Thank you for treating your audience like adults. This is the right amount of fast-paced and insightful. I appreciate a video that is meant to inspire people to dig deeper.
The content and presentation is top notch. I also like your eloquent speech: I'm interested in low level programming, but even if i was not, I would probably listen to this vid.
kay..thanks for this awesome series. Im a react native developer who is interested to learn thee nitty gritty of coding. but actually lacks the core concepts. will be following ODES series as i find the videos very enlightening👍
I randomly saw this on my RUclips feed and so glad I watched. I don't even code in C but it was interesting learning something new! I hope to see more videos! :)
I really enjoy your videos as i had some experience with c++ and i almost understand all of just C. I think this video format is great for people who are already informed about basics, pointers and memory concepts. Love from Moscow, Russia
I liked the video, but honestly probably a bit too hard for beginners. Honorable mission though. In particular, the attempt to simplify some topics may actually end up misleading: on discussion about returning strings, I can see how you really try to explain variables' lifetimes but without mentioning them explicitly only focusing on the example. I would try to instead insist on the notion of lifetime and then go on explaining why callee variables on stack cannot outlive their lifetime and be accessible from parent (i.e. they "expire" upon return, and thus caller's attempt to access that now potentially overwritten stack memory is flawed). I would also showcase how a struct wrapping the fixed size array is possible due to return semantics.
Thank you for the video. Please review the memory dump you're showing. Each line should display 16 bytes, but yours only shows 8. Additionally, please verify the byte order. On most modern architectures, it should be little-endian, meaning the smaller byte appears first.
The presentation and whole video overall was very smooth, even though I already knew this stuff, I watched until the end. Subscribed, very good content.
The for-loop is syntactic sugar for a while loop with the following setup: initialization; while (condition) { increment; } and its equivalent for-loop for (initialization; condition; increment) { } EDIT: i think it's easier to think of a for-loop as: for (before; during; after) { } but that can be confusing too. "During" is the condition that is technically checked *before* running the , but one can say that it's constantly being checked *during* the loop. "Before" is self-explanatory, but "after" seems confusing. "After" is code run after the has done its thing.
I have a background in java and actually started in C last week, was openai to help me show examples etc but this video was very helpful, thanks so much!
Hey Kay, this is really good class. I hope you take more of such classes, maybe involving more of gdb and valgrind and other mess which we usually face 🙂 Really really good work.
i like your way of giving the information good work really i have a little knowledge of C Language so i may ask you to teach me and others like me C Language like this way Thank you in Advance 😍😍😍😍😍
I've been having a crisis of meaning lately; I love my work as a web dev, but every day I'm just writing bloat. I was looking for a good intro to C or embedded development; grateful the algorithm sent your channel to me 👍
I enjoyed this video. I liked how you connected everything together. I am a less than beginner level in C. I have been in Assembly for quite a while now. I have have bounced into all of these concepts during my own writes and you lay it out so nicely and I like how it doesn't have to be fast and skippy. The whole concept of a function returning data has bothered me for some time now. The concept is an abstraction from what is really happening. A function call is really just a sugary goto but with a fancy way to return to the next line after the goto. And when the function does return; it just cannot carry data with it - LOL. But it could have altered the values in the registers. And our ancestors created a fancy trick by altering the RAX (EAX, AX, AL, AH) register; then copying that new value into a variable on the caller side. This gave the illusion in our beloved gen 3 languages that a function may return a value. But it really didn't, it just changed a register value. My point is that now everyone wants to return big stuff from functions and it has just gone too far. Ok - thats my rant for the day. And your video supports my thoughts with realistic examples. Love it!
Thanks for the comment! Really enjoyed reading it. I'm looking forward to getting more into building up the 'function' abstraction through assembly in the next couple videos. Was just playing around with it today actually. Exciting!
I really love your style of presenting the source, the structure of your story, and just the easy going nature of it all. I will gripe about using two spaces for indentation, so -1 point for that, but +2 points for showing simple macros so that I now no longer feel that I need to avoid them. Thanks! Oh and, printf is super slow because I believe it's monstrously complex, so that might be worth mentioning (plus security risks I believe). And is make really worth using?
