or Ridiculously Rousing Reckless Repositioning
Five versions is a pretty big period for DCSS. Going from 0.11 to 0.16 is like playing a different game. Even when the development was much slower, like in the 0.20s, the gap between 0.20 and 0.25 was huge. I noticed several things when I returned into 0.35a from from 0.30. There is the usual: new species and backgrounds, reworked gods, new items and spells. But some changes stood out more than the others. The very first thing I noticed is a better responsiveness of WebTiles, which brings the experience closer to playing Tiles locally. 0.30 wasn’t as slow as much earlier versions, but the improvements are noticeable still: 0.35a is comfortable to play even without turning animations off.
The next big change is the unlimited inventory™, the best thing since sliced bread. Finally I don’t have to worry about picking up scrolls, potions, and evokables. Although, a new problem appeared: I carry too much junk - weapons, artefacts, rings - and throw them away only when there is no room left. Surprisingly, the game gives you more artefacts now, although the number of consumables feels the same or slightly lower.
But the biggest shock was the talismans: I’ve tried the first batch when talismans replaced Transmutations spells, but the number of forms has tripled since! The talismans have added a whole new dimension to building your character. And it took some time to learn what they all do. Some of them are straightforward, but there is, for example, the Lupine talisman, which packs as many game mechanics as Usk or WJC.
The last but not the least big change is new monster perception. Enemies no longer unconditionally notice you when you cast spells, zap wands, or do some other actions, which makes Stealth more relevant. Although, it feels a bit NetHack’y when you fight one monsters and everyone else in the room don’t auto-aggro on you.
Also, there are a lot of small improvements, like the detailed potions and scrolls usage tables in morgues, which are amazing.
Gale Centaurs are the most complex iteration of centaurs, which strives to support four primary play styles. Their aptitudes are pretty decent across the board:
Magic apts were inherited from Armataurs with minimal changes: Air -1 → +2, Translocations 0 → +1, but Alchemy got -1 → -2. A pleasant surprise is a non-terrible Ranged apt: the original Centaurs had +3 Bows, but Palentongas and Armataurs had -2 and -3 Ranged apt respectively.
Their key game mechanic is four winds, which enhance your Stampede/Rampage. The current wind depends on your highest skill, and you can change it throughout the game. There is a melee wind (N), stabber wind (S), ranged wind (W), and caster wind (E). You could find the details in the CrawlWiki article, but I would like to focus on the usefulness of the winds.
It’s hard to pick the strongest: both caster and ranged winds are S-tier. The stabber wind is in the C-tier (plaCebo), and the melee wind gets a D (Don’t).
- The ranged wind gives you a penetrating shot, which does extra damage. You shoot at the movement speed even if you use a heavy triple crossbow at low skill. Your Stampede can push a line of enemies back. And on top of that you get Salvo shots when you stop Rampaging. This play style reaches its peak on V:5: you could just Stampede back and forth in the central corridors until there are no enemies left!
- The caster wind does even more things. First of all, you regain MPs, which stays relevant throughout the whole game. Nearby enemies get exposed: your spells become more accurate and deal more damage to them. Also, it lowers their Willpower, which is helpful for hexing monsters via spells and wands. Unlike other winds, your Stampede status doesn’t expire when you use spells. In my games, there were a lot of moments when I ran out of mana and had to “trust the wind”. I Rampaged even deeper into enemy crowds to get just enough MPs for one more spellcast, and it usually worked!
- On paper, the stabber wind benefits the corresponding play stile. But in practice it feels more like a placebo. You get extra Stealth when Rampaging, but when a monster doesn’t wake up, is it because of the bonus, or because you just got lucky? Stabs give you Tailwind, which allows you to move instantly. But unlike WJC’s Serpent’s Lash, it’s barely noticeable. I had to double check it in wiz-mode to make sure it even exists! I had a GCEn that was a perfect character for this wind: I worshipped Usk, had a rapier of speed, blade talisman, and double Rampage. But the char felt pretty unimpressive still.
- The melee wind is the simplest: 70% of the damage taken during Rampaging returns back to you as bonus regeneration. If you have decent defences, you don’t take much damage, and because of that the wind does exactly nothing most of the time. This effect feels like a half of the wind, it is less fun, and it is clearly the weakest of the four. Also, some talismans don’t work well with the melee wind. For example, if you use a spore talisman, then burstshrooms will constantly break your Stampede as they appear behind enemies and you can’t move them.
