Podcasts about compose

There is a compose coffee place here in South Korea poster of BTS V

An airhacks.fm conversation with Tomasz Ptak about: discussion about the guest's path from an Atari and a 486 to professional Java development, loading games from cassette tapes, building a clock with the Logo programming language, making websites with PHP for a community, studying data management and computer science, learning Perl, Bash, Pascal, Python, C, C++, Ruby and Java, Java 1.4 and Java 5 with generics and annotations, an island optimization algorithm switching from Python to Java for memory control, preference for strictly typed languages, first job at motorola Solutions building a server-side Java configuration system with SNMP and SNMP4J, moving from Tomcat to Netty, using Ant and Maven, managing a Jenkins server, rebuilding a buggy no-code Spring CRUD generator, rewriting an application with Apache Wicket for stateful web development, comparing Wicket structure coupling with Jakarta Faces, event sourcing with the Axon Framework and domain objects, bitemporal awareness and Hibernate Envers versioning, the Naked Objects pattern and object-oriented UI generation, third job at Open Market stabilizing a legacy Java SMS gateway, weekly outages and same-day retrospectives, containerizing bare-metal systems with Testcontainers and docker Compose, near zero-downtime deployment with Ansible, migrating from Maven to Gradle and removing the Buck build tool, upgrading legacy systems from Java 1.4 to Java 8, minimalistic Maven usage, a zero-dependency Java builder zb and zero-dependency unit runner zunit using only built-in compiler and jar tools, Java 25 as an automation tool replacing Python scripts, executable JARs without external dependencies, shebang instance-method scripting, reactive or infinite streams and stream gatherers, Git-tag-based versioning for monorepos, the AWS DeepRacer and AWS AI community, the mediocris blog Tomasz Ptak on linkedin: https://www.linkedin.com/in/tomasz-ptak

