Toilets and race conditions!

A co-worker asked me what issues multi-threading and shared memory can have. So I explained him that stuff with the lock. He wasn't quite sure whether he got it.

Me: imagine you go to the toilet. You check whether there's enough toilet paper in the stall, and it is. BUT now someone else comes in, does business and uses up all paper. CPUs can do shit very fast, can't they? Yeah and now you're sitting on the bowl, and BAMM out of paper. This wouldn't have happened if you had locked the stall, right?

Him: yeah. And with a single thread?

Me: well if you're alone at home in your appartment, there's no reason to lock the door because there's nobody to interfere.

Him: ah, I see. And if I have two threads, but no shared memory, then it is as if my wife and me are at home with each a toilet of our own, then we don't need to lock either.

Me: exactly!

Knock Knock. Multi-threading. Who's there?

I guess my best AHHA moment was back when I learned that good code is simple code.

When I started out I wanted to prove myself by showing of how good of a programmer I was(and which I retrospectively wasn't) , which basically meant to use every high level concept I was aware of whenever possible. Multi threading where linear execution would have been totally okay, polymorphism with x meta classes where a switch would have been enough, all that shit.

It wasn't until I had to guide the first person through that mess of useless ego stroking that I found out how much time and money I wasted by not going with the easiest approach that solves the problem.

Took me some time to fully lay off that attitude but it surely was one of the most influential moments of my career.

Doing linguistic research where I need to parse 2000 files of a total of 36 GB. Since we are using python the first thing I thought was to implement multi threading. Now I changed the total runtime from three days to like one day and a half. But then when I checked the activity monitor I saw only 20 percent of the CPU usage. After a searching process I started to understand how multi threading and multi processing works. Moral of the story: if you want to ping a website till they block you or do easy tasks that will not use up all power of one core, do multi thrading. If you need to do something complicated that can easily consume all the powers of a single CPU core, split up the work and do multi processing. In my case, when I tried to grab information from a website, I did multi thrading since the work is easy and I really wanted to pin the website 16 times simultaneously but only have 4 cores. But when it come to text processing which a single file will take 80 percent of cpu, split it up and do multi processing.

This is just a post for those who are confused with when to use which.

Oh boy some mutex deadlocks inside the 16 year old, unmaintained, company application framework.

Time to look at the stack traces of 24 different threads and try and guess which one fucks it up

(Send help)

Trying to run a process on each CPU thread in our vps only to find that our vps has just one CPU and no multi threading. :/

We need an upgrade.

Multi vs mono threading
😂😂

There's a lil tool someone wrote that runs as a post build event for one of our projects. I got bored of watching that tool hog a single CPU thread for like half a minute each time I had to build, so I tried making the thing multi-threaded...

...and it worked! Shit runs on all threads for just a few seconds now 💪🏻

Multi-threading in PHP be like

Yeaaahhh that moment when the program flawlessly crunches through ten thousands of files, only limited by the slowish HDD! :-)

In full multi-threading, tons of dynamic buffer resizing, pointer shit left and right, also two star programming, and everything written in raw C!

Best:
Really getting into Rust. It has taught me so many things.
1. Null is evil
2. Sum types are amazing
3. Compiler can actually have good error output
4. Multi threading is actually really scary if you don't have a compiler to back you up

Worst:
I had to deal with SSIS. It has also taught me many things:
1. No matter how 'mature' a product is, it can be awful. Simply dump a random error code, the user can figure out what went wrong, no need for good error messages.
2. The modern concept of the database is crap. It's a gigantic global state that is used by everyone and owned by no one.
3. Don't use tools that aren't made to be used with version control.
4. Even when you tell your team that it's bad, you will be ignored.

** this means words are muted **

Friday:
I send a mail the client a Google doc with elaborate details about evaluation of an Android tablet from a Chinese manufacturer.

Monday:

The client is upset, he says "You say there is no GPS chip on the tablet while the manufacturer says otherwise"

Me- "I have clearly mentioned that it has a GPS chip"

Client- Opens the Google doc, points to a sentence. Looks at me like I did something horrible.

Me - **This guys is either word blind or something else is wrong with him, the line reads 'GPS chip available'**

Me- "Look, it says 'GPS chip available'.

Client- **Blinks n blinks again** "Alright, but why did you share a Google document, why not PDF, docx"

Me-**Politely** "You can download the document in any format, look I will show you..."

Client- "It should have been in the mail itself ideally"

Me- **WTH** "We normally maintain a document for such things to keep everything organised, but if you want I will put everything in mail itself"

Client- "Hmm.. do both from next time"

Me- "Alright" **BS**

Client- "Why is the new feature taking so much time"

Me- "As planned earlier, we going to deliver it tomorrow"

Client- "Why not today??" **Gives a strange look.**

Me thinking - **Enough**

Me- "See, I am trying to integrate a smarten with a socket connection, reading it's data via exposed APIs that are hardly documented, we need faster performance so I need to implement caching, multi threading, offline handling, multiple processes to avoid memory fluctuations, sync adapter to sync data...."

Client- "Ok ok ok, it's fine if you give working build tomorrow"

Me- "Ok, fine"

#limit

It's always so funny when a person starts using multiprocessing in Python, because if there's "multi", obviously THIS is the thing that should free the person from a headache of having GUI frozen. You know, because it does "multiple" ehm... stuff..... at once....... yeah. And it's popular, it must work for me too! Oh how often I see this. :D

Stupidly enough that's not entirely a user's fault, but Python's as naming things with "multi" doesn't end up well basically with anything. I bet if there was such thing as multipointer in C half of the beginners would be totally fucked and the other one would just break their machines beyond repair with a joy.

