The best Hacker News stories from Show from the past day
Latest posts:
Show HN: TutoriaLLM – AI Integrated programming tutorials
Hello! I’m So Tokumaru ( <a href="https://tokumaru.work" rel="nofollow">https://tokumaru.work</a> ), high school student from Japan working on ways to leverage large language models (LLMs) for K-12 programming education. Recently, I built a project called "TutoriaLLM," a self-hosted application designed to support educators and students through a multi-modal LLM that offers guided tutorials.<p>TutoriaLLM integrates voice (voice-to-voice) and text-based AI interactions into Blockly, a block editor widely known for its use in Scratch. It enables server-side code execution and allows students to test their code directly in Minecraft (Bedrock or Education Edition) without any setup.<p>I’m actively developing this project and aim to bring it to a level ready for use in programming education by next year. I’d love to hear any feedback!<p>Documentation: <a href="https://tutoriallm.com" rel="nofollow">https://tutoriallm.com</a><p>Demo: <a href="https://demo.tutoriallm.com" rel="nofollow">https://demo.tutoriallm.com</a><p>Repository: <a href="https://github.com/TutoriaLLM/TutoriaLLM">https://github.com/TutoriaLLM/TutoriaLLM</a>
Show HN: TutoriaLLM – AI Integrated programming tutorials
Hello! I’m So Tokumaru ( <a href="https://tokumaru.work" rel="nofollow">https://tokumaru.work</a> ), high school student from Japan working on ways to leverage large language models (LLMs) for K-12 programming education. Recently, I built a project called "TutoriaLLM," a self-hosted application designed to support educators and students through a multi-modal LLM that offers guided tutorials.<p>TutoriaLLM integrates voice (voice-to-voice) and text-based AI interactions into Blockly, a block editor widely known for its use in Scratch. It enables server-side code execution and allows students to test their code directly in Minecraft (Bedrock or Education Edition) without any setup.<p>I’m actively developing this project and aim to bring it to a level ready for use in programming education by next year. I’d love to hear any feedback!<p>Documentation: <a href="https://tutoriallm.com" rel="nofollow">https://tutoriallm.com</a><p>Demo: <a href="https://demo.tutoriallm.com" rel="nofollow">https://demo.tutoriallm.com</a><p>Repository: <a href="https://github.com/TutoriaLLM/TutoriaLLM">https://github.com/TutoriaLLM/TutoriaLLM</a>
Show HN: TutoriaLLM – AI Integrated programming tutorials
Hello! I’m So Tokumaru ( <a href="https://tokumaru.work" rel="nofollow">https://tokumaru.work</a> ), high school student from Japan working on ways to leverage large language models (LLMs) for K-12 programming education. Recently, I built a project called "TutoriaLLM," a self-hosted application designed to support educators and students through a multi-modal LLM that offers guided tutorials.<p>TutoriaLLM integrates voice (voice-to-voice) and text-based AI interactions into Blockly, a block editor widely known for its use in Scratch. It enables server-side code execution and allows students to test their code directly in Minecraft (Bedrock or Education Edition) without any setup.<p>I’m actively developing this project and aim to bring it to a level ready for use in programming education by next year. I’d love to hear any feedback!<p>Documentation: <a href="https://tutoriallm.com" rel="nofollow">https://tutoriallm.com</a><p>Demo: <a href="https://demo.tutoriallm.com" rel="nofollow">https://demo.tutoriallm.com</a><p>Repository: <a href="https://github.com/TutoriaLLM/TutoriaLLM">https://github.com/TutoriaLLM/TutoriaLLM</a>
Show HN: Draw.Audio – A musical sketchpad using the Web Audio API
Show HN: Draw.Audio – A musical sketchpad using the Web Audio API
Show HN: Draw.Audio – A musical sketchpad using the Web Audio API
Show HN: Draw.Audio – A musical sketchpad using the Web Audio API
Show HN: BemiDB – Postgres read replica optimized for analytics