We've been running a bit of an Agent Cloud series surveying all the top inference/compute/cloud providers, from Databricks to Daytona to Railway and, even further back, E2B, but we're excited to conclude this series returning to Modal, which has just raised a monster $355M Series C.The cloud was built for developers. But agents are now changing that.The old infra stack was designed for a human who could read docs, reason through YAML, and understand dashboards to figure out what they need when something broke. While this was painful for developers, it worked since they could fill in missing context in their heads.However, agents don't have that luxury. Now in this new era of agents, everything has to be tighter.They need a place to write code, run it, inspect the output, change the environment, debug failures, and try again. Fast iteration and feedback loops with all the necessary context are crucial for agents to operate properly. Furthermore, sandboxes are a clear representation of this shift as agents can easily spin up isolated environments. This programmatic infra even extends to research:Two years ago, we were one of the first to cover Modal with CEO Erik Bernhardsson and Alessio designed our favorite LS thumbnail of all time:At the time, Modal was just a teeny little company with a $17M Series A.Today, fresh off their $355M Series C, Modal is one of the clearest examples of the agent cloud future being built in real time: a cloud platform moving past traditional web app assumptions toward the workloads AI actually creates such as elastic inference, sandboxes, GPU burst, post-training, background agents, and infrastructure that agents themselves can operate.In this episode, Modal CTO Akshat Bubna joins swyx and Vibhu to unpack why AI applications don't fit traditional cloud assumptions, why Kubernetes was never designed for bursty compute-heavy workloads, and why Modal is now shifting from developer experience to agent experience.We go deep on Modal's AI infra stack: serverless functions, decorator-based infrastructure, elastic inference for custom models, GPU snapshotting, DeFlash, speculative decoding, Auto Endpoints, sandboxes, persistent storage, networked containers, private IPv6, RDMA, multi-node training, and Modal's capacity pool across 17 cloud providers. Akshat also explains why RL rollouts can require 100,000 sandboxes, why production agents need hard guardrails, why observability may matter more than reading code, and why AI has made infrastructure exciting again.We discuss:* Why Kubernetes wasn't built for bursty AI workloads* How Modal started as a better runtime before becoming an AI cloud* Why Modal added GPUs before ChatGPT* The shift from developer experience to agent experience* Why observability matters when agents are writing the code* Elastic inference for custom models across audio, video, robotics, and comp bio* GPU snapshotting, cold starts, and why inference workloads are so bursty* Why RL rollouts can require 100,000 sandboxes* DeFlash, speculative decoding, and frontier-level inference performance* Auto Endpoints and making optimized inference easier to deploy* What Modal adds beyond vLLM, SGLang, and raw GPU rental* Modal's 17-cloud capacity pool and supercloud strategy* Networked sandboxes, sidecars, private IPv6, and RDMA* Serverless multi-node training for post-training and research workloads* Auto-research, model-guided sweeps, and agents launching GPU experiments* Compute strategy, capacity planning, and batch tiers* Why production agents need specialized sandboxes and hard guardrails* Modal's take on managed agents, CI, Gitpod/Ona, Python, TypeScript, and Modal BenchAkshat Bubna* LinkedIn: https://www.linkedin.com/in/akshat-bubna-188885103* X: https://x.com/akshat_bModal* Website: https://modal.comTimestamps00:00:00 Introduction00:00:39 Modal's origin and why Kubernetes wasn't enough00:04:32 Developer Experience → Agent Experience00:06:21 Modal's AI cloud primitives00:09:14 Sandboxes, agent loops, and proto-Cognition00:12:12 Elastic inference, GPU snapshotting, and 100,000 sandboxes00:15:24 DeFlash, speculative decoding, and Auto Endpoints00:19:59 Production-grade inference beyond raw GPUs00:22:00 Background agents, Ramp Inspect, and the agent lifecycle00:24:08 Modal's 17-cloud supercloud strategy00:26:40 Networked sandboxes, private IPv6, and RDMA00:32:48 Multi-node training, post-training, and auto research00:37:36 Compute strategy, capacity planning, and batch tiers00:40:55 Open models, real-time AI, and production agent infra00:43:06 Hard guardrails, managed agents, and specialized sandboxes00:46:06 Why AI made infrastructure exciting again00:48:30 Model APIs, differentiated products, and agentic video00:51:50 CI, coding-agent infra, SDKs, and Modal Bench00:57:28 Closing ThoughtsTranscriptIntroduction: Modal, Series C, and the Art PartySwyx [00:00:00]: We're here with Akshat, CTO of Modal, together with Vibhu. Congrats on your Series C.Akshat [00:00:10]: Thank you.Swyx [00:00:11]: Your party yesterday was amazing.Akshat [00:00:15]: Yeah.Swyx [00:00:15]: From all the photos and all the swag.Akshat [00:00:17]: We had a bunch of art installations, which was fun, seeing, like, our products on pedestals next to, like, Rodin.Swyx [00:00:25]: Very nice. Very nice. When you started, it was not the GPU inference company. Maybe it was in your mind. Take us back to the origin story.Modal's Origin: A New Runtime Beyond KubernetesAkshat [00:00:39]: I first met Eric, who's the CEO, through an investor. Back then Eric was already thinking about building, a new runtime, and he got there thinking through why are workflow orchestration products so hard to use. It's because you have to run them on Kubernetes. Kubernetes is hard to manage. It's not built for burstiness and, custom images,Swyx [00:01:03]: YeahAkshat [00:01:03]: It has a terrible developer experience.Swyx [00:01:05]: And I'll, I'll interjectAkshat [00:01:06]: YeahSwyx [00:01:07]: For listeners, who are new, we interviewed Eric two years ago, and there's a bit more of the story there from Spotify and all those things.Swyx [00:01:14]: And I came across Eric through Data Council because he did that talk on the serverless container stack that you guys did, which was like, that was my first like, “Okay, I need to take Modal very seriously” moment.Akshat [00:01:26]: Yeah.Swyx [00:01:26]: But it was still very unclear, like, do I need all this for just my data pipelines?Akshat [00:01:33]: Yeah. initially what we were thinking about was if we build a better runtime, it's a very useful primitive in itself. It's There's a lot of things that, get solved by serverless functions, like you can do, ETL stuff, you can do job queues, you can do all this, like, bursty processing, which it turns out every company had needs for. but then we also were thinking about this as like, this is a primitive that we can build a whole collection of products on, which are very verticalized. So perhaps data engineering would've been the first one, but we were thinking about inference. Back then it was more classical inference, like computer vision stuff and running XGBoosts and whatnot. But we added GPUs to the product a year before ChatGPT came out.From Serverless Containers to GPU WorkloadsSwyx [00:02:19]: Nice.Akshat [00:02:19]: We just didn't think it would be that big of a deal.Swyx [00:02:22]: Yeah, just like add A100.Vibhu [00:02:23]: Was there any, like, early key problem that really sparked off why you built it?Akshat [00:02:28]: Yeah. Primarily it's just, none of the tooling that was out there was built for, one, a really great developer experience, and also there's a general trend of, a lot of the workloads that we were seeing were very. I wish there was a better word for it, but compute-heavy. Like, they need, one, like, need a lot more resources, so you need to burst up and down a lot, versus like Kubernetes designed for, like, slow scaling and, more for, like, web server use cases. And also there's just a lot more specialization in, like, what kinds of environments these workloads run in. Like, we had sometimes they need accelerators, sometimes they need different kinds of images, and this is just like a consistent thing that we saw across a lot of companies. That would be the next step.Software-Defined Infrastructure and Decorator-Based DXSwyx [00:03:13]: Yeah. Yeah. Be nice. I don't know how much this factored into the early story, but I wrote a post when I was at Temporal about infrastructure, software-defined infrastructure or something like that.Akshat [00:03:22]: Yeah, the self-provisioningSwyx [00:03:23]: Self-provisioning.Akshat [00:03:24]: Yeah.Swyx [00:03:24]: Yeah. I can't even remember my own post.Swyx [00:03:26]: And then you put me on the landing page.Akshat [00:03:28]: Yeah. We really like, the term and so we stole it.Swyx [00:03:32]: Because you had the insight that everything can just be in decorators co-located with the code, right?Akshat [00:03:37]: Yeah.Swyx [00:03:37]: Was that a big part of the originalAkshat [00:03:39]: YesSwyx [00:03:39]: Story or it was just like a DX layer?Akshat [00:03:41]: That was, really important because we really didn't want people to spend, so much time, writing YAML, and it seemed like you could really condense the surface area of what you're doing, put it in code so you can operate on it just like you operate on other code, and like build stuff that's more expressive and dynamic. and so yeah, that was always a very important part.Swyx [00:04:04]: Then the pushback is this is a DSL.Akshat [00:04:07]: Yeah.Swyx [00:04:07]: It's you're closed source. I am locked into Modal.Akshat [00:04:11]: Yeah. We never really got pushback for that because the nice thing about Modal is you can bring whatever code you have, and sure, the DSL is at the configuration layer for, what hardware you're using, how you're scaling things up, but you still own the code.Akshat [00:04:27]: And that's, that's been an important, part of our story, even as we do inference now.Swyx [00:04:32]: Yeah.Vibhu [00:04:32]: How much of do you think still stays the same today? Like if you were to build something today, DevX very important, but I feel like, a lot of this has been changed with just hook it up to an agent, have Claude Code, have Codex implement a tool. there's very agent native primitives that are different than if I'm doing this myself, right?Developer Experience → Agent ExperienceAkshat [00:04:54]: We've changed our SDK team to think about agent experience instead of, developer experience and we think that the same benefits that apply for DX also apply for AX, which is why would you have an agent read through hundreds of Kubernetes files and like write YAML that's not even typed when it can make a couple of changes in a decorator and it gets this self-provisioning runtime of, being able to see its changes live in action? yeah, it just seems from the customers we talk to, they find Modal is much faster for agents to use versus operating on a different substrate.Swyx [00:05:34]: Yeah, because like you, again, you co-locate the infrastructure requirements to the code that runs it.Akshat [00:05:38]: Yeah.Swyx [00:05:38]: Well, the negative thesis now is that nobody's looking at their code anymore, so there's no point.Akshat [00:05:44]: Yeah, people aren't looking at code. one thing we still see is really important is observability.Swyx [00:05:51]: Yeah.Akshat [00:05:51]: Like how good is your dashboard? And of course, like we have, we push a lot of it to the CLI so the agents can do their own investigation, but you still need humans to go interpret what's going on and, make judgment calls and whatnot. and that's I feel like, Maybe more important now than looking at the code itself.Swyx [00:06:11]: Yes, because like, you can try to treat the code as a black box and then use, see the observable action that comes out of it, and then just prompt a change.What Modal Is For: AI Cloud PrimitivesAkshat [00:06:21]: Yeah.Swyx [00:06:22]: So I think it takes a bit of restraint to not specialize, to say, “I want to ship a new primitive,” and then just be general purpose.Swyx [00:06:31]: People ask you, “What are you for?” You're like, “ I don't know. We can do this, we can do that.”Vibhu [00:06:36]: Well, I'd be curious to see, like, okay, if we were to ask you, like, what is Modal for even at a high level? There's a lot you guys do, sandboxes, GPUs, everything. How do you answer?Akshat [00:06:46]: Modal is a cloud platform that's built for, where we've built the primitives from scratch for AI applications. and right now it covers, inference, training, batch processing, and sandbox workloads.Akshat [00:07:00]: But we're building a lot moreSwyx [00:07:02]: I noticed you didn't say web server, so there is still a role for, like, the always-on large-scale Kubernetes type things.Akshat [00:07:09]: Yeah, absolutely. We're, we're not trying to compete with the renders of the world, because yeah, we think the differentiator for us is the, are the workloads that need specialized compute, need to scale up and down a lot. yeah, they're, they're, they're just shaped differently.Working Alongside Frontier StartupsVibhu [00:07:26]: I think you're building a lot of it alongside the startups, right? They're innovating quite a bit, even in your, like, latest blog post. Like, even in the series C, the customers that you mention here, the cognitions, technical ones, ramps and whatnot, they're, they're innovating with you, right? And that's not something AWS is doing directly with.Akshat [00:07:45]: Yeah, absolutely. I think, this is again classic. We're a small team. We can move really fast. our engineers are working with our customers and figuring it out. Yeah.Swyx [00:07:54]: So my first week at Cognition, I walked in, there was someone wearing a Modal shirt. I was like, “What are you doing here?” They're like, “Yeah, I just. I am embedded inside of Cog.”Akshat [00:08:05]: Yeah, I think that was Peyton. We sent him overSwyx [00:08:07]: Yeah.Akshat [00:08:07]: Because, the latency of communication was too high otherwise.Swyx [00:08:12]: Yeah, distributed node, you have to - you have to place one and collocate.Vibhu [00:08:16]: Yeah.Swyx [00:08:16]: So I had a, I had direct personal experience, right? So I worked on smol developer three years ago. it was inspired by Claude 1. I think you onboarded me at some point, like, just before, and I was like, “Oh, like, I need some bursty compute. Like, I was just gonna try using Modal.” And it was a, it was a pretty pleasant experience. apparently, I showed up in the board meeting, like the analytics.smol developer, Sandboxes, and Proto-CognitionAkshat [00:08:39]: Yeah, you blew up on Hacker News and,Swyx [00:08:41]: YeahAkshat [00:08:41]: We got a big traffic spike. I. I think the way you used smol developer was Modal functions for running stuff, which was. Like, the, that was a good use case. but then, yeah.Swyx [00:08:53]: Yeah. That - So to me, that was proto-cognition.Akshat [00:08:55]: Right.Swyx [00:08:56]: If only I had, like, stuck to it.Swyx [00:08:58]: Like, that was like, if - did you say draw the tech treeAkshat [00:09:00]: AbsolutelySwyx [00:09:00]: You're just like, “Yeah, like, probably this will happen.”Akshat [00:09:02]: Yeah. Like, he was so close. You were just rebuilding upon usSwyx [00:09:04]: I just didn't realize.Akshat [00:09:05]: But the funny story there is at the same time, we were talking to a bunch of customers who needed something like sandboxing.Swyx [00:09:14]: Yeah.Akshat [00:09:14]: This is like twenty-three.Swyx [00:09:15]: Yeah.Akshat [00:09:16]: So we builtSwyx [00:09:17]: You introduced a new API right after that.Akshat [00:09:18]: Yeah.Swyx [00:09:19]: Yes.Akshat [00:09:19]: Like, we built sandboxes in May of twenty-three before anyone was even knew this was gonna be a thing. And the first example we published was, we took smol developerSwyx [00:09:28]: Smol developerAkshat [00:09:28]: And put it in a loop, so the agent can iterate on itself.Swyx [00:09:33]: Loops are hot these days.Vibhu [00:09:34]: It's the looper.Akshat [00:09:34]: Yeah.Vibhu [00:09:35]: Loops in. When was this, twenty-three?Akshat [00:09:38]: Yeah.Vibhu [00:09:39]: A small check.Akshat [00:09:39]: Yeah.Swyx [00:09:39]: It's like twenty-three. so the. the, those for listeners, like, the problem was the models are not built for any of this, right?Swyx [00:09:46]: Like, you're just trying to like. They're not post-training to understand, like, looping and, like, self-correction and tool calling was there, but, like, also not that great.Akshat [00:09:55]: Yeah.Akshat [00:09:55]: I don't remember if you used tool calling in this one, but yeah, the models would just diverge after like ten iterations and not produce anything meaningful.Swyx [00:10:03]: Yeah. But like, then. So okay, like now talking to myself three years ago, the answerVibhu [00:10:08]: Of course they will get betterSwyx [00:10:09]: Collect all the failures, build benchmark, and then collect all the, examples, build the RL environmentAkshat [00:10:15]: RightSwyx [00:10:15]: Sell it for like ten billion dollars to Meta.Swyx [00:10:17]: And then also train a model and then sell that for sixty billion dollars to Elon. And this isAkshat [00:10:23]: Yeah, of courseSwyx [00:10:23]: The funny machine. Like, it's like, it's about the hardware.Akshat [00:10:28]: It's hard to have that inherent conviction that the stuff will get that much better.Swyx [00:10:33]: In retrospect, it's so f*****g obvious.Akshat [00:10:36]: Fair enough.Swyx [00:10:37]: Like, what else were we doing back then? I don't know. anyway. Yeah. So this. That was the start of your sandboxing journey, right? I feel like it didn't blow up until, like, last year.Akshat [00:10:49]: Yeah.Swyx [00:10:50]: So there was like a couple years of quietness.Akshat [00:10:52]: Exactly, yeah. We wereVibhu [00:10:53]: I think very underrated product value. Like, my experience with Modal, Charles, before he had joined Modal, met this guy at a hackathon, and he really insisted we wanted to run some small model, not hosted anywhere, and he's like, “ there's this cool company, Modal. They'll like spin up a GPU sandbox, we can throw it on there. They'll take a Hugging Face link.” And like there's so much value just right there, right? Like instant hosting, spin it up, spin it down. It'll stay cold, but we run the demo a few days later, it'll come back up and like all this stuff in retrospect, like it's still what we needed like today.Akshat [00:11:27]: Yeah, it's still needed today. workload shapes have changed a lot as, we run stuff for people with really massive production scale and, there it's it's not about scaling from zero to one, but it's how do we scale really elastically, from like thousand to fifteen hundred GPUs very quickly in a given region. It's the same shape problem.Elastic Inference, GPU Autoscaling, and Custom ModelsVibhu [00:11:50]: Okay. So you look at, say, Cursor Composer, right?Akshat [00:11:53]: Yeah.Vibhu [00:11:53]: They had a. “We'll do RL on a model every couple hours.” you guys have a whole version of RL inference gym and whatnot.Vibhu [00:12:01]: When you look at workloads like that, you're doing train runs where you need to scale up, scale down every hour thousands of GPUs, right? That's the example for we do need it, right?Akshat [00:12:12]: Yeah. Well, so I'll, I'll take a step back and, maybe talk about like how people use Modal today. because our biggest use case is, elastic inference. And the thing we first found product market fit, with was inference for custom models. So we stayed away from the LLM space, and we were serving companies like Suno for audio, Runway for video, robotics, comp bio companies that train their own model elsewhere. But Modal is the best black box that for deployment, scaling to however many GPUs you need as your traffic pattern changes. And we saw all of them like have a very unpredict- predict- predictable, traffic pattern. it's like diurnal. It's Some days, like the company will do a launch and, they'll need like, way more. And it's not just one model that they deploy. They-- all these companies deploy, lots of different models in different regions, and so the autoscaling problem becomes even harder because then you have to scale within a certain region, and those cycles are offset. So different times you scale up in different regions.Akshat [00:13:20]: So that's like our sortVibhu [00:13:22]: And thatAkshat [00:13:22]: YeahVibhu [00:13:22]: That in and of itself is a huge category. There's a bunch of inference providers which, provide this fireworks, does this as a service together, whatnot, Base10. that's carved into its own niche for language models, at least right now.Akshat [00:13:36]: Yeah. the thing that we have specialized in is the autoscaling aspect.Vibhu [00:13:41]: Yeah.Akshat [00:13:41]: Because we found that it's not universally true that everyone else can autoscale, and we've gone deeper into it on the tech side by, we've incorporated GPU snapshotting into the product so we can take the GPU state, like your torch.compile model, snapshot it, and the next cold start is way faster. And so going back to your question, it's That's why you need a lot of burstiness for inference. But then people also do a lot of demand training, like for RL stuff, your rollouts are bursty, as you said. People also do a lot of batch jobs. So we'll see, a lot of companies, before they have a training run, they'll need thousands of GPUs to run encoding or something like that. And I think those things are much more bursty than. I agree that agents are not that bursty. sandboxes are, except when you're doing RL. RL is justRL, Batch Jobs, and 100,000 SandboxesVibhu [00:14:28]: Or commerceAkshat [00:14:28]: Insanely bursty.Vibhu [00:14:29]: Yeah.Akshat [00:14:30]: Yeah. Like when you're doing, rollouts, you sometimes need a hundred thousand sandboxes in your sandboxes.Vibhu [00:14:37]: Yeah. I'm curious if you've seen early sparks of continual learning. There are some people, like our friends, ngram, recently announced thisAkshat [00:14:45]: YeahVibhu [00:14:45]: They're, they're trying to do training. That also seems like a different workload, right? If you're doing training twenty-four/seven per se, there's a very weird dynamic of how you're using GPUs between people and whatnot, but seems like something you guys would work for.Akshat [00:15:00]: As you said, we're, we're fortunate to work with a number of, customers at the frontier and grab some of our customers. and they are taking the primitives we have, and trying to use them in very interesting ways, like continual learning. It's possible as the stuff gets better, some of that will be part of, our offering as well if, more people need it. but we're, we're just waiting to seeVibhu [00:15:23]: YeahAkshat [00:15:23]: How it shakes out.Vibhu [00:15:24]: Is there a primitive that you added after sandboxing that was the next step in the story?LLM Inference, DeFlash, and Speculative DecodingAkshat [00:15:32]: I guess we've been going much deeper into LLM inferenceVibhu [00:15:35]: YeahAkshat [00:15:35]: Because we realized that some of the advantages we have with like autoscaling, again, especially in different regions and whatnot, are, not present elsewhere. and the place where we had a gap was we weren't, working on the model layer itself. Like we were a black box. And, we realized that, we can get to frontier-level model performance, with, by having great people who work on this. And, we've been open sourcing a lot of our work, in terms of, Recently, we, shared our work on DeFlash, which is a block-based, speculator, and we've open sourced, all of it. So, you can - By using open source DeFlash, you can get the same performance as you would with one of the proprietary providers. And the next thing we're thinking about hereVibhu [00:16:23]: I thought this wasAkshat [00:16:24]: YeahVibhu [00:16:24]: An interesting blog post as well, right? Like, I think in here you make a claim that. Not a claim, just that how effective speculative deco-decoding really just get to.Akshat [00:16:33]: Yeah.Vibhu [00:16:33]: Anything you wanna point out from this around, what people should know?Akshat [00:16:39]: Yeah, absolutely. the high-level summary is, it would help to describe what speculative decoding is.Vibhu [00:16:44]: Yes.Akshat [00:16:44]: I will, yes.Vibhu [00:16:45]: I think, likeAkshat [00:16:46]: YeahVibhu [00:16:46]: So we've covered like Eagle and all thisAkshat [00:16:47]: YeahVibhu [00:16:47]: Like Hydra and all those things, but it was like two years ago.Akshat [00:16:51]: Yeah.Vibhu [00:16:51]: I think it doesn't hurt, right?Akshat [00:16:52]: Yeah. Speculative decoding is you have a smaller model, called a draft model, predict tokens ahead of the bigger model, and then you have the bigger model, verify all of this, all the tokens are predicted. And the reason it's faster is if you're predicting, one token at once, you're bound by memory bandwidth. But if you can batch the verification of, the draft model, then you're much more efficient using compute, and it's faster, and as long as your draft model is producing a lot of tokens that can get accepted, which is called the accept length, you can get a speed up that's, multiple times of, the original model speed. and well, that's what we highlight here. It's Like people talk a lot about we made these kernels faster and whatnot, but improving kernel will only give you like few percentage points of improvement, and, increasing accept length, literally is a multiplicative decreaseVibhu [00:17:47]: Like two to four X.Akshat [00:17:48]: Yeah, exactly.Vibhu [00:17:48]: Without much head-on performance.Akshat [00:17:50]: Yeah. I think it may - you are running a second model, right? So it may be something more expensive in the compute,Vibhu [00:17:57]: I meant quality performanceAkshat [00:17:58]: Probably not by muchVibhu [00:17:58]: But yeah. I thinkAkshat [00:17:59]: So there's no drop in quality performanceVibhu [00:18:01]: YeahAkshat [00:18:01]: Because you're always. You're never accepting a token that the big modelVibhu [00:18:04]: It's strictly betterAkshat [00:18:05]: YeahVibhu [00:18:05]: Or it's same.Akshat [00:18:06]: Exactly.Vibhu [00:18:07]: Right. Yeah.Akshat [00:18:08]: And so we've been working a bunch on DeFlash, which is a block-based speculator. so it's instead of predicting, one token at a time, it's predicting a block. And we've been open sourcing our work with it. The next thing for us here is for helping people train speculators and custom models. it's it's something that traditionally is very forward-deployed engineering driven, support deployed, engineer driven, like you work with customers and help them do that. And our vision for. This is why we launched Auto Endpoints, is we want to make frontier-level performance available to everyone. And so, we mentioned this in the announcement, we teased it. The next thing we're, we're launching is, as you run an auto endpoint, we shadow trafficAuto Endpoints and Frontier-Level PerformanceVibhu [00:18:54]: Do you want to explain what auto endpoints are?Akshat [00:18:57]: Yeah.Vibhu [00:18:57]: I lovely, yeah.Akshat [00:18:58]: Yeah. So, this is, I guess, going back to your Modal is you touch the code, but, sometimes people don't wanna touch the code, and they wanna get started with an endpoint that works and has all the great performance and, scalability that Modal has. So we've made that easier with, a way to create an endpoint from our UI, from the CLI, that has all of our optimizations that we talked about, like the DeFlash stuff already baked in, and there's full transparency. So we give you the code, you can go run it yourself, and if you want, you can eject out into the full Modal experience, which we see as people get sophisticated, they do wanna tweak the models, they wanna, fine-tune stuff. You can still do all of that. It's it's not a black box. And yeah, the next thing, as we teased later in the post, is how do we give you value even beyond this in terms of having your draft models evolve as your data distribution evolves, again, without having to talk to a person and, yeah.Vibhu [00:19:59]: I guess just to understand it directly, you have the GPUs, you have an endpoint that's compatible, you serve open model. If someone was to do this themselves, what's the delta that you guys provide? So you do a lot of open source great work on effective inference. how does it compare to, say, I take the same model, 5.2 FP8, take shelf inference engine, vLLM, SGLang, get compute of similar capacity, similar cost. What's the delta that plugging into something this, like this offers outside of the benefit of, scaling?Production Inference Beyond Raw GPUsAkshat [00:20:34]: It's interesting because we've taken the approach of open sourcing our contributions and upstreaming them. we work closely with the SGLang team. We want the improvements that our team, comes up with to be, there in open source for others to use, even outside of Modal. The benefit to us is we have a team that has significant expertise in terms of if you do have something that is not there, our team can help you get that performance, first. the other thing is with these endpoints, we are way more elastic, as you said, than, anyone else, and you have true scaling to zero. you have true, burstiness, and in practice, that matters a lot more to people than just finding, the GPU and, running Modal code on something.Vibhu [00:21:20]: Yeah. And I will say it's not that straightforward to just. like what I said is easier said than done, right?Akshat [00:21:26]: Yeah.Vibhu [00:21:27]: It's I think still for the average person, still hard to just gut check using different. There's, there's quite a bit of combinations you can make there. the trade-offs aren't really known at face value.Akshat [00:21:40]: Yeah. it's it's not just that. I think it's it's that running production-grade inference is a hard infer problem.Vibhu [00:21:49]: YeahAkshat [00:21:49]: Even if you subtract out the autoscalingVibhu [00:21:50]: YeahAkshat [00:21:51]: Is controlling things like tail latency and, making sure every, request is delivered at least once and whatnot.The Model and Agent LifecycleVibhu [00:22:00]: There's a lot of innovation that you can do here. I think, it's very interesting that you're starting to encroach on, like as you become a full cloud, you're starting to encroach on other people's turf.Vibhu [00:22:09]: What will you not do?Akshat [00:22:13]: Well, we wanna follow our users and, make sure they get like a platform that has everything that works well together. so right now we're focused on the model lifecycle and the agent, lifecycle. so both like going from data prep to training to inference, and then also if I want to deploy a background agent, let's say, sandbox, do persistent storage, a whole bunch of other stuff.Vibhu [00:22:38]: We talked to Cole, who did, OpenInspect. Yeah.Akshat [00:22:42]: Yeah.Vibhu [00:22:42]: And RealInspect also is on Modal.Akshat [00:22:44]: Yeah. So Ramp Inspect was a great example of a background agent that was really successful because they, were able to use some of the primitives like snapshotting and fast scaling to just have something that feels really reactive and works well.Ramp Inspect and Background AgentsVibhu [00:23:02]: Yeah. That's the new CTO of, Ramp right there.Akshat [00:23:05]: Yeah, Rahul.Vibhu [00:23:08]: It was really fun. yeah, okay, I think, all very bullish. Like, one of my reflections was also I did not originally. So when I met you guysThe Inference Inflection: CPU, GPU, and Co-LocationVibhu [00:23:19]: You weren't that much in the GPU game, and now you're all about, inference. And one of the points that I hinged on for Jensen's keynote at GTC this year was, what we're calling like the inference inflection, right? That let's say in AI workloads or machine learning workloads, it used to be like, let's call it eight to one GPU to CPU, and now it's more like one to one, which is like a interesting. Like, - because of how much agents are blocked or call out to this, to CPU heavy stuff the actual, like, limiting factor, like, swings back and forth from GPU to CPU a lot more than it used to be all GPU and then occasional CPU.Akshat [00:24:01]: Yeah.Vibhu [00:24:02]: GPU, CPU. And now it's like just constantly, and you just have to locate everything.Seventeen Clouds and the Supercloud StrategyAkshat [00:24:08]: Yeah. And that's one of the things that, again, we see as, something appealing about Modal, which is we've built this capacity pool that spans, 17 cloud providers, so we're, we're very good at Running on various kinds of cloud capacity across the worldSwyx [00:24:24]: You don't have your own data centers?Akshat [00:24:25]: We don't have our own data centers. We just run across a lot of neo cloudsSwyx [00:24:29]: Yeah. AreAkshat [00:24:30]: Metal providers.Swyx [00:24:30]: Yeah. Question mark.Swyx [00:24:31]: Yeah. You're, you're running the math, and you're like, “What's the cutover point where you're like.”Akshat [00:24:36]: Yeah, it's a good question. part of it is we see our differentiator in the software layer, and, being capital light and focusing on the software helps us move really fast. so far it's worked out well because there are so many other people building data centers that we're able to work effectively with them, and again, focus on what makes us, special.Swyx [00:24:55]: Yeah.Swyx [00:24:56]: 17 gets you into, like, the local providers sometimes. LikeAkshat [00:25:00]: The,Swyx [00:25:01]: Which was the most interesting one?Akshat [00:25:02]: There are a lot more neo clouds than you expect, and they all have various degrees of, various levels of reliability. And, that's why it's something we've invested a lot of time in, is building our own reliability layer on top. so if the GPU falls off the bus or something happens, we user workloads are not affected, and that lets us use a lot more capacity than,Swyx [00:25:30]: YeahAkshat [00:25:30]: You as a user would be able to.Swyx [00:25:32]: It's a useful thing to have because like now everyone knows, like, what layer you are and, like, you optimize for being the super cloud of all clouds.Akshat [00:25:41]: Yeah. That's, that's, that's the idea. and so I guess when you mentioned colocation, that's, that's another interesting thing where, one thing we've seen is people come to us when they want, very specifically located, CPUs or GPUs, like they wantSwyx [00:25:57]: Oh, they pin it in likeAkshat [00:25:58]: YeahSwyx [00:25:58]: EU?Akshat [00:25:59]: Exactly. Or EU, US.Swyx [00:26:01]: Right. Data resiliencyAkshat [00:26:02]: AustraliaSwyx [00:26:02]: Locality thing or performance or what?Akshat [00:26:04]: It's either data locality or latency, yeah.Swyx [00:26:07]: Yeah.Akshat [00:26:07]: Like, you want your. They're running sandboxes and model. They want them to be right next to aSwyx [00:26:10]: Yeah, it's easy thenAkshat [00:26:11]: YeahSwyx [00:26:12]: To. That is important in all those things. and so, like, you've accidentally, I don't know if it's accident, but, like, you've built the perfect primitive for agents to express themselves. And then, like, it's almost very funny how every extra development just involves more file system, just involves more CPU.Akshat [00:26:30]: Yeah.Swyx [00:26:31]: Just like the things that you already have. I don't know much about, if there's any, like, networking usages that are interesting, but you've also done some good work on networking.Networking, Sidecars, Private IPv6, and SandboxesAkshat [00:26:40]: Yeah, that's exactly right. Like, we're just taking compute storage and networking and building stuff on that layer, for, again, the stuff people need.Swyx [00:26:49]: YeahAkshat [00:26:50]: We see a few interesting networking things coming up. one is people want networked sandboxes. so we haveSwyx [00:26:57]: For like a Docker cluster type thing.Akshat [00:26:59]: Yeah.Swyx [00:26:59]: Sorry, Docker Swarm. Oh, f**k. What is it called?Akshat [00:27:02]: Compose.Swyx [00:27:03]: Compose type thing.Akshat [00:27:04]: Yeah. So if you want Docker Compose, our sandboxes now support, this thing called sidecars. So you can. A sandbox is a pod of containers, and you can run multiple containers in, a sandbox. also useful because, going back to networking, people want a lot of control over, outbound networking from a sandbox.Swyx [00:27:23]: Yeah.Akshat [00:27:23]: Like, they might wanna run a middle proxy for, like, maybe logging stuff for RL or, controlling how egress can happen to a domain, injecting credentials. and yeah. So we've, we've had to build a lot of that stuff ourselves.Swyx [00:27:38]: Yeah.Akshat [00:27:39]: But then also sometimes people want, sandboxes spanning multiple nodes to talk to each other, which is an emerging thing we're seeing. We have support for that for a different reason, and yeah, we'll see if that becomes stable.Swyx [00:27:52]: Like, just an open socket. It's a. This is directly like mTLS.Akshat [00:27:56]: We do support that, which is you can, expose a tunnel inside a sandbox.Swyx [00:28:01]: Yeah.Akshat [00:28:01]: And then you can either expose it to public internet or it can be, you can add like a HTTP, auth layer above it. But we have this thing called I6PN, which we haven't talked about, which is this, like, overlay network using IPv6 addresses. so if Modal containers, within the same workspace, when this is enabled, can address each other using this private IPv6 address, and no one else can.Akshat [00:28:28]: So it's like private networking, for containers. We built it because we needed it as a primitive for our distributed training product. so we have this other feature, which is you can add a decorator to a function, and you get a cluster of GPUs. and they have RDMA networking. so you can run a distributed training job, that's truly serverless. and we did the overlay network for that. But then we've seen that people are using it for other reasons, and, I'm intrigued to yeah, what would people do with it.Swyx [00:28:59]: Build primitives and let people figure it out, right?Akshat [00:29:01]: Yeah, exactly.Swyx [00:29:02]: You put out a pretty interestingAkshat [00:29:03]: They're like, they read the docs webpage. Let me use thatSwyx [00:29:06]: YeahAkshat [00:29:06]: Something they never intended to work. This is literally not even in our docs page. People somehow found it, and they're using it.RDMA, Memory Movement, and Distributed TrainingSwyx [00:29:12]: Huh.Swyx [00:29:14]: The way you portrayed it with, like, RDMA versus TCP, like, very well laid out, but just the transfer speed change at scale for RL, like yeah, you have it, you have it built in. I'm sure someone found it. It's found it to be a lot more efficient before you made a thing out of it, right?Akshat [00:29:32]: Yeah. And not to split hairs, I guess the overlay network is the TCP overlay network.Akshat [00:29:39]: The reason we have that is you need that to do the key exchange for RDMA before you set up the RDMA network on top of that. but then people found the TCP part.Swyx [00:29:48]: Can I tell you, this is like a big aha moment for me becauseAkshat [00:29:51]: YeahSwyx [00:29:51]: So I review 2,200 submissions for the World's Fair.Akshat [00:29:56]: Yeah.Swyx [00:29:57]: And then I got this from John OsterhoutAkshat [00:29:58]: HuhSwyx [00:29:59]: Who I don't know if. Do John Osterhout by name?Akshat [00:30:01]: The name sounds familiar.Swyx [00:30:02]: He published a. He's a well-known professor, published a lot of interesting software design books, and this is the talk he chose to submit, is on RDMA at Inference. And I'm like, you wouldn't think that this guy, who is like operating systems guy, would care about RDMA.Akshat [00:30:20]: I, it makes sense to me because I,Swyx [00:30:24]: This is the cloud, right? YeahAkshat [00:30:25]: Like, the way you move around your KV cache and how efficiently you can do it, how efficiently you move, your weights from your training GPUs to your inference GPUs in RL is there's a lot of degrees of freedom, and it is a systems problemSwyx [00:30:41]: YeahAkshat [00:30:41]: Moving memory aroundSwyx [00:30:42]: YeahAkshat [00:30:43]: Scheduling.Swyx [00:30:44]: This shows you how primitive my understanding of networking stuff is.Swyx [00:30:46]: Is this like the domain of WireGuard as well?Akshat [00:30:50]: Not quite.Swyx [00:30:51]: It's adjacent?Swyx [00:30:53]: Explain everything.Akshat [00:30:54]: Sure.Swyx [00:30:56]: How do we move memory around GPUs?Akshat [00:30:58]: Well, so sorry. Yeah, that is memory. Sorry, I was talking more, and maybe I was talking like five minutes back, about the private IPv6, addressing that you've set up.Swyx [00:31:09]: Yeah.Akshat [00:31:09]: Is it like it's a VPN?Swyx [00:31:10]: Yeah, it is like a VPN, and yeah, WireGuard is, yeah, you're right. It is,Akshat [00:31:16]: Right. Yeah, you already moved on to new topicsSwyx [00:31:17]: A similarAkshat [00:31:18]: OkaySwyx [00:31:19]: In the same space, WireGuard is, encrypted and this is,Akshat [00:31:23]: And you don't need encryption.Swyx [00:31:23]: Yeah.Akshat [00:31:24]: Yeah.Swyx [00:31:24]: This is not encrypted. that's the main difference. This is TCP and we have eBPF programs that will reject or allow the TCP connection based on whether you're allowed to do it.Akshat [00:31:35]: Used to involve a full sidecar, but now you have eBPF in the Linux kernel.Swyx [00:31:39]: Yeah.Akshat [00:31:40]: Yeah. I don't know if this is a natural follow-on to the topic of like my skepticism on distributed training is that while, like, people spend a lot of money on, like, cables to hook up GPUs, and even that is not, like, fast enough, and that's the bottleneck, is your networking fast enough?Swyx [00:31:59]: Yeah. So I guess you're talking about fully distributed training like, Dialog or something which is like cross data centerAkshat [00:32:06]: That would be, yes.Swyx [00:32:07]: That's the extreme.Akshat [00:32:08]: Yeah.Swyx [00:32:08]: You're in the middle, and then other people would have like the Mellanox cables up in, like, their actual data center.Akshat [00:32:14]: When you run multi-node training on Modal, RDMA, I think Mellanox, is, or InfiniBand is like a, is all seen as RDMA. but it's a way to bypass the TCP networking stack and, transfer, stuff much faster, between one node, to the other. And we have I think like 3 terabit per second, internal networkingSwyx [00:32:40]: OkayAkshat [00:32:40]: Which is the standard that's needed.Swyx [00:32:42]: Okay. So I misunderstood whatAkshat [00:32:43]: 50Swyx [00:32:43]: What part of the stack you wereAkshat [00:32:44]: 50 gigs overSwyx [00:32:45]: YeahAkshat [00:32:45]: If you wentSwyx [00:32:45]: YeahAkshat [00:32:46]: RDMA.Swyx [00:32:46]: Okay.Swyx [00:32:48]: Yeah. I, very impressive work.Multi-Node Training, Post-Training, and Auto ResearchSwyx [00:32:52]: So effectively you're extending like the model philosophy to the training cluster, like, yeah.Akshat [00:32:59]: Yeah. And we're, we're not going for like large scale training runs. the thing that we've built multi-node training for is, we see a lot of, smaller scale post-training. like, people are post-training like medium sized fund models, so they can, get higher quality on inference. this is a perfect fit, for something like that.Swyx [00:33:21]: Yeah. That is my impression of how a lot of these labs explore branches in post-training and then eventually merge whatever they find in.Akshat [00:33:31]: Yeah. The other use case we've seen for multi-node training is even if you have a big cluster, your researchers are still doing small runsSwyx [00:33:38]: YesAkshat [00:33:39]: Having elasticity thereSwyx [00:33:40]: Right, sureAkshat [00:33:40]: Matters a lot more.Swyx [00:33:41]: Yeah. the, like, this is like the current limiting factor for auto research, which is like you need to give your model some GPUs in order for it to completely run.Akshat [00:33:51]: We have a blog post on auto resource and model is,Swyx [00:33:55]: YeahAkshat [00:33:56]: Yeah, like, turns out to be pretty good substrate for that.Swyx [00:33:59]: So my impression is auto research means many things, likeAkshat [00:34:01]: YeahSwyx [00:34:01]: Anything that Andrej coins. Right now it's still science fair, right? Like not like, I don't know how many people are doing this.Akshat [00:34:08]: We're having a golf.Swyx [00:34:08]: Yeah.Akshat [00:34:09]: I thought the same thing.Swyx [00:34:11]: Yeah, you would know.Akshat [00:34:12]: We, like, our internal both training and inference teams use this the general shape of this quite a bit. like we have this one internal repo called auto inference, which essentially we've automated our own forward-deployed engineering efforts using, this harness, which is, the agent will just spin up a sweep of different things. It'll even run like, NVIDIA inside profiler and it'll like tweak configs and it'll arrive the right thing. it'll change your GPUs both from H200 to B200, and works really well.Swyx [00:34:47]: Nice.Akshat [00:34:47]: So yeah.Swyx [00:34:48]: By the way, I enjoy that your forward-deployed engineering is so technical that you have to do these things.Swyx [00:34:52]: It's very different from forward-deployed engineering from other people.Akshat [00:34:54]: Yeah. For our forward-deployed engineering team is, essentially they're like applied inference researchers or applied training researchers.Swyx [00:35:02]: Someone told me like they have to be able to build, but they also have to be able to sell. do they have to sell or are they like they're good, they're just like post-sale type of thing?Akshat [00:35:09]: It does, being able to talk to a customer and engage effectively with themSwyx [00:35:13]: YeahAkshat [00:35:13]: Matters a lot.Swyx [00:35:14]: They want the same thing.Akshat [00:35:15]: Yeah.Swyx [00:35:15]: ?Akshat [00:35:15]: But it's it's not really a sales, thing. We pair them with-- We have solution architects as well that are more on the sales side.Swyx [00:35:23]: Okay. Let's spend a bit more time on auto research. This is a big focus for for this year. Where does this go? like, have people explored enough? Like, there's all these beautiful charts of like improve and then level off a bit and then you find the next thing. Is this one abstraction up from normal training? Is that how we think about it, or do you think about it differently? Like model level training versus high, like driven hyperparameter search.Auto Inference and Modal BenchAkshat [00:35:51]: Yeah, like,Swyx [00:35:51]: Someone, some people call it like neural architecture search or whatever, right? Like.Akshat [00:35:54]: Yeah, - So the stuff I've seen people do with it is nowhere on the architecture level. It's pretty much tweaking parameters, but it's it's a hyperparameter sweep that's guided by some model intuition, so it's like much more efficient than, whatever other, sweep you would have.Swyx [00:36:12]: Yeah, it's just, it's just a question of where you want to spend your compute?Akshat [00:36:16]: Right.Swyx [00:36:16]: ‘Cause yeah, you can just throw infinite amounts of money on this and somehow you'll bang out Shakespeare?Akshat [00:36:22]: Yeah, infinite monkey.Swyx [00:36:24]: Yeah, so like the very good for model. and I think it's also very important that agents can spin up other agents, can spin up their infrastructure. Like very good for you. how good is our LLMs at generating model code? Like the benefit of existing LLMs is that you are in the data.Akshat [00:36:42]: Yeah. They're, they're surprisingly good. I think like pre Cloud 4 they were not, and then now they're able to shot, stuff out of the box. But we're playing around with releasing like a Modal Bench for like the harderSwyx [00:36:55]: YeahAkshat [00:36:55]: Things, that the LLMs cannot do yet and maybeSwyx [00:36:59]: What's an example of that?Akshat [00:37:01]: I think the things that- Sometimes agents struggle with, without right guidance and a skill is, how to, use the rest of our observability. Like how to. Something is failing, like how do you look at the logs and then update the right thing? It's reasoning about that. But they're able to shot, likeSwyx [00:37:23]: Yeah. You can just add a skill to it?Compute Strategy and Capacity PlanningAkshat [00:37:26]: Yeah. So we have a Modal skill now that. Which is why we built this Modal Bench. It's to find things like that, so we can address them in our tool.Swyx [00:37:35]: Tune a skill. Yeah.Akshat [00:37:36]: Yeah.Swyx [00:37:36]: No. it's it's good. are you facing any shortages? like we talk a lot about GPU shortages, but also CPU, also memory.Swyx [00:37:44]: Yeah.Akshat [00:37:45]: We have had a lot of growth, which means that, there's - we've had to be much better aboutSwyx [00:37:53]: PlanningAkshat [00:37:54]: Proactive capacity planning.Swyx [00:37:55]: Yeah.Akshat [00:37:55]: So we have,Swyx [00:37:57]: Which by the way, like it's like a MBA's like dreamAkshat [00:38:00]: YesSwyx [00:38:00]: Is like just planning this stuff. I think last time you and I talked about something maybe about this.Akshat [00:38:03]: Yeah. we have a really competent team of people that we call, The role is called compute strategy. so yeah, if anyone listening here or wants to work on thatSwyx [00:38:13]: Compute strategy?Akshat [00:38:13]: Yeah.Swyx [00:38:14]: I think,Akshat [00:38:14]: I feel like,Swyx [00:38:15]: I think the normies call it FP&A or something.Akshat [00:38:18]: Well, it's more It's it's not FP&A. It's it's There's a lot of interesting financial questions of like what is the blend between one year and three-year reservations? how do we forecast our own capacity? how do we. especially since our capacity is very fungible across different GPU types and different regions, like you have to model a lot of it. and you also have to have an opinion on how the supply chain is gonna evolve, and then you have to like, take bets,Swyx [00:38:49]: YeahAkshat [00:38:49]: Based on that.Swyx [00:38:50]: Tokenomics.Akshat [00:38:50]: Yeah.Swyx [00:38:51]: This is like probably a not a real point, but, I was trying to think about like what other industries. I was trying to think about like, we cannot be first to like these kinds of problems.Akshat [00:38:59]: Yeah.Swyx [00:39:00]: And what other industries have had this? And I was like, airlines with fuel and like they have to hedge their fuel and like, I think for a long time Southwest because they made like a hero fuel bet, they like were like super low cost becauseAkshat [00:39:12]: OhSwyx [00:39:12]: Compared to everyone else.Akshat [00:39:14]: Yeah. I hadn't thought about that.Vibhu [00:39:16]: We're at a fun time too?Akshat [00:39:18]: Yeah. It's. A lot of the compute business in general, for us is also about being very good about capacity management. That is how you have great unit, economics. but also over time it's how you can unlock more value for customers. Like, one of the things we're building now is like a way for customers to get, If they don't care about latency, like get much cheaper pricing and they'll get results back in like next 24 hours or something, like a batch tier essentially.Batch Tiers and Latency-Insensitive WorkloadsSwyx [00:39:47]: Yeah.Akshat [00:39:47]: And those are levers we have because we control the whole stack and scheduling and whatnot to give people a sufficientSwyx [00:39:53]: Yeah. I feel like they're not as popular. Like those, like the Frontier Labs have all those APIs. They're not as popular as they should be.Akshat [00:40:00]: The demand that we see for something like that is not for LLMs. although sometimes people wanna run evals andSwyx [00:40:08]: OkayAkshat [00:40:08]: Synthetic data prep and there it makes sense.Swyx [00:40:10]: Okay.Akshat [00:40:11]: But it's from a lot of LLM companies, like people who are doing computational bio, like they have to run really big batch jobs and they don't care about when they get it back.Swyx [00:40:22]: Yeah. And like they have a reasonable. It's it's also like a cousin to the stopping problem of like, will this finish in time?Akshat [00:40:30]: Yeah. You can bound it.Swyx [00:40:33]: Yeah.Akshat [00:40:33]: Like you can give peopleSwyx [00:40:34]: YeahAkshat [00:40:34]: SLAs on it.Swyx [00:40:35]: Yeah. I think what's, what's interesting is like the next phase of model.Swyx [00:40:38]: Like what, do people expect from you, now that you're established and you're like well-known compute player among all these leading companies. You had an inference launch week, and we talked a little bit about the launches. like what else? Like what else should people know?What Modal Builds NextAkshat [00:40:55]: We are building primitives that make our users' lives much easier. So, I think for example, with LLM inference, thousands more companies are gonna post-train their own models and, deploy open source models for inference. so we're thinking a lot about what is the best product shape for that. And, that involves everything from our training gym to, then, endpoints that get frontier-level performance. again, but I haven't talked to anyone. It looks somewhat different on other verticals. Like, we're also seeing a lot of real-time, audio-video stuff in there, which is why like, we're working on things like regional routing, with fallbacks. So you can get GPUs that are as close to users as possible. so you get like low latency for video streaming and whatnot. And then on the agent side, it's,Akshat [00:41:52]: We're still working very closely with our customers because stuff is changing so fast in terms of what they need. And, I think beyond sandboxes and persistent file systems, there's a lot of other things people will need from this agent stack as they build production agents. So yeah, we're thinking about those other things that fit in there.Swyx [00:42:13]: I want to ask what the other things are.Akshat [00:42:15]: Yeah. I probably should share right now.Swyx [00:42:17]: I think-- I think, okay, so, I do think a lot about the principal components of cloud, and you do talk about compute storage networking.Akshat [00:42:25]: Yeah.Swyx [00:42:25]: Because so far for me, it's fine. so far for the. the first couple generations of cloud, it's fine. What's different, qualitatively different about agents that you need some new permission level? Like a lot of people, okay, and I'll just kinda spew tokens at you until it like hopefully sparks something.Akshat [00:42:43]: Yeah.Swyx [00:42:44]: Like the new level now is whatever Claude Code does, which is dangerously scope permissions or like allow list by command or like whatever, right? And sometimes they're like, “Well, okay, we have like this adaptive thinking mode where like, just trust me, bro. I will make the calls for you.” Is that it? like mediated permissions.Hard Guardrails vs. LLM-Mediated PermissionsVibhu [00:43:03]: Now you're looping it with a goal and letting it roll.Akshat [00:43:06]: Yeah, I'm, I'm skeptical of LLM media permission for stuff that is at the sandbox level because you do want hard boundaries.Swyx [00:43:16]: Yeah.Akshat [00:43:16]: Otherwise, someone can exfiltrate stuff.Swyx [00:43:20]: But likeAkshat [00:43:20]: YeahSwyx [00:43:20]: Maybe that's old school thinking. Maybe we're the dinosaurs.Swyx [00:43:23]: Maybe the AI OS or the LLM OS is really the kernel is a goddamn LLM.Swyx [00:43:30]: Like it makes you feel uncomfortable.Akshat [00:43:31]: Yeah, I'm, I'm toldSwyx [00:43:32]: But that's what trusting the LLM is. Like imagine a spherical cow perfect LLM.Akshat [00:43:36]: Right.Swyx [00:43:37]: That it.Akshat [00:43:39]: Maybe.Swyx [00:43:41]: I wanna test the boundaries, right?Akshat [00:43:42]: Yeah.Swyx [00:43:42]: Like, and I don't believe that, but I wanna see where I'm wrong ‘cause that's, that's the consensus.Akshat [00:43:49]: Yeah. I think you always need hard guardrails when you want, And you can pair those with softer guardrails, right? And that's gonna be a lot of mediated.Managed Agents and Specialized SandboxesSwyx [00:44:00]: There. I'll also get you a end with a couple of your commentary on like the ecosystem outside of Modal. Manage agents. Everyone has one. Gemini, OpenAI, Claude, very useful for you, but also like it is their way of starting to edge into your space.Akshat [00:44:17]: Yeah.Swyx [00:44:17]: What's going on?Akshat [00:44:19]: Yeah, we're, very excited to partner with Anthropic and some of the other foundation labs, will not name who we're also working with. the way we see it is the manage agent thing is a great place to start if you're starting out building an agent and, But then when you get to, building something more production grade, like you're a company that's like Ramp that's building their own, Ramp also runs their accounting agent on us, so their external-facing agent. You need a lot more control over, your compute primitive on things like, what sort - how do you persist different files that the agent has access to, and how do you snapshot and restore? How do you control the networking? maybe you want GPUs. When you get to that point, you kinda want, a specialized sandbox provider, that gives you those things, and that's the role that we are trying to play.Swyx [00:45:15]: YeahAkshat [00:45:16]: We don't really have an opinion on the harness, whether it runs - it's a cloud-managed agent, and you hook it up to Model Sandbox, or you run the harness in Model Sandbox. We'll see where people converge with that.Swyx [00:45:26]: Yeah. Do you any opinions on like the meta harnesses, or just another layer on top of these things?Akshat [00:45:31]: You mean like the OpenPipeSwyx [00:45:33]: OpenPipe is one. I think Vercel had one, which I can't remember the name of right now. Fredshot had one. and then, to me, most recently was Data Databricks that had Omnigen. All these are meta harness. Like it's kinda pseudo agent cloud type things.Akshat [00:45:50]: I personally have not played around with them.Swyx [00:45:53]: Yeah.Akshat [00:45:53]: Build agents with them.Swyx [00:45:54]: Everything's bullish Modal, as long as it consumes more infra.Akshat [00:45:57]: That's why we're focusing on the infra layer. It's somewhere where our, relative competence is and, also it's a hard problem to solve.Swyx [00:46:06]: Yeah. I will say like just generally reflecting on that, I don't know if - if there's other topics on Modal, but like just generally reflecting as an infra person, not as intense as you, but in that field, this has like been the most exciting time in infra. Like it was boring for a while, and you couldn't really get people excited about data infrastructure. Like Eric would get on Data Console, everyone just watched the video and like say, “Look at how many sandboxes I can spin up,” and no one gave a crap.Why Infrastructure Became Exciting AgainAkshat [00:46:39]: Yeah.Swyx [00:46:40]: And like now everyone gives a crap.Akshat [00:46:42]: That's true. It is a very exciting time, and I think a lot of that's driven by just the amount of scale all of this stuff needs.Swyx [00:46:50]: I think the, like a lot of your initiatives or a lot of your like product directions make sense in retrospect, which is like the best kind, but I wouldn't necessarily have thought about it myself, which.Akshat [00:47:00]: We need the predictions.Swyx [00:47:02]: I think there's a lot that you just don't even see, right? Like you have the batch, you have the voice, you have the multimodal, but what else?Akshat [00:47:10]: What else is coming up for usSwyx [00:47:11]: Yeah. Where do you see things going?Akshat [00:47:13]: Yeah. I, in generalBiotech, Robotics, and Non-LLM AI WorkloadsAkshat [00:47:15]: It's it's clear that there's there's a huge shift happening. I think one thing that's not as obvious to people because LLM inference gets talked about so much and is also we work a lot of companies that are, doing things like drug discovery and computational bio, like the Chai Discoveries of the world. Big things are probably gonna happen there. we work a lot of robotics companies that are putting robots in like active deployments and getting good results out of them.Swyx [00:47:45]: Is there Air Gap Modal? Is there a version that is like prem air gapped whatever?Akshat [00:47:50]: No. We,Swyx [00:47:51]: You should cloud only.Akshat [00:47:51]: Yeah.Swyx [00:47:52]: Yeah. Okay. But yeah, so what you're saying is like because you're focused on primitives and they're good primitives, you find use cases in all these kinds of things.Akshat [00:48:01]: Yeah.Swyx [00:48:01]: Probably diversifies you a little bit away from LMS all the time.Akshat [00:48:05]: Yeah, absolutely. We're, we'- our goal isn't to only serve the LLM inference market.Swyx [00:48:10]: There are a lot just on the website, the audio,Akshat [00:48:12]: Yeah. We said both onSwyx [00:48:14]: Computational bio images. Yeah, there's a lot here. There's QTA TTS, customizing. Oh, Chatterbox. there was customizing Whisper.Akshat [00:48:24]: Okay. Yeah.Swyx [00:48:25]: This screen reminds me of a fallen competitor, which Replicate.Model APIs vs. Differentiated AI ProductsSwyx [00:48:31]: What's your postmortem on what happened?Akshat [00:48:34]: This is one thing we've stayed away from is providing an API for models because I think providing model APIs is some of it ends up serving like a really hobbyist market, which is much less sticky.Swyx [00:48:50]: Yeah.Akshat [00:48:50]: And we've always wanted to build for companies that are building products and need more flexibility that's not just an API.Swyx [00:48:57]: Which you can build an API for a model and this is clearly what it is. But you - but what you're saying, you can wrap it into a more fully functioning back end that you run.Akshat [00:49:06]: Yeah. So all of our examples, it's not that spin up this model, here's an API token, use it. They're all code.Swyx [00:49:13]: Okay.Akshat [00:49:13]: And so the point is that this is just an example.Swyx [00:49:16]: Starter code.Akshat [00:49:17]: Yeah. But you can tweak it however you want.Swyx [00:49:20]: Yeah.Akshat [00:49:21]: And if you're like a company building a product, like, computational bio whatnot, yeah.Swyx [00:49:26]: I guess I'm trying to tease out for listenersAkshat [00:49:28]: YeahSwyx [00:49:28]: When does it stop becoming, oh, you're just an API call and you're just a wrapper on API to becoming what you call a product, right?Swyx [00:49:36]: Like, what is that layer? Like what-- Like, more lines of code, but like beyond that, what is the substance that people add that qualifies it to be something more?Akshat [00:49:46]: I think there's a little bit of like a selection effect of like a lot of the companies who do wanna get deeper into that level are probably building something that's more differentiated. And, I think, an example is like - with LLM inference, originally we, worked with companies that were building their own post-training frameworks or they were, - Ramp early in the day was training their own tokenizer and like swapping out the tokenizer in Llama and whatnot. I'm not saying that's, that successful, in that case. But a better example is like, let's say Suno. because Suno, does not use Modal for training.Swyx [00:50:26]: Mikey on the pod. Yeah.Akshat [00:50:27]: But they use Modal for all their inference and that's because they have like a custom-- They have completely custom model architecture and that means that they have to be at the code level and tweak things that are not, just an API.Swyx [00:50:41]: It's interesting as well, like we had, Ethan, most recently on the xAI Groq team make a prediction that like the next tier in video gen is not a better video model, it's a better model or agent that orchestrates video models.Video Agents and Production WorkflowsAkshat [00:50:56]: Oh, interesting.Vibhu [00:50:56]: Language model backbone that can use toolsAkshat [00:50:58]: RightVibhu [00:50:59]: And write code.Akshat [00:51:00]: Like, yes, I can make my second video or my second video from Groq, but I want my minute video.Akshat [00:51:06]: And I'm not going there through normal video gen.Swyx [00:51:10]: Yeah, that's interesting. I - So we have GPU sandboxes and recently have seen a few companies doing agents that do video manipulation or,Akshat [00:51:22]: Yeah. Give it FFmpeg and just do it.Swyx [00:51:23]: Run FFmpeg. But likeAkshat [00:51:25]: That's not enough.Swyx [00:51:25]: Yeah.Akshat [00:51:26]: You need to give it Adobe.Swyx [00:51:27]: Yeah, I hadn't put it together with like it would be a video production thing. in my mind these things were going more towards editingAkshat [00:51:36]: Yeah.Vibhu [00:51:36]: Well, shout out Mantis.Akshat [00:51:37]: I think about this a lot.Swyx [00:51:38]: .Akshat [00:51:41]: Yeah. Sorry.Vibhu [00:51:41]: Luma. Luma Agent is a version of this for video production, but it's a off.Swyx [00:51:46]: I was gonna get your quick takes, on some other stuff that happensGitpod/Ona, CI, and Runtime SandboxesSwyx [00:51:50]: In recent news and just-just see if you have anything interesting. Gitpod, very li

