Undisputed World Record

2023-10-15

To the best of my knowledge, I am now the undisputed Dungeons of Drakklor any% @ 4000 cycles speedrun world record holder!!!1one come at me bro!

Wooo!!!

Okay, I think I've got an MVP for this site and PoCs for the rudimentary processes of all the associated activities. The "mise en place" is done, if you will. I can actually start sinking my teeth into this very soon, which is an exciting prospect. Some housekeeping first, primarily in the form of bulleted lists for the sake of brevity/laziness.

The Runs

I've probably done a dozen or so start-to-finish runs, but only just the two timed/recorded since getting that set up and debugged - I wanted those done so I could soft launch this actual site which has been sitting undeployed for too long. Anyway, some thoughts I've been having on running itself:

• Even without much practice/optimization, the "any%" runs are really quick - like 7-8 minutes. I initially thought they would take much longer.
• As I've been saying, a large chunk of the "midgame" does feel kind of skillless/linear which isn't an inspiring prospect.
• I still think optimizations will mainly be on level 1 and level 8, at least for this run type.
• I may end up migrating to my idea of "100%" runs that include getting all the powerups because I think that would definitely amp up the complexity, but TBD after I get a better impression of how much optimization potential there is for any%.

The Logistics

Primarily for run tracking, which I ended up getting through faster than I expected:

• Getting LiveSplit and OBS set up to do the timing/recording was actually really easy;
• but there is no "autosplit" support for Drakklor (obviously) so I have to do time splits manually, which feels awkward and imprecise;
• and developing an autosplit script for Drakklor seems complex, primarily due to the abstraction layer that is DOSBox, although I did find a thread about it to get me started.
• I'm really half-assing how I'm hosting the runs but will likely put off anything YouTube/Twitch related at least for a while -that would be lower on my priority list than the autosplit script.
• I don't think I want to record every run, which poses a problem: if I beat my PB AKA the WR then I won't have it recorded. Not that this really matters. I just don't like the idea of basically writing to my drive non-stop for hours, especially this early on in the process. Instead, I'll probably just post a new time each time I beat my PB/WR until things level off, then start recording again? This also would maybe be mitigated just by streaming to Twitch w/ VODs and then scraping out from the VODs as needed - TBD on all this again re: effort/etc.

Next Up

Short-term priorities at this point are:

• Get this site deployed (if you're reading this and you aren't me, I did it!)
• Do a good chunk of timed runs to establish a baseline and feel out optimization potential.
• Do some theory crafting re: optimization potential instead of just following my current strat. I think a theory crafting blog post once that's done would be a good idea also.
• Finish mapping out all of the levels - which is tedious. I still have to do the linear middle levels but they will become more relevant for 100% runs.
• Take a kick at writing an autosplit script - at least get an impression of how difficult/feasible it is.
• Keep posting to this blog occasionally, and update runs (at least the timing grid if I'm not doing recordings).
• Try to grok The Secret Hint.
• Chip away at (or at least think about) the more long-term goals and refinements that are pending.

That's all for now folks.