Why Modern Math Education Is Obsolete

a reply to: ChaoticOrder

readable, that's for sure. hardly optimal performance-wise, but that's js, so the engine takes care of most stuff, and i'm kinda obsessed with code performance anyway.

i just hope you've used something else than js for more complicated stuff. also, i won't try to extrapolate that small piece of js into a larger picture - it really doesn't prove much.

anyway, you may want to take a look at "thinking forth" - ebook is available for free. i kinda doubt the language itself will be your cup of tea, but you may like some of the ideas, especially generic ones about building algorithms.

a reply to: jedi_hamster

i just hope you've used something else than js for more complicated stuff.

I always use PHP, JavaScript, and MySQL for website development. Although I will create my own file system based database code instead of MySQL when I need it to be super fast because MySQL is quite inefficient when it comes to handling large amounts of data.

When it comes to desktop software I typically just stick with good old C++ because it tends to be fast and powerful and it's just what I know best. I can code in Object Pascal but I don't do that very much anymore because I prefer C++ for most things. The old Delphi IDE is still good for creating some things though.

I also know quite a bit of Java because I watched an excellent university lecture series on Java programming but I don't really use it for anything because the popular IDE's seem to make everything twice as complicated as it needs to be. The language its self seems quite attractive but I just never really had a use for it.

a reply to: ChaoticOrder

i was talking about desktop software - it can be written in js, but results tend to vary

java equals to "boilerplate clusterf..." - if you're fine with C++, don't bother, unless you're into android or other mobile stuff. it's still used on servers, but i guess not as much as some time ago, thanks to node.js. for desktop it's pretty pointless, same with the web.

as to mysql speed - i never really bothered to use it, because i hate SQL, but isn't the main approach to store really large amounts of data outside of the database, while storing paths to those files inside? replacing whole mysql database with your own solution seems to be an overkill - i know, i did that over a decade ago.

a reply to: jedi_hamster

but isn't the main approach to store really large amounts of data outside of the database, while storing paths to those files inside?

I don't mean large amounts of data as in filesize, I mean a large number of database entries. I was able to get about a 50x speed increase in one of my applications by replacing MySQL with my own system.

originally posted by: ChaoticOrder
a reply to: jedi_hamster

but isn't the main approach to store really large amounts of data outside of the database, while storing paths to those files inside?

I don't mean large amounts of data as in filesize, I mean a large number of database entries. I was able to get about a 50x speed increase in one of my applications by replacing MySQL with my own system.

that sounds like you're doing something wrong then.

stackoverflow.com...

I highly doubt I would get anywhere near a 50x speed up even if I tried all the optimizations suggest in your link. If that guy thinks 7000 queries is a lot he should try doing several hundred thousand at once. MySQL just has a lot of overhead and is not exactly known for it's speeds, it's not a good option if you need maximum speed. But if you need lots of nice features and don't really care about speed very much because you're not inserting or updating a huge number of entries at one time, then using MySQL is a good option.

a reply to: ChaoticOrder

you doubt it. yet you didn't test it. mysql-level optimizations, database server configuration, it all has effect on the overall performance. and as you've said, mysql has tons of features.

it was also tested in tons of scenarios and works well. your own solution, even if relatively bug-free, certainly has a lot of limitations compared to mysql.

in that regard, comparing it to mysql in terms of speed is simply silly. even more so considering the fact that you can't even prove that it's impossible to reach the same speed with mysql alone.

and if it is possible - then wether your choice is good, depends on the quality of your solution. can you code a database server that matches mysql quality-wise, when you can't even optimize your usage of mysql?

somehow, i doubt it.

The point I was making is that if one is willing to give up the extra features they can get the extra speed. Obviously I was not trying to recreate a fully featured database server, I was trying to solve a specialized problem involving a very large amount of queries. It was extremely obvious to me that MySQL was orders of magnitude too slow and nothing I did was going to make it work as fast as my custom solution. I think most of the slow down was caused just by communicating with the MySQL server. None of that overhead existed with my custom solution because I was writing data directly to disk and using a fast seek operation to quickly read from files. When you have several hundred thousand queries to worry about it becomes important to cut out as much overhead as possible so that's what I did, and it made all the difference in the world.

a reply to: ChaoticOrder

communicating equals to two things - connection itself and command parsing. first thing can't be a problem, second thing can be significantly improved.

as for disk access, mysql has caches for that.

if you would get 50x speed increase - or even 10x - vs heavily optimized mysql queries - fine. but, as you've said, "i highly doubt [...] even if i tried".

in simple terms, you've traded good solution for worse solution, just because it was easier to make that worse solution work fast enough for you. good - and lazy - programmer would do whatever he can to speed up mysql, because it's less work than coding your own database solution - unless the whole scenario is so simple that using mysql for it was an overkill from the very beginning.

but we're going offtopic here, and my earlier points stand. this mysql subject may at best prove that you lack technical expertise when it comes to database server configuration and/or lack will to research the best possible solution, going with the easiest possible one that you know how to do already. it doesn't matter, it has nothing to do with math.

did you/could you code a raytracer? and if yes, what was/you think would be the speed (and with what set of features)? more or less, an order of magnitude.

There is too much unnecessary math in one course first place. 2nd math should be taught and used at the same time, especially in engineering/construction. This makes most math obsolete and become useless homework stressing for kids.

Teachers expect way too much. This is the true reason for society downfall.

"Do not expect the expected."

Just look at immigrants coming to learn in college/uni. I've seen too many teachers expecting immigrants to know English for example. That is like bringing Aliens over to our planet and expect them to be able to talk to us.

Teach what we came here to learn for, not what they don't need to learn or use.

I've been in 2 different programs so far and College has turned into # with high cost and useless courses. If only the Teacher have taught what I signed up for like in High school I would probably be done by now.

Today, Schools are unlike any past/ancient schools. We use to have 1 teacher who teachers 1 specific thing. Like if I wanted to be a black smith, a blacksmith would teach me all the way. Now it is just dumb courses, full of random homework and many random teachers stressing kids out with different time schedules.