Hi HN! We're Evgeny and Arjun, and we’re building a better way to do analytics with Postgres.<p>We love Postgres for its simplicity, power, and rich ecosystem. But engineers have to still get bogged down with heavyweight and expensive OLAP systems when connecting an analytics data stack.<p>Postgres is amazing at OLTP queries, but not for OLAP queries (large data scans and aggregations). Even in this case, we’ve still heard from countless scaling startups that they still try to use only a read replica to run analytics workloads since they don’t want to deal with the data engineering complexity of the alternative. This actually works surprising well initially, but starts to break for them as they scale or when integrating multiple data sources. Adding lots of indexes to support analytics also slows down their transactional write performance.<p>When growing out of “just use Postgres”, companies have to understand and wrangle complex ETL pipelines, CDC processes, and data warehouses — adding layers of complexity that defeat the simplicity that undermines their initial choice for Postgres as their data storage in the first place.<p>We thought there had to be a better way, so we’re building BemiDB. It’s designed to handle complex analytical queries at scale without the usual overhead. It’s a single binary that automatically syncs with Postgres data and is Postgres-compatible, so it’s like querying standard Postgres and works with all existing tools.<p>Under the hood, we use Apache Iceberg (with Parquet data files) stored in S3. This allows for bottomless inexpensive storage, compressed data in columnar files, and an open format that guarantees compatibility with other data tools.<p>We embed DuckDB as the query engine for in-memory analytics that work for complex queries. With efficient columnar storage and vectorized execution, we’re aiming for faster results without heavy infra. BemiDB communicates over the Postgres wire protocol to make all querying Postgres-compatible.<p>We want to simplify data stacks for companies that use Postgres by reducing complexity (single binary and S3), using non-proprietary data formats (Iceberg open tables), and removing vendor lock-in (open source). We'd love to hear your feedback! What do you think?
Show HN: BemiDB – Postgres read replica optimized for analytics
Hi HN! We're Evgeny and Arjun, and we’re building a better way to do analytics with Postgres.<p>We love Postgres for its simplicity, power, and rich ecosystem. But engineers have to still get bogged down with heavyweight and expensive OLAP systems when connecting an analytics data stack.<p>Postgres is amazing at OLTP queries, but not for OLAP queries (large data scans and aggregations). Even in this case, we’ve still heard from countless scaling startups that they still try to use only a read replica to run analytics workloads since they don’t want to deal with the data engineering complexity of the alternative. This actually works surprising well initially, but starts to break for them as they scale or when integrating multiple data sources. Adding lots of indexes to support analytics also slows down their transactional write performance.<p>When growing out of “just use Postgres”, companies have to understand and wrangle complex ETL pipelines, CDC processes, and data warehouses — adding layers of complexity that defeat the simplicity that undermines their initial choice for Postgres as their data storage in the first place.<p>We thought there had to be a better way, so we’re building BemiDB. It’s designed to handle complex analytical queries at scale without the usual overhead. It’s a single binary that automatically syncs with Postgres data and is Postgres-compatible, so it’s like querying standard Postgres and works with all existing tools.<p>Under the hood, we use Apache Iceberg (with Parquet data files) stored in S3. This allows for bottomless inexpensive storage, compressed data in columnar files, and an open format that guarantees compatibility with other data tools.<p>We embed DuckDB as the query engine for in-memory analytics that work for complex queries. With efficient columnar storage and vectorized execution, we’re aiming for faster results without heavy infra. BemiDB communicates over the Postgres wire protocol to make all querying Postgres-compatible.<p>We want to simplify data stacks for companies that use Postgres by reducing complexity (single binary and S3), using non-proprietary data formats (Iceberg open tables), and removing vendor lock-in (open source). We'd love to hear your feedback! What do you think?
