[Ohrrpgce] SVN: james/12694 New enemy bitset "Controlled by Player"

Ralph Versteegen teeemcee at gmail.com
Thu Jan 27 01:33:54 PST 2022


On Thu, 20 Jan 2022 at 14:56, James Paige <Bob at hamsterrepublic.com> wrote:

>
>
> On Wed, Jan 19, 2022 at 6:49 PM Ralph Versteegen <teeemcee at gmail.com>
> wrote:
>
>>
>>
>> On Wed, 19 Jan 2022 at 13:49, James Paige <Bob at hamsterrepublic.com>
>> wrote:
>>
>>> So I do have a few other changes related to this one planned.
>>> * An option to make heroes controlled by (random) AI
>>> * A concept of "traitor" which will affect targeting classes when an
>>> attacker is targeting
>>> * A concept of "turncoat" which will affect targetting classes when an
>>> target is being targeted
>>> * Attacks that can turn these effects on and off or set-to-default
>>>
>>> So an enemy with all 3 of Controllable, Traitor, and Turncoat would
>>> function as a hero for that one battle.
>>>
>>> To simulate a classic "Confuse" status, you would have an attack that
>>> turns Controllable off, and traitor on, but don't touch turncoat. Then to
>>> end that status, use an attack that sets Controllable and Turncoat back to
>>> default.
>>>
>>
>> I was hoping this meant you were going down this direction :)
>>
>> I'm not sure whether "Traitor" is proposed to swap foes and allies of a
>> target, or just makes everyone count as a foe. Those are two different ways
>> that you might want a Confused status to work, and it seems that these bits
>> would only allow one or the other.
>>
>> What I was thinking was to give each combatant a team (default 1 for
>> heroes, 2 for enemies) and an "acting" team. A target is considered an ally
>> by an attacker if their team is the same as the attacker's acting team,
>> else they're a foe. Also team 0 could mean "independent", with no allies.
>> You probably wouldn't use more than a third team, for "Nature", say when a
>> clan of hyenas opportunistically attack while you're fighting someone
>> else).
>>
>> So Confuse to make someone attack anyone indiscriminately would change
>> their acting team to 0 (so two confused targets still hit each other), and
>> to swap sides you'd change their acting team (although now I realise that
>> means the attack would need to be specific to use by heroes or enemies,
>> unless there was an attack bit like "swap target's acting team" that just
>> set it to the attacker's).
>>
>> Maybe I've overcomplicated it again, while still not adding all that much
>> utility/flexibility (really should work on allowing script hooks for things
>> like this) vs just adding a third Independent bit.
>>
>
> Yeah, I think teams will overcomplicate it for now-- and yes, having
> scripting hooks so people can customize this behavior will be the best way
> to get advanced fancy effects
>

> I do kinda like the idea of being able to make a confused enemy target
> all, rather than only the opposite side, but I'll have to think if there is
> a nice simple way to do that.
>

Going down the route of bitsets then I don't really see another option but
adding another bitset to make everyone an enemy.

Now, I'm not pushing for the team IDs idea, but I just wanted to write
something about complexity. Say you add a third bit, or even a fourth ("Foe
to all"). I think that arguably two integer-valued settings are simpler
than 3 bits, because 3 bits is 8 possible combinations, a lot to think
about. And even an 8-way setting could be simpler to reason about than 3
bits if you don't have to think about any interactions. Complexity of
implementation is usually also secondary.
But in fact after looking at the new version of get_valid_targs I realised
team IDs would actually have been simpler in implementation too. The
bitsets are more complex... in fact I see some mistakes in the code, which
I'll fix: "Dead-ally (hero only)" and "Dead foe (enemy only)" were meant to
be informative only, to warn that those settings didn't make sense for
enemies/heroes, but not to intentionally restrict the targets.



