What exactly is he hoping to achieve? Cool though and undoubtedly a challenging project!

It's just cool learning material.

"looks good on your cv"

A candidate showing this would definitely have cookie bonus in our hiring process.

It's a very valuable thing to do to teach a new generation of programmers.
In the same vein, there is Handmade Hero, which is a great project.

However, I have an issue with that statement:

if you were ever interested in [...] what went "behind the scenes" at id studio roughly 20 years ago

In my opinion, the following 2 things are what made the development of Quake uniquely interesting:
1. The chaotic team dynamic.
2. The fact the team tackled several big technical challenges at once and had to figure out a bunch of things the really hard way (the client-server network architecture, building the PVS using an analytic method, etc).
You won't find that again since the project is handled by 1 guy and the "big" problems have been solved.

If you're interested in what went on at id, check out Masters of Doom: How Two Guys Created an Empire and Transformed Pop Culture.
If you're interested in the technical challenges, check out Michael Abrash's Graphics Programming Black Book's chapters about Quake. The book is now free.

I belive it's also interresting for level designers as they need to know how the engine work to optimise the creations. Specifialy the second vid remind me of some tutorial I/ve been reading about optimising levels in GTKradiant.

here is also a completely unrelated video about making a doom-style engine in C: https://www.youtube.com/watch?v=HQYsFshbkYw

in the first 30 minutes he rambles on and removes some files

not a single line of code

So, he's a normal dev?

Interesting concepts.

1.3 - custom strcmp and atoi with bugs. This will take forever.

Yeah, it's an extremely odd video.

Stuff like atoi, strcmp and his argument parsing is a material for complete programming newbies. Yet he proceeds to implement and explain this stuff.

If the target audience needs this kind of explanation, getting to a level where working code can be produced will literally take years.

So the guy will have to make a huge jump difficulty-wise, rendering the thing even more odd.

Various things he says also seem wrong, e.g. his claim that the arg table was zeroed because he is running stuff in the debugger.

Don't get me wrong, it's cool people work on stuff like this, but this looks like another bad tutorial.

be up the creek when he reaches the ASM for video cards

Various things he says also seem wrong, e.g. his claim that the arg table was zeroed because he is running stuff in the debugger.

Not watched the video so don't know exactly what he was trying to say, but debug builds from Visual C++ (but not just in the debugger) do initialise integer stack variables with 0 even though it's not required for C/C++ compliance. It is not done on release binaries so it's useful for newbies to know that they shouldn't trust that debug and release behave the same.

It does sound like you're right and he hasn't worked out what level his target audience is though.

Watch 2 minutes starting from here:

https://youtu.be/TdUszxsJ98A?t=21m02s

largv has static storage duration, and unless windows is really weird, it should be guaranteed to be zeroed.

Further, I may have trouble parsing what he said, but he seems to claim char *foo = "" is equivalent to char *foo = NULL. (Around :35)

I don't know anything about windows programming. Is not zeroing stuff on the stack intentionally weak? I would expect stuff to be filled with known junk instead, it is too easy to get zeroes by accident.

Apparently he assumes that his average viewer knows C++ or some higher level language, so he's explaining pointer arithmetic thoroughly. For such viewers it won't be "huge jump difficulty-wise".

There is no way you go from not being fluent with pointers to writing working C code in few days. You learned how to solve basic equations in elementary school, but that's nowhere near sufficient to even start doing higher lever math. And that kind of jump is approaching.

I don't know, I heard good oop programmers struggle with pointers if they haven't used them before. Kinda hard to believe that C++ programmers may have such issues though. You can learn abstract math without "basic equations" btw, read about "new math" (although it is ultimately considered a failure). You surely can learn high level programming language without any low-level preparation as most modern programmers do I think.

C code in Quake is pretty straight-forward procedural programming most of the time. At least the parts I know.

Unclear what you are trying to say.

The analogy was to illustrate the difference in difficulty.

Anyway, you are not switching from a higher level language to C in few days either. When you screw something up in C (and it is way easier to screw up than in other languages), and you find about the existence of some screw up, it is typically in the form of a crash. What's worse, seemingly correct code suddenly stops working after unrelated changes.

