HNHacker News
TopNewBestAskShowJobs

quizotic

401 karma · joined June 2, 2011

submissionscomments
quizotic··on Gödel's Incompleteness Theorems (2015)
Why? I honestly don't see how it's out of scope at all (majored in mathematics, minored in physics)
quizotic··on Gödel's Incompleteness Theorems (2015)
When I rant about it, I claim his theorem says "any system that reference parts of itself is either incomplete or inconsistent," and that the universe is such a system, as we're a part of it, and refer to parts of it.

So therefore, the universe is either full of magic (in a system with inconsistency, anything can happen), or mystery (there are true things that we can never prove).

My belief is that the quest to find a unified physics that describes everything is provably impossible due to Godel's theorem.

And on the rare occasions when I look for evidence of God, the fact that one of the things we can know for sure - is that we can't know everything - provides about as much comfort as I need.

quizotic··on [dead]
For those who aren't riveted by the gorgeous ... sweater ... the link to the relevant TED talk is

https://www.youtube.com/watch?v=Lp7E973zozc

TL;DR: when tempted by procrastination count backward from 5

quizotic··on What's new in C++17, with lots of examples
Most of it seems pretty real-world to me. But if IIUC that std::filesystem means I don't have to write separate code for windows and linux, then it's hard not to imagine EVERYONE using that.
quizotic··on SHA1 collisions make Git vulnerable to attacks
Interesting hypothetical ... even more interesting if you're implying it might not be so hypothetical
quizotic··on Warren Buffett's Annual Letter to Berkshire Hathaway Shareholders [pdf]
> the [antebellum] American south was probably the most advanced agriculture civilization ever

I don't follow. Certainly the current farming system in the American mid-west is orders of magnitude more productive. But even if we're focussed on pre-industrial history, are you sure that the ancient South American agri-economies weren't better? And for productive land use, what about Japan and south-east Asia, where every square inch ran like clockwork.

quizotic··on Warren Buffett's Annual Letter to Berkshire Hathaway Shareholders [pdf]
Yes, we took land from indigenous peoples, and enslaved Africans. And yes we did awful things.

But isn't it obvious that neither of these things created the wealth we have today? To my thinking, our current wealth derives from technological innovation, process engineering, psychological insight, most of which would not have been possible without welcoming people from other lands, and being willing to work with those who are different than we are.

quizotic··on Recommendations about coding in C++
If you're an experienced C++ coder, I'd skip this article.

While it's well structured, with real-world examples, it's very basic. I'm not an excellent C++ coder by any stretch of the imagination. I make lots of mistakes. But not these. The article is fairly thorough and long, so if you're experienced you'll invest non-trivial time for little or no reward.

quizotic··on What does Atari mean?
Sorry, I posted the same before seeing your response.
quizotic··on What does Atari mean?
I thought it was a term from Shogi, meaning "Check!", equivalent to the term in chess.
quizotic··on Why LINQ beats SQL
In the 70s, when I was all bushy-tailed, I cornered Gerry Sussman at MIT to tell him about "this new relational data model and SQL language". About two minutes into the conversation I saw him looking at me like "God, why have you chosen to torture me with this idiot."

Or maybe it was the look you have when you are SURE that someone is telling you an elaborate joke and suddenly realize that no punch line is coming.

LINQ: I'm looking at YOU :-/

quizotic··on Why LINQ beats SQL
Sigh. It's hard to know where to begin critiquing this.

I guess the TL;DR would be something like: "For constrained but not atypical data models (no N:M relationships, foreign keys explicit in the schema, etc), it's possible to have a more concise syntax than SQL, with less impedance mismatch."

But even that underwhelming statement is generous.

In the 1980s, the object database claim to fame was less impedance mismatch, and ODI's "coding by deletion", but at least ODBs coupled that with claims to superior performance for network data models.

