Ming Ying

@paradedb, professional tennis wannabe, stoked on Stok coffee

[Ming Ying @themingying]

@zhoujinjing09 @philippemnoel @fulmicoton @paradedb While our indexing throughput is satisfying for the majority of production use cases we've seen, improving throughput is one of our top engineering priorities.

[Ming Ying @themingying]

@zhoujinjing09 @philippemnoel @fulmicoton @paradedb We do commit on every transaction. If there are lots of small transactions, this would indeed lead to many small commits, which Paul Masurel explains leads to higher segment merge times (see "Indexing Latency vs Throughput, search performance" in https://t.co/q5rYrrJ8wC).

[Ming Ying @themingying]

@zhoujinjing09 @philippemnoel @fulmicoton @paradedb That being said, we've done thorough benchmarking around this and have found that, while waiting for a Tantivy commit does increase per-transaction time, both search and indexing times are stable as the number of small commits increases.

[Paul Copplestone - e/postgres @kiwicopple]

this is cool:

https://t.co/D9xfS5an0v

◆ Postgres extension for OLAP data
◆ Stores data as parquet
◆ Uses DeltaLake with plans for Apache Iceberg

tldr: turns Postgres into data warehouse
by: @paradedb

[Ming Ying @themingying]

@roshanpadaki @philippemnoel You can’t spell jail without aji

[Ming Ying @themingying]

@roshanpadaki 52

[ParadeDB @paradedb]

We're excited to unveil pg_bm25, a Postgres extension that significantly enhances Postgres' full text search capabilities. It's our first step in eliminating the need to bring ElasticSearch into the data stack.

https://t.co/nUqRKEKHn6

[Ming Ying @themingying]

@lessin @lessin I find Calendly irritating but much less so when offered as an alternative to traditional scheduling options. e.g."Let's find time to chat, when are your availabillities? Or, if it's easier, here is my Calendly" sounds 100x better than "Please book time on my Calendly."

[Ming Ying @themingying]

@_wenwei_ @mehgcap What if Chrome never used more than 0.3GB of RAM? Check out @whisthq

[Ming Ying @themingying]

@KyleAJablonski Hey Kyle, check out @whisthq, which we built to solve this exact problem.

[Ming Ying @themingying]

@AudreLawdAMercy We built @whisthq to tackle Chrome's tab memory usage problems. Below: a side-by-side comparison of how much memory Chrome uses vs. Whist when you have ten tabs open.

[Ming Ying @themingying]

@CCuster19 @AudreLawdAMercy In case she's interested...we built https://t.co/vMJnciScC8 to solve this exact problem!

[Ming Ying @themingying]

@jostewart @mattolwell We built @whisthq to help silence the horror and shame your computer's battery and fan experience when you have all those tabs open :)

[Ming Ying @themingying]

@nxify We built @whisthq to never use more than 300MB of RAM, no matter how many tabs you have open.

[Ming Ying @themingying]

@salleeharrison We built @whisthq to use 10x less memory and energy than Chrome!

[Ming Ying @themingying]

@msmodernme @whisthq lets you keep all 162 tabs open without slowing down your computer or using more than 300MB of memory. We've been building it for the last few years, check us out!

[Ming Ying @themingying]

@ThatguyFalcone @Lord_Caboose @MintBlitz We built @whisthq to be a Chrome browser that uses 10x less memory than Chrome by offloading tabs to the cloud.

[Ming Ying @themingying]

@arpitrage @HoffProf @brave @googlechrome We built @whisthq to use 5-10x less battery and memory than any Chromium-based browser by offloading tabs to the cloud.

[Ming Ying @themingying]

@sukiissanoob We built @whisthq to never slow down your computer no matter how many tabs you have open by offloading tabs to the cloud.

[Ming Ying @themingying]

@nocrownking97 Check out @whisthq -- we're building a faster version of Chrome that uses 10x less power and memory.