I’m a big fan of the non-advancing Tab, but GCs are all about (recklessly) running into the unknown. Also, GCs are very splat-friendly: if you start splatting them, it is pretty hard to stop. For example, I won 12 GCs on the first try, but it took me 45 games to win a GCFE. As a result, my GC win rate was only 11%, which is a significant drop from the overall 25% and the recent-ish 30-something percent win rate. I even had to play a few other species to confirm that I still remember how to tab.
Overall, Gale Centaurs are amazing and I fully enjoyed playing them. They are quite complex, sometimes tilt-inducing, but extremely fun.
I started my first GC - an Alchemist - immediately after Anemocentaurs were renamed into Gale Centaurs. I played it for a few minutes and then parked it in the Temple. I was in the middle of playing Healers, but I couldn’t resist an opportunity to get a trunk-only high score. I returned to this char after almost 6 months, when I finished megaprogress. sekai - a legendary CNC player - won this combo with 4 runes, so stealing this high score wasn’t a problem. Sequell and DCSS Stats scoreboard show this run as a GCAl game, but CAO has the original abbreviation: see the 2nd row in the list of all combo high scores.
There were several memorable games and epic moments.
Do you know what is better than three orc friends?
The answer is, obviously, six orc friends! It works as long as you can handle apostles. For example, I had a promising GCWn, but lost the char in the first challenge. Also, worshipping Beogh is the fastest and most consistent way to win. Such combos were faster than even ranged chars of Oka/Hep.
GCDe proved once again that Ash is still ridiculous after all the nerfs. I was able to get a Servitor with Bombard online on Lair:1:
After clearing Orc, the Servitor was casting LCS.
But my most OP centaur was a GCHW. It all started on D:3 where I encountered a hobgoblin with an amazing artefact:
And it wasn’t the only +Int item - this run had an abundance of such artefacts. After clearing Orc, I had 54 Int, which is ridiculous for a non-Chei/non-Dg char.
Also, speaking about Chei, the combo of Gale Centaurs and Chei feels even more illegal than Palentongas/Armataurs of Chei. Stampeding back and forth with double speed is not especially deliberating, is it? Chei forbids Rampaging, but GCs are exempt from this conduct, which is unique across all species and all god conducts. But it’s even weirder than that: GCs of Chei are not allowed to use Rampage-branded items! So if you find an artefact armour with Rampage, you won’t be able to wear it.
Playing DCSS is always a roller-coaster. One char gets the worst malevolent teleports:
Or you get a really big pan lord on your orb run :) :
On the other hand, you could get the best entrance into the Zot lungs, with zero traps:
Or you get an amazing trove for free (Swamp has plenty of open spaces, so the bane of Claustrophobia does almost nothing):
I can’t resist mentioning a couple of exploits and long standing bugs. Did you know that there was a reliable way of getting next to a Tesseract without waking it up?
All you need is 12+ Evocations and a good stack of vulnerability scrolls and polymorph wand charges. If you kill polymorphed orb guardians, you won’t trigger the alarm. Unfortunately, it was fixed at the end of the 0.35 dev cycle.
The following Ash-related bug is very old, but I always find it amusing. It is a weirdly familiar monster, right?
Currently, we live in an era of abundance thanks to the currently active devs. The number of new and reworked features the players get with each new version is huge. In this sense, the early 0.30s are as far away from the early 0.20s as they could be. It also means that the current problems are also quite different.
There is nothing new about stairdancing, but some recent changes made it even more useful. In old versions, you could stairdance too, but respawning monsters added some risk to that. Also, we had a couple of weak clocks: the hunger clock and piety decay, which limited the wait-on-stairs tactic, especially if you worshipped a fast piety decay god or played permanently hungry Trolls, Ghouls, or Vampires. The removal of piety decay deleted the last real limiting factor, because the Zot clock gives more than enough time for waiting.
Also, the game got more player tools with cooldowns, which benefit from stairdancing. The orb of stardust deserves a special mention here, because it is the ultimate stairdancing tool. Of course, you can’t win the game by waiting on stairs, but there are more than enough moments where stairdancing, waiting, and luring are the safest and the most boring methods of fighting monsters. Something should be done about this. Luckily, the devs are aware about the problem, and there are some ideas how to address this, like Level Feelings.