Maybe it's just the grandiose title "LINQ beats SQL" that's so annoying. Simpler syntax would certainly be welcome for the relatively rare cases of humans generating SQL (I think most SQL is machine generated these days)... but only with a host of etceteris parabis conditions, prominently including performance.

quizotic··on What no one talks about when running and selling a business
It's ironic that you see him this way ... cause he's trying to say that seeing him this way is one of the things he hates and wishes was otherwise.

And by the way, it doesn't help if he goes quiet, never talking about this wealth or experience. Then he's just seen as aloof an heartless.

It's just true. Becoming wealthy changes things for most people. Maybe not Warren Buffet.

quizotic··on What no one talks about when running and selling a business
Interesting. I didn't have that reaction at all. I grew up in a ridiculously wealthy community, and saw the exact problems with wealth differences that he described. And I think he identifies an important truth: most of time, successful businesses are built by people who really love doing the business, and cashing out means giving that up. Sure, some people are lucky enough to recreate that feeling in a new situation. But lightning doesn't always strike twice. The author is just advising us to realize what we give up as well as what we gain.
quizotic··on How Do You Know a Developer Is Doing a Good Job?
In situations where agile/story-points make sense, I suppose developers that accomplish more story-points are "better".

In academia, articles that are referenced more frequently are "more important". For page ranking, pages that are referenced more frequently, are more important. I wonder if there's an equivalent with infrastructure programming? If a developer builds a subsystem that is used frequently, does that indicate the developer has provided substantive value?

quizotic··on Being black in tech can cost you $10k a year
Just so disturbing. We pride ourselves in tech on being a meritocracy. But it's not really true, even now. I'm a starchy-assed white guy, who was born in Asia and married a dark-skinned woman, so you'd think maybe I was different. But when I ran engineering organizations 25 years ago, I interviewed and hired precisely 0 African Americans and 0 Hispanics. I promoted two white women, and two white men, and nobody I recall of Asian, Indian or Middle-Eastern background.

But articles like this help. Keep pounding in the message, and eventually people will change ... or die off.

quizotic··on RethinkDB versus PostgreSQL: my personal experience
should also point out that Mongo is more than a jsonb store, and postgres is, at least currently, less than a distributed, "scale-out" database. Mongo supports problems that postgres cannot currently address. period. (and I'm a huge postgres fanboi)
quizotic··on RethinkDB versus PostgreSQL: my personal experience
And for those who may not know, but appreciate irony, Mongo now uses ... wait for it ... a somewhat stripped down, key-value-ish, embedded ... RDBMS as its underlying data store. Yes friends, "it's a dessert topping AND a floor wax"
quizotic··on RethinkDB versus PostgreSQL: my personal experience
> "Saving" 15 minutes of dev time because you don't know what your schema is going to look like is going to cost you orders of magnitude more time.

While I agree with the spirit of this, sometimes you'll never know what your schema is, because it's not under your control to begin with. That's why there are triple-stores and "document" databases and key-value stores now, and why there were deductive databases 30 years ago.

The world is wide, and there's room for more than one tool in the chest.

quizotic··on The Fallacy of Premature Optimization (2009)
I've seen Hive take minutes (OMG!) to count a table with 5 rows ... but (other) people still think it's OK, because it scales well. It's latency sucks for small data sets, but it can handle very large data sets.
quizotic··on The Fallacy of Premature Optimization (2009)
And I've heard it said that lowering latency is harder than increasing throughput ... though whether it's more important probably depends on the context.
quizotic··on The Fallacy of Premature Optimization (2009)
I've spent much of my career trying to optimize latency of long analytic-type queries; and I've been wrong on both sides of the issue. While it sounds stupidly tautological to say, I've come to think the important thing is being right in what I choose to optimize and when. With that bit of "wisdom" out of the way, there's a lot of ways the "premature optimization" phrase can be helpful:

* I might think I know what's important/critical-path and be wrong (it's happened)

* Users might be willing to live with slowness in some areas I didn't expect and demand performance in others

* More performant solutions might come to light later

* Particularly before product-market fit, features and their optimizations might be discarded