In 1782, Mozart was corresponding with a fellow composer about the art of composition. Mozart made the point that the true essence of music lies not only in the notes themselves but also in the spaces between them. That thought was echoed a century later by Claude Debussy who said, "Music is the silence between the notes."  Silence is powerful. One part of the power of silence is its ability to define the notes around it. The spaces between the notes create rhythm. Lengthening spaces create a ritardando. Shortening spaces create an accelerando. Actual silence, created by a rest in the music, causes tension that is only relieved, or possibly heightened, by the next note. For us harpists, creating silence as we play is challenging. Our music naturally rings and resonates as each string is played. In those moments when we need to stop the sound, it becomes a physical and intentional act. We create silence on purpose.  But what of those moments between the notes, when we want a note to sustain and to hold the tension into the next one? Do those moments need special attention? I believe they do.  So today, I'd like to explore the two different kinds of silence we harpists have to consider: what each means and what we need to do to observe those silences with artistry and musicianship. And as we get started, here is another quote to let roll around in your mind. It's from Kate Kennedy's book, Cello: A Journey Through Silence to Sound. She writes this: "Silence is the opposite of music, but it is also its lifeblood — the breaths between the phrases, the drama, the anticipation, and the quality of the breathless hush between final note and applause."   Links to things I think you might be interested in that were mentioned in the podcast episode:  Join the Compose for Christmas Challenge Register for the Chrysalis Music Workshop  Harpmastery.com Get involved in the show! Send your questions and suggestions for future podcast episodes to me at podcast@harpmastery.com Looking for a transcript for this episode? Did you know that if you subscribe to this podcast on Apple Podcasts you will have access to their transcripts of each episode? LINKS NOT WORKING FOR YOU? FInd all the show resources here: https://www.harpmastery.com/blog/Episode-268  

