Since then the language design has evolved and the compiler has gotten much more solid. Assorted extra info: - The C3 name is a homage to the C2 language project ( http://c2lang.org ) which it was originally inspired by. - Although C3 mostly conforms to C syntax, the most obvious change is requiring `fn` in front of the functions. This is to simplify searching for definitions in editors. - There is a comparison with some other languages here: https://ift.tt/rfQYvuG - The parts in C3 which breaks C semantics or syntax: https://ift.tt/XyKlDzN - Aside from the very C-like syntax, one the biggest difference between C3 and other "C competitors" is that C3 prioritizes C ABI compatibility, so that all C3 special types (such as slices and optionals) can be used from C without any effort. C and C3 can coexist nicely in a code base. - Currently the standard library is not even alpha quality, it's actively being built, but there is a `libc` module which allows accessing all of libc. Raylib is available to use from C3 with MacOS and Windows, see: https://ift.tt/mRzilS4 - There is a blog with assorted articles I've written during the development: https://ift.tt/TxIcFJL July 7, 2022 at 01:01AM
Show HN: Graphsignal – ML profiler to speed up training and inference Hi, Graphsignal founder here. We've launched Graphsignal earlier this year to make machine learning profiling practical and easy to use. Basically, it enables the profile-optimize-benchmark loop. For example, making inference faster by optimizing an ML model, while still maintaining accuracy. We've make a lot of progress that I wanted to share. The profiler now natively supports TensorFlow, Keras, PyTorch, PyTorch Lightning, Hugging Face, XGBoost and JAX frameworks along with built-in support for distributed workloads. Profiles now include tracing information in chrome trace format. Process and GPU utilization data has been extended as well. It is now possible to monitor all run metrics. Useful for long runs. Profiled workloads are now sharable across teams and publicly (if enabled). I'm excited to show it here and appreciate any thoughts, comments and feedback! https://ift.tt/PoULa7h July 4, 2022 at 05:10PM
Show HN: Trane, an automated system for learning complex skills Hi HN, I released Trane over the weekend: https://ift.tt/m1QxKyh . Trane is an automated system for learning complex skills. Think of it like defining a skills tree (technically a graph) of all the smaller skills you need to master a complex skill and having an automated system to automatically traverse the graph as you master them. The seed for Trane was planted after my frustration trying to learn music, and jazz in particular. There are simply too many things you need to master first (e.g. knowing the names of a note, knowing where the notes are in your instrument, timing, etc) and it becomes difficult to track what it is that you should focus on, and there is a process of constant atrophy, even if you practice consistently. Trane is an early state, but is already usable. I have released a command line interface at https://ift.tt/oiDr8pI and some music courses at https://ift.tt/wS5MyRH . I would like to get some ideas in regard to what other skills could be a good fit for Trane. I am thinking chess, programming, or languages could be a fit. I am wondering if Trane could be applied to something like learning pure mathematics. I would love to hear any suggestions. Perhaps there's some of you who have found a similar issue while practicing your own hobbies. https://ift.tt/m1QxKyh July 4, 2022 at 11:34PM
Show HN: WebReducer – micro event sourcing cloud function Hey everyone, WebReducer is a project I've been working on since last December. You can read about my process here: https://ift.tt/XfTxHUB Send some data to WebReducer. Then, write a reducer function to retain some state. You can change the reducer function as much as you want and recreate that state. Still working on how to explain it. It's a tiny FaaS platform. It's a tiny database. It's micro event sourcing. It's "redux" on the server as a service. It's a backend for tiny personal projects. It's a place to send your webhooks. I'm looking for feedback on the following: - Do you get what it is? - What would you use it for? - What would I need to iron out for you to seriously consider trying this? https://webreducer.dev July 2, 2022 at 11:47PM
संदेश
