On Mon, Jun 27, 2011 at 04:56:02PM +0200, Raphael Langella wrote:
> First, we won't get rid of victory dancing in 0.9.
That's a major bummer. This is the part of 0.9 I hoped for the best.
However, if we follow Elliptic's advice and start the tourney on Aug 13
instead of Aug 1, the branching could be shifted by that amount. Would that
be enough?
> * Base HP has been boosted.
> HP from fighting reduced and scaled to species HP. Big nerf to frail
> species with high fighting like SpBe.
And a big boost to Demigod and Ogre casters. Even DE casters have been
boosted, so there's far smaller need for everyone to train fighting.
> * tornado's damage has been reduced by about 40% (and recently changed to
> conj/air).
Plus code changes, like (implemented) tossing around monsters, (to do)
tossing items, fix for the translocations abuse and possibly other changes.
> * skill costs (how many XP for a skill point) are no more dependent on skill
> level.
> The best players have reported that to reliably beat D:1, advanced pillar
> dancing techniques were often needed. Knowledge of the energy randomisation
> system and how to abuse it doesn't seem like the most interesting skill for
> a good crawl player. We can change the randomisation, but more importantly,
> we can start by making it less relevant. So it has been suggested to raise
> initial HP.
> I suggest we give +3HP to everyone. It's a bigger buff for casters with
> their low HP which balances out the skill training nerf.
It will do very little to fix stair dancing, but can at least produce less
random deaths. Let's do that.
> Then there's tornado. Adam changed it back to conj/air because he wants to
> balance the spell against firestorm. From what I've heard, it still more
> powerful than firestorm,
Uhm, how come? Rapid fire firestorm gives about thrice the damage, while
single-cast firestorm when you let them simmer is almost as powerful. This
is altered by resists which are hard to directly compare, but outside of
Ziggurats you rarely have big open areas so tornado can't operate at full
capacity either.
> But many of us don't want a third storm spell.
That's why it operates differently, and could be made even more different.
Distinct mechanics is good, much weaker power is not.
The only argument I can agree with is that "fire has no utility spells, so
it deserves to be stronger". So an interesting nerf could be added, but a
straight change to damage would be bad.
> Here is a proposal for a nerf:
> * Radius increase progressively from 2 to full range.
Sounds good, although what is that "turn" you're talking about? At that
time 10 aut turns are quite rare. It's not that straightforward to implement,
but it could be done by having a radius ramp and calculating an integral
radius affecting a given square over the turn's duration.
For example, if the radius at t=0 would be 0, t=10aut 2, t=20aut 4, t=30aut
5, effective radius during 10 aut turns with my formula would be 1/3/4.5/5/5
respectively.
> * A cooldown timer of 3 turns. The time it takes for all the clouds to
> dissipate.
Bad idea, primarily for thematic reasons. I'd go the other way, and promote
recasting it while it lasts -- pumping more mana into sustaining the winds.
If you insist on a nerf, it could be somewhat stretched in time.
> * Change school back to pure air.
Terrible idea, for reasons I outlined so many times. L9 spells in a 3 runer
simply can't work without being hardly stronger than a two-school ~L7.5.
And there are two already.
* Fix translocations abuse.
It turns out some people use controlled blink to jump from one target to
another, positioning themselves to wipe most of low-hp opposition and do
damage to as many harder targets. You can even use controlled teleport for
that! This is what needs nerfing badly. Theme-wise, how exactly would a
new vortex of winds spawn around you if you put energy spinning up one on
the other end of the level?
Proposed nerf: reduce remaining duration by 10 aut for every unit of
(Euclidean) distance translocated. In all real cases this aborts the spell.
This adds implementation issues with blinking from above water / to above
water.
* Fix the two other high-end air spells.
CBL and Chain Lightning are better fit for that casting difficulty niche.
* Implement tossing items around.
Not needed for balance, can cause some implementation issues, but badly
needed for theme, if the spell is supposed to have anything in common with
its description.
* Better mixing of clouds.
Not that vital, current pure randomness could be argued to work good enough
as air mixes so much faster than things caught in the wind. It might be
good to allow overwriting the visual clouds with new actual ones, though.
Other changes for 0.9:
* Okawaru piety.
I hoped to get at least that in, even if bigger changes would need to wait
(especially that most of dpeg's ideas would need radical rethinking). It
does change the flow a lot. Being effectively exp/factor(xl) has the
potential to be saner than what we have currently, but it's probably better
to not introduce drastic changes at this moment. I think I'll disable it
completely for 0.9.
* Unicode fallbacks
Ie, replacing "é" with "e" rather than "?" when needed. Support for
fontconfig didn't make it for 0.9, so this could be nice. Not vital, but I
realized how ugly Windows char dumps look -- and they wouldn't be improved
with fontconfig too. Quite easy to add, I'll see if I'll manage this.
* corrupted maps
I'm afraid this is a severe but neglected bug. I fixed database builds just
before 0.8 but didn't do maps as a fix would be more intrusive. Would be
good to get this done.
* new Abyss
Let's see what bmh says.
* webtiles
Uh oh. A huge, intrusive feature. We don't have a choice though, and it
_has_ to be integrated or players will flay us alive.
* tile maintenance in general
I'm afraid that with no tile developers alive, no one bothered with even
easy bug fixes and improvements. Since most of the userbase (by numbers)
use tiles, I'm afraid this needs a look over.
* Mac maintenance
Not a single person can do Mac builds or fix bugs. Baaad.
* catoblepas
A non-important feature, but also a totally unintrusive one. I don't think
there is a reason to reject it if Wensley finishes it before the branching.
Disabling generation is trivial...
* shotgun targetting, improved beam targetter
Not for 0.9.
* lightning bounces
I think we should leave the physics as-is (since it works well in other
cases and is realistic!), but artificially limit beams to affect the same
square at most twice. Playing an AE is currently too spoily, and massive
unexpected damage from elf casters and storm dragons forces players to be
both spoiled and on guard for such quirks.
* wand of polymorph other
There's a suggestion to disable it in hands of monsters, like they don't use
random effects, digging or enslavement. A random gnoll or Ijyb giving you
six bad mutations just makes people quit. Neqoxecs being able to do the
same is ok since at that time the player is supposed to have means to deal
with the downfall somehow.
"Polymorph Other" and "Malmutate" could be separated in 0.10.
* rune display improvements
Mostly, I have a lack of good ideas. I pondered improving the '%' screen,
but if one already knows to look there, the data is already present. Not
sure what else to do, but avoiding confusing players would be nice.
* makefile improvements
Existing state, while quirky, works good enough for 0.9.
* reST -> text converter
This matters only for the release itself, but still needs to be done. Could
be implemented after the branching, though.
--
1KB // Microsoft corollary to Hanlon's razor:
// Never attribute to stupidity what can be
// adequately explained by malice.
------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security
threats, fraudulent activity, and more. Splunk takes this data and makes
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
Crawl-ref-discuss mailing list
Crawl-ref-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/crawl-ref-discuss