Thanks! And all good points - I can definitely feel the complexity of printf coming through when I reach for it :) Re: make, with these videos questions like that are often a balancing act. Could use a plain shell script but then learners might not remember make at all and that can be confusing later, or could go for something more fully featured and then you're going to also have to include setup instructions and risk diverting attention. My angle is often trying to put people in a position where they've encountered the right set of things to then go build on the basics later - rather than taking a sophisticated approach early on and getting muddled. Could easily argue for other tradeoffs though! Anyway, thanks for the comment :)
@@neoeno4242 You're right about make and your approach. This was more of a personal curiosity, and not a comment on your content. I got inspired by Handmade Hero that a batch file or shell script will suffice because usually it's just a few lines of mkdir, copy, gcc and run. So I have come to rely on a simple VS Code "skeleton project" as I clumsily called it that does these things: build, debug, and test. Whatever language I'll use is not even important because the concept works for whatever I'm making. You done any video on git yet? VS Code and git are a match made in software heaven.
If the memory is allocated with malloc and returned, if we call printf it wont be corrupted no? That problem was because the array was on the stack? Very interesting videos an great teacher!
Yes that's correct! Though if we use malloc in the function then we will have to decide when to free or 'give back' that memory (since in C nothing will free it up automatically and if the function is called repeatedly this will eventually use up a lot of memory). So string functions in C instead typically take a pointer to a buffer as an argument and then alter that buffer. That way the caller is responsible for allocating and freeing memory however they see fit.
Working on it :) At the moment I'm working towards creating binary executables from scratch as a little milestone. That will involve spending a bit more time working with binary data, file formats, syscalls, and ASM. I think my planning will solidify in the next few months - will share more as that solidifies!
TL;DW Noob: Why does the program segfault? I'm too used to OOP! Expert: The stack does too good a job cleaning itself up, which is why you can't just return an array like that in your string concat function. Noob: Seriously, why doesn't it work like I expect it to? Expert: Because with great memory control comes great responsibility. Noob: So I can just make a big array at the beginning of the program and use that instead of worrying about free andMalloc? Expert: That's what I do in my prototypes anyway until I find it necessary to allocate such memory by hand. Noob: I'll just stick with JS and hope for the best. Expert: Wait a minute, you can still leak memory in J- Noob: _Sees browser crash due toOOM error_ Thanks anyway, I'll just stick with my abacus. At least that doesn't break on me when I least expect it to...
at 8:32 i believe there is something off in how the sentence is composed logically. "...it just gets inlined straight into the file, and everything in the other C file gets chunked into the C file we are including from..." "...chunked into the file we are including from..." does it? Or am i missing something?
You are right, that might be a small mistake. For anyone looking at this later, it just does a simple copy and paste (from the other file into your file). By convention, we also generally separate code into two files "header files" (.h) contain code that you want to be copied into other files (you can #include them). Implementation files (.c) contain code that you don't want to (or can't) copy around. One thing to note is that you cannot copy function implementations, so leave the implementation in a .c file, and the declaration in the .h file. This lets you use the function in other places without having multiple functions with the same name causing a linking error.
Excuse me, young fellow. May you please provide me with some advice on how to become a professional (or better yet, dangerous) C programmer. Thanks a bunch
By way of advice on going pro with C I'm not sure I have anything really revelatory - it does appear like safer systems languages like Zig, Go, Rust are gaining popularity lately. If you're looking into C though, common pro books I see recommended are: C Programming, A Modern Approach by K N King and The C Programming Language by Kernighan, Richie. On top of that, write a lot of code and ideally run it yourself day to day :) Why use the built in `ls` when you can make your own and feel all the pain when it doesn't work? Wishing you good fortune.
19:13 - what the hell each 4 bits have its own address? x86 cpus can only address minimum a one byte(8 bit)... and other cpus, except microcontrollers( where part of memory can be 1-bit addressed for gpio)
Ah yes, my mistake - thanks for flagging. I mistakenly gave each digit its own address, which isn't correct! Should be something more like this: link.excalidraw.com/readonly/KooN4ScH7IO4QuRsZtEG
Is it possible that the assert without printf could also fail at 25:10 because buffer variable is reallocated to something else? I am not sure why it is stated here that the program fails only after printf is added.
