Hermit crabs adjust nutrient absorption not intake

24 09 2026
Hermit crabs adjust nutrient absorption not intake

Researchers found that hermit crabs constrained by undersized shells do not eat less to slow their growth. Instead, they increase defecation frequency, indicating a lower efficiency in converting consumed food into new tissue. This shifts the mechanism from reduced appetite to altered internal nutrient assimilation.

The finding suggests that growth regulation can occur at the level of metabolic processing rather than just external consumption. While this does not directly explain human weight variation, it highlights that intake and tissue growth are distinct biological processes.

more: https://www.futurity.org/hermit-crabs-growth-shells-3345862/?utm_source=rss&utm_medium=rss&utm_campaign=hermit-crabs-growth-shells-3345862

my other side: ivo.blogspot.com





Assertiveness Drives Promotion, Not Effectiveness

24 09 2026
Assertiveness Drives Promotion, Not Effectiveness

Research published in the Journal of Applied Psychology shows that employees who are assertive are more likely to be promoted into leadership roles. However, assertiveness does not correlate with the traits that make leaders effective. Effective leaders tend to exhibit high achievement motivation, a preference for order, strong self‑discipline, and optimism.

Organizations that rely on assertiveness as a hiring criterion risk appointing leaders who lack the skills to guide teams well. The study suggests re‑evaluating selection processes to focus on motivation, order, self‑discipline, and optimism instead. Still, the findings are based on aggregated past studies and may not capture all contextual variables that influence leadership success.

more: https://www.futurity.org/personality-leaders-3346102/?utm_source=rss&utm_medium=rss&utm_campaign=personality-leaders-3346102

my other side: ivo.blogspot.com





MilleMiglia generates realistic middle‑mile logistics data

24 09 2026
MilleMiglia generates realistic middle‑mile logistics data

MilleMiglia is a C++ library that produces synthetic middle‑mile logistics networks with up to 10,000 nodes and 100,000 edges. It runs on any POSIX system with a C++17 compiler and requires only standard STL, no external dependencies. The generator outputs JSON‑like files that match the format used by existing VRP solvers, effectively replacing proprietary datasets that are normally inaccessible.

It is useful for researchers and practitioners who need realistic benchmark instances without accessing commercial data. However, it abstracts away real‑world constraints such as traffic variability and carrier contracts, so it may not capture all operational nuances. If you need fine‑grained vehicle scheduling or integration with live routing engines, this tool alone won't suffice.

more: https://research.google/blog/millemiglia-a-realistic-instance-generator-for-middle-mile-logistics/

my other side: ivo.blogspot.com





MaxText config reproduces OLMo 3 7B on TPUs

24 09 2026
MaxText config reproduces OLMo 3 7B on TPUs

This is the MaxText configuration for reproducing OLMo 3 7B pre-training on Google Cloud TPUs. It covers stage 1 and stage 2, matching AI2's reference run on 5.93 trillion tokens. The setup uses JAX/XLA and a specific data pipeline to replicate the PyTorch training curve.

It is valuable if you want to verify that a PyTorch recipe translates faithfully to TPU hardware. The catch is that it requires access to Google Cloud TPUs and the specific OLMo 3 data mix. It is not a generic GPU tutorial, but a concrete validation of MaxText's numerics against a known baseline.

more: https://developers.googleblog.com/reproducing-olmo-3-7b-pre-training-in-maxtext-case-study-of-large-scale-training-on-tpus/

my other side: ivo.blogspot.com





Migrating MuJoCo Simulation to GPU Batches

24 09 2026
Migrating MuJoCo Simulation to GPU Batches

This guide walks through migrating an SO-101 robot arm simulation from standard CPU-based MuJoCo to MJWarp, which runs on NVIDIA GPUs. It batches 2,048 parallel environments using NVIDIA Warp kernels, allowing large-scale simulation steps without moving data off the device.

