Cain on Games - Tim Cain's new YouTube channel

[Russia is over. The end.]

It's just Cain is not some Carmack as people seem to think sometimes in this thread.

I've written about this kind of thing in this very thread in response to you https://rpgcodex.net/forums/threads/cain-on-games-tim-cains-new-youtube-channel.147060/post-8498467

Bugs aren't programmers' fault, but a byproduct of programmers' productive process. All types of manufacturing have an acceptable percentage of defects. Monitors with defective pixels used to be sold, as long as it was less than 3. CPUs are often manufactured with defects - if there's 2 or less defective cores, an i7 is rebranded as i5 and its cores are deactivated. If more cores are defective, it becomes an i3. And so is the case with programming - this type of production comes with an acceptable percentage of defects, i.e. bugs. The fact that some bug caused a crash is completely random, bugs appear in random places and it's nobody's fault.

 

[NoMoneyNoFameNoDame]

If anyone here thinks Tim is some kind of genius programmer, I invite you to take a look at his navmesh implementation for the world map in Fallout 1.
https://github.com/alexbatalov/fall...b9ca76a43a404f4c/src/game/worldmap_walkmask.c

absolutely nothing wrong with this
in fact its the best solution possible!
its not a navmesh either, its a ...mask
instant lookup for the low price of 1500*133 chars

He could have pre-sized static array for the mask data then loaded the real data from file to overwrite the static array.
Since he didn't (trusting Bester), I suspect it would be because the dimensions must be known to the compiler, he decided the data may as well be known too.

If the testing pipeline required a fresh compile each time anyway, then inserting up to date data for the array may be all the same
a .c export as any other file choice. This suggests two possible policies:

1) every tester always got a fresh build from their local repo of the game code. and local repo was frequently updated.
2) local repos got rebuilt for each tester, including possibly out of date code, and it was enitrely up to whoever made the wall data to distribute
.c file containing new mask data when it mattered.

Until Tim makes a video about this we will not know.

Psychologically the tendency to include things native to executable when possible speaks to the old times
when all you had was assembly anyway.

 

[Butter]

When ever I see someone make a post like that to discredit a programmer (especially in a completely meaningless way) I see their post as this:

Nobody's trying to discredit anyone. It's just Cain is not some Carmack as people seem to think sometimes in this thread. Read his Fallout 1 code, the repo is available. He seems to be on the level of Swen Vincke. He can script things, he's okay. If asked about how virtual or atomics work under the hood, he'd draw a blank.

I didn't realize Swen Vincke writes engines from scratch.

 

[Russia is over. The end.]

I didn't realize Swen Vincke writes engines from scratch.

 

[Butter]

I didn't realize Swen Vincke writes engines from scratch.

So not engines then.

 

[Russia is over. The end.]

So not engines then.

He made games that didn't run on any engine? Wow.

An engine is a renderer with a for loop, and optionally with some systems like audio/physics/animation/etc. All games run on engines.

 

[std::namespace]

If anyone here thinks Tim is some kind of genius programmer, I invite you to take a look at his navmesh implementation for the world map in Fallout 1.
https://github.com/alexbatalov/fall...b9ca76a43a404f4c/src/game/worldmap_walkmask.c

absolutely nothing wrong with this
in fact its the best solution possible!
its not a navmesh either, its a ...mask
instant lookup for the low price of 1500*133 chars

He could have pre-sized static array for the mask data then loaded the real data from file to overwrite the static array.
Since he didn't (trusting Bester), I suspect it would be because the dimensions must be known to the compiler, he decided the data may as well be known too.

If the testing pipeline required a fresh compile each time anyway, then inserting up to date data for the array may be all the same
a .c export as any other file choice. This suggests two possible policies:

1) every tester always got a fresh build from their local repo of the game code. and local repo was frequently updated.
2) local repos got rebuilt for each tester, including possibly out of date code, and it was enitrely up to whoever made the wall data to distribute
.c file containing new mask data when it mattered.

Until Tim makes a video about this we will not know.

Psychologically the tendency to include things native to executable when possible speaks to the old times
when all you had was assembly anyway.

bro, its obviously anno 1996,
its a win32 rig, there is no aslr, i would not be surprised if there is code that uses the raw, fixed virtual address + offset where the mask is always stored by the win pe memmapper!

90% chance that that mask was generated once and never touched again, there is zero reason to bother with any file loading here

testing pipeline? kek, if it doesnt bsod the win95 from userspace - its fiiiiiiiiiiiiiine!

 

[std::namespace]

https://github.com/alexbatalov/fallout2-re/blob/main/src/game/worldmap.c

oh look! i found an exploit in 30 seconds!
SUCH CODE! MUCH WOW!

send_a_guy_a_mask_file_wiht_a_name_longer_than_40_characters_and_you_own_his_rig_bester_wanna_test_my_f2_mods_and_metasploit_sk1llz?

