AtariAge + Atari Q&A

[SIO2]

On 10/2/2023 at 3:42 AM, Random Terrain said:

 

I nominate the Atari Logo.

 

 

 

I guess that explains the rainbow logo?

[Raiu]

I think Bentley Bear would make a good mascot, both because he's a classic character, and because the character is gameplay-agnostic. You could transport Bentley and the other characters into different scenarios and different situations without too much difficulty: Berthilda The Witch and her magically-animated minions stir up trouble somewhere, and here comes Bentley Bear and his magic hat to save the day. (Like Mr. Do, if Universal still existed and needed a mascot)

 

And even if he's not that well-known right now, that's no problem that a really good and popular game couldn't solve.  (The trick is not to half-ass it, it has to be a genuinely good game)

[famicommander]

1 hour ago, Raiu said:

I think Bentley Bear would make a good mascot, both because he's a classic character, and because the character is gameplay-agnostic. You could transport Bentley and the other characters into different scenarios and different situations without too much difficulty: Berthilda The Witch and her magically-animated minions stir up trouble somewhere, and here comes Bentley Bear and his magic hat to save the day. (Like Mr. Do, if Universal still existed and needed a mascot)

 

And even if he's not that well-known right now, that's no problem that a really good and popular game couldn't solve.  (The trick is not to half-ass it, it has to be a genuinely good game)

There has been an Atari in my home literally every single day of my life and I couldn't pick Bentley Bear out of a lineup. 

[jerseystyle]

I think we are kind of past mascots now- Microsoft and Sony don’t really have one (but have many noticeable IPs for sure) and the “mascot platformer” has been dead as a concept for a while. I think if they wanted a symbol they should just use their logo. 

[GraffitiTavern]

1 hour ago, Raiu said:

I think Bentley Bear would make a good mascot, both because he's a classic character, and because the character is gameplay-agnostic. You could transport Bentley and the other characters into different scenarios and different situations without too much difficulty

I know they basically used him that way, he was the only returning character in Atari Karts. Then there is the Crystal Quest homebrew platformer for the 7800, no clue on what will become of those now that AtariAge is bought. I hope they do an improved reissue of it for the 2600+, Crystal Quest seems like a better sidescroller for the 7800 than Scrapyard Dog.

Edited October 3, 2023 by GraffitiTavern

[JPF997]

12 minutes ago, jerseystyle said:

I think we are kind of past mascots now- Microsoft and Sony don’t really have one (but have many noticeable IPs for sure) and the “mascot platformer” has been dead as a concept for a while. I think if they wanted a symbol they should just use their logo. 

This right here, Microsoft's best selling games are Forza and Flight Simulator (and now Starfield thanks to Bethesda), Sony's best selling games are also non mascot games like Gran Turismo, Uncharted, Last of Us etc, besides Nintendo and Sega no major developers make mascot platformers they're flagship ip's, that era has been over since the PS2's last day's and it's not coming back anytime soon.

[GraffitiTavern]

19 minutes ago, famicommander said:

There has been an Atari in my home literally every single day of my life and I couldn't pick Bentley Bear out of a lineup. 

Well he's a Bear...the other Atari characters are giant centipedes, triangular spaceships, and a freaking paddle.

[famicommander]

1 minute ago, GraffitiTavern said:

Well he's a Bear...the other Atari characters are giant centipedes, triangular spaceships, and a freaking paddle.

I meant if you lined up a bunch of bear characters from various media, I wouldn't be able to pick out Bentley. I can't picture Bentley in my mind's eye. Crystal Castles is a game I've played at some point but it didn't make a huge impression on me and I honestly didn't know the character's name until people here started proposing him as a possible Atari mascot. I've played Atari Karts but it's an awful game and I wasn't aware any of its characters were shared with other games.

[GraffitiTavern]

7 minutes ago, JPF997 said:

This right here, Microsoft's best selling games are Forza and Flight Simulator (and now Starfield thanks to Bethesda), Sony's best selling games are also non mascot games like Gran Turismo, Uncharted, Last of Us etc, besides Nintendo and Sega no major developers make mascot platformers they're flagship ip's, that era has been over since the PS2's last day's and it's not coming back anytime soon.