It is useful if you are scaling reinforcement learning workflows and need raw simulation throughput rather than single-step precision. The migration requires writing custom CUDA-centric logic and does not cover policy training, so it is strictly an infrastructure setup task for developers comfortable with GPU programming.

more: https://huggingface.co/blog/nvidia/how-to-use-nvidia-warp-and-mjwarp

my other side: ivo.blogspot.com





GoLive skill deploys agent apps to real infrastructure

24 09 2026
GoLive skill deploys agent apps to real infrastructure

This npm package installs a skill for Codex or Claude Code that takes your local app live. It handles hosting, database, auth, domain, email, and payments on your own accounts. The current alpha covers Vercel, Netlify, Supabase, Neon, Porkbun, GoDaddy, Resend, and Stripe test mode.

It is worth trying if you use a coding agent that supports skills and want to automate the post-build setup. The catch is that it is an early alpha, so you must verify the results yourself. It is not a hosted service or a replacement for manual configuration of unsupported providers.

more: https://github.com/mikehasa/golive-skill

my other side: ivo.blogspot.com





Caffeine Lowers Deep Sleep Even With Normal Hours

23 09 2026
Caffeine Lowers Deep Sleep Even With Normal Hours

A recent study using EEG found that caffeine consumption reduces the depth of slow‑wave sleep, the brain’s restorative phase, even when people still sleep the same number of hours. The researchers measured electrical activity in the brain and observed less slow‑wave activity after caffeine intake. The reduction occurs despite no obvious change in how long people stay asleep or how often they wake.

This suggests that people who drink coffee, especially late in the day, might not feel as rested even if they think they’re getting enough sleep. The effect varies among individuals and depends on dose, timing, and metabolism, so the findings don’t apply uniformly. Those who rely on caffeine for alertness should consider how it might affect the quality of their overnight recovery.

more: https://www.sciencedaily.com/releases/2026/09/260922005719.htm

my other side: ivo.blogspot.com





Bird feeding lifts well‑being

23 09 2026
Bird feeding lifts well‑being

A Virginia Tech‑led study found that 95% of participants reported a positive effect on their well‑being from feeding and watching birds, compared with 55% for other life events. The research used a standard mental‑well‑being questionnaire over a six‑month period and asked respondents to reflect on how the activity influenced them. The authors identified three key pathways: emotions, engagement, and a sense of meaning.

The results suggest that a simple, low‑cost activity can boost mood and reduce stress for many people. However, the study relies on self‑report from volunteers who already enjoy birdwatching, so it may not apply to the general population. Anyone who wants to feed birds should still follow best‑practice guidelines to minimise risks to wildlife.

more: https://www.futurity.org/feeding-birds-can-make-you-feel-better-3345882/?utm_source=rss&utm_medium=rss&utm_campaign=feeding-birds-can-make-you-feel-better-3345882

my other side: ivo.blogspot.com





cleanupper CLI for safe macOS disk cleanup

23 09 2026
cleanupper CLI for safe macOS disk cleanup

This Node.js tool scans your Mac for rebuildable files like Xcode DerivedData, npm caches, and browser logs. It moves these items to the Trash rather than deleting them, ensuring nothing is permanently lost. The installation requires Node.js 18 and runs entirely locally without network requests.

It is a useful alternative for developers who want to reclaim gigabytes without trusting a graphical suite. The main limitation is the dependency on Node, which may be unnecessary for some users. Avoid it if you need to clean system-level directories outside your user home folder.

more: https://github.com/SewCabinSpout/cleanupper

my other side: ivo.blogspot.com





K16s runs real kubeadm Kubernetes exam clusters

23 09 2026
K16s runs real kubeadm Kubernetes exam clusters

This self-hosted simulator provisions a full kubeadm cluster using Incus LXC containers as worker nodes. It supports three deployment modes: a disposable Lima VM on your laptop, a lightweight Docker kind setup, or a direct VPS install with 4 vCPUs and 8 GB RAM. Each mock exam question includes an automated setup script and a validator that grades your fixes against live cluster state.

It is worth trying if you are preparing for CKA or CKS and need practice with systemd and crictl on real nodes. The main catch is that AppArmor testing and specific worker node debugging are unavailable in the lightweight Docker mode, so you should use the VPS or laptop VM options for a complete experience.

more: https://github.com/immanuwell/k16s-exam-simulator

my other side: ivo.blogspot.com