 

[std::namespace]

and now apologize ^^

 

[StrongBelwas]

Source control wasn't used in the 80s, at least in the places Cain worked at.
The one programmer every code module had was pretty much the controller of the source, and any new build was made on that machine.
Rags to Riches may have been the first time they had a dedicated build machine that no particular programmer in charge of it.
Fallout was the first time they used software controlled source management. SourceSafe, from Microsoft.
SourceSafe was very good, you could check things in and out of the network, quickly download read only copies of the assets and make a new build. Making changes would require checking it out, making it writeable on your machine and can't be touched by anyone else.
The main drawback was that if you had a very large set of code, multiple programmers trying to make changes could run into trouble over who 'owned' it. Required making the code chunks smaller so one programmer can make edits to a section without causing trouble for anyone else. Interface programming was broken up into different chunks for the journal and the inventory to avoid this.
Linear history of changes helped them identify where some bugs had been introduced.
For Arcanum, they probably used Perforce, but the problem was that the world map needed multiple people to edit it. Obviously couldn't have one person lock up the whole world at a time. The editor would try to lock specific sectors once someone started working on it in the network, used file creation to create a lock. They had no file history.
Definitely had Perforce by ToEE, non exclusive check out was very useful. Mergers could create problem for some art assets. Sometimes people would try to merge assets and realize they were working on the same area or doing the same fix and then have to figure out how to make the changes work together. Generally, ToEE programming was well segregated enough this wasn't a constant issue.
One very helpful feature in Perforce was Shelving, which let you create a private little section you could put into the main branch when you decided on it. Something similar is branching, which lets you take all of the assets and split them into another branch to can choose when to work on, very helpful for demos. Patches, DLCs, and console ports can all be branches.

 

[Remax]

If anyone here thinks Tim is some kind of genius programmer, I invite you to take a look at his navmesh implementation for the world map in Fallout 1.
https://github.com/alexbatalov/fall...b9ca76a43a404f4c/src/game/worldmap_walkmask.c

https://youtu.be/qBKd2z43K7E?si=gVLVlUFIzgD8f-3x&t=638

Tim said on youtube that Kevin Wetherman coded the worldmap. You making shit up, Bester?

 

[Viata]

If anyone here thinks Tim is some kind of genius programmer, I invite you to take a look at his navmesh implementation for the world map in Fallout 1.
https://github.com/alexbatalov/fall...b9ca76a43a404f4c/src/game/worldmap_walkmask.c

Real genius programmers hate him because he (used to) releases great games with ugly code.

 

[Viata]

absolutely nothing wrong with this

https://github.com/AceLewis/my_first_calculator.py/blob/master/my_first_calculator.py

Nigga could easily write a code to generate that just for the joke. He is clearly trying to imitate Indian programmers.

 

[StrongBelwas]

IP should have a distinct core essence you can express in a single sentence
Arcanum was 'fantasy world that is undergoing industrial revolution.' Outer Worlds was Fallout Meets Firefly.
Cain think it's important the IP can be reduced to/easily recognizable by one symbol. Fallout would be the vault door or vault boy. For the Outer Worlds, they didn't go out of their way to do that, but the Spacer's Choice slogan and Moonboy mascot seem to have become that.
IP should be distinct, does not necessarily have to be unique. Arcanum cribbed a lot from Tolkien, but the industrial revolution made it distinct.
Should be able to write a lot of simple rules about what does and doesn't appear in the setting. Rules should be held to.
Eventually, if you have to keep using the same IP instead of something new, you end up exploring new areas or digging deeper into the old areas and probably overanalyzing them. Leads to dilution of the IP. It becomes unclear what can and cannot appear in it.
Feels like the Fallout IP is being diluted, he is at fault somewhat. He introduced the ghost sidequest into Fallout 2, Chris Avellone warned him not to, but he liked the idea of introducing supernatural elements. Realizes now Fallout doesn't and shouldn't have supernatural ideas, pushes the idea of the IP and makes it too broad. Introducing ghosts questions why you shouldn't bring in more undead and the afterlife. Rereferences his doubts about Synths becoming so important to Fallout 4. Feels like the 50s props in later games are being overused, makes it feel like the war happened in the 1950s, not the future the 1950s predicted to be.
The summary of the video could be IPs should be simple and distinct, don't dilute them, and sequels and DLCs often lead to dilution.

Recommends this video : Cain feels like it does a better job than him of explaining IP dilution over time.