Topics covered in this episode: Backup Docker volumes locally or to any S3 Pyodide 314.0 Release nb-cli: A Command-Line Interface for AI Agents and Notebook Automation Hindsight Agent Memory That Learns Extras Joke Watch on YouTube About the show Sponsored by us! Support our work through: Our courses at Talk Python AWS Community Day Midwest tomorrow Wednesday the 24th in downtown Indianapolis, Six Feet Up is sponsoring and there are 2 Sixies presenting Connect with the hosts Michael: Mastodon / BlueSky / X / LinkedIn Calvin: Mastodon / BlueSky / X / LinkedIn Show: Mastodon / BlueSky / X Join us on YouTube at pythonbytes.fm/live to be part of the audience. Usually Tuesday at 7am PT. Older video versions available there too. Finally, if you want an bonus digest of every week of the show notes in email form? Add your name and email to our friends of the show list, we'll never share it. Michael #1: Backup Docker volumes locally or to any S3 Via Bryan Weber (thanks Bryan!), who spotted it over on Virtualization HowTo. Find Bryan at bryanwweber.com. offen/docker-volume-backup is a lightweight companion container that backs up the volumes your apps actually depend on, then ships them somewhere safe. It's tiny: written in Go and about 25MB compressed, roughly 1/20th the size of the shell-based image (jareware/docker-volume-backup) that inspired it. Drop it into your docker compose file as a backup service, mount the volumes you care about as read-only, and you're off. Push backups to a pile of destinations: a local directory, plus any S3, WebDAV, Azure Blob Storage, Dropbox, Google Drive, or SSH-compatible target. Mix and match as many as you want in one run. Recurring cron-style backups in a Compose setup, or one-off backups straight from the Docker CLI. Production-friendly touches worth calling out: Rotates away old backups so you don't quietly fill the disk. GPG encryption for your archives. Notifications on finished and failed runs (so you find out about failures before you need the backup). Stop a container during backup for a consistent snapshot using a simple docker-volume-backup.stop-during-backup=true label, then auto-restart it. Run custom commands during the backup lifecycle (great for a database dump before the file copy). Docker Swarm support, plus arm64 and arm/v7 builds. Hello, Raspberry Pi homelab. Fun aside from Bryan: he searched our back catalog for this tool and the search came back so fast he thought it hadn't run. Love to hear it. Calvin #2: Pyodide 314.0 Release PEP 783 is the real news — Pyodide maintainers used to hand-build 300+ packages. Now anyone can publish Pyodide wheels to PyPI with cibuildwheel. The version jump from 0.29 to 314.0 is intentional — it now tracks the Python version, so 314.x = Python 3.14. Binary compatibility is locked per Python cycle, meaning packages you build today won't break on the next Pyodide release. sqlite3, ssl, and lzma are back in the default stdlib — no more await pyodide.loadPackage("sqlite3"). Bigger download, but a much smoother experience for newcomers. bigint precision bug is fixed — values above 2^53 were silently losing precision when crossing the Python/JS boundary. The new JsBigInt type makes the roundtrip correct. Worth flagging if anyone is doing numeric work in a browser app. Experimental TCP sockets in Node.js — you can now connect Pyodide to a real database (MySQL, PostgreSQL, Redis tested) when running server-side. Blurs the line between "Python in the browser" and "Python runtime anywhere Wasm runs." Michael #3: nb-cli: A Command-Line Interface for AI Agents and Notebook Automation From Piyush Jain (Jupyter and LangChain maintainer) on the Jupyter blog: nb-cli: A Command-Line Interface for AI Agents and Notebook Automation. nb-cli is an experimental, Rust-based CLI to read, write, execute, and search Jupyter notebooks. The premise: agents are great at CLIs but terrible at hand-editing the nested JSON in an .ipynb, so let them operate on the notebook from the outside instead of running inside it. Works with or without a Jupyter server. No server? It reads/writes .ipynb files directly and talks to kernels over ZeroMQ. Connected to a live JupyterLab, your edits show up instantly via Y.js (the same CRDT Jupyter uses). Smart output format: instead of token-heavy JSON or ambiguous plain markdown, it uses @@cell / @@output sentinels with inline metadata. Less wasted context, unambiguous structure, and it degrades gracefully on truncation. The payoff is composability. "Add a summary section and run it" becomes one shell pipeline instead of six agent tool calls. And nb search notebook.ipynb --with-errors returns only the failing cells, so the agent skips the cells that worked. Claude Code tie-in: it ships as an agent skill. npx skills install jupyter-ai-contrib/nb-cli and your agent can drive notebooks via nb. Out of jupyter-ai-contrib, which aims to become an official Jupyter AI subproject. Still early (crates.io is at v0.0.5), so kick the tires before anything load-bearing. See also marimo-pair. Calvin #4: Hindsight Agent Memory That Learns AI agents forget everything between sessions — Hindsight gives them persistent memory that learns over time Simple three-method API: retain(), recall(), reflect() — store, retrieve, and reason over memories TEMPR retrieval runs semantic, keyword, graph, and temporal search in parallel for accurate results Automatically consolidates related facts into durable observations instead of piling up duplicates pip install hindsight-all runs the entire server in-process; integrates with LangChain, LlamaIndex, Pydantic AI, CrewAI, and more Extras Calvin: Clanker: A Word For The Machine **Ponytail — You know him. Long ponytail. Oval glasses. Has been at the company longer than the version control** **Klangk: Multi-User AI Sandboxing, Collaboration and Coding Platform** Cursor announces Origin performative-ui to quick start your new idea Michael: Astral Joins OpenAI: The Interview SpaceX to acquire Cursor And OpenAI renews Open Source support Portuguese subtitles are now available for Talk Python courses DSF is hiring including Six Feet Up support Joke: Oh Babe…

