malisper

39 ms·

391.
by malisper 10y ago
Your queries are executing different plans. The first one is executing a nested loop join which filters out 1,246,035,384 intermediate rows. The second one is executing a index join which doesn't filter out any intermediate rows at all
392.
by malisper 10y ago
> but it does bring the query response time down to the same order of magnitude as the non-JSON table. In the specific example given it might, but you will still wind up with a handful of queries that are planned wrong and are orders of
393.
by malisper 10y ago
Using @> instead of ->> only causes the selectivity estimate of the predicate to be a different hard coded estimate. It doesn't fix the underlying problem of Postgres not keeping statistics on JSONB.
394.
by malisper 10y ago
The assumption that the computer you are modeling has an infinite amount of memory finite number of caches is a useful assumption as it dramatically simplifies the analysis and still allows you to use big-O to somewhat accurately analyze th
395.
by malisper 10y ago
The article demonstrates that memory access is an instruction that actually has O(n^1/2) performance.
396.
by malisper 10y ago
The hypothesis of the statement doesn't hold. O(n log n) isn't greater than O(n log n).
397.
by malisper 10y ago
If you make the assumption that there is an upper bound on the maximum time a read can take (e.g. reading from disk), reading from memory becomes an O(1) operation regardless of cache misses.
398.
by malisper 10y ago
The definition of big-O is what happens as n approaches infinity. You are free to model your computer as having O(1) memory access as it's a much simpler model and close enough in practice.
399.
by malisper 10y ago
In part two, the author demonstrates that if you packed information as densely as possible, i.e. your computer is a black hole, the best you could do is O(n^1/2).
400.
by malisper 10y ago
This advice applies a ton to databases. At my job, our application is heavily disk i/o bound, with a database query typically taking several seconds. I benchmarked how different sorting the rows in our db by time would be compared to o
401.
Three Optimization Tricks for C++(facebook.com)
3 points by malisper 10y ago|0 comments
402.
by malisper 10y ago
> you should be prepared to implement a red-black tree in C, followed by an A* pathfinding algorithm on the whiteboard. I don't know about that, but you should at least know the algorithms necessary to traverse the DOM.
403.
by malisper 10y ago
> The BRIN index should be ideal for situations where there is a huge storage of data (TBs) inserted in sequential order and rarely queried - a BRIN index for such a thing would be only a few MB in size. I say should because when I tried
404.
by malisper 10y ago
From the abstract: > The synthesis algorithm takes a generate-and-test approach: it generates an infinite stream of all the expressions having the same type as f; then, they are tested against the given predicate. I'm really impress
405.
by malisper 10y ago
I tried out Smalltalk for a day and I was completely amazed by the some of the features that still haven't made it into other languages. One of the features that I remember in particular was the ability to search for functions by examp
406.
by malisper 10y ago
Algorithm W, the algorithm used to do type inference in Hindley Milner is effectively just unification (the process by which Prolog infers facts, as mentioned in the post you referenced), applied to types.
407.
by malisper 10y ago
How much of a speedup do you get by adding JSDoc type annotations?
408.
by malisper 10y ago
In one of the later chapters, the book Lisp in Small Pieces (great book), shows you how to write a a Scheme compiler to C. The general idea is you create a structure that contains a union type of all types in the dynamic language, and also
409.
by malisper 10y ago
Index only scans work in some cases but not all. In order for an index only scan to be more efficient than a regular index scan, the visibility map needs to be up to date. This is a problem because (currently) the only thing that updates th
410.
by malisper 10y ago
Do you have any tips for how a company should approach building its own CTF?
411.
by malisper 10y ago
You typed miles per second and not miles per hour.
412.
by malisper 10y ago
> If each firing requires 25Mw, then 1000 shells would require 25Gw That's not how power works. Power is energy/second, so if one each firing requires 25Mw, firing every round requires 25Mw. Edit: Although after doing some calc
413.
by malisper 10y ago
If anyone's curious, I wrote a blog post a while ago on how to write programs that obtain their own source code: http://malisper.me/2016/04/20/writing-self-referential-progr...
414.
by malisper 10y ago
> I've heard debugging Lisp macros can be quite hard It's much easier now that there is slime-macrostep[0]. Hmm. Maybe I should write another debugging lisp post on it. [0] http://kvardek-du.kerno.org/2016/
415.
by malisper 10y ago
I posted this on the blog but I'm copying it here because I have to wait for approval on the blog. If you didn't know, Clojure's dispatch mechanism is based on the Common Lisp library, filtered functions[0], which is descri
416.
by malisper 10y ago
I'm sorry, but I'm having a hard time understanding why the constraints are contradictory. For example, can't you just have a thin countertop such that the corner is within 18 inches of the edge and put an outlet there?
417.
by malisper 11y ago
What were some of the query features that caused you to choose Elasticsearch over pg_trgm, because pg_trgm supports regular expressions out of the box.
418.
How to Write a Quine(malisper.me)
2 points by malisper 11y ago|0 comments
419.
by malisper 11y ago
My reason is that there are eight different folding functions: foldl, foldr, foldl1, foldr1, foldl', foldr', foldl1', and foldr1'. There are twelve if you count: scanl, scanr, scanl1, scanr1. Also, the implementation of
420.
by malisper 11y ago
From the source you gave: > Full Year 2015: > Net Income of $197.5 million > Adjusted Net Trading Income of $500.7 million His numbers seem pretty accurate.

More ›