 

[NecroLord]

IP should have a distinct core essence you can express in a single sentence
Arcanum was 'fantasy world that is undergoing industrial revolution.' Outer Worlds was Fallout Meets Firefly.
Cain think it's important the IP can be reduced to/easily recognizable by one symbol. Fallout would be the vault door or vault boy. For the Outer Worlds, they didn't go out of their way to do that, but the Spacer's Choice slogan and Moonboy mascot seem to have become that.
IP should be distinct, does not necessarily have to be unique. Arcanum cribbed a lot from Tolkien, but the industrial revolution made it distinct.
Should be able to write a lot of simple rules about what does and doesn't appear in the setting. Rules should be held to.
Eventually, if you have to keep using the same IP instead of something new, you end up exploring new areas or digging deeper into the old areas and probably overanalyzing them. Leads to dilution of the IP. It becomes unclear what can and cannot appear in it.
Feels like the Fallout IP is being diluted, he is at fault somewhat. He introduced the ghost sidequest into Fallout 2, Chris Avellone warned him not to, but he liked the idea of introducing supernatural elements. Realizes now Fallout doesn't and shouldn't have supernatural ideas, pushes the idea of the IP and makes it too broad. Introducing ghosts questions why you shouldn't bring in more undead and the afterlife. Rereferences his doubts about Synths becoming so important to Fallout 4. Feels like the 50s props in later games are being overused, makes it feel like the war happened in the 1950s, not the future the 1950s predicted to be.
The summary of the video could be IPs should be simple and distinct, don't dilute them, and sequels and DLCs often lead to dilution.

Recommends this video : Cain feels like it does a better job than him of explaining IP dilution over time.

People still complaining about the ghost in The Den and the Talking Deathclaws?
I thought we were over it already...

 

[Roguey]

Seems awfully limiting to think that intellectual properties should only be high concepts. Let some things be a bit more complex than that.

 

[Russia is over. The end.]

Maybe he's right that it should be reducible to an elevator pitch, but only because otherwise people will have a hard time understanding what they're playing, and recommending it to others, etc.

I have an idea for a game that's not reducible to an elevator pitch, and as a result I don't even know how I would market it.

 

[NecroLord]

I have an idea for a game that's not reducible to an elevator pitch

Do tell.

 

[Butter]

You need a way to distinguish your IP from the competition. How are you going to convince people to try your new online shooter instead of sticking with Call of Duty or Battlefield or whatever else? That's where the elevator pitch comes in.

 

[StrongBelwas]

Only get one chance to make that first impression
Really likes the Lynch Dune movie, knows he is in the minority.
Argues the 'beginning' of the game begins before the player even starts up the game. This is less relevant today, but how you presented the box and manual of the game could be important for the player's first impression.
Careful with the Fallout box to put a lot of screenshots on it. Tried to make the manual look like a found item from the world. Even the installation could be used to make an impression as the player would take that time to go through the manual and you could play music or visuals during the installation. Nowadays you're pretty much just watching a bar load in (And it's a lot faster than old installs) so it's less important.
Cain has never returned a game. Feels like you have 10 minutes to grab someone's attention, if you can't get them interested in that time and express the game, you are probably going to lose them. Used to say you had 30 minutes, wonders if it's getting even shorter than 10.
Cain feels he didn't do that very well. Even before Fallout launched, people on the team and QA and general Interplay staff were complaining the start was too slow and too brutal. Cain thinks now that the rat cave was the wrong way to go about it. Would have opened with a cinematic of the player going outside into a very bright desert and then moving into the isometric view where the player can see nearby that rats are chewing on a body and that there is some equipment nearby the body. You can fight the rats for the gear or walk/sneak away. Doesn't like that you can make a talking character but the first thing you are doing is fighting rats, but they thought it was important people knew how to fight before they went out into the world map and got dropped into a random encounter. Believes Fallout 1's poor attempt to teach the player led to Fallout 2's Temple of Trials mandated by marketing, although he still thinks the Temple was a bad idea.
Doesn't like having to play through growing up in the vault in Fallout 3 after the first time. Generally doesn't like intros that are too long and/or unskippable.
Thinks they did a good job with Arcanum's opening cinematic of condensing the idea behind the game.
Your opening should be fun to watch and make people want to figure out what your game is, make them think of what kind of character they want to make, make them think about what location they should go to first.
Done well, this is the best way of pulling people into your game.

 

[agris]

Cain thinks now that the rat cave was the wrong way to go about it. Would have opened with a cinematic of the player going outside into a very bright desert and then moving into the isometric view where the player can see nearby that rats are chewing on a body and that there some equipment nearby the body.

 

[Roguey]

Interested in a companion video to this regarding endgames. Tim's record on these is also a mixed bag with Fallout, ToEE, and Outer Worlds being quite good and Arcanum/Bloodlines being rather lousy.

 

[0sacred]

Really likes the Lynch Dune movie, knows he is in the minority.

how prestigious of Tim. I agree.

Lynch's movie was much more stylish, and they did a much better job with casting than nu Dune

 

[Bulo]

My only conclusion after watching both Dune films is that I don't much like Dune. That said, Lynch's version is more interesting visually than Villeneuve's. Also this:

they did a much better job with casting than nu Dune

 

[Wesp5]

My only conclusion after watching both Dune films is that I don't much like Dune. That said, Lynch's version is more interesting visually than Villeneuve's. Also this:

they did a much better job with casting than nu Dune

I agree. I mean they are targeting the typical Star Wars/Marvels blockbuster audience and then we get Poe/Aquaman/Thanos/Spider-Man's girlfriend/Thor's scientist and whomever I forgot in the same movie! Why not more new actors?