As I look back on my childhood and my very first music studies, I realize that I was incredibly fortunate, blessed, actually. Not because I was gifted or because my parents were not only willing but were able to support my harp lessons. Not because I was in one of the music capitals of the world and had access to world class teachers. Not because I had so many opportunities and people that encouraged me along my journey. Well, yes, I was blessed because of all those things, certainly. But I want to talk today about a different gift that was given to me, one I didn't realize the value of until much later. The gift was this: I was always encouraged to play music outside the box. Let me explain. I started piano lessons when I was four years old, and my piano teacher was not only a fabulous musician, but a creative and generous teacher. My lessons included all the usual piano exercises - which I hated - and sonatinas - which I loved - and she also made sure that she gave me performance-worthy arrangements of popular music. I played arrangements of "Blue Moon" and "People" from Funny Girl and "Somewhere Over the Rainbow" that were written by pianists like Roger Williams. These were fancy arrangements that took quite a bit of technique and a lot of practice to play well, and I enjoyed them.  But my teacher never stuck exactly to the printed page. She always had some alterations to add a little more pizzazz to the arrangement. We were always taking bass notes down an octave or playing another part up in the high register like a music box, or changing dynamics and tempo. What she taught me was that music was self-expression. It was about so much more than just playing the written notes. It was about making the notes say what you thought they should say. Please understand; we didn't take liberties with the classics. She made sure I played absolutely every note that Bach wrote on the page. But I learned very early on how to be creative in my music-making, and that has been a tremendous gift. I've never shied away from arranging, and in fact, it's one of my favorite things to do. I'd love every harpist to learn the freedom of making their music their own, and that's what this show is about. I actually went back about two years into the podcast archives to pull out this episode which speaks directly to this. If you've ever wondered about what it takes to make an arrangement, or how to go about it, this is the episode for you. And afterwards, I want to tell you about an opportunity - a new challenge, in fact -  to work with me on your own arrangement.  Links to things I think you might be interested in that were mentioned in the podcast episode:  Get creative with the Compose for Christmas Challenge  Go even further at the Chrysalis Music Workshop  Harpmastery.com Get involved in the show! Send your questions and suggestions for future podcast episodes to me at podcast@harpmastery.com Looking for a transcript for this episode? Did you know that if you subscribe to this podcast on Apple Podcasts you will have access to their transcripts of each episode? LINKS NOT WORKING FOR YOU? FInd all the show resources here: https://www.harpmastery.com/blog/Episode-267

durée : 00:25:26 - par : Judith Chaine - Aujourd'hui, Francesco Filidei revient sur son parcours d'organiste, ses années de formation et les rencontres qui ont façonné sa pensée. Entre sons cachés, vanités et structures invisibles, il dévoile un univers où la musique devient un art du temps, de l'écoute et de l'imaginaire. - équipe : Marie-Christine Ferdinand, Adrien Roch - invités : Francesco Filidei organiste et compositeur italien contemporain Vous aimez ce podcast ? Pour écouter tous les épisodes sans limite, rendez-vous sur Radio France

"C'est important que je comprenne le code qui a été généré." Le D.E.V. de la semaine est Guillaume Lours, Software Engineer chez Docker. Avec lui, on plonge dans les coulisses de Docker Sandbox, cette solution pensée pour sécuriser l'automatisation et l'exécution de code par IA. Guillaume nous explique comment limiter les risques sans sacrifier l'efficacité, et détaille l'architecture technique : micro VM légère, man-in-the-middle pour la gestion des credentials, profils de sécurité personnalisables, le tout pour une expérience fluide. Il partage aussi des astuces d'usage au quotidien pour tester, reviewer ou itérer plus sereinement. Un éclairage concret sur le futur du développement outillé par l'IA.Chapitrages00:00:53 : Docker et Magie Noire00:01:32 : Présentation de Guillaume00:03:15 : Docker et les Générations00:04:50 : Résumé de Docker00:07:12 : Utilisation de Docker pour Tous00:08:51 : Écosystèmes et Docker00:12:43 : Complexité vs Simplicité00:17:20 : Introduction à Docker Compose00:20:23 : Fonctionnement de Docker Compose00:22:43 : Responsabilités de Compose et Engine00:28:19 : Ordonnancement et Réconciliation00:32:45 : Conteneurisation vs Virtualisation00:37:31 : Introduction à Docker Sandbox00:40:04 : Fonctionnement de Docker Sandbox00:45:58 : Usages de Docker Sandbox00:52:28 : Philosophie de Docker sur le Code00:58:44 : Recommandations et Conclusion Liens évoqués pendant l'émission YT Devoxx France Construire une application indépendante de la tech US en 2025 | Eventuallycoding2025, Europe Vs USA : la tech à l'heure des choix | EventuallycodingLe mythe de la neutralité : quand la tech devient politique

Fredrik snackar Kotlinconf 2026 och språket Kotlin i allmänhet med Johan Blomgren och Emil Kantis. Hur var konferensen? Hur fungerar utvecklingen av Kotlin, och vad är på gång i språket? Det blir tips på intressanta presentationer värda att se när de släpps på nätet, och en förklaring av varför Kotlinconfs officiella app inte känns helt hemma på Appletelefoner. Vi snuddar också - inte helt oväntat - vid språkmodeller. Vi pratar om AI, teknikutvecklingen, och presentationen av saker som oundvikliga kontra att bygga en bättre värld genom att helt enkelt prata mer med andra människor. Gärna öga mot öga också. Det är en mänsklig superkraft! Som avslutning bjuds på en snabb genomgång av anledningar att byta till Kotlin från Java. Ett stort tack till Cloudnet som sponsrar vår VPS! Har du kommentarer, frågor eller tips? Vi är @kodsnack, @thieta, @krig, och @bjoreman på Mastodon, har en sida på Facebook och epostas på info@kodsnack.se om du vill skriva längre. Vi läser allt som skickas. Gillar du Kodsnack får du hemskt gärna recensera oss i iTunes! Du kan också stödja podden genom att ge oss en kaffe (eller två!) på Ko-fi, eller handla något i vår butik. Länkar Johan Emil Java Kotlin Komma igång med Kotlin som Java-utvecklare Att övertyga andra om Kotlins storhet Helping decision makers say yes to Kotlin React Vue ATG Junit Kotest Kotlin-test Kotlinconf 2026 Keynoten för Kotlinconf 2026 Jetbrains utvecklar både IDE:er och Kotlin Javazone i Oslo Kotlin på Youtube - inklusive inspelninigar från Kotlinconf 2026 när de släpps KEEP - Kotlin evolution and enhancement process Local lifetimes i Kotlin Value semantics i Kotlin Rich errors i Kotlin Sum types Union types Javas projekt Leyden och projekt Valhalla Virtual threads i Java Kotlin multiplatform Compose multiplatform Kotlinconf-appen Stöd oss på Ko-fi! Erik Hellman Eriks presentation på Kotlinconf 2025 om IOT MQTT Matter Spec-driven development Jake Wharton - pratade om composebaserat terminal-UI-bibliotek Jesse Wilson Okhttp - Squarebyggt ramverk Lena Reinhard snackade om utvecklares roll i AI-världen Professional development - presentation från Google IO 2026 Clean code Kotlins LSP Junie - Jetbrains kodagent Lars Wikman Coursera-kurs om Kotlin för Javautvecklare Kotlin-övningar Builder pattern Bygga DSL:er med Kotlin WASM Uber snackar Kotlin Titlar Min Kotlinbana Kotlin på många olika sätt (Min upplevelse av) Sex år i Kotlin Bypassa hela stdout Jag är ju redan i utlandet Här slutade nullpointers En konstruktor som har alla parametrar

Compose effective words. Enjoy the process and let's maintain a sense of balance.