Show HN: BemiDB – Postgres read replica optimized for analytics
Hi HN! We're Evgeny and Arjun, and we’re building a better way to do analytics with Postgres.<p>We love Postgres for its simplicity, power, and rich ecosystem. But engineers have to still get bogged down with heavyweight and expensive OLAP systems when connecting an analytics data stack.<p>Postgres is amazing at OLTP queries, but not for OLAP queries (large data scans and aggregations). Even in this case, we’ve still heard from countless scaling startups that they still try to use only a read replica to run analytics workloads since they don’t want to deal with the data engineering complexity of the alternative. This actually works surprising well initially, but starts to break for them as they scale or when integrating multiple data sources. Adding lots of indexes to support analytics also slows down their transactional write performance.<p>When growing out of “just use Postgres”, companies have to understand and wrangle complex ETL pipelines, CDC processes, and data warehouses — adding layers of complexity that defeat the simplicity that undermines their initial choice for Postgres as their data storage in the first place.<p>We thought there had to be a better way, so we’re building BemiDB. It’s designed to handle complex analytical queries at scale without the usual overhead. It’s a single binary that automatically syncs with Postgres data and is Postgres-compatible, so it’s like querying standard Postgres and works with all existing tools.<p>Under the hood, we use Apache Iceberg (with Parquet data files) stored in S3. This allows for bottomless inexpensive storage, compressed data in columnar files, and an open format that guarantees compatibility with other data tools.<p>We embed DuckDB as the query engine for in-memory analytics that work for complex queries. With efficient columnar storage and vectorized execution, we’re aiming for faster results without heavy infra. BemiDB communicates over the Postgres wire protocol to make all querying Postgres-compatible.<p>We want to simplify data stacks for companies that use Postgres by reducing complexity (single binary and S3), using non-proprietary data formats (Iceberg open tables), and removing vendor lock-in (open source). We'd love to hear your feedback! What do you think?
Show HN: Firecrawl-Simple – Stable fork of Firecrawl optimized for self-hosting
Firecrawl Simple is a stripped down and stable version of firecrawl optimized for self-hosting and ease of contribution.<p>The upstream firecrawl repo contains the following blurb:<p>>This repository is in development, and we're still integrating custom modules into the mono repo. It's not fully ready for self-hosted deployment yet, but you can run it locally.<p>Firecrawl's API surface and general functionality were ideal for our Trieve sitesearch product, but we needed a version ready for self-hosting that was easy to contribute to and scale on Kubernetes. Therefore, we decided to fork and begin maintaining a stripped down, stable version.<p>Fire-engine, Firecrawl's solution for anti-bot pages, being closed source is the biggest deal breaker requiring us to maintain this fork. Further, our purposes not requiring the SaaS and AI dependencies also pushes our use-case far enough away from Firecrawl's current mission that it doesn't seem like merging into the upstream is viable at this time.
Show HN: Term-Lisp – A Lisp, based on pattern matching and term rewriting
Show HN: Term-Lisp – A Lisp, based on pattern matching and term rewriting
Show HN: Term-Lisp – A Lisp, based on pattern matching and term rewriting
Show HN: SuperSplat – open-source 3D Gaussian Splat Editor
Show HN: SuperSplat – open-source 3D Gaussian Splat Editor
Show HN: SuperSplat – open-source 3D Gaussian Splat Editor
Show HN: Aide, an open-source AI native IDE
Hey HN, We are Sandeep and Naresh, the creators of Aide. We are happy to open source and invite the community to try out Aide which is a VSCode fork built with LLMs integrated.<p>To talk through the features, we engineered the following:<p>- A proactive agent<p>Agent which iterates on the linter errors (powered by the Language Server) and pulls in relevant context by doing go-to-definitions, go-to-references etc and propose fixes or ask for more files which might be missing in the context.<p>- Developer control<p>We encourage you to do edits on top of your coding sessions. To enable this, we built a VSCode native rollback feature which gets rid of all the edits made by the agent in a single click if there were mistakes, without messing up your changes from before.<p>- A combined chat+edit flow which you can use to brainstorm and edit<p>You can brainstorm a problem in chat by @’ting the files and then jump into edits (which can happen across multiple files) or go from a smaller set of edits and discuss the side-effects of it<p>- Inline editing widget<p>We took inspiration from the macos spotlight widget and created a similar one inside the editor, you can highlight part of the code, do Cmd+K and just give your instructions freely<p>- Local running AI brain<p>We ship a binary called sidecar which takes care of talking to the LLM providers, preparing the prompts and using the editor for the LLM. All of this is local first and you get full control over the prompts/responses without anything leaking to our end (unless you choose to use your subscription and share the data with us)<p>We spent the last 15 months learning about the internals of VSCode (its a non-trivial codebase) and also powering up our AI game, the framework is also at the top of swebench-lite with 43% score.