I mostly agree, but I think Atari does suffer a bit from a lack of recognizable symbols besides the logo. Even if mascots aren't big, Halo, Last of Us, Uncharted, Splatoon, Xenoblade, Final Fantasy all have very recognizable characters, which Atari still doesn't. Now these are different companies, but I still think Atari trying to build a small character stable wouldn't be a bad thing.

[Wildstar]

10 hours ago, Lord Mushroom said:

Why not? If you are making a Recharged game and selling it on Steam and modern platforms, I don´t see why you can´t. Perhaps you can´t develop games for old platforms like the 2600 that way, but I don´t think they plan to get much of their income from new games on cartridges for old platforms.

You can't use Unity or most modern game engines to make games for NES or SNES or similar era hardware. It would exceed memory resources of said systems and for a lot of genuine reasons. Simple fact of hardware. When I wrote that, I was talking real hardware or games to run under the emulation not running on modern. Context of what I said matters. Making a "recharged" (in other words, a remake) of games on new systems are not the same as making games FOR the respective original computers and consoles (using real hardware or an emulator). Making a new game for an actual Commodore 64 or game running inside a C64 emulator is not the same thing as remaking a game made for a Commodore 64 and remade to natively run on a new system. That was what I was referring to.

 

10 hours ago, Lord Mushroom said:

I think $10 million is more than enough, but of course it depends on how ambitious they are (how much they are investing in games). After a quick search on Steam, Recharged games seem to me to bring in very roughly $5 per game after the 30% cut to the platforms (not $1-2 as you are suggesting). And then there are bigger games and bundles, which bring in more.

 

They have released a lot of Recharged games now, and if making Recharged games was unprofitable, they would have stopped doing it. They have also made their original games available on new platforms for decades. We must assume they have crunched the numbers based on these experiences, and others, and come to the conclusion that buying IP and releasing games based on it is profitable.

That's true. I was just running some general number but even then, it still not going to make much difference. They still need to make $20-$30M in revenue from games to make sustainable income. None of the board of directors are the kind to run a retro computer boutique hoppy shop business. People who tend to run these retro computing shops don't do it for a living. They usually are operated by 1-3 individuals and net income after expenses is maybe about the income of a public school teacher/teaching assistant. They aren't going to make executive salaries... in most retro businesses because you are relying on maybe 5,000 to 15,000 customers world wide. People like the board of directors, they do basically a pro forma analysis. Hobbyist / homebrew guys tend to do such. It's all about passion not investing. Many game developers making games for retro computers like making games for a real Commodore 64... have a day job. 

 

I looked at that myself. Looked at all the people actively using a Commodore 64 or emulator is maybe 10,000 or so (worldwide). That's the people who would buy the games. They would be the ones that would know about it. Here's the reality, you are lucky to get 10% of that customer base to buy it. Cartridges, for example would cost ~$50 or so and then you have to tack up $15-$30. Market price ceiling for games is about $80 because people don't want to pay more than that amount for games.... period except maybe 20-50 individuals per 10,000. So you see, you're luck to get 300-500 purchases. Revenue is likely to be in the $9,000 to $15,000 range. You might be especially lucky to get 1000-1500 purchases for any game. So, a game may get you $15,000 to $45,000. That's not enough to pay a team three spending 40 hours a week for 9 months or the equivalent at 20 hours a week for 18 months even at minimum wage. That's a far cry from making the kind of living to professionally make a game that runs on real hardware like a real C64 or Atari 400 or 800 or ST or NES, or SNES. 

 

When you make games that actually runs on a Super Nintendo in 65c816 machine language, you are going to be coding the game in an ML monitor or 65c816 assembly language tools. You can't use Godot or Unity for that. Those tools don't target such cpus in the compiler and they aren't meant to be used for that. I know because I have done computer programming for these systems back in the 80s. 

 

You may use some C compilers. However, Godot, Unity, or UE4/5 are not designed for the types of video and sound chips of these classic systems. You must use the right tools. 

 

Making retro-style game on modern systems is different than making games FOR the retro platforms of the 1980s/90s. 

 

 

Edited October 3, 2023 by Wildstar

[Zoyous]

1 hour ago, Wildstar said:

You may use some C compilers. However, Godot, Unity, or UE4/5 are not designed for the types of video and sound chips of these classic systems. You must use the right tools. 

 

Making retro-style game on modern systems is different than making games FOR the retro platforms of the 1980s/90s.