Compose effective words. Enjoy the process and let's maintain a sense of balance.

Chaque année, chacun d'entre nous produit près de huit kilos de déchets électroniques. À l'échelle mondiale, cela représente 62 millions de tonnes en 2022. Un volume colossal… et surtout en constante augmentation. Le problème, c'est que cette masse croît cinq fois plus vite que les capacités de recyclage. Résultat : une grande partie de ces déchets finit enfouie ou incinérée, avec des conséquences environnementales bien réelles.Recycler ces objets reste un défi technique. Nos appareils sont conçus comme des assemblages complexes : plastiques, métaux, composants électroniques… souvent imbriqués de manière indissociable. Et c'est encore plus vrai pour les robots dits “souples”, de plus en plus utilisés en agriculture ou en médecine. Ces machines combinent des matériaux avancés comme des polymères élastiques, des alliages métalliques et des semi-conducteurs, le tout difficile à séparer en fin de vie.Mais une équipe de chercheurs sud-coréens, issue de l'Université nationale de Séoul et de l'Université Sogang, propose une piste radicalement différente : concevoir des robots… entièrement biodégradables. Leur étude, publiée dans la revue Nature Sustainability, présente un robot souple capable de se décomposer sans laisser de trace toxique. Pour y parvenir, les scientifiques ont utilisé un matériau structurel particulier, un polymère biodégradable appelé poly(sébacate de glycérol), ou PGS. Ce type de matériau, que l'on appelle un élastomère, possède des propriétés proches du caoutchouc tout en étant capable de se dégrader naturellement.À cela s'ajoutent des composants électroniques eux aussi biodégradables, fabriqués à partir de matériaux comme le magnésium, le molybdène ou encore le silicium, choisis pour leur capacité à se dissoudre progressivement dans l'environnement sans danger. Malgré cette conception inédite, les performances sont au rendez-vous. Le robot peut embarquer des capteurs de température ou d'humidité, produire de la chaleur ou même administrer des médicaments. Et surtout, il reste fonctionnel après un million de cycles d'utilisation, preuve de sa robustesse. Une fois son rôle terminé, il peut être placé dans des conditions de compostage industriel et se décomposer en quelques mois seulement. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

In this episode of Memoirs of an LDS Servant Podcast, Maurice Harker expands on the idea of “composing your life”—teaching how to intentionally design who you become instead of drifting into habits and patterns you don't want.Using powerful examples like vision boards, creative processes, and even studying how artists build their work piece by piece, Maurice explains how personal growth is not random—it's constructed through intentional choices, reflection, and repetition.You'll learn how to identify who you want to become, gather influences that support that identity, and begin building your life step by step—just like composing music, choreography, or any creative work.

Send us Fan Mail

durée : 00:27:04 - par : Gabrielle Oliveira Guyon - Dans son dernier disque "Waves of Light", Paul Lay fait dialoguer jazz et musique chorale dans une fresque lumineuse et aux côtés du chœur Les Éléments dirigé par Joël Suhubiette. Le pianiste continue ainsi son voyage à la croisée des styles, des époques et des imaginaires. - réalisation : Yassine Bouzar, Julia Macarez, Côme Jocteur-Monrozier, Morgane Tourreilles, Maxime Laporte, Valentin Lévy-Chaudet - invités : Paul Lay Pianiste, compositeur jazz (22 juillet 1984, Orthez) Vous aimez ce podcast ? Pour écouter tous les épisodes sans limite, rendez-vous sur Radio France

¿Es posible mantener la extrema sencillez de Dockge mientras aprovechamos la robustez y seguridad de Podman? La respuesta es un rotundo sí, y en este episodio te explico exactamente cómo lo he configurado en mi propia infraestructura.Llevo semanas explorando alternativas para la gestión de contenedores, pero siempre acabo volviendo a Dockge. Su capacidad para levantar un stack simplemente pegando un Docker Compose es imbatible para quienes disfrutamos probando nuevas herramientas cada día. Sin embargo, mi migración a Podman planteaba un reto: no quería perder esa agilidad ni tampoco comprometer la seguridad del sistema.En este podcast detalle mi "fórmula ganadora":Quadlets: Cómo he encapsulado Dockge y Traefik para que se comporten como servicios nativos del sistema.Seguridad Rootless: La ventaja de correr Dockge bajo un usuario sin privilegios, eliminando riesgos de escalada de privilegios.Persistencia: La gestión de volúmenes y cómo Dockge almacena los archivos Compose de forma transparente en el sistema de archivos.Hibridación: Mi estrategia para decidir qué servicios van en Quadlets y cuáles se quedan en Dockge.Además, comentamos características fundamentales como el terminal web interactivo incorporado, ideal para solventar problemas rápidos (como borrar un volumen rebelde) cuando estás fuera de casa y solo tienes una tablet a mano. Si te interesa el self-hosting, la administración de servidores Linux y quieres simplificar tu flujo de trabajo con contenedores, este episodio es para ti.Capítulos del episodio:00:00:00 Introducción y el dilema de la gestión de contenedores00:01:41 El miedo a la migración: De Docker a Podman00:02:46 La gran noticia: Dockge funcionando como Quadlet00:03:09 ¿Qué es Dockge? La alternativa sencilla a Portainer00:05:14 Características clave: Editor interactivo y terminal web00:06:09 Gestión remota: El uso de agentes y múltiples VPS00:07:33 Funciones avanzadas: De comandos Docker a Compose00:08:55 La ventaja competitiva: Podman Rootless y seguridad00:09:41 Anatomía de un Quadlet para Dockge00:10:45 Configuración de volúmenes y persistencia de Stacks00:11:24 Integración con Traefik y Health Checks en Podman00:12:22 Cómo gestionar tus archivos Compose y Dotfiles00:13:58 El gran debate: ¿Cuándo usar Quadlets vs Dockge?00:15:53 Conclusiones: Seguridad, simplicidad y futuro00:17:12 Despedida y comunidad Atareao con LinuxÚnete a la conversación en nuestro grupo de Telegram y descubre más en atareao.es.Más información y enlaces en las notas del episodio

Comment avons-nous rendu la terre de moins en moins habitable ? Comment arrêter la course vers l'abîme ? Et si les ontologies étaient au cœur des enjeux ? Philippe Descola, anthropologue, professeur émérite au collège de France, revient ici sur le fil conducteur de ses recherches, depuis sa rencontre avec les Indiens Achuar jusqu'à sa mobilisation dans la ZAD de Notre-Dame-Des-Landes : le changement de paradigmes indispensable pour sortir de l'impasse. Il nous explique la nécessité de dépasser le dualisme nature / culture, les rapports entre les humains et les non humains variant selon des modes d'identification qu'il regroupe en quatre ontologies (totémisme, animisme, analogisme et naturalisme). Si chaque humain compose son monde à partir de la culture dans laquelle il baigne, des expériences et des socialisations vécues, il reste qu'en occident, le naturalisme est déterminant, pesant sur les comportements de destruction du vivant. Les mots, les concepts, inadaptés aux réalités, enferment nos pensées et guident notre manière d'habiter le monde. Et même si le grand partage est de moins en moins clair, du chemin reste à parcourir pour traiter avec respect le vivant qui nous entoure. Heureusement, l'espoir vient des marges, et de la capacité qui s'y déploie d'inventer rapidement de nouveaux mondes. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

Send us Fan MailIn this episode of the FuturePrint Podcast, Marcus Timson welcomes back Dario Urbinati, CEO of Gallus Group, for a timely and far-reaching conversation about the structural changes reshaping print, packaging and the wider industrial economy.Two years ago on this podcast, Dario predicted that the market was not simply going through a temporary downturn, but entering a more fundamental reset. In this follow-up discussion, he explains why he still believes that this is happening now - and why the old pre-2019 model of global stability, easy growth and predictable supply chains is not coming back.The conversation explores the long-term macro forces behind this reset, including demographic decline, labour shortages, geopolitical fragmentation, inflation, supply chain disruption and the end of the old globalised operating model. Dario explains why these are not cyclical challenges, but structural ones, and what that means for converters and label printers trying to plan for the future.Marcus and Dario also discuss what the most successful print businesses are doing differently, why mindset matters as much as machinery, and how technology must now be viewed not only as a productivity tool, but as a resilience tool. Key themes include modularity, smart connected printing, workflow, total cost of ownership, operational flexibility and long-term investment security.Dario also reflects on the thinking behind Gallus's System to Compose strategy and why modular production environments may be increasingly important in a more volatile world.A thoughtful, realistic and ultimately optimistic conversation about what print businesses need to do now to adapt, invest and grow.Listen on:Apple PodcastGoogle PodcastSpotifyWhat is FuturePrint?FuturePrint is a digital and in person platform and community dedicated to future print technology. Over 20,000 people per month read our articles, listen to our podcasts, view our TV features, click on our e-newsletters and attend our in-person and virtual events. We hope to see you at one of our future in-person events:FuturePrint Packaging, Labels & DTS, 29-30 September '26, Valencia, SpainFuturePrint Leaders Summit, 29 September '26, Valencia, SpainFuturePrint Industrial Print, 14-15 April '27, Munich, Germany

«Правда ли, что тесты в мобильных приложениях — это пустая трата времени? Почему Xcode называют “тихим ужасом” и почему iOS-разработчики до сих пор его терпят? Сегодня в подкасте Java Swag мы погружаемся в мир Android, Compose и Kotlin Multiplatform. У нас в гостях человек, который знает о мобильной экосистеме JetBrains всё — Константин Цховребов, разработчик в команде Compose Multiplatform. Мы обсудим путь Кости от первого Android-приложения на слабом нетбуке в 2010 году до техлида в JetBrains. Поговорим о том, как Kotlin захватил мобильный мир, почему “галера” — это идеальный старт для новичка, и как магия expect/actual позволяет писать код сразу под все платформы. 00:00 Старт 01:16 Путь в Android 03:01 Почему работа в аутсорсе — отличная школа для разработчика 09:36 После Extension-функций не хочется возвращаться в Java 13:30 Плюсы и минусы Extension-методов 17:30 Что такое Compose и как выглядела UI-разработка до него 21:00 Почему Compose «зашел» 23:30 Проблема списков в Android 26:40 Особенности мобильной разработки: батарейка, ресурсы и «отсутствующий» интернет 29:26 Навигация в Android: история библиотеки Cicerone 39:00 Google Navigation 3 40:35 Kotlin Multiplatform (KMP) 46:25 Как работает магия expect/actual и почему это лучше, чем дефайны в C++ 49:30 LSP-сервер для VS Code: Kotlin теперь не только в IntelliJ IDEA 01:00:50 Compose Multiplatform на iOS 01:03:30 Проблема нативности: должен ли UI выглядеть «как родной»? Кейс Duolingo 01:06:33 Flutter и React Native 01:18:20 Глубокий интероп и Swift 01:42:40 «Xcode — это тихий ужас» 01:44:15 Будущее: Compose for Web (Wasm/JS) 02:00:50 Чиним скролл в вебе 02:12:05 Непопулярное мнение №1: Gradle — прекрасный фреймворк 02:15:15 Непопулярное мнение №2: В большинстве мобильных приложений тесты не нужны Гость https://www.linkedin.com/in/terrakok/ Ссылки https://bell-sw.com/blog/what-is-crac-a-guide-to-cutting-java-startup-and-warmup-from-minutes-to-milliseconds/ https://blog.jetbrains.com/kotlin/2026/01/the-journey-to-compose-hot-reload-1-0-0/ https://openjdk.org/projects/leyden/ https://openjdk.org/jeps/516

Download the free guide “5 Spells Every Composer Needs” and learn interval-based techniques you can use immediately in your compositions: https://musicintervaltheory.academy/spells/ In this episode, Frank explains why many composers feel slow: they try to generate and refine at the same time. By separating the creative phase (draft one) from the development phase (draft two), composers can work faster, think more clearly, and reduce creative friction. The key is to write the simple core idea first and shape it later. Speed in composing is not about talent—it's about process.

Send us Fan MailLights! Microphone! Podcast -- Episode 2  What goes into creating a piece of music that defines a cultural moment? Nick Thorburn, the composer behind Serial's unforgettable theme, sits down with host Shana to tell the story. Spoiler: it all happened in a weekend, and he had no idea the show would become a global phenomenon. Nick walks through his creative process: how he approached the project as a straightforward freelance job, why he drew inspiration from Twin Peaks, and what it means to work as a "live wire" where the unconscious mind does the heavy lifting. He also gets into the business side of things, explaining why podcasters need to stop treating music like something you slap on at the end. Sound design matters. It shapes how people experience your show. Whether you're launching a podcast or just curious about what happens behind the scenes in audio storytelling, this episode offers a rare look at the intersection of creativity, business, and cultural impact. Created inside a podcasting special topics course (DG 8010: MDM Podcast Lab) within the Master of Digital Media program at The Creative School at Toronto Metropolitan University, this six-part series explores what it really takes to start and grow a podcast. I'm all about interesting projects with interesting people! Let's Connect on the web or via Instagram. :)

¿Te preocupa tener tus claves y contraseñas en texto plano? En este episodio 770 de Atareao con Linux, te explico por qué deberías dejar de usar variables de entorno tradicionales y cómo Podman Secrets puede salvarte el día. Yo mismo he pasado años ignorando este problema en Docker por la pereza de configurar Swarm, pero con Podman, la seguridad viene de serie.Hablaremos en profundidad sobre el ciclo de vida de los secretos: cómo crearlos, listarlos, inspeccionarlos y borrarlos. Te mostraré cómo Podman gestiona estos datos sensibles fuera de las imágenes y fuera del alcance de miradas indiscretas en el historial de Bash. Es un cambio de paradigma para cualquier SysAdmin o entusiasta del Self-hosting.Pero no nos quedamos ahí. Te presento Crypta, mi nueva herramienta escrita en Rust que integra SOPS, Age y Git para que puedas gestionar tus secretos de forma profesional, permitiendo incluso la sincronización con repositorios remotos. Veremos cómo configurar drivers personalizados y cómo usar secretos en tus despliegues con MariaDB y Quadlets.Capítulos destacados:00:00:00 El peligro de las contraseñas en texto plano00:01:23 El problema con Docker Swarm y por qué elegir Podman00:03:16 ¿Qué es realmente un Secreto en Podman?00:04:22 Ciclo de vida: Creación y muerte de un secreto00:08:10 Implementación práctica en MariaDB y Quadlets00:12:04 Presentando Crypta: Gestión con SOPS, Age y Rust00:19:40 Ventajas de usar secretos en modo RootlessSi quieres que tu infraestructura sea realmente segura y coherente, este episodio es una hoja de ruta esencial. Aprende a ocultar lo que debe estar oculto y a dormir tranquilo sabiendo que tus tokens de API no están al alcance de cualquiera.Más información y enlaces en las notas del episodio