Because it's on the stack and the stack is changed after calling printf(). Literally any function which makes use of some stack space would have the same effect.
@@anon_y_mousse but even without modifying our program, it is possible that stack memory could be allocated or used by some other program and there is no gurantee this program would work everytime (without printf Or any other change) right?
@@vasantheee I certainly wouldn't depend on it being valid after the function call, but OS's typically will allocate a stack per thread or process, depending on configuration. However, depending on what the compiler does to your code, even if you don't make a function call, it's likely that data higher up on the stack could be instantly invalidated upon function return. For older chips and microcontrollers, if you have a stack at all, then all parameters would be passed on the stack, generally, and then upon return, anything allocated off the stack inside the function would be instantly invalidated. For newer chips, it's pretty much a given that parameters will be passed by register if you have few enough parameters, but the compiler might very well finagle your code such that it has to push and pop some data around the function call which would thus invalidate the array. In short, you don't have to worry about other programs running on the system invalidating your stack because your own program will likely do it without you directly seeing the effects.
@@vasantheee I would say we shouldn't ever count on it. The thing to remember, if you want to work with an array, pass it into the function. If you want it to be resizable, then pass the address of the pointer. If you want to do dynamic arrays, then consider using a struct and passing the address of that struct, say typedef struct { size_t length; int data; } array_t; and just replace int with whatever type you wish to have an array of. You could even make the array type something like int_array_t or array_int_t or whatever name you feel like.
What version of gcc are you using that it does that? I'm currently using 10.3.1 and it definitely does not erase the source file. If this is said in the video somewhere, then what's the time index?
@@anon_y_mousse waaaait, is this just a me thing? HAHA $ gcc --version gcc (GCC) 14.1.1 20240522 Copyright (C) 2024 Free Software Foundation, Inc. This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. $ nvim test.c $ gcc test.c -o test;./test Hello, World! $ gcc test -o test.c;./test /usr/bin/ld: cannot use executable file 'test' as input to a link collect2: error: ld returned 1 exit status Hello, World!
@@anon_y_mousse thinking the issue might be that the executable already exists, I ran it after rm it: $ gcc test -o test.c;./test /usr/bin/ld: cannot find test: No such file or directory collect2: error: ld returned 1 exit status bash: ./test: No such file or directory
One of the best C classes that i have seen in my whole life, way better than the college classes i attended. I apreciate
Thanks :)
Thank you for treating your audience like adults. This is the right amount of fast-paced and insightful.
I appreciate a video that is meant to inspire people to dig deeper.
Great job. I have to thank the youtube algorithm for suggesting your channel. Thanks a lot. Greetings from South America.
i agree with you and my greetings from Egypt 👏💪
The best explanation of how memory works that I've ever seen. Very clear. Thanks for your effort.
The content and presentation is top notch.
I also like your eloquent speech: I'm interested in low level programming, but even if i was not, I would probably listen to this vid.
kay..thanks for this awesome series. Im a react native developer who is interested to learn thee nitty gritty of coding. but actually lacks the core concepts. will be following ODES series as i find the videos very enlightening👍
I randomly saw this on my RUclips feed and so glad I watched. I don't even code in C but it was interesting learning something new! I hope to see more videos! :)
Gold mine of content. This deserves much more audience
I really enjoy your videos as i had some experience with c++ and i almost understand all of just C. I think this video format is great for people who are already informed about basics, pointers and memory concepts. Love from Moscow, Russia
C really is fun! And because it's been around for so long, used textbooks on Amazon are dirt cheap.
I don't see you making the same argument for Fortran and Pascal ;)
I've recently started learning C, and this video is perfect. I really like the format and presentation. Thank you!
I need to learn C++ for work, and its also really complicated in my opinion. What do you think?
the explanation of the function pointer in the memory space is great. well done.
Poor hand writing is a sign of intellect':] You rock!
Just had to comment again and I’m a new subscriber. This was so good. You’re very intelligent and explain things well. Made all this seem very easy
What a calm person you are !:) First time I've watched your video. So let me check your playlists :)
I liked the video, but honestly probably a bit too hard for beginners. Honorable mission though. In particular, the attempt to simplify some topics may actually end up misleading: on discussion about returning strings, I can see how you really try to explain variables' lifetimes but without mentioning them explicitly only focusing on the example. I would try to instead insist on the notion of lifetime and then go on explaining why callee variables on stack cannot outlive their lifetime and be accessible from parent (i.e. they "expire" upon return, and thus caller's attempt to access that now potentially overwritten stack memory is flawed). I would also showcase how a struct wrapping the fixed size array is possible due to return semantics.