Still, there are modern tools for retro game creation, like GB Studio. Something made with a toolset like that may never be able to produce the best-optimized, highest-performance game, but it will certainly set at least a few designers on the path to learning ways to more directly access the target platform, and for the others who just want the enjoyment of making something that works on classic hardware, it's a good way to get started. Also, as AI and LLMs continue to improve in generating code, I think that will be another avenue that helps would-be designers be able to use contemporary development tools to target classic platforms.

[roots.genoa]

2 hours ago, jerseystyle said:

I think we are kind of past mascots now- Microsoft and Sony don’t really have one (but have many noticeable IPs for sure) and the “mascot platformer” has been dead as a concept for a while. I think if they wanted a symbol they should just use their logo. 

For some reason I find this video, which is now at the beginning of PlayStation movies, kind of embarrassingly funny:

They try to make it look like they have tons of memorable characters, but most of them have less than ten years and are very generic anyway. But hey, the Microsoft equivalent would be even sadder I guess.

[Matt_B]

You're talking about the company that just remade a nine year old game that already ran on their latest consoles. In Sony's eyes, ten years old is practically retro.

 

As for Microsoft, I guess this is their character showreel. 🙂

 

 

[Giles N]

Strange that its mostly japanese gaming companies that have strong, enduring, likeable characters: Nintendo (of course) , Sega (yeah, the blue critter but also others) Capcom, Square Enix, Taito (yellow endangered species bird, Bubbly dragons=snotty kids, etc), Namco (quiz: can anyone here at AA, guess at a likeable, fameous character often associated with Namco? Hint: he isn’t a red or blue F1 Car), Konami (The Belmont family).

 

And then… an Anonymous landing pod landing on some surface somewhere in space, a cuddly anti-nuke ground-to-air missile launch-site, 2 nameless elvish fairies… who happen to be so invisible only their anti-centipedian guns can be seen, a car, your car… meeting other strange cars in the darkness exchanging glances, etc etc… ya’ll get my point.

 

There’s no competetion here.

 

And Atari must just get on with making their gaming franchises come across as less generic and give gamers something thats fun, something they don’t get easily by playing all-the-other things that took the initial genre-builds and expanded them ever-so-much in all directions… for 40 years straight…

 
Nothing in Atari’s most fameous IPs are novel or exciting in itself anymore.

 
And dealing with that means getting to grips with makes for a good game.

Atari, if you want to

entertain us, 

Here we are now.

 

We’re not stupid

(ok, sometimes we are)

but contagious …

I forget just what

the Atari logo meant,

or why I should bring my

guns and my friends,

 

why some games make me feel

overbored,

is Atari too self-assured ?

(Oops, sorry couldn’t help it)

 

- - -  

 

(This, yes this^ bs-lyrics crap at the end here actually alludes to something we’re discussing, even something official from Atari)

 

Edited October 3, 2023 by Giles N

[+Stephen]

3 hours ago, jerseystyle said:

I think we are kind of past mascots now- Microsoft and Sony don’t really have one (but have many noticeable IPs for sure) and the “mascot platformer” has been dead as a concept for a while. I think if they wanted a symbol they should just use their logo. 

I agree - as long as it's not the awful "bell bottom" logo that looks stupid.

[Giles N]

Unless I only sounded harsh in my prior comment:  

 

I’m really happy for seeing Atari trying to do the most out of the 2600+ and the Gamestation Pro.

 

I understand that big tables aren’t turned overnight.

 

Just, the new software must be much richer in every aspect of design. Even that which is meant as simplistic, must be that in manner which makes it stand out from the crowd.

[Wildstar]

1 hour ago, Zoyous said:

Still, there are modern tools for retro game creation, like GB Studio. Something made with a toolset like that may never be able to produce the best-optimized, highest-performance game, but it will certainly set at least a few designers on the path to learning ways to more directly access the target platform, and for the others who just want the enjoyment of making something that works on classic hardware, it's a good way to get started. Also, as AI and LLMs continue to improve in generating code, I think that will be another avenue that helps would-be designers be able to use contemporary development tools to target classic platforms.

There are tools, yes. There are cross-platform development tools. However, you have to use tools for developing for the specific platforms. I did also state, modern game engines. There weren't such "game engines". The term and concept barely even existed at the time. You used collections of subroutines and apply it. You build your conditional logic routines, your routines to set the registers of video chip for the particular video modes, especially the VIC-Ii chip and the video chip used in the MSX AND TI and other chips. GTIA/TIA and ANTIC in Atari 8-bit very unique. Still, you had to program and still do. Game playability requires some optimization in routines to perform the tasks in a timely manner. You have to pay attention to the clock cycles used and available per frame cycle. 