Dans cet épisode, Emmanuel interview Arnaud Giuliani. Arnaud est dans l'écosystème Kotlin et est le créateur de Koin, la solution de Dependency Injection. On discute de la genèse de Kotlin, de son alignement avec Android puis de son évolution multiplateforme. On discute coroutine, impact de K2, de développement mobile. On finit en discutant de Kotzilla et de l'entrepreneuriat sur un projet Open Source. Enregistré le 7 janvier 2026 Téléchargement de l'épisode LesCastCodeurs-Episode-336.mp3 ou en vidéo sur YouTube. Interview Ta vie ton oeuvre (présentation de l'interviewé) ton historique de développeur Koin d'où est venu l'idée, pourquoi difference vs Dagger, Hilt, CDI? fondateur de Kotzilla Introduction à la techno (5 à 10 mins max) Kotlin en 4 phrases nombre de développeurs usages (front, mobile, backend) Compose, K2 en une phrase La techno en concepts Kotlin le langage Quel sont ses particularités et spécificités pourquoi il a pris sur Android ? Kotlin multiplateform comment ça marche concretement WASM en beta, tu as eu des retours? pour les devs de framework, c'est transparent? Co-routines et concurrence structurée fais nous un point de ce que c'est son usage dans l'ecosystème vs loom, des ponts ? Kotlin et le backend connu pour le support Android, quid du back end? travaux avec Spring Ktor les autres plateformes Java genre Quarkus et micronaut, utilisées ? La competition de Kotlin c'est quoi ? Comment on l'utilise en pratique pour un dev je me lance, je faisais du Java et du Spring, je pars comment pour faire un projet Kotlin moderne IDE, outil de build, frameworks migrationd e code Java? des anti patterns des choses qui "ressemblent à du code Java" des comportement de perf ou de memoire differents du monde Java? c'est quoi ta feature préférée? Et l'IA, Kotlin as Koog notamment, tu vois quoi emerger ? Sous le capot K2 est le nouveau compilateur Qu'est-ce qui a changé des cassages de compatiblitiés ca change des choses pour les utilisateurs ? Et pour les editeurs de framework comme Koin ? Koin ne fait pas de generation de code à la compil Dagger, Arc (le moteur CDI de Quarkus) et Micronaut sont passé au pre travail à la compil quels ont été les critères de choix un mot sur Kotlin Symbol Processing les coroutines, c'est implémenté comment, vous avez 3 heures machine a etat continuation apssing style etc Kotlin multi platforme que fait le compilo code commun / code specifique interop avec les platformes cibles (object structure etc) La communauté, le futur comment va la commuanuté aujourd'hui grossis ? et les francais là dedans? La gouvernance de Kotlin travaux dominés par JetBrains comment cela a évolué (ecoute, autres acteurs etc) Kotlin foundation futurs fonctionalités de Kotlin qui t'interesse de Koin? autre ? Monter une boite Tu as fondé Kotzilla. Peux-tu nous expliquer ce que Kotzilla apporte à l'écosystème Kotlin ? Quels problèmes tu cherches à résoudre pour les entreprises qui adoptent Kotlin ? ton experience de fonder une boite d'editeur quelle mouche t'as piqué votre business model, comment vous en etes arrivé là de maniere generale discussion sur le lancement de boites techs Nous contacter Pour réagir à cet épisode, venez discuter sur le groupe Google https://groups.google.com/group/lescastcodeurs Contactez-nous via X/twitter https://twitter.com/lescastcodeurs ou Bluesky https://bsky.app/profile/lescastcodeurs.com Faire un crowdcast ou une crowdquestion Soutenez Les Cast Codeurs sur Patreon https://www.patreon.com/LesCastCodeurs Tous les épisodes et toutes les infos sur https://lescastcodeurs.com/

¿Sigues usando Docker Compose para todo? Es hora de descubrir la verdadera potencia de Podman: los Pods. En este episodio te acompaño en la migración de un stack completo de WordPress, MariaDB y Redis para que veas cómo simplificar radicalmente la gestión de tus contenedores.Aprenderás por qué el concepto de "vaina" (Pod) cambia las reglas del juego al permitir que tus contenedores compartan la misma red y dirección IP, atacando directamente a localhost. Veremos desde el funcionamiento técnico del contenedor Infra hasta la automatización profesional con Quadlet y systemd.¿Qué es un Pod?: El origen del nombre y por qué es la unidad lógica ideal para tus servicios.Adiós a los problemas de red: Cómo conectar WordPress y base de datos sin crear redes virtuales, usando simplemente 127.0.0.1.Seguridad y Sidecars: Blindar servicios como Redis dentro de la misma vaina para que sean inaccesibles desde el exterior.Gestión unificada: Cómo detener, arrancar y monitorizar todo tu stack con un solo comando.Persistencia y automatización: Generar archivos YAML de Kubernetes y convertirlos en servicios nativos de Linux con archivos .kube.Si buscas soluciones prácticas para "cualquier cosa que quieras hacer con Linux", este episodio te da las herramientas para profesionalizar tu infraestructura.Notas completas y comandos utilizados: https://atareao.es/podcast/766

Season 2 of Compose Like a Girl will launch later this week!   Follow us: @composelikeagirl on Instagram and Facebook Learn more: Compose Like a Girl

Voici peut-être l'idée la plus simple… et la plus efficace pour démocratiser l'open source. Un projet indépendant baptisé Github Store vient de transformer GitHub en véritable magasin d'applications, à la manière d'un App Store ou d'un Google Play, mais dédié exclusivement aux logiciels libres. Disponible sur Android et sur ordinateur — Windows, macOS et Linux — Github Store propose une interface claire et familière : catégories, captures d'écran, descriptions détaillées et surtout un bouton d'installation en un clic. Fini la chasse aux fichiers au fond des dépôts ou la peur de télécharger la mauvaise archive. Ici, tout est pensé pour l'utilisateur final, pas uniquement pour les développeurs.Le fonctionnement est astucieux. L'application analyse automatiquement les dépôts GitHub publics qui publient de vraies versions installables dans leurs “releases”. Elle filtre les formats pertinents — APK, EXE, MSI, DMG, PKG, DEB, RPM — et écarte les simples archives de code source. Résultat : seules les applications réellement prêtes à être installées apparaissent dans le catalogue. L'utilisateur peut ensuite naviguer par popularité, mises à jour récentes ou nouveautés, et même filtrer par système d'exploitation pour ne voir que les logiciels compatibles avec sa machine. Chaque fiche application va plus loin que de simples captures d'écran. On y retrouve le nombre d'étoiles, de forks, les problèmes signalés, le README complet, les notes de version et le détail précis des fichiers téléchargeables. Une transparence fidèle à l'esprit open source.Côté technique, Github Store repose sur Kotlin Multiplatform et Compose. Sur Android, l'installation passe par le gestionnaire de paquets natif. Sur ordinateur, le fichier est téléchargé puis ouvert avec l'outil par défaut du système. Il est possible de se connecter avec un compte GitHub, optionnel mais utile : cela permet d'augmenter fortement les limites d'accès à l'API pour explorer sans contrainte. L'application est distribuée via les releases GitHub du projet et sur F-Droid pour Android, sous licence Apache 2.0. Autrement dit, libre, modifiable et réutilisable. Une précision importante toutefois : Github Store n'a pas vocation à garantir la sécurité des logiciels proposés. Il facilite la découverte et l'installation, mais la responsabilité reste entre les mains des développeurs… et des utilisateurs. En rendant l'open source aussi accessible qu'un store grand public, Github Store pourrait bien changer durablement la façon dont nous découvrons et utilisons les logiciels libres. Une petite révolution, sans marketing tapageur, mais avec une idée redoutablement efficace. Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.

In this episode, Court lays out a practical foundation of photography “rules” — not as rigid constraints, but as dependable starting points. These principles show up again and again because they work: they help you handle light, composition, focus, and technical settings with more confidence. Once you understand them, you can follow them when they serve the image, and break them deliberately when the scene calls for something different. The goal is simple: more intentionality, and better photographs in the field.The 14 Fundamental Rules of PhotographyFocus on the eyes.Use the rule of thirds.Follow the inverse focal length rule for handholding.Prioritize side lighting.Shoot during golden hour.Compose with odd numbers of subjects.Keep horizons straight.Include foreground, mid-ground, and background.Expose for the highlights.Leave space in the direction your subject is looking (eye-line rule).Avoid lines cutting through faces/heads (face-and-line rule).Use the 500 rule for astrophotography.Create subject/background separation.Simplify the frameCourt's Websites Check out Court's photo portfolio here: shop.courtwhelan.com Sign up for Court's photo, conservation and travel blog at www.courtwhelan.com Follow Court on YouTube (@courtwhelan) for more photography tips View Court's personal and recommended camera gear Sponsors and Promo Codes: ArtStorefronts.com - Mention this podcast for free photo website design. BayPhoto.com - 25% your first order (code: TWP25) LensRentals.com - WildPhoto15 for 15% off ShimodaDesigns.com - Whelan10 for 10% off Arthelper.Ai - Mention this podcast for a 6 month free trial of Pro Version

La Coalition des forces pour la République – le CFR – avec à sa tête  l'Imam Mahmoud Dicko, ancien président du Haut Conseil islamique espère un retour à l'ordre constitutionnel. Quelles sont les autres figures du mouvement qui prône un dialogue avec les groupes armés ? Étienne Fakaba Sissoko, Porte-parole de la coalition était notre invité.

Le camp de Bitche, en Moselle. Une forteresse de béton et de barbelés posée à la lisière des Vosges du Nord. C'est là que s'entraîne le 16ᵉ bataillon de chasseurs à pied. Une unité d'infanterie légère de l'armée de Terre française, forte d'environ 1 200 hommes, réputée pour sa discipline et ses exercices exigeants. C'est là que le caporal Tristan Sprunck, marqué par une enfance chaotique, a trouvé un refuge devenu obsession. Maniaque de la discipline, il s'acharne sur le première classe guyanais Sulyvan Flora, son exact opposé : jovial, détendu, populaire. Leurs affrontements se multiplient, jusqu'au 30 juin 2022. Ce jour-là, la promesse de fraternité militaire vole en éclats, révélant les failles d'un univers où l'ordre règne jusqu'à la rupture.L'air est froid dans l'armurerie. Une odeur d'huile, de métal et de poussière se mêle à celle du béton humide. La lumière pâle du matin filtre à travers la vitre renforcée. Sur la table, un Glock 17. À côté, un chargeur encore plein. L'homme respire lentement. Ses gestes sont mécaniques, précis. Il vérifie la culasse, enlève le chargeur, referme l'arme. Tout est rangé selon la procédure. Puis il prend son téléphone. Compose un numéro.Crimes • Histoires Vraies est une production Minuit. Notre collection s'agrandit avec Crimes en Bretagne, Montagne et Provence.

How to Collect Memory-Only Filesystems on Linux Systems Getting forensically sound copies of memory-only file systems on Linux can be tricky, as tools like dd do not work. https://isc.sans.edu/diary/How%20to%20collect%20memory-only%20filesystems%20on%20Linux%20systems/32432 Microsoft Azure Front Door Outage Today, Microsoft s Azure Front Door service failed, leading to users not being able to authenticate to various Azure-related services. https://azure.status.microsoft/en-us/status Docker-Compose Vulnerability A vulnerability in docker-compose may be used to trick users into creating files outside the docker-compose directory https://github.com/docker/compose/security/advisories/GHSA-gv8h-7v7w-r22q

How do I write a message on the walls of my heart? What is the language of my heart? How do I engage in “heart prayers” rather than prayers from my head? Is there a difference?I encounter people who have been in counseling for months or years to process their hurts and are still in pain! Or they have been praying for a miracle of healing, and it is still delayed. So I ask them to share with me what they have done, and listen to their story. Often I am able to help them see that they have NOT been using the language of the heart to heal their hearts. They have been using the language of their minds (logic and analytical reason). The heal the heart, we must use the language of the heart.Read more here: https://www.cwgministries.org/blogs/heart-prayers-employing-language-your-heartSupport the show

Python in 2025 is different. Threads really are about to run in parallel, installs finish before your coffee cools, and containers are the default. In this episode, we count down 38 things to learn this year: free-threaded CPython, uv for packaging, Docker and Compose, Kubernetes with Tilt, DuckDB and Arrow, PyScript at the edge, plus MCP for sane AI workflows. Expect practical wins and migration paths. No buzzword bingo, just what pays off in real apps. Join me along with Peter Wang and Calvin Hendrix-Parker for a fun, fast-moving conversation. Episode sponsors Seer: AI Debugging, Code TALKPYTHON Agntcy Talk Python Courses Links from the show Calvin Hendryx-Parker: github.com/calvinhp Peter on BSky: @wang.social Free-Threaded Wheels: hugovk.github.io Tilt: tilt.dev The Five Demons of Python Packaging That Fuel Our ...: youtube.com Talos Linux: talos.dev Docker: Accelerated Container Application Development: docker.com Scaf - Six Feet Up: sixfeetup.com BeeWare: beeware.org PyScript: pyscript.net Cursor: The best way to code with AI: cursor.com Cline - AI Coding, Open Source and Uncompromised: cline.bot Watch this episode on YouTube: youtube.com Episode #524 deep-dive: talkpython.fm/524 Episode transcripts: talkpython.fm Theme Song: Developer Rap

Facilitator: NedTopics:  How to make your Iphone vibrate?; Has anyone used Ryoko Mobile Wi-Fi?;   A lot of Issues with the IOS 26 update i.e. VO and the timer going on;    Using filters for text messages;  Are people have issues with their Mics?  Suggested new app, Live Read;  Not showing a search box at top in Contacts with new update; Issues using VO since update;  Compose button is now at bottom;  Can choose go back Classic or stay with Unified view in certain messaging apps;  Any good apps to help with scam calls?;    Screening calls;   Changes with delete and trash switching;  What is Magic Tap and Rotor;  Issues with getting out of music library;  assuring new users that their hope with iphone;  How to create separate mail boxes for class and regular mail;  IOS 26, Can you change back to Classic Vs New view in mail?;  How do you delete conversation history in the Ally App ?;  Suggested App, One Step;  Buzz Byte:  Sandhya:  Using the threads feature settings in mail messages

This show has been flagged as Clean by the host. Hello, this is your host, Archer72 for another episode of Hacker Public Radio. In this episode, I continue to fall for the AI trap. Here I was, minding my own business, when I was bothered by the icon only showing a generic icon for the Beeper app. Now, I'm not saying that Duck.ai is not useful, but be very careful what you ask for. It was probably a combination of the early morning, and not reading completely through the AI suggestions, but I ended up losing all icons on the Gnome desktop except for a few like Firefox. I won't leave the problematic command so I don't trip up the listener, but it involved updating a desktop database. This in turn left a dash or blank where the icons should be. If that wasn't bad enough, it was suggested to reset Gnome settings, and nothing was as it seemed before. Things that I had taken for granted were not there. You forget what custom settings are there when mistakes like this are made. So the short answer is that the icons directory, located on my Debian system should be located in .local/share/icons. Instead it was in a sub-directory .local/share/icons/icons Correcting the directory location solved everything, but I was still left to reset my custom Gnome keybindings. • Swap Escape and Caps lock key I used this because I am a Vim user, and this feels more natural when I need to hit Escape to change modes. In Gnome, the setting is under Gnome Tweaks > Keyboard > Additional Layout Options > Swap Esc and Caps Lock Key As of this show release the current stable version is Trixie. Gnome Tweaks - Debian Trixie can be installed by sudo apt install gnome-tweaks on any Debian based system. • Compose key • Compose key shortcuts The Compose key is found at Settings > Keyboard > Compose Key. I selected the Menu key, because this is rarely used, and can still be accessed by the track pad. • Shortcut to open MPV with a clipboard URL from Youtube This can be found in Setting > Keyboard > View and Customize Shortcuts > Custom Shortcuts Shift+Ctrl+P Code placed in /usr/local/bin/ #!/bin/bash ## mpv-url url=`xsel -o -b` echo $url mpv $url Now I can get back to what I started in the first place, creating a .desktop file for Beeper. I created a beeper-desktop.desktop file in `~/.local/share/applications' with the follow contents. [Desktop Entry] Name=Beeper Desktop Exec=/home/mark/AppImages/Beeper-4.1.169.AppImage Icon=/home/mark/.local/share/icons/beeper.png Type=Application Categories=Network;InstantMessaging; Terminal=false StartupWMClass=Beeper The last part of the config file can be found by xprop | grep WM_CLASS Provide feedback on this episode.