Such an interesting perspective - thanks for sharing, appreciate it and I'm sure others will too.
@neoeno4242 do you have a Discord server/are you a Discord user?
Thank you for the video. Please review the memory dump you're showing. Each line should display 16 bytes, but yours only shows 8. Additionally, please verify the byte order. On most modern architectures, it should be little-endian, meaning the smaller byte appears first.
The presentation and whole video overall was very smooth, even though I already knew this stuff, I watched until the end. Subscribed, very good content.
Appreciate it!
youre awesome, kay! learning a lot from you. thanks!
Good explanatory content for beginning C programmers.
The for-loop is syntactic sugar for a while loop with the following setup:
initialization;
while (condition) {
increment;
}
and its equivalent for-loop
for (initialization; condition; increment) {
}
EDIT: i think it's easier to think of a for-loop as:
for (before; during; after) {
}
but that can be confusing too. "During" is the condition that is technically checked *before* running the , but one can say that it's constantly being checked *during* the loop. "Before" is self-explanatory, but "after" seems confusing. "After" is code run after the has done its thing.
perfect for my reintroduction to the c programming language.
I have a background in java and actually started in C last week, was openai to help me show examples etc but this video was very helpful, thanks so much!
Hey Kay, this is really good class. I hope you take more of such classes, maybe involving more of gdb and valgrind and other mess which we usually face 🙂 Really really good work.
Great content. Appreciate your hard work. Btw, could you make a video about C++ ?!
thanks for this crisp, to the point video.
Such an amazing talk...pure code and explanations.
i like your way of giving the information good work really i have a little knowledge of C Language so i may ask you to teach me and others like me C Language like this way
Thank you in Advance 😍😍😍😍😍
I've been having a crisis of meaning lately; I love my work as a web dev, but every day I'm just writing bloat.
I was looking for a good intro to C or embedded development; grateful the algorithm sent your channel to me 👍
This rules, someone lovely making nice videos about C what could be better
this is a great video! thanks for sharing
bro is majestic
bro is trans
she is slaying physically and intellectually 💅
Agreed, what a fierce flow.
I don't even know why I'm watching this since I've been using c++ for like 3 years now, still nice video
Thank you, your videos are very enjoyable to watch.
I enjoyed this video. I liked how you connected everything together. I am a less than beginner level in C. I have been in Assembly for quite a while now. I have have bounced into all of these concepts during my own writes and you lay it out so nicely and I like how it doesn't have to be fast and skippy. The whole concept of a function returning data has bothered me for some time now. The concept is an abstraction from what is really happening. A function call is really just a sugary goto but with a fancy way to return to the next line after the goto. And when the function does return; it just cannot carry data with it - LOL. But it could have altered the values in the registers. And our ancestors created a fancy trick by altering the RAX (EAX, AX, AL, AH) register; then copying that new value into a variable on the caller side. This gave the illusion in our beloved gen 3 languages that a function may return a value. But it really didn't, it just changed a register value. My point is that now everyone wants to return big stuff from functions and it has just gone too far. Ok - thats my rant for the day. And your video supports my thoughts with realistic examples. Love it!
Thanks for the comment! Really enjoyed reading it. I'm looking forward to getting more into building up the 'function' abstraction through assembly in the next couple videos. Was just playing around with it today actually. Exciting!
Thanks for the lesson. I hope you keep up with producing awesome content.
Very, very good !
Great video, thanks! I'm using C to build a little hobby os for fun, and this is a nice refresher video :)
i really had fun watching this
I really love your style of presenting the source, the structure of your story, and just the easy going nature of it all. I will gripe about using two spaces for indentation, so -1 point for that, but +2 points for showing simple macros so that I now no longer feel that I need to avoid them. Thanks! Oh and, printf is super slow because I believe it's monstrously complex, so that might be worth mentioning (plus security risks I believe). And is make really worth using?