 

GB Studio is for Game Boy. It is not for many of the other platforms. While there are tools to develop for these other platforms but if you are going to get people willing to buy, you are going to need to get to machine language level and manually code optimized routines or it will result in sluggish joystick controls, graphic glitches, and general slowness. No commercial video game for Commodore 64 was being made using BASIC by any major studio in the late 80s. They would compile the BASIC code to P-code or machine language in 1984-1986. From 1987, the core of video games were machine language. As for this AI stuff, whose going to pay for an AI generated video game. However, ChatGPT isn't going to make more optimized subroutines and machine language code than demo scenes have already done. They achieved and produced the most optimized performing routines possible to be done with the absolute smallest number of clock cycles as well as invented subroutines to create entire games/demos on absolutely the smallest memory footprint possible. 

 

Any commercial machine language game would be expected to perform at least 75% as performant on each routine and subroutine as the most optimal by clock cycle as the most optimized routine or subroutine of the same type ever made for the platform. In some older retro systems, you have to be attentive to memory footprint more. It is part of the culture, you can say. If I make a game for C64 to sell for $20-$90, it must be made at the level of quality, depth, etc. of the best commercial video games from 1987 to 1995. That's the base line. That isn't something you put together on the weekends for a couple months. You have a LOT of work to do. Game would need to fill a 1581 floppy disk 70-85% at the very least or multiple 5.25" floppy disks (1541 formatted). 

 

No amount of AI is going to make a game created in GODOT 4 for a Commodore 64. There are tools to make cross-platform development easier. I do agree on that. Many traditional games don't have stuff like camera objects. There's no 3d space. Even Godot is underneath, a 3d game engine simulating 2d. It is that which blows the amount of memory in a C64. You couldn't possibly make that work in the memory footprint. There's a different process and tools needed. I am aware of those. A commercial video game for C64 and NES, would require 3000 to 5000 or so hours of labor for a game created by a team of 1-3. An SNES game would take about 1.25x that amount for a project typically produced by a team of 1-3. Games like Chrono Trigger would take a team of 25 to put about 12-18 months. Many commercial games for SNES in 1994-1996 required project teams of 15-25. At least half the team was full time on that project.  Modern tools and computers aren't going to reduced labor time by more than 15-20%. Only about 10-15% of the time was time spent waiting for files to load or be saved. Most of the time is limited by the humans typing code, artist making the graphics, etc. 

 

So, yes, game development process is still limited by the human element. 

 

People in the retro games just aren't going to spend money on games that were 80-100% computer generated code, graphics, etc. People want the art of the coder and the art of the graphic artist. Computer generated games are worth $0.01 or less. If someone is going to buy a game that I developed, they want that that I created it, not some bot generated video game. 

 

[Zoyous]

3 hours ago, Wildstar said:

However, ChatGPT isn't going to make more optimized subroutines and machine language code than demo scenes have already done. 

Maybe not this month or next month. Are you willing to take that bet out to next year? Five years from now? Ten? If it's within the realm of possibility, AIs are going to get there.

3 hours ago, Wildstar said:

People in the retro games just aren't going to spend money on games that were 80-100% computer generated code, graphics, etc. People want the art of the coder and the art of the graphic artist. Computer generated games are worth $0.01 or less. If someone is going to buy a game that I developed, they want that that I created it, not some bot generated video game.

I'm not talking about something that's completely "bot generated." But in my own experience I've found ChatGPT helpful for writing code for specific aspects of larger projects. It does make mistakes and it requires some established knowledge to identify and correct those mistakes. I also have found AIs helpful as sounding boards in developing ideas, as assistants in developing visual designs, etc.

 

You really think a time won't come when an AI could analyze assembly code and find some optimizations that no human found? Or a time won't come when someone could design a game in some high level design app that tracks system resources, and then a number of AI agents working together can adapt it to code that will run on a 2600? I'm not sure when that may become possible but I think it will. It's a pretty niche interest so there's not going to be many people working on that angle. I think most of the focus will be on high-end mainstream gaming moving forward, as usual.

