There has been some interest in additional categories. For example:

100% (all secrets and treasures) Boss1% (Any% run, up to boss1 aka The Defiled Church)

My question: What difficulty would you be interested in running 100% on? And should I make a leaderboard for Boss1%?

IMO Boss1% might be better left as a race category, but if people want a leaderboard for it I'll happily make one.

To be fair, it seems SDA has definitely made some good updates to its rules since I last looked at them (which would have been in 2013 or so), and the ruleset as a whole seems a lot more reasonable than what I remember. Then again I may have been thinking about their verification guidelines:

https://kb.speeddemosarchive.com/Verification_Guidelines

which I still think are asinine. I welcome additional opinions on this matter; I've been running on 1.00 all along, and am happy to continue doing so. I will leave 1.42d as an optional game version for now, pending further input from people who would prefer that it isn't allowed.

P.S. You're right, original game discs seem to be easier to get than I thought. But the merits of letting people get into running the game via the steam version shouldn't be understated. Maybe we should see if downpatching the steam version to 1.1 would help.

Your objection seems not to be against the changes made by the patch, but against the fact that it isn't in line with what the communities of other speedgames have decided. This patch is more readily available than version 1.00, and is almost guaranteed to be faster than 1.32 by virtue of having a working Heinrich skip, so why shouldn't it be allowed? Is there some general speedrunning authority that we all answer to that I'm not aware of?

Fixed, didn't copy it all :D

We now have a Discord, to serve as a central hub for discussion/asking questions. Here's a permanent invite link:

https://discord.gg/BUW2drb

EDIT: Fixed link

Does anyone know which version of the game GOG distributes?

That's a quirk of the id Tech 3 engine. If you set com_maxfps to some value that isn't an integer division of 1000 (e.g. 100, 125, etc.), the engine will cap fps at the next highest integer division of 1000. So in this case:

com_maxfps = 85 1000 / 11 ~= 90.90909090909091

Thus cg_drawfps indicates a framerate of 90.

EDIT: removed a gross generalization EDIT2: improved explanation

Unless there's a compelling argument to be made against it in terms of fairness, I see no reason not to allow people to run on the unofficial patch if they wish.

Exciting news!

After doing some testing, it appears that the 1.42d unofficial patch for RTCW is suitable for speedrunning! This is particularly important because it means that people can now install the steam version of the game, then apply the unofficial patch and run on equal footing with 1.00. Some of the game crashes have been fixed, which should improve quality of life for runners quite a bit.

I have added a value for 1.42d in the Game Version variable, and the win32 binary distribution of the patch is available on the Resources page as a .ZIP archive.

See "Patch 1.42d readme.txt" and "1.42d_Changes.txt" in the .ZIP archive for author and change information.

Update:

I've gone ahead and implemented the proposed changes above after soliciting feedback from a couple of runners. These seem to be reasonable compromises for everybody.

Happy running!

Alright, given the input so far, I propose the following:

ISSUE 1 PROPOSAL: A new category will be made to allow for this glitch.

ISSUE 2 PROPOSAL: The use of stock player movement cvars will be added to the rules.

ISSUE 3 PROPOSAL: All game versions will be allowed, and a variable will be added to indicate which version of the game was used in a run.

EDIT: Feedback on these changes would be appreciated.

Exosum generously shared lots of useful links with me, so I'll make the informational ones available here.

Informational article on connection-related cvars on funender: http://web.archive.org/web/20140427054244/http://www.funender.com/quake/console/q3connection.html

Framerate-dependent jump height calculator: http://dash9.ro/q/jumpheightcalculator/

NOTE: The following are cheat-protected, so you need to load a map with "spdevmap <mapname>"

Here are the map names, in order (not counting cutscenes): escape1 escape2 tram village1 crypt1 crypt2 church boss1 forest rocket baseout assault sfm factory trainyard sub norway xlabs boss2 dam village2 chateau dark dig castle end

Useful cvars for routing: cg_showAIstate 1/0, g_scriptdebug 1/0, cg_debugposition 1/0, aicast_debug 1/0, g_debugdamage 1/0, s_debugmusic 1/0, g_debugaudibleevents 1/0, r_shownormals 1/0, r_novis 1/0, r_nocull 1/0

Useful binds for routing: bind i "god" bind o "noclip" bind p "nofatigue" bind j "notarget" bind k "give all"

These are the ranges (in units) that enemies will be alerted for each weapon: Knife 64 Walking 256 Luger without silencer 700 Luger with silencer 64 Colt 700 Mp40 1000 Thompson 1000 Sten 64 K43 2000 K43 silenceer 128 FG42 1500 Grenade 1500 (upon throwing, not blast) Dynamite 3000 (upon throwing, not blast) Panzer 1000 Venom 1000 Flame 1000 Tesla 1000

My thoughts:

ISSUE 1: This changes the dynamic of the run pretty drastically. As such, I reckon this should get a glitched category or something. Then again, eliminating forest as an annoyance in the run can't be bad, can it?

ISSUE 2: If there's a significant majority in favor of changing these cvars, I'm fine with it. That said, it feels like a bit of an arbitrary change over the stock game. My gut tells me to run with stock values.

ISSUE 3: I don't feel strongly about this one either way. As it stands, 1.00 is at a rather significant advantage with the Heinrich skip, but it's not easy to get your hands on it (legally) unless you own an old copy of the game. As such, mandating a later version of the game would allow new runners to pick up the steam version and start on equal footing, which can only be a good thing.

Cheers.

EDIT: Misused "however"

Hello all,

There are a few decisions to be made regarding the rules for this game going forward, so I'm making this thread for runners who intend to continue running the game to share their thoughts. Please weigh in on as many of the following issues as you can.

¤¤¤ISSUE 1: Save/Load glitch¤¤¤ The following video was brought to my attention:

Basically, if you save your game and name it some (positive) integer, then load that save, the actual loading is delayed by the same number of seconds as you named your save (for example, the name "35" in the video means 35 seconds of delay before the save loads). The AI stops running as long as you're in this deferred loading state, allowing you to move through the level and kill things without being detected or attacked.

The question: Should this be allowed? Does it warrant a new category?

¤¤¤ISSUE 2: Player movement cvars (com_maxfps, pmove_fixed)¤¤¤ Anyone familiar with the q3 engine will know that these cvars affect jump height and distance.

The question: Should we allow these to be changed (e.g. com_maxfps [125 or 333], pmove_fixed 1), or run with the defaults (com_maxfps 85, pmove_fixed 0)?

¤¤¤ISSUE 3: Game version¤¤¤ Currently, the preferred game version for running is 1.00 (or 1.0.0 as it prints in the console) since the Heinrich skip was patched out. The steam version of the game is 1.32, and there have since been unofficial patches including the 1.42c version of the game, as well as iortcw, which is an open-source version in the same vein as ioquake3 which adds various enhancements to the engine.

The question: Should we enforce game version in the rules? If so, which version?

Thanks in advance for helping out.

Cheers, Mouse

EDIT: expanded save glitch explanation