Welcome to Now in Android, your ongoing guide to what's new and notable in the world of Android development. Dan covers Android 16 QPR2 beta 1, the Android Studio Narwhal feature drop, Jetpack Compose 1.9, and more!   Chapters: 0:00 - Introduction 0:24 - Android 16 QPR2 beta 1:38 - Android Studio Updates 2:42 - Jetpack Compose August ‘25 released to stable 3:27 - Media3 1.8 Released 4:18 - Recently published Articles 4:55 - Recent published videos 5:26 - Android Developers Backstage 5:50 - AndroidX 6:45 - Recap Resources:

Dr. Noel Rousseau and Trevor Jones have combined forces to form a partnership whose sole focus is to help you beat the Yips. Both Trevor and Noel are PGA Golf Professionals and they bring their coaching/teaching insight and their extensive research into the Yips to the #OntheMark show. Suffering from the Yips is an awful malady and shockingly about 50% of the world's golfing population suffer from, or have suffered from the Yips.  Hence the urgency in Trevor and Noel's work and they share tips, tricks and thoughts to help you back to golf freedom.  They discuss: The Yips - What, How and Why? The fact that having the Yips is no longer a death penalty for golfers Solving the problem with more than just technical solutions Neuroscientific influences in the Yips The differences between the Yips in Putting and Chipping Contrasting and dealing with Type 1 and Type 2 Yips  FOPO - Fear of Other People's Opinions Exposure Therapy for success, and  Diffusion exercises and skills to compose the mind. Trevor also highlights the two ways a golfer will deal with the yips - the B.A.D. Way (Blind Spots, Avoidance and Distraction)  and the A.C.E. Way (Acknowledge, Compose and Engage). This podcast is also available for viewing on YouTube.  Search and subscribe to Mark Immelman.

Preview: AI at School. Colleague Professor Tim Kane of the University of Austin reports the adaptation needed in light of AI ability to compose like a student and fool the faculty. More later.

@PermissionToStanPodcast on Instagram (DM us & Join Our Broadcast Channel!) & TikTok!NEW Podcast Episodes every THURSDAY! Please support us by Favoriting, Following, Subscribing, & Sharing for more KPOP talk!Comebacks: AMPERS&ONE, YOUNG POSSE, MARK (GOT7), JOY (RED VELVET), KEP1ER, BOYNEXTDOOR, STRAY KIDS, CHANYEOL (EXO), IVE, TWICE, GIRLSET (VCHA)Music Videos: AMERS&ONE, KEY (SHINEE), JEON SOMI, JISOO (BLACKPINK)HAYLEE seeing BOYNEXTDOOR on Thursday at the Grammy MuseumBOYNEXTDOOR teaser for Japanese single comebackNew Girl Group by US based agency & SM producers: ATHEARTBABYMONSTER After Inkigayo Snack Bar variety show2nd Annual SBS GAYO DAEJEON SUMMER lineupLOLLAPALOOZA JYP Groups finally on YouTube, we recap: KICKFLIP & TWICEBTS V x Compose Coffee: V COMPOSED STRAY KIDS CHANGBIN birthday live w/ rewritten profile and tidbitsSTRAY KIDS 'KARMA' album unveil track CREEDSTRAY KIDS FELIX selected to be the next MC host for Fridge Interview after SANA & DEXSupport this podcast at — https://redcircle.com/permission-to-stan-podcast-kpop-multistans/donationsAdvertising Inquiries: https://redcircle.com/brandsPrivacy & Opt-Out: https://redcircle.com/privacy

In this episode of Android Back Stage, Hosts Tor, and Chet are joined by Melba and Ralston, Software Engineers on the Compose Team, to talk about Autofill in Compose. Learn about Autofill services, best practices, and more! Chapters: 0:00 - Intro 0:40 - What is Autofill? 1:21 - Autofill team journey and semantics 2:15 - Defining semantics in Compose 3:29 - Bridging views and Compose for Autofill 4:23 - Developer Experience: Early autofill in Compose 7:02 - Autofill services and communication flow 7:59 - User authentication and multiple autofill services 12:53 - The Autofill flow: From tap to fill 21:56 - Handling list types and saving credentials (commit) 30:11 - The “Asteriks” bug and semantics refractor 31:09 - Performance improvements in Autofill 37:48 - Compose's architectural advantages (composition over inheritance) 40:30 - Best practices and future improvements  

durée : 00:25:09 - Pauline Viardot, Cendrillon - par : Anne-Charlotte Rémond - Pauline Viardot a 83 ans en 1904 lorsqu'elle compose ''Cendrillon'', un opéra-comique miniature d'après le conte de Charles Perrault. Anne-Charlotte Rémond vous invite aujourd'hui à suivre la célèbre chanteuse et compositrice dans les salons parisiens de Mathilde de Nogueiras où l'œuvre fut créée. - réalisé par : Claire Lagarde Vous aimez ce podcast ? Pour écouter tous les autres épisodes sans limite, rendez-vous sur Radio France.

Let's start episode 737 of Secrets of Organ Playing Podcast. This question was sent by Kathrin and she writes:Do you also compose for others, after their wishes? Do you get such requests sometimes?For what instruments do you compose beside the organ or piano?Hope you will enjoy it!

Send us a textIn this special episode, Sven talks with Trevor Morris, a client, friend, and the creator of this podcast's theme music. The 2-time Emmy-Award winning composer has now completed a short film, "Butterfly On A Wheel," which he wrote, produced, and directed. Although Trevor is an incredible success by any measure, his personal issues are just as critical as anyone who has appeared on this show. Sven walks Trevor through the stages of his career, his challenges, and his success with Sven's help in healing his soul. Explicit content.

Something I teach to all composition students is the idea of a compositional pantry and creating music like a chef.  In this episode, I'll share what's in my pantry and how to use it as an example of how you can develop your own voice as a composer and write music that fits you to a tee! Musicianship Mastery is formerly known as The Musician Toolkit. Let me know your thoughts on this episode as a voice message to possibly share on a future episode at https://www.speakpipe.com/MusicianToolkit If you enjoyed this, please give it a rating and review on the podcast app of your choice.  You can find all episodes of this podcast at https://www.davidlanemusic.com/toolkit You can follow David Lane AND the Musician Toolkit podcast on Facebook @DavidMLaneMusic, on Instagram and TikTok @DavidLaneMusic, and on YouTube @davidlanemusic1

本期节目我们和《二分电台》的主播 2BAB 探讨了移动应用开发领域的技术趋势。AB 详细介绍了原生与非原生开发的区别，以及 Flutter、ReactNative 和 Kotlin Multiplatform (KMP) 等跨平台框架的特点。嘉宾们还分析了各种技术选型的优劣，例如 ReactNative 的热更新优势和 Flutter 的 UI 一致性，以及 Kotlin 作为 Android 官方语言的崛起。最后，节目还探讨了 On-Device 模型在移动设备上的应用前景，例如图像语义搜索和离线推理，并对 AI 技术在移动开发领域的潜在影响进行了展望。 嘉宾 2BAB (AB) 主播 laike9m Manjusaka 章节 00:14 移动端开发框架介绍与原生/非原生定义 07:03 ReactNative 的兴起、问题与 Flutter 的挑战 14:19 Kotlin Multiplatform (KMP) 与 Jetpack Compose 的发展 23:22 KMP 的流行度、ReactNative 的价值与未来发展 30:05 Electron 的妥协与热更新的重要性 37:43 入门移动端开发的建议与 Flutter 的未来 42:57 Flutter 的风险与 Kotlin 的竞争 48:45 On-Device Model 的应用与发展 55:10 On-Device Model 的功耗与应用场景 1:03:08 On-Device Model 的隐私与安全 1:10:03 总结与推荐 链接 React Native Flutter Kotlin Programming Language Jetpack Compose Kotlin Multiplatform (KMP) Compose Multiplatform (CMP) SkiaSkia is an open source 2D graphics library which provides common APIs that work across a variety of hardware and software platforms. It serves as the graphics engine for Google Chrome and ChromeOS, Android, Flutter, and many other products. The Truth About React Native - YouTube google/XNNPACK: High-efficiency floating-point neural network inference operators for mobile, server, and Web React Native Panel hosted by Jamon Holmgren - Chiara Mooney, Eli White, Keith Kurak, Chris Traganos - YouTube Gemini Nano litert-community/Gemma3-1B-IT · Hugging Face OpenAIDoc | 开发者友好的文档中心，一站式解决您的技术文档需求 《mono 女孩》

Entertainment Weekly Reports Alexandre Desplat to join Jurassic World Rebirth!~BUY PODCAST MERCH~https://www.jurassicparkpodcast.com/store~SOCIAL MEDIA | FOLLOW US~Website: https://www.jurassicparkpodcast.com/Twitter: https://twitter.com/jurassicparkpodInstagram: https://www.instagram.com/jurassicparkpodcast/Facebook: https://www.facebook.com/jurassicparkpodcast ~DON'T MISS OUR WEEKLY JURASSIC PODCAST~iTunes: https://apple.co/2VAITXfGoogle Play: http://bit.ly/2uV4kGRSpotify:  https://spoti.fi/2Gfl41T ~CHECK OUT OUR PLAYLISTS~Podcast Episodes: http://bit.ly/2P0Mqf0Toys & Merch: http://bit.ly/2VziQ2ETheme Parks: http://bit.ly/2UtOGBpJurassic World Live Tour: http://bit.ly/2IcRQmGLive Streams: http://bit.ly/2IdhxDhEvents: http://bit.ly/2UsXBD6~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Don't forget to give our voicemail line a call at 732-825-7763!Share this post and comment below! Enjoy.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Catch us on YouTube with Wednesday night LIVE STREAMS, Toy Hunts, Toy Unboxing and Reviews, Theme Park trips, Jurassic Discussion, Analysis and so much more.Your weekly podcast source for all things Jurassic Park, The Lost World, Jurassic Park 3, Jurassic World, Fallen Kingdom, Battle At Big Rock, Jurassic World Dominion, Jurassic World Live Tour, Camp Cretaceous and more! The Jurassic Park Podcast covers the films, the video games, live shows, theme park lands and rides, television shows, Mattel, LEGO, Michael Crichton, Steven Spielberg, Colin Trevorrow, Michael Giacchino, John Williams, Don Davis, Chris Pratt, Bryce Dallas Howard, Jeff Goldblum, Sam Neill, Laura Dern and everything else surrounding the Jurassic franchise.

Hello! We're starting a new project this month. We have almost 50 bonus episodes on our Patreon. The goal is to get these older bonus episodes out from behind the paywall. So travel back in time with us for poor recording setups, bad editing, out of date opinions, and so much more!Original Show Notes:Support Night Clerk Radio on Patreon Well friends, it seems like nothing gets us into a long conversation like the historical and sociopolitical landscape of corporate art, aesthetics, and marketing. We talk about everything from the post-WW2 rise of popular luxury to the current hellscape of flat purple vector people.Definitely check out the CARI link below and let us know your favorite styles.As always thank you so much for your support.Music Sampledｓａｔｕｒｎ　ａｕｄｉｏ　ｓｙｓｔｅｍ　ｄｅｍｏｎｓｔｒａｔｉｏｎ　ｃａｓｓｅｔｔｅ　茨ヱめ畝ヹ by wilson arcadeAdditional LinksThe Consumer Aesthetics Research Institute - CARINote: The original “Your Face Here” page on Airbnb Design is gone.Blame Corporate MemphisPopuluxe on WikipediaWhy do "Corporate Art Styles" Feel Fake?Corporate Art Style - Know Your MemeWill that ‘big tech art style' become the comic sans of art styles?Corporate Music - How to Compose with no SoulWe Are Sears 1986.mpgCreditsMusic by: 2MelloArtwork by: Patsy McDowellNight Clerk Radio on Bluesky

Get a free download of Amy Carmichael's poem here.  --------Thank you for listening! Your support of Joni and Friends helps make this show possible. Joni and Friends envisions a world where every person with a disability finds hope, dignity, and their place in the body of Christ. Become part of the global movement today at www.joniandfriends.org. Find more encouragement on Instagram, TikTok, Facebook, and YouTube.

Remember when nightlife was about being present instead of posting about it? Flash and Thaddeus Park take us on a nostalgic journey through Hawaii's golden era of clubbing, when every night of the week featured packed venues and genuine human connections.This conversation reveals how nightlife promotion evolved from guerrilla marketing tactics (running from security while distributing flyers in parking lots) to today's digital-first approach. Flash explains how successful promoters developed exceptional social skills by learning to connect with strangers face-to-face—a stark contrast to today's social media marketing. Both hosts lament how smartphone culture has fundamentally changed the clubgoing experience, with patrons more focused on documenting moments than living them.The evolution of DJing provides another fascinating lens, from DJs lugging crates of vinyl to today's digital setups. Flash recounts witnessing 8-Track demonstrate one of the first Pioneer CDJs—a revolutionary moment when digital began replacing analog. They discuss how veteran DJs like Taco, Compose, and Technique developed skills that some modern performers bypass entirely.Between sharing celebrity encounters and concert memories, the conversation turns heartfelt when discussing Super C-Dub's lasting legacy through the Aloha Cancer Project. Her remarkable ability to find the best in everyone exemplifies the authentic connection that defined Hawaii's nightlife at its peak.Ready for a dose of nostalgia that might just inspire you to put your phone down next time you're out? This episode reminds us what we've gained in convenience—and what we've lost in authentic connection—as technology has transformed how we experience nightlife and music. 

Jeeyoon Kim is a professional concert pianist who has achieved things that many classical musicians could only dream of, having performed in venues like Carnegie Hall and Stradivari Society. Jeeyoon Kim shares the personal story of her life and how, in spite of seemingly impossible odds, she was able to compose the life of her dreams. Subscribe for ad-free interviews and bonus episodes https://plus.acast.com/s/the-unmistakable-creative-podcast. Hosted on Acast. See acast.com/privacy for more information.

Avec Casse-noisette, ballet féérique qui nous plonge dans la magie de Noël, Tchaïkovsky parachève sa trilogie composée pour la danse après Le Lac des cygnes et La Belle au bois dormant. Mention légales : Vos données de connexion, dont votre adresse IP, sont traités par Radio Classique, responsable de traitement, sur la base de son intérêt légitime, par l'intermédiaire de son sous-traitant Ausha, à des fins de réalisation de statistiques agréées et de lutte contre la fraude. Ces données sont supprimées en temps réel pour la finalité statistique et sous cinq mois à compter de la collecte à des fins de lutte contre la fraude. Pour plus d'informations sur les traitements réalisés par Radio Classique et exercer vos droits, consultez notre Politique de confidentialité.Hébergé par Ausha. Visitez ausha.co/politique-de-confidentialite pour plus d'informations.