So far Iran looks like the aggressive nation (who also funded Hamas and all acts of terror in the middle east instead of investing in drinking water infrastructure) thats going down, with the love of god, after 50 years since its revolution. Israel seems more resilient at the moment, only responding to savagery from Gazans. Remember October 7th and the hostages that could've ended it so much time ago
What do you mean by "could have ended it"? Hamas wanted to make a deal to release the hostages since October 10th in return for Israel releasing the 5,200 Palestinian prisoners. The entire point of the hostages was to exchange them. Israel is the one that didn't want to make a deal. That's why there were so many protests within Israel to "return the hostages". Those protests weren't about continuing the war, they were about pushing Netanyahu to end it and make a deal
A deep dive into Redis and Intel partnership that introduced a new Vector search compression SVS-Vamana. Discussion and benchmarks on various new quantization production-ready techniques
Very exciting!
I sympathise with the Redis "lego" way. But don't you think that hybrid search ha meaningful use cases? E.g searches that combines BM25 or piplned search in more sophisticated way than simple filter + knn / hnsw then filter?
Or maybe its more for a search engine than a cache
I think that hybrid is useful in many contexts, but that providing hybrid as a bundle with other filtering techniques, API-side, is not necessarily the best way. Just a trivial example, you want to filter by documents year for similarity. With Redis vector indexes you didn't even care about hybrid, you create different keys for different years. This is the usual Redis tradeoff, it's very similar to the past but applied to a novel field.
However I don't deny that in the future maybe, if elements of vector sets happen to be JSON documents (maybe just with the attributes you want to filter with, and a document ID), it may be useful to have VSIM options to filter by such scalars.
They permitted larger numbers of civilians being killed in the pursuit of terrorist barbarics who raped and murdered and dragged babies with their mothers into underground dungeons, and praised themselves for that while hiding between civilians.
Rapes have been officially recognized as having happened by the UN, i don’t see how that is hard to believe given Hamas filmed most of their atrocities on GoPros. Oct 7 was a brutal terror campaign regardless of your position on the conflict, denying that is absurd
> Following a mission lasting just over two weeks, the UN team found 'reasonable grounds to believe' that 'rape and gang rape' took place during the attack on October 7, 2023, while acknowledging the limitations of its own investigation
Read the report - there is zero evidence of rape that hasn't been thoroughly debunked. That "reasonable grounds" line was apparently added to placate Israel by the report's author (who has a history with Israel).
Even the NYT now admits (albeit quietly) that their "expose" contains no evidence.
And, as I'm sure you know, there are exactly zero videos showing rape. Quite frankly, I find it incredible that some people still believe the atrocity propaganda, even when it's so obvious and of such poor quality.
Dresden wasn't full of guerilla army disguised as civilians, and there weren't even the tools to target specifics as we have today. Any civilian was evacuated from north Gaza 5 months ago, secured by IDF under Hamas attacks. Those who stayed there insist on sticking to a war zone and risk to be labeled as terrorists. blowing up mosques and houses is nothing compared to raping innocents with no one to defend them miles away. And while they hold hundreds of the same innocents underground and keep provoking Israel's destruction, I see no reason to care about their precious houses. Israel is judged by Western morals in a barbaric medieval war that it was dragged into.
Vector search is no replacement for complex logic, its merely another data type like bm25 ir regular range filter.
Searchbase looks to me as better fit for intuitive data exploration, rather than replacing existing robust high throughput searches
the interest in vector database rise as an external, specialized service in a system that runs in addition to the "single source of truth" data lake where the data actually resides.
like Redis and Memcached before, it solves a specific problem. Redis started to act like a fully fledged DB (with weird persistency method and guarantees) only after it was really wide-spread.
sure, every DB will support vector and every vector search engine will act like a DB. But that's missing the point that vector search is an expensive problem with tradoffs that justify a specialized design
We recently added vector support to FeatureBase: https://cloud.featurebase.com/. What's interesting about using it with our existing b-tree storage layer is that I can now do set operations on the record's stringset fields before I do the similarity calculations on the vectors. I've been experimenting with laying out the vector space in a b-tree structure using this technique along with keyterm extraction, allowing for a type of semantic graph to be applied by the LLM to the vector space.
I think only 1 to 0, as the cosmic ray cause electrons to "discharge" hence lowering the voltage, assuming that 1 is a little bit higher voltage than 0. But far from an expert in electronics and physics :)
One flip of a bit in the memory of an onboard computer appears to have caused the change in the science data pattern returning from Voyager 2, engineers at NASA’s Jet Propulsion Laboratory said Monday, May 17. A value in a single memory location was changed from a 0 to a 1.
On May 12, engineers received a full memory readout from the flight data system computer, which formats the data to send back to Earth. They isolated the one bit in the memory that had changed, and they recreated the effect on a computer at JPL. They found the effect agrees with data coming down from the spacecraft. They are planning to reset the bit to its normal state on Wednesday, May 19.
Hi all, I have a bit off topic question but seems related.
I'm trying to write sort of a SQL compiler. The current goal is to analyze queries and find similarities, later maybe to translate between sql dialects.
I found Uber's QueryParser[1] but it's in haskell, so I started wrapping the python sqlparse[2] library and implement a Visitor to traverse their weird AST.
1. How close is it to implementing a compiler?
2. Is there theory you can suggest further reading for that matter?
3. Would you use a different language/library then I picked?
Depends on the complexity of your queries, but if you have a narrow subset that you're interested in, implementing a recursive descent parser for just those parts of the syntax that fits your problem like a glove could be a better solution.
I'm aiming to analyze all the BI queries in my organization, some of them are quite complicated and most of them are hundreds lines of sql.
Thanks for the direction, I'll dive into it
ZetaSQL[1] seems like it could be a fit for your use case. I've worked with Apache Calcite in the past and found it to be very complex to work with. I found ZetaSQL to be a little easier to use.
Then you should focus on the optimization phase. Where indexes, table statistics, etc are involved. The parsing and even translating sql from one dialect to another wouldn't give much for you to analyze.
I'm assuming you want to learn about query performance and execution plans?
Frankly, if you want to analyze queries, just install the rdbms, set up a data warehouse and use the tools available to analyze query plans/etc. There are so much that affects queries beyond the actual sql since the hardware and data load affects the queries. You could run the same sql query and end up with wildly different query/execution plans depending on how the data/tables/indexes are changed on your system.