On top of this, since the whole AI side of the logic runs locally on your machine you have complete control over the data, from the prompt to the responses and you can use your own API Keys as well (can be any LLM provider) and talk to them directly.<p>There’s still a whole lot to build and we are at 1% of the journey. Right now the editor feels robust and does not break on any of the flows which we aimed to solve for.<p>Let us know if there’s anything else you would like to see us build. We also want to empower extensibility and work together with the community to build the next set of features and set a new milestone of AI native editors.
Show HN: Aide, an open-source AI native IDE
Hey HN, We are Sandeep and Naresh, the creators of Aide. We are happy to open source and invite the community to try out Aide which is a VSCode fork built with LLMs integrated.<p>To talk through the features, we engineered the following:<p>- A proactive agent<p>Agent which iterates on the linter errors (powered by the Language Server) and pulls in relevant context by doing go-to-definitions, go-to-references etc and propose fixes or ask for more files which might be missing in the context.<p>- Developer control<p>We encourage you to do edits on top of your coding sessions. To enable this, we built a VSCode native rollback feature which gets rid of all the edits made by the agent in a single click if there were mistakes, without messing up your changes from before.<p>- A combined chat+edit flow which you can use to brainstorm and edit<p>You can brainstorm a problem in chat by @’ting the files and then jump into edits (which can happen across multiple files) or go from a smaller set of edits and discuss the side-effects of it<p>- Inline editing widget<p>We took inspiration from the macos spotlight widget and created a similar one inside the editor, you can highlight part of the code, do Cmd+K and just give your instructions freely<p>- Local running AI brain<p>We ship a binary called sidecar which takes care of talking to the LLM providers, preparing the prompts and using the editor for the LLM. All of this is local first and you get full control over the prompts/responses without anything leaking to our end (unless you choose to use your subscription and share the data with us)<p>We spent the last 15 months learning about the internals of VSCode (its a non-trivial codebase) and also powering up our AI game, the framework is also at the top of swebench-lite with 43% score.
On top of this, since the whole AI side of the logic runs locally on your machine you have complete control over the data, from the prompt to the responses and you can use your own API Keys as well (can be any LLM provider) and talk to them directly.<p>There’s still a whole lot to build and we are at 1% of the journey. Right now the editor feels robust and does not break on any of the flows which we aimed to solve for.<p>Let us know if there’s anything else you would like to see us build. We also want to empower extensibility and work together with the community to build the next set of features and set a new milestone of AI native editors.
Show HN: Aide, an open-source AI native IDE
Hey HN, We are Sandeep and Naresh, the creators of Aide. We are happy to open source and invite the community to try out Aide which is a VSCode fork built with LLMs integrated.<p>To talk through the features, we engineered the following:<p>- A proactive agent<p>Agent which iterates on the linter errors (powered by the Language Server) and pulls in relevant context by doing go-to-definitions, go-to-references etc and propose fixes or ask for more files which might be missing in the context.<p>- Developer control<p>We encourage you to do edits on top of your coding sessions. To enable this, we built a VSCode native rollback feature which gets rid of all the edits made by the agent in a single click if there were mistakes, without messing up your changes from before.<p>- A combined chat+edit flow which you can use to brainstorm and edit<p>You can brainstorm a problem in chat by @’ting the files and then jump into edits (which can happen across multiple files) or go from a smaller set of edits and discuss the side-effects of it<p>- Inline editing widget<p>We took inspiration from the macos spotlight widget and created a similar one inside the editor, you can highlight part of the code, do Cmd+K and just give your instructions freely<p>- Local running AI brain<p>We ship a binary called sidecar which takes care of talking to the LLM providers, preparing the prompts and using the editor for the LLM. All of this is local first and you get full control over the prompts/responses without anything leaking to our end (unless you choose to use your subscription and share the data with us)<p>We spent the last 15 months learning about the internals of VSCode (its a non-trivial codebase) and also powering up our AI game, the framework is also at the top of swebench-lite with 43% score.
On top of this, since the whole AI side of the logic runs locally on your machine you have complete control over the data, from the prompt to the responses and you can use your own API Keys as well (can be any LLM provider) and talk to them directly.<p>There’s still a whole lot to build and we are at 1% of the journey. Right now the editor feels robust and does not break on any of the flows which we aimed to solve for.<p>Let us know if there’s anything else you would like to see us build. We also want to empower extensibility and work together with the community to build the next set of features and set a new milestone of AI native editors.