#test | Logs for 2025-04-04

« return
[00:01:54] <kolie2> ai https://hackaday.com
[00:03:47] <devbot> A Hackaday article delves into the history of the SILLIAC computer, a pivotal machine built for the University of Sydney during the late 1950s.
[00:03:49] <devbot> Based on the design of the US-based ILLIAC, SILLIAC marked a significant development in Australia's early computing landscape.
[00:03:50] <devbot> Illustrating the enormous scale typical of computers from that era, the article highlights its physical requirements:
[00:03:51] <devbot> > **"It occupied three rooms – one for the main cabinets (3 meters long), one for the power supply, and one for air conditioning needed to cool the vacuum tubes (valves)."**
[00:03:52] <devbot> The project's budget significantly increased from initial estimates, eventually costing around AU£75,000 – comparable to the price of ten Sydney homes at the time. Interestingly, part of this funding came from a donation related to a horse race.
[00:03:54] <devbot> For its operation, SILLIAC relied on 2C51 vacuum tubes. Though more expensive than common alternatives, these components contributed to the computer's reputation for reliability, allowing it to perform important work effectively until 1968.
[00:03:55] <devbot> Although SILLIAC was eventually dismantled, its legacy endures. Parts of the original machine are preserved in the Powerhouse Museum and at the University of Sydney, securing its place in computing history alongside other pioneering Australian computers like CSIRAC.
[00:10:28] <kolie2> ai https://arxiv.org
[00:12:26] <devbot> ```markdown
[00:12:27] <devbot> The arXiv paper "Higher multipoles of the cow" by Benjamin V. Lehmann critiques the widely used but often simplistic "spherical cow approximation" in physics analysis.
[00:12:29] <devbot> Lehmann highlights that this common approximation is frequently unjustified and proves insufficient for physical processes sensitive to non-spherical geometry.
[00:12:30] <devbot> > **"Here, I propose several schemes for extending the spherical cow approximation to a full multipole expansion, in which the spherical cow is simply the first term."**
[00:12:31] <devbot> This proposed multipole expansion enables a significantly more detailed and accurate representation of a cow's geometric structure and its resulting physical properties.
[00:12:32] <devbot> The paper details the methodology for computing these higher multipole coefficients, demonstrated using a "benchmark cow". It illustrates applicability to problems where spherical symmetry fails, like rotational spindown via gravitational waves.
[00:12:34] <devbot> This work provides a valuable framework for calculating bovine potentials and interactions with greater precision, moving beyond symmetry limitations and offering a way to define the cow's geometry more accurately.
[00:12:35] <devbot> ```
[01:29:25] <chromas> I forgot the model but I'm not at home now so I can't Czech
[01:30:17] <chromas> Using open-webui's baked in rag feature
[01:30:52] <chromas> Maybe it couldn't load the ny page or something
[05:04:35] <chromas> deepseek-r1:32b