>
>
>>
>>>
>>> Or a classic Berzerk could be implemented with Controllable=Off and
>>> could end with controllable set to default (this would work for heroes, but
>>> wouldn't do anything meaningful on an enemy)
>>>
>>> This should allow a lot of possibilities, and is all pretty easy to
>>> implement.
>>>
>>> And yes, someone could totally fake 5 or 6 heroes in the party with
>>> this, by using an instead-of-battle script, and adding hero enemies to the
>>> formation with a script before the battle starts. Definitely not ideal, but
>>> fine if people want to try it.
>>>
>>> Actually increasing the size of the active party > 4 and increasing the
>>> number of enemies in a formation > 8 is something I definite;ly want to do,
>>> but it will require lots and lots of cleanup, which is outside of the scope
>>> of what I am trying to do right now. In particular, there are tons of
>>> places where the ID range within the bslot() array defines what a
>>> BattleSprite Instance does, so the first step of that cleanup will probably
>>> be to convert all access to bslot() to a set of accessor functions for
>>> heroes, enemies, attack sprites, and weapon sprites. Then those different
>>> ranges can be split apart into different arrays, which can be dynamically
>>> sized when you load a battle formation with 15 enemies in it, or something
>>> like that. But that is for later. I want to keep the scope of what I am
>>> working on broken down into bite-sized baby-steps to mix a metaphor :D
>>>
>>
>> I don't think we would want to split bslot() into separate arrays for
>> heroes and enemies: being able to index across all of them with a bslot()
>> index is very useful and widely used (eg. targeting) so it would be a lot
>> of work to remove that. Why not just add is_hero and is_enemy attributes.
>> There's a lot of lines of code to change, but each would then be an easy
>> change. Could also start using polymorphism.
>>
>
> Yes, you are right. is_hero and is_enemy attributes are much better than
> what I was thinking of with the accessor functions for bslot. Glad you said
> it :)
>
>
>
>> On the other hand, I do want to remove attacks and weapons from bslot()
>> and was considering doing it soonish. Almost all of the BattleSprite data
>> is irrelevant for them, and nearly all of the advantages of having them in
>> bslot are (or will be) gone now that battles are converted to slices.
>>
>
> Ah, right! Those only get used in animations, so the slice is all that
> really matters :)
>
>
>>
>>
>>> Fortunately I think the current features I am adding will not make any
>>> of that later work harder, and might even lead to a little helpful cleanup.
>>>
>>> ---
>>> James
>>>
>>>
>>>
>>>
>>>
>>> On Tue, Jan 18, 2022 at 8:22 AM Ralph Versteegen <teeemcee at gmail.com>
>>> wrote:
>>>
>>>> Wow! That's not a feature I was expecting to see for a long time. A
>>>> nice surprise!
>>>>
>>>> I suppose this is particularly useful for giving the player extra
>>>> actions they can perform in battle. People are going to inevitable think to
>>>> use it to get around the 4 hero limit, but it seems really problematic for
>>>> that. Or is time to add team numbers to battles, so you can define which
>>>> combatants are "foe" or "ally"?
>>>>
>>>> On Mon, 17 Jan 2022 at 14:01, <subversion at hamsterrepublic.com> wrote:
>>>>
>>>>> james
>>>>> 2022-01-16 17:01:32 -0800 (Sun, 16 Jan 2022)
>>>>> 39
>>>>> New enemy bitset "Controlled by Player"
>>>>> ---
>>>>> U   wip/bmodsubs.bas
>>>>> U   wip/enemyedit.bas
>>>>> U   wip/loading.rbas
>>>>> U   wip/udts.bi
>>>>> U   wip/whatsnew.txt
>>>>>
>>>>> _______________________________________________
>>>>> Ohrrpgce mailing list
>>>>> ohrrpgce at lists.motherhamster.org
>>>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>>>
>>>> _______________________________________________
>>>> Ohrrpgce mailing list
>>>> ohrrpgce at lists.motherhamster.org
>>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>>
>>> _______________________________________________
>>> Ohrrpgce mailing list
>>> ohrrpgce at lists.motherhamster.org
>>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>>
>> _______________________________________________
>> Ohrrpgce mailing list
>> ohrrpgce at lists.motherhamster.org
>> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>>
> _______________________________________________
> Ohrrpgce mailing list
> ohrrpgce at lists.motherhamster.org
> http://lists.motherhamster.org/listinfo.cgi/ohrrpgce-motherhamster.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.motherhamster.org/pipermail/ohrrpgce-motherhamster.org/attachments/20220127/33157d9f/attachment-0001.html>


More information about the Ohrrpgce mailing list