 

Someone excels at writing machine language code, that's a great accomplishment. There were people who were great steam engine designers as well, nothing's going to take that accomplishment away from them. Technology changed and knowledge built upon knowledge. The fact is, there are going to be a lot more people designing video games in the future and there is a very high likelihood that they'll be able to do it in ways that are comparatively effortless. It's not going to take away the enjoyment and satisfaction from people who still do it manually. Newbies might learn machine language (at least partially, and soon if not already) by chatting with an LLM. Some people, a relative few, start off buying furniture from IKEA and then eventually learn woodworking.

[Lord Mushroom]

9 hours ago, jerseystyle said:

I think we are kind of past mascots now- Microsoft and Sony don’t really have one 

Mascots are still important to Nintendo. It is one of the reasons why they are the most profitable of the three main console makers. They don´t just help in terms of selling games, but also in selling merchandise and stuff. But yes, with an aging gaming population, they are becoming less and less important.

[jerseystyle]

2 minutes ago, Lord Mushroom said:

Mascots are still important to Nintendo. It is one of the reasons why they are the most profitable of the three main console makers. They don´t just help in terms of selling games, but also in selling merchandise and stuff. But yes, with an aging gaming population, they are becoming less and less important.

Sure. But other than Nintendo I don’t think they are critical. I don’t think Atari would need one and it would probably be a waste of resources to create one.

[Lord Mushroom]

8 hours ago, Wildstar said:

Making retro-style game on modern systems is different than making games FOR the retro platforms of the 1980s/90s. 

If you think Atari´s plan is to make games for the original hardware, I completely understand your skepticism. But that is not their plan. It is not what they have done with their existing IP, so there is no reason to think they will do that with the IP they are buying.

[+S.D.]

8 minutes ago, jerseystyle said:

Sure. But other than Nintendo I don’t think they are critical. I don’t think Atari would need one and it would probably be a waste of resources to create one.

Totally agree.  Sony has like 70% of the global console market share, and around 50% in the US (with just over the other half going mostly to Xbox, a big chunk to Nintendo).

 

What is Sony's gaming mascot?  More to the point, who cares.  Atari has solid branding already, I just wish the jingle got used more (like in Albert's youtube ads for new homebrews).  I mean, the words were in the 2600+ launch trailer, but I missed hearing a synth-ey jazzed up version of it.  Oh well.

 

As for the topic of this thread... I'm still holding out hope that some of the unlicensed "similar but legally distinct" titles we lost could maybe somehow become actual licensed releases.  How cool would it be to get a Sega logo on the box front of our favorite blue porcupine!

 

I'm thinking of putting a 2600+ into a rental of mine with a box of dirt cheap fun classics, stuff like Combat, Yar's, Pitfall, Demon Attack etc.  Seems like a nice little feature for the guest to pick up & play whenever they want.  If I can figure a way to nail down the console itself somehow so it doesn't walk off, I could replace any missing carts every so often. 🤔

 

[Lord Mushroom]

6 hours ago, Giles N said:

Strange that its mostly japanese gaming companies that have strong, enduring, likeable characters:

It is probably because Japanese companies were dominant in the golden era of mascots. A lot of the popular mascots of today were made back then.

[Lord Mushroom]

33 minutes ago, jerseystyle said:

I don’t think Atari would need one and it would probably be a waste of resources to create one.

I agree that Atari should focus on making good games, but if they happen to be making a cartoony game, I think they should try to make the main character a potential mascot.

[Wildstar]

10 minutes ago, Zoyous said:

Maybe not this month or next month. Are you willing to take that bet out to next year? Five years from now? Ten? If it's within the realm of possibility, AIs are going to get there.

I'm not talking about something that's completely "bot generated." But in my own experience I've found ChatGPT helpful for writing code for specific aspects of larger projects. It does make mistakes and it requires some established knowledge to identify and correct those mistakes. I also have found AIs helpful as sounding boards in developing ideas, as assistants in developing visual designs, etc.

 

You really think a time won't come when an AI could analyze assembly code and find some optimizations that no human found? Or a time won't come when someone could design a game in some high level design app that tracks system resources, and then a number of AI agents working together can adapt it to code that will run on a 2600? I'm not sure when that may become possible but I think it will. It's a pretty niche interest so there's not going to be many people working on that angle. I think most of the focus will be on high-end mainstream gaming moving forward, as usual.

