News for hackers. Stuff that matters.

erthink

erthink on Hacker News

Re:Roblox October Outage Postmortem on Monday January 31, 2022 @04:26PM (#30154655)
These issues partially solved in the libmdbx (a deeply revised and extended descendant of LMDB).

So BoltDB and LMDB affected users may switch to libmdbx as the Erigon (Ethereum implementation) does year ago https://github.com/ledgerwatch/erigon/wiki/Criteria-for-tran...

For now this is (relatively) easy since bindings for GoLang, Rust NodeJS/Deno, etc are available and the API is mostly the same in general.

---

The ideas that MDBX uses to solve these issues are simple: zero-cost micro-defragmentation, coalescing short GC/freelist records, chunking too long GC/freelist records, LIFO for GC/freelist reclaiming, etc.

Many of the ideas mentioned seems simple to implement in BoldDB. However the complete solution is not documented and too complicated (in accordance with the traditions inherited from LMDB ;)

Compact and fast double-to-string conversion

Posted by erthink on Tuesday February 04, 2020 @09:44AM
from the self-post dept.

Ryu algorithm by Ulf Adams (https://github.com/ulfjack/ryu) known to fastest in much cases, especially for short string representations. However, I would like to draw attention to my own speed-competing but more compact GRISU-based implementation.Briefly, about this implementation:1. it is a GRUSU-based, but not exactly the GRISU2;2. for now produces only a raw ASCII representation, e.g....

( Read More... | no comments | 4 points | on HN )
Re:Ask HN: What is good enough C/C++ benchmark suite for sorting algor... on Monday January 20, 2020 @08:11AM (#22098533)
For now I found the Christopher Swenson's collection (https://github.com/swenson/sort) . This is more than nothing, but it is clearly not enough.

Ask HN: What is good enough C/C++ benchmark suite for sorting algorithms?

Posted by erthink on Monday January 20, 2020 @08:10AM
from the ask-hn dept.

Of course, this is a very broad topic. Nevertheless, is there a ready-made set of tests that shows the advantages and flaws of well-known sorting algorithms in the context where the std::sort or qsort() are used?Actually I created some improvements to a quicksort-based algorithm and it work pretty for target scenarios (i.e. faster than std::sort, that based on introsort). So I want to...

( Read More... | 1 comment | on HN )
Re:LMDB – Lightning Memory-Mapped Database Manager on Wednesday November 14, 2018 @02:23PM (#18452976)
All improvements over LMDB are listed here = https://github.com/leo-yuriev/libmdbx#improvements-over-lmdb
Re:T1ha: even more Faster and Better hash functions family on Tuesday August 14, 2018 @10:12AM (#17758055)
t1ha aka "Fast Positive Hash" - In most cases up to 15% faster than StadtX, xxHash, mum-hash, Metro-hash, etc.

Simple benchmark is included (git close && make). The quality and speed could be checked by Reini Urban's (https://github.com/rurban/smhasher) and Yves Orton's (https://github.com/demerphq/smhasher) forks of SMHasher suite.

Reasonable to note: t1ha have two injection points of each data word inside the loop of entropy sponge.