Thanks! And all good points - I can definitely feel the complexity of printf coming through when I reach for it :) Re: make, with these videos questions like that are often a balancing act. Could use a plain shell script but then learners might not remember make at all and that can be confusing later, or could go for something more fully featured and then you're going to also have to include setup instructions and risk diverting attention. My angle is often trying to put people in a position where they've encountered the right set of things to then go build on the basics later - rather than taking a sophisticated approach early on and getting muddled. Could easily argue for other tradeoffs though! Anyway, thanks for the comment :)
@@neoeno4242 You're right about make and your approach. This was more of a personal curiosity, and not a comment on your content. I got inspired by Handmade Hero that a batch file or shell script will suffice because usually it's just a few lines of mkdir, copy, gcc and run. So I have come to rely on a simple VS Code "skeleton project" as I clumsily called it that does these things: build, debug, and test. Whatever language I'll use is not even important because the concept works for whatever I'm making.
You done any video on git yet? VS Code and git are a match made in software heaven.
Great teacher!
This comment works roughly as you would expect it: you make an awesome video, I comment nice stuff.
The clip cuts are nice to keep things moving, keep making educational videos plz
thank you yt! thank you Kay! can I request for a holy-C playlist !
Just added playlists to my todo list!
This was wonderful. Thanks!
Thank you for this video
If the memory is allocated with malloc and returned, if we call printf it wont be corrupted no?
That problem was because the array was on the stack?
Very interesting videos an great teacher!
Yes that's correct!
Though if we use malloc in the function then we will have to decide when to free or 'give back' that memory (since in C nothing will free it up automatically and if the function is called repeatedly this will eventually use up a lot of memory).
So string functions in C instead typically take a pointer to a buffer as an argument and then alter that buffer. That way the caller is responsible for allocating and freeing memory however they see fit.
Wonderful job!
do you have a map of topics you plan to cover in the next10-15 videos?
Working on it :) At the moment I'm working towards creating binary executables from scratch as a little milestone. That will involve spending a bit more time working with binary data, file formats, syscalls, and ASM. I think my planning will solidify in the next few months - will share more as that solidifies!
this series is grt.
TL;DW
Noob: Why does the program segfault? I'm too used to OOP!
Expert: The stack does too good a job cleaning itself up, which is why you can't just return an array like that in your string concat function.
Noob: Seriously, why doesn't it work like I expect it to?
Expert: Because with great memory control comes great responsibility.
Noob: So I can just make a big array at the beginning of the program and use that instead of worrying about free andMalloc?
Expert: That's what I do in my prototypes anyway until I find it necessary to allocate such memory by hand.
Noob: I'll just stick with JS and hope for the best.
Expert: Wait a minute, you can still leak memory in J-
Noob: _Sees browser crash due toOOM error_ Thanks anyway, I'll just stick with my abacus. At least that doesn't break on me when I least expect it to...
This is my fave one: ruclips.net/video/GcDshWmhF4A/видео.html
enjoy your video, thanks!
the first this I do in an empty main is allocate 5k array just for the sake of it. Feels great to allocate arrays and then free it
At 14:00 i think it should either be nums[0] = 0x10; nums[1]=0x20; etc or 0a 14 1e in memory
incredible
12:46 unsigned char MAX should be +255 , am I wrong ?
Nice video. And I like your shirt. It looks good on you.
Thanks!
Lord Farquaad
Hi, I subscribed yesterday.
What text font are you using?
Largely: Monaspace Xenon for the code font, and Mona Sans for the rest. Sometimes I mix up the monaspace font families though :)
at 8:32 i believe there is something off in how the sentence is composed logically.
"...it just gets inlined straight into the file, and everything in the other C file gets chunked into the C file we are including from..."
"...chunked into the file we are including from..." does it?
Or am i missing something?
You are right, that might be a small mistake. For anyone looking at this later, it just does a simple copy and paste (from the other file into your file).
By convention, we also generally separate code into two files "header files" (.h) contain code that you want to be copied into other files (you can #include them). Implementation files (.c) contain code that you don't want to (or can't) copy around. One thing to note is that you cannot copy function implementations, so leave the implementation in a .c file, and the declaration in the .h file. This lets you use the function in other places without having multiple functions with the same name causing a linking error.