* New development might change the critical path

OTOH, I've found it's often beneficial to organize data and processing in a way that can take advantage of parallelism like multi-threading and vectorization. Often that's the kind of thing that's harder to do later.

quizotic··on How Does the SQL Server Query Optimizer Work?
Ah. Thank you for the correction!
quizotic··on How Does the SQL Server Query Optimizer Work?
As a "newer" engine, SQL Server has a couple of tricks that aren't found in most other systems, including intra-query parallelism across multiple threads. This capability requires occasional reshuffling of data between threads, which makes SQL Servers internals a little more like a distributed DBMS (think Netezza/Teradata) than the wonderful, but ancient, Postgres. All of which means that SQL Server has a larger palette of execution options and some different optimization decisions than PG.
quizotic··on How Does the SQL Server Query Optimizer Work?
I think you can just prepare them again, no? Of course the onus is then on poor you to know when that might make sense because of changes to the referenced tables. Though postgres is so incredibly open, you could probably do that with a trigger or a hook into analyze or vacuum or copy-from.
quizotic··on How Does the SQL Server Query Optimizer Work?
Exactly! We're so stuck in the implementation language of the 1980s: index, tablespace, partition, hash-vs-merge-vs-nestloop join .... that we lose sight of what matters. "I want these queries to be fast" ... genius.
quizotic··on How Does the SQL Server Query Optimizer Work?
> It's still quite hard to optimize queries especially if there are numerous JOINs involved

Sigh. In the real world (meaning 99+% of applications) there are a few big sets of data and lots of smaller sets of reference data. The thing that matters for performance is reducing that cardinality of the big stuff as quickly as possible. Which is a huge constraint on the search space.

Sure, in some theoretical sense, join ordering needs an A* algorithm or simulated annealing or some other esoteric approach to search. But in practical sense, it is almost a non-issue.

And here's the thing. Those systems with advanced search and costing for join ordering STILL make incredibly stupid decisions more than 3% of the time. And a bad decision can take hours to days and dominate all the good decisions.

DeWitt made incredible contributions to database architectures. He's a great guy and I'm a huge fan. But that was 30 years ago. Where are today's DeWitts?

quizotic··on How Does the SQL Server Query Optimizer Work?
No. Not true at all. At around a billion records, you start getting into distributed/scale-out systems, and distributed indexes are performance death whether for OLTP or analytics. Even before you cross into distributed query processing, the random access nature of indexes levies a heavy tax. Indexes are a tiny piece of this issue. They address the problem of finding a single needle in a haystack, but if you need to aggregate lots of data, indexes are largely worthless.

What about different compression schemes for different sets of "column" values - with the idea that you execute against the compressed values, rarely ever decompressing. Even column store systems that allow you to specify a compression scheme aren't adaptive. Why the hell not? This is almost 2020.

quizotic··on How Does the SQL Server Query Optimizer Work?
A profoundly important observation. Yes, using a limited set of physical structures simplifies the code and, in some cases, streamlines execution. But that's putting lipstick on a pig. With a handful of exceptions, database architectures were set in stone in the mid-1980s in Wisconsin. Which is ridiculous. The machines that they were designed to optimize don't even exist any more. Take simple storage tier-ing as an example. Why don't infrequently accessed column values migrate to slower cheaper storage automatically? Sure, databases have concepts of tablespaces and partitions, but those are WAY too coarse grained for today's environments. And there are more than 100 issues like this. In short order, NVMe SSD will be ubiquitous, with 3Dxpoint on the horizon. The data structures that make sense for spinning disk are just not equally optimal for the performance characteristics of SSD, Ram, and cache.

Regarding your point about reads and writes on the same store, you might take a look at the new "HTAP" systems (hybrid transactional and analytic). The VectorWise guy has a "new" (probably old by now) project called HyPer

quizotic··on Stealing passwords from McDonald's users
Curious! Ten days after a public disclosure, and no statement from McDonald's? Have they fixed the problem? Do they not know?
← PreviousPage 2 of 5Next →