From the posts about megaprogress, you know that I like spotting bugs in the game. When I returned to playing trunk, I’ve started making notes each time I noticed something broken, missing, or requiring an improvement. This list includes only things from the last two versions: everything I missed playing megaprogress games. I have no idea how many items were on the list in total: the devs fixed some bugs while I was playing GCs, then I found more bugs, the devs added more content with brand new bugs, and so on. But currently there are 65 bullet points left. If I had to quantify the number of bugs in 0.35a, I’d say it’s 10% - 20% more than what I’ve seen at the start of 0.30s. I would like to mention three annoying bugs to illustrate this. All of them were reported months ago and are known to the devs already.
The first one is related to the invisibility rework, and you’ve probably seen it when playing ranged chars. The game refuses to shoot at monsters if there is an invisible monster in front of them:
The second one is kinda similar, but it affects not only ranged chars, but blaster mages too. The game doesn’t auto-target auxiliary monsters like briar patches and shooting stars. Like, if a thorn hunter is standing behind a briar patch, you have to manually shoot each arrow while spells can target the thorn hunter just fine. Shooting stars are worse: you could destroy them with attack spells or push them back with the Mystic Blast. But you need to force-cast these spells, since the game refuses to auto-target stars.
The third one is related to auto-pickup. The pickup of unknown magical staves is just broken. Auto-explore would bring your char to a staff, step-identify it, and move further without actually picking the staff up, since it’s a known item now. This makes playing mages a bit more annoying than it should. There’s a similar problem with unidentified jewellery too.
It’s unlikely that the currently active devs forgot how to write good code. But if each new big thing adds a few bugs, then having more features means having more bugs in total. Another source of bugs is the complexity of new features, especially when they interact with each other. For example, when item-melding forms meet slot-granting unrands.
Another problem is a bit unusual, but maybe not a completely new one. I don’t remember it being as prominent in old versions, though. It is unfinished and unexplored ideas.
Grave Claw is a good example. This Necromancy spell relies on harvesting souls, which give you charges. You get more damage out of the spell, but the number of charges is limited, so you can’t spam it. This is an original and fitting game mechanic for the Necromancy spell school, but it’s used only for one spell.
Another example is the diversity of Beogh’s apostles and orcs in general. We have a cosmopolitan and flavourful god, but it’s not reflected neither in monster orcs nor in apostles. Blorkula is the only exception, but where are dual-wielding Exo-Orcs or glass-cannon Deep Orc mages or Big Orcs with giant clubs or Minotorcs with a headbutt attack? There is so much unexplored design space here!
The final example is baubles. Imagine having a whole new type of consumables with only one item in it! Now, after 18 months since the creation of the first bauble, there is the second one.
Another issue is the lack of removals. As a player, I like new stuff, but removals are an essential part of the development process. It’s debatable what should be the next thing to remove, but it’s clear that the current development is primarily focused on adding more and more features.
The time between releases steadily grows in recent years. 0.33 took 8 months, 0.34 took 9 months, and for 0.35 it will be almost 10 months. At the same time, the feature freeze period, when the devs stop adding new features and focus on balancing things, fixing old and accumulated bugs, and reviewing PRs, is the same two weeks and it is not getting longer.
If you have a mega-talented dev who is amazing at creating new things, it would be a waste of their time if they had to fix bugs full time instead. DO is against using AI for coding, so delegating bug fixing to AI or just adding “make no mistakes” is not an option :). Cloning DO is not going to work too: according to DO herself, DO-2, DO-3, etc. would rather add even more features, instead of, say, rewriting the graphical backend of the game.
Hopefully the community would be a bit more active and contribute more PRs and fixes. Maybe some old devs would come back, or maybe we would see new ones - the more the merrier. In any case, it is going to be interesting to see how the dev team handles these challenges while creating new features as well as maintaining balanced gameplay and polished experience for all of us to enjoy.
I’d like to end this post with a screenshot of three steam-cooked chonchones. This feels like a perfect metaphor for Gale Centaurs in particular and 0.35a in general: they are original, flavourful, but a bit raw.