Yes it's not super clear - thanks for flagging and also to @BoardGameMaker4108 for clarifying. Appreciate it.
Spec requirement: high tech title card. ☑️
Excuse me, young fellow. May you please provide me with some advice on how to become a professional (or better yet, dangerous) C programmer. Thanks a bunch
By way of advice on going pro with C I'm not sure I have anything really revelatory - it does appear like safer systems languages like Zig, Go, Rust are gaining popularity lately. If you're looking into C though, common pro books I see recommended are: C Programming, A Modern Approach by K N King and The C Programming Language by Kernighan, Richie. On top of that, write a lot of code and ideally run it yourself day to day :) Why use the built in `ls` when you can make your own and feel all the pain when it doesn't work? Wishing you good fortune.
whaf js yonr genper?
Tests in C I didn't know they existed
19:13 - what the hell each 4 bits have its own address? x86 cpus can only address minimum a one byte(8 bit)... and other cpus, except microcontrollers( where part of memory can be 1-bit addressed for gpio)
Ah yes, my mistake - thanks for flagging. I mistakenly gave each digit its own address, which isn't correct! Should be something more like this: link.excalidraw.com/readonly/KooN4ScH7IO4QuRsZtEG
great! video
watching this quite cool video feels a little bit like stealing for the goodness received for free!
Is it possible that the assert without printf could also fail at 25:10 because buffer variable is reallocated to something else? I am not sure why it is stated here that the program fails only after printf is added.
Because it's on the stack and the stack is changed after calling printf(). Literally any function which makes use of some stack space would have the same effect.
@@anon_y_mousse but even without modifying our program, it is possible that stack memory could be allocated or used by some other program and there is no gurantee this program would work everytime (without printf Or any other change) right?
@@vasantheee I certainly wouldn't depend on it being valid after the function call, but OS's typically will allocate a stack per thread or process, depending on configuration. However, depending on what the compiler does to your code, even if you don't make a function call, it's likely that data higher up on the stack could be instantly invalidated upon function return. For older chips and microcontrollers, if you have a stack at all, then all parameters would be passed on the stack, generally, and then upon return, anything allocated off the stack inside the function would be instantly invalidated. For newer chips, it's pretty much a given that parameters will be passed by register if you have few enough parameters, but the compiler might very well finagle your code such that it has to push and pop some data around the function call which would thus invalidate the array.
In short, you don't have to worry about other programs running on the system invalidating your stack because your own program will likely do it without you directly seeing the effects.
@@anon_y_mousse thank you. So we cannot count on the program being successful like it happened at 25:10 all the time, right?
@@vasantheee I would say we shouldn't ever count on it. The thing to remember, if you want to work with an array, pass it into the function. If you want it to be resizable, then pass the address of the pointer. If you want to do dynamic arrays, then consider using a struct and passing the address of that struct, say typedef struct { size_t length; int data; } array_t; and just replace int with whatever type you wish to have an array of. You could even make the array type something like int_array_t or array_int_t or whatever name you feel like.
Sehr gut
what code editor are you using here?
VS Code :)
Bro choose the arch linux way
But what if want to live dangerously in C? 😂
Then use after free all you like 😈
gcc will erase the source file if you don't call it in this format "gcc file.c -o executable" lol
Very strange
@@w花b I'd say it's just different implementations. :D
What version of gcc are you using that it does that? I'm currently using 10.3.1 and it definitely does not erase the source file. If this is said in the video somewhere, then what's the time index?
@@anon_y_mousse waaaait, is this just a me thing? HAHA
$ gcc --version
gcc (GCC) 14.1.1 20240522
Copyright (C) 2024 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
$ nvim test.c
$ gcc test.c -o test;./test
Hello, World!
$ gcc test -o test.c;./test
/usr/bin/ld: cannot use executable file 'test' as input to a link
collect2: error: ld returned 1 exit status
Hello, World!
@@anon_y_mousse thinking the issue might be that the executable already exists, I ran it after rm it:
$ gcc test -o test.c;./test
/usr/bin/ld: cannot find test: No such file or directory
collect2: error: ld returned 1 exit status
bash: ./test: No such file or directory
You is a dream waifu for many including me :3
bruh you pretty as hell T_T