Yet... reading the damn documentation should be a requirement before using threading or multiprocessing to prevent the confusion, because there's this funny difference between multiple threads and multiple processes which will haunt you unless you see what's what and use it correctly.

Multi-rant incoming!

1.
Stereotyping.

When did that shit become the norm?

2.
I'm lost in ROS smach. Does anyone do multi-threading in ROS services, or should I flip a table on smach? 😒

When some colleague ask for help in a project that you're supposed to know everything but you really don't (damn multi threading in Java )

When I was on my first internship, I started developing an Android app, while my friend developed a C# program that read a .txt with info and references from a mail service (in my country it's CTT).

The damn .txt files got really really big, na she had to display all of the data in a listbox (it was a PoC) and when he pressed the item, it had to fill some fields at the left of the listbox.

Needless to say, he didn't learn of multi-threading yet, and I had, so I taught him how to multithread so the app wouldn't lock up while loading the massive .txt file.

The listbox filling made a cool animation (like CMD executing commands from a bat file) and we even implemented a progressbar.

I felt like a badass Dev after that.

A third party jar did nit support multi threading in our project. After the module was multi threaded it got stuck after few runs.

I felt like superman after finding the bug.

wrote dumb threading and sockets system to make a bunch of https calls

extracted the working code into nice objects and methods to make it look sane and so I can re-use the code for more complex functionality that builds on itself

now suddenly the threads are locking and not multi-threading anymore

turns out?
the http(s) library expects the tls / https thing in an Arc / atomic reference count
but despite Arc being literally intended and designed for threadwork it seems the library in question throttles / locks itself if they are all using the same Arc (I don't even know how that is possible?)
if I clone the tls / https thing, no throttle / lock issues 😒
why did they even try that lol, they didn't test?!

I really didn't expect to be better at multi-threading than others already. I'm newbie. pls

So funny thing happen yesterday night. I was attending a small talk at a meeting here in our town where one guy had to present some unconventional "React" methods and the other one had to present "ClosureScript".

The "React" guy didn't show up, and the "Closure" guy told us that this compiler is multi-threading but in fact by his examples was single-threading.

So instead of learning new stuff in there I just laugh my ass off because of this event.

Programmer looking for a new language

I have been a JavaScript developer for a few years now (non professionally) and I really like the language. I mainly program for execution with NodeJS rather than web, because I feel like I get more freedom (e.i. the ability to use a computer file system).

I normally never use other people's libraries and instead either write my own library/ies for the specific task or use an old one. I only ever use someone else’s if I need a quick frame work to test an idea, never for something I will actually use.
I prefer to work object / class orientated.

I have worked on distributed servers with NodeJS before, however trying to distribute a load across one computer across it's multiple threads has proved problematic due to the heavy delays of standard io transfer speeds.

Why do I want to switch?:
•Because JavaScript is not at all created with multithreading in mind, and pretty much any multithreading solution is a bodge and allot of the time it is more efficient to work single threaded.
•Also, I get the sense that JavaScript + NodeJS is not used often in the programming industry comparison to other languages like; ruby, python, and I don't want to get stuck in a nesh language of which would decrease my employment chances heavy.

Side Note: I have been working on a pet project to have a distributed database (made with nodejs), and so far, there are no language specific problems, but I feel like it would be more efficient if I used a programming language designed more to cater for multi threading.

Ive been running a brute force program on kali linux virtual machine which was using multi threading by sending Hundreds of requests per 10 seconds 24/7 my laptop overheated and shut down and now my kali linux vm wont boot up something got broken heres a screenshot help please

I have to change the model of my application but I don’t like the way it was written because it was written with multi threading in mind which I don’t mind using and It makes sense to use but now Idk how I can make any changes.

And I don’t even know where the fuck to start

Also it relies on a lot of OOP stuff and it annoys me. (I don’t really enjoy OOP)

Hey guys I've a problem I've been trying to solve for a while. Also I'm a college student so my knowledge isn't going to be the greatest so go easy on me if it's simple to solve😂. So I'm creating a real time licence plate detector using yolo lite, my own deep learning ocr and plan to add the model to fast api. So as an input to the rest api, the user will submit a IP camera link for openCV to get individual frames for preprocessing before yolo predictions. The problem I have is how to I handle multiple real time IP camera feeds at once?. Ive been researching multi threading but read that it can cause issues with async definitions in fast api. Any advice will be greatly appreciated and if more information is needed just shout!.

Python3's asyncio is awesome.

In an IoT server we used python2 socket programming with multi-threading. Recently I have changed the socket layer to python 3 asyncio and performance was far better. I am not gonna use multi-threading anymore.

What do you guys think about asyncio?

I love ruby, except the official implementation is super slow and bad at multi threading. But that's where Jruby comes in :)

(Still relatively slow though :( )

Yeah... I spent way too much fucking time thinking my multi threading was fucked up. When in actuality I had a parameter in a JSON payload for an API call I neglected to check for changes.

A fucking form id! Just one parameter had me frustrated for 2 hours.

few days back i created real life example of live lock in multi threading

... and the whole night i was dreaming like i would have created dead locks in my real life... and i couldn't find how to remove those dirty synchronized blocks