 

Someone excels at writing machine language code, that's a great accomplishment. There were people who were great steam engine designers as well, nothing's going to take that accomplishment away from them. Technology changed and knowledge built upon knowledge. The fact is, there are going to be a lot more people designing video games in the future and there is a very high likelihood that they'll be able to do it in ways that are comparatively effortless. It's not going to take away the enjoyment and satisfaction from people who still do it manually. Newbies might learn machine language (at least partially, and soon if not already) by chatting with an LLM. Some people, a relative few, start off buying furniture from IKEA and then eventually learn woodworking.

There already exist C compiler optimization but even then it sometimes not as good as hand crafted machine languages for simpler processor with as simplistic instruction set. 

 

C and C++ was intended for more complex processors like the 68K processors. Trying to clean up some of the messiness of C compilers kind of is more work than having hand coded the optimized code. Perhaps, someone can make some AI gadget to some of that. The problem is using a GAME ENGINE, which means using something like Godot, Unity, Unreal Engine 4 or 5, you got all this library for a bunch of stuff that won't even be used. You are better off using SDL.

 

Your best using the graphic tools to make bitmaps and sprites that are designed for the specific platform and use maybe cc65. That's what people been doing for the past 20 years. For SNES/Super Famicom, I would use something like those suggested here: https://wiki.superfamicom.org/  - guess what, it uses the tools of cc65.... namely the assembler.

 

There is also a host of tools like linkers and packers that would be used. These tools are purposefully made for these systems so the whole toolchain is optimized for the workflow. In a C compiler, it has to have libraries needed to target the compile for a particular computer, not just the CPU but the computer. Otherwise, you'll run into a nightmare of errors and failed compiling. Likewise, also the macro assemblers. These are what you use.

 

For Commodore 64/128, if you want to make graphics, cross-platform (and not use an actual C64 graphic program inside the emulator) - there's examples here: https://commodore.software/downloads/category/79-graphics-tools

For some video modes to use like in a title splash screen, you are better off coding it by hand crafted ML / Assembly. This way you got fine tooth control and manipulation of the raster interrupts. You know, like UIFLI.

 

For SNES, there are tools like YY-CHR for graphics/tiles. Suitable for a number of consoles. The tools were intentionally are made for these platforms. If you done game development for these systems and modern, you will know what I am talking about. For music on SNES: https://megacatstudios.com/blogs/retro-development/creating-music-and-sound-for-snes-games-a-crash-course-in-snes-gss

and https://megacatstudios.com/blogs/retro-development?page=5

 

These guys have modern tools for cross-platform development of NES/SNES and possibly some others in the retro game development process. What you will notices these are collectively an SDK toolset. With these consoles and computers, you need tools for graphics, music/sfx, and programming. There's no drivers. You make the code to read the inputs from joysticks, gamepads, light pens, graphic tablets, mouse, keyboard as well as code to load content from disks/discs, cassettes, etc. as well as store or write to like how you will record the save data. You must code the business end of a game.... the core.... the logic, input/output handling, etc. You may have to design how save data is recorded and structured and how it would be read back. There's no OS. The OS is the game more or less. It's 'bare metal' programming as it is sometimes called. Literally. This isn't something "modern game engines" like Godot or Unity is designed for and is more likely to cause more headache and trouble than it solves or addresses... for this kind of application. When you make a retro-style game to be used on modern platforms natively, then that's fine. That's what these tools are intended for. 

 

I think you should understand where I am coming from. When it comes to graphic design for these old platforms.... maybe something simple....like graph paper and color pencils might work. That's how Mario was designed.

 

 With a 65c816, you have 256 and only 256 opcodes. There's only 256 hexadecimal values that represents every operation in every addressing mode provided for each type of operation. In Assembly language for 65c816, there is even fewer mnemonics. The 6502, 6510, 6507, 65c02, 6502C, 65EC02, 7501, 8502, etc. has even fewer legal opcodes and a number of illegal opcodes but only with 8-bit hexadecimal representation for each opcode.

 

These are simple enough processors that a reasonable person could reasonably learn to do machine language programming. It is a unholy nightmare to learn how to do machine language programming for recent generations of Intel x86 processors from as far back as the Pentium to current. Instruction set architecture on these more newer CPUs are far too large to memorize. It isn't something you would want to do on 32/64 bit Intel/AMD CPUs. 

 

Jim Butterfield wrote some of the best books on learning 6502 machine language programming. God bless that man.