Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
mastodon.gamedev.place is one of the many independent Mastodon servers you can use to participate in the fediverse.
Mastodon server focused on game development and related topics.

Administered by:

Server stats:

5.1K
active users

mastodon.gamedev.place: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.7

#compsci

34 posts18 participants2 posts today
Public
Compsci Weekly

Optimizing Barrett Reduction: Tighter Bounds Eliminate Redundant Subtractions

blog.zksecurity.xyz/posts/barr

Discussions: discu.eu/q/https://blog.zksecu

blog.zksecurity.xyzOptimizing Barrett Reduction: Tighter Bounds Eliminate Redundant Subtractions - ZKSECURITYBarrett reduction is a widely used algorithm for reducing a value modulo $m$. Our analysis, conducted during the Rust p256 crate audit, shows that the error bound for Barrett reduction can be tighter than traditionally assumed. For most moduli used in cryptography (e.g., NIST curves), the quotient approximation error is at most 1 (not 2). This improvement eliminates the need for a second subtraction in practice. By adopting this optimization, RustCrypto p256 achieves a 14% performance improvement in scalar multiplication.
#compsci#cryptography
Public
Compsci Weekly

Methods for IP Address Encryption and Obfuscation

datatracker.ietf.org/doc/draft

Discussions: discu.eu/q/https://datatracker

IETF DatatrackerMethods for IP Address Encryption and ObfuscationThis document specifies methods for encrypting and obfuscating IP addresses, providing both deterministic format-preserving and non-deterministic constructions. These methods address privacy concerns raised in [RFC6973] and [RFC7258] regarding pervasive monitoring and data collection. The methods apply uniformly to both IPv4 and IPv6 addresses by converting them into a 16-byte representation. Two generic constructions are defined—one using a 128-bit block cipher and the other using a 128-bit tweakable block cipher—along with three concrete instantiations: * *ipcrypt-deterministic:* Deterministic encryption using AES128 (applied as a single-block operation). * *ipcrypt-nd:* Non-deterministic encryption using the KIASU-BC tweakable block cipher with an 8-byte tweak. * *ipcrypt-ndx:* Non-deterministic encryption using the AES-XTS tweakable block cipher with a 16-byte tweak. Deterministic mode produces a 16-byte ciphertext (enabling format preservation), while non-deterministic modes prepend a randomly sampled tweak (which MUST be uniformly random when generated, as specified in [RFC4086]) to produce larger ciphertexts that resist correlation attacks. Discussion Venues This note is to be removed before publishing as an RFC. Source for this draft and an issue tracker can be found at https://github.com/jedisct1/draft-denis-ipcrypt.
#compsci#cryptography
Public
Rachel Wil Singh ~ Moos-a-dee

Something new I'm trying this semester in my classes: I just have one big cumulative test, with it split into 3: multiple choice concept questions for 70%, then some coding and debugging questions for up to 90%, then a couple coding questions that combine topics for the last 100%.

Every question comes out of a pool, so the test is random.

Students can retake the test once a week, and when I grade an attempt I tell them what kinds of questions they got wrong.

(Continued)

#Education#CompSci
Public
Compsci Weekly

Qwen3: Think Deeper, Act Faster

qwenlm.github.io/blog/qwen3/

Discussions: discu.eu/q/https://qwenlm.gith

Qwen · Qwen3: Think Deeper, Act FasterQWEN CHAT GitHub Hugging Face ModelScope Kaggle DEMO DISCORD Introduction Today, we are excited to announce the release of Qwen3, the latest addition to the Qwen family of large language models. Our flagship model, Qwen3-235B-A22B, achieves competitive results in benchmark evaluations of coding, math, general capabilities, etc., when compared to other top-tier models such as DeepSeek-R1, o1, o3-mini, Grok-3, and Gemini-2.5-Pro. Additionally, the small MoE model, Qwen3-30B-A3B, outcompetes QwQ-32B with 10 times of activated parameters, and even a tiny model like Qwen3-4B can rival the performance of Qwen2.
#compsci#machinelearning
ExploreLive feeds
About