In order to diagnose stuff which is not blatantly trivial you need to be familiar with actual debugging tools, concepts like syscalls, pages. At least rudimentary knowledge of assembly for your target cpu.

People starting with C often drown in the sea of crashes.

As such, I find it downright impossible to write a 3D engine by people who "learned" pointer arithmetic yesterday.

Instead, as it usually goes, viewers who decide to do anything will parrot the code, try to do meaningless modifications which will result in crashes they will have no idea how to debug and most of them will say "fuck it, 1v1 thunderstruck". Some may take the hint and decide to properly learn basics first.

And the tutorial really talks about bullshit which you are supposed to to be able to do before you even try reimplementing any real-world program.

This tutorial so far caters to people who lift 10kg and will require the ability to lift 200kg in few episodes. You don't get to that level by trying to lift that much straight away.

EDIT: now I feel the need to mention that most technical tutorials having anything to do with C are at least questionable and often straight up wrong. In fact I find stuff like this actively harmful - from beginner perspective stuff written in this language works and stops working completely randomly. Of course it seems that way because said beginner does not yet have the expertise to understand how they screwed up and why it led to given symptoms. That's a terrible put off for anyone learning programming.

anyone who has visual express uses pointers in all code, from c to .net

there are nothing hard about a classes pointing to its class functions

in C you (very often) have to do so explicitly, while in .net you (very often) don't.
At least not when you are used to object oriented programming.

As a result a lot of programmers that 'only' learned OOP often have problems with pointers.


I didn't really watch any of the vids though so I can't really go into the rest of the comments.

its part of intelisense like on

https://code.visualstudio.com/docs/editor/editingevolved

if you don`t use visual studio code, then no editors on linux or mac will help you out, so coding is just all pure text like using notepad

when you start typing in visual studio, all class functions will be listed somewhere in the droplist along with all other functions, they are not sorted or grouped, by they are listed, which saves many hours, because you don`t have to remember the names of everything, you wil just get a droplist as soon you do type ->

ye, but what I mean is that in OOP teachings these concepts are not so much explained as being pointers but rather as "functions belonging to a class" or "properties belonging to a class" and thus making the pointer references implicit rather than in C where you have to make them explicit.
Thus in C you are more conscious of pointers 'as being pointers' while in .net people often view them slightly different.

It's a different way of saying the same thing but it leads to slightly different thinking patterns I guess.

I had to learn both as if they were separate things...

implicit and explicit, both words mean nothing to me

if a pointer does`nt work, then in any code it would be more todo with the code in the class or a include not being included in one of the classes

or whether the function is public or private, if a function is private then you can`t use it within another class

Oh man I crashed so many C programs due to bad alignment when I was first learning and had no idea what was going on.

should try this

https://www.visualstudio.com/en-us/products/code-vs.aspx

it has intelisense on linux and mac

https://code.visualstudio.com/docs/editor/editingevolved

book da fuck marked

This should be pretty awesome :)

Hopefully more awesome than his suite of tutorials on Q2 - which he appeared to abandon after 2 days.

:(

The goal is interesting, but it seems incomplete to me. The design of Quake was basically to create a three-dimensional DOOM game. Physics and gravity, various types of polygonal brushes, 3D actor models, 3D sound, client/server communication design, AI filters, dynamic lighting, and so forth.... The challenges of the Quake project were mainly due to target OS and hardware restrictions. Optimizing the BSP tree, reducing the client rendering processing, minimizing the client/server packet flow, single-threaded or raw device calls, so the game could actually functionally run on a DOS Pentium 1 computer. Building an FPS game from the ground up on Windows10 with Visual Studio 2013 sounds like a much different goal than "let's tackle the challenges that idsoftware did to make Quake".

It's obviously impossible to tackle the entirety of the challenges idsoftware did, it would like going back in time, and then you might as well include dealing with Romeros personality and cheap gas prices. it's a very good idea to isolate the generic game creation process in modern, relevant circumstances. that way its actually useful and practical for modern day kids.

so he is going to make a game that has already been made?

So you're going to make a comment that has already been made?

You covered that )

GL to all 11 viewers!

I loled when he explained the GPL licence and deleted it just after :-)