Occupant Combat, Redux

Occupant Combat, Redux

[Elijah Meeks]

Hey guys, if somebody gets a chance, can there be some
kind of overhaul of the occupant combat system?  I was
searching through the old archives and I noticed a
post by Hans stating that the current system more
represents units in a transport, like an APC.  Is it
possible to add a separate routine for units in a
city?  Otherwise, I think it's a real problem that
when units attack a place, they get free attacks on
all the occupants in that place.

I know, I know, but I promise I'll pick up Lisp for
Dummies next week and see how much I can contribute to
the actual code.



__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com

Re: Occupant Combat, Redux

[Hans Ronne]

>Hey guys, if somebody gets a chance, can there be some
>kind of overhaul of the occupant combat system?  I was
>searching through the old archives and I noticed a
>post by Hans stating that the current system more
>represents units in a transport, like an APC.  Is it
>possible to add a separate routine for units in a
>city?  Otherwise, I think it's a real problem that
>when units attack a place, they get free attacks on
>all the occupants in that place.

You can set protection to 0, i.e. full protection, and the occs will never
be hit when you hit the transport. Furthermore, if  you leave
occupant-escape-chance at its default 0 they will be killed when the
transport is killed (but remain intact until then).

With respect to the city, bombarding it should damage the occupants, so I
see nothing wrong in principle with this.

What I find strange in this code is rather what happens if you miss the
transport. You must all have seen the frequent message: "Your unit a misses
enemy unit b and hits its occupants c, d and e". But this is how things
work. If we change this scheme we would also have to change every game that
uses the standard combat model. Which I doubt would be a good idea right
now.

Hans

P.S. occupant protection in combat model 1 works quite differently. See
table.def and combat.c for details.

Re: Occupant Combat, Redux

[Elijah Meeks]

> You can set protection to 0, i.e. full protection,
> and the occs will never
> be hit when you hit the transport. Furthermore, if 
> you leave
> occupant-escape-chance at its default 0 they will be
> killed when the
> transport is killed (but remain intact until then).

Yes, but with full protection, a place can have no
capture chance (As my fortresses are set up in
Specula).  Otherwise, if they can be captured, then
the units within the cities, which are supposed to
defend it, instead are wiped out or captured.

> With respect to the city, bombarding it should
> damage the occupants, so I
> see nothing wrong in principle with this.
> 

Most units aren't bombarding a city in the sense you
describe (Artillery or strategic bombing), they fight
the units within they city, and when they do the way
the current system is set up, their fire is multiplied
by the number of occupants.  Even with bombarding
units, this is only realistic if you're trying to
simulate some kind of troop density in a hex, so that
a more heavily occupied hex is vulnerable to strategic
weaponry.  That's situational, and should be handled
as an exception and not a rule.

> What I find strange in this code is rather what
> happens if you miss the
> transport. You must all have seen the frequent
> message: "Your unit a misses
> enemy unit b and hits its occupants c, d and e". But
> this is how things
> work. If we change this scheme we would also have to
> change every game that
> uses the standard combat model. Which I doubt would
> be a good idea right
> now.

You could add it separately, so that xconq still
recognizes and handles "occupant" and "transport"
relationship as it does now, while we can add a
"resident" and "dwelling" relationship which would be
subject to the revised system.  This way we can keep
the old occupant-transport rules, which are suitable
in many cases, and utilize a more realistic system for
units in a place.


__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com

Re: Occupant Combat, Redux

[Bruno Boettcher]

On Wed, Jul 02, 2003 at 11:27:15AM -0700, Elijah Meeks wrote:
> > With respect to the city, bombarding it should
> > damage the occupants, so I
> > see nothing wrong in principle with this.
yep but there's somehow a lack of consistency between an attack in open
field (i attack with a bomber a hex filled up to the brim with units,
    and my bombs hit only 1 unit) and one onto a place (where all units
      get hit..)

i absolutely agree that units should be damaged if their container is
damaged but its still annoying if a lone armor is able to take a city
with 5 armor in it...

> Most units aren't bombarding a city in the sense you
> describe (Artillery or strategic bombing), they fight
true, there should be a disctinction between shooting and bombing, means
point weapons and surface weapons

> > What I find strange in this code is rather what
> > happens if you miss the
> > transport. You must all have seen the frequent
> > message: "Your unit a misses
> > enemy unit b and hits its occupants c, d and e". But
yah that one kept me wondering for years now :D

> > work. If we change this scheme we would also have to
> > change every game that
> > uses the standard combat model. Which I doubt would
> > be a good idea right
> > now.
hmmm but i think that if we manage to move to something more consistent
and logical it would be a necessary step to take IMHO sooner than later
;)

> You could add it separately, so that xconq still
> recognizes and handles "occupant" and "transport"
> relationship as it does now, while we can add a
> "resident" and "dwelling" relationship which would be
> subject to the revised system.  This way we can keep
> the old occupant-transport rules, which are suitable
> in many cases, and utilize a more realistic system for
> units in a place.
i can only and absolutely second this !! coming to combat, cities should
be regarded more as a type of terrain than as a unit.

-- 
ciao bboett
==============================================================
bboett@adlp.org
http://inforezo.u-strasbg.fr/~bboett
===============================================================

Re: Occupant Combat, Redux

[Hans Ronne]

>yep but there's somehow a lack of consistency between an attack in open
>field (i attack with a bomber a hex filled up to the brim with units,
>    and my bombs hit only 1 unit) and one onto a place (where all units
>      get hit..)

>> You could add it separately, so that xconq still
>> recognizes and handles "occupant" and "transport"
>> relationship as it does now, while we can add a
>> "resident" and "dwelling" relationship which would be
>> subject to the revised system.  This way we can keep
>> the old occupant-transport rules, which are suitable
>> in many cases, and utilize a more realistic system for
>> units in a place.
>i can only and absolutely second this !! coming to combat, cities should
>be regarded more as a type of terrain than as a unit.

This is how combat model 1 works. The attacker is pitted against one
defender at a time, and has to defeat all of them in order to capture the
city. The city is in a way reduced to a battle field (just as in Civ).

This kind of scheme suffers from the opposite problem, however: it is very
difficult to capture and adequately defended city.

Hans

Re: Occupant Combat, Redux

[Bruno Boettcher]

On Wed, Jul 02, 2003 at 08:59:03PM +0200, Hans Ronne wrote:
> This is how combat model 1 works. The attacker is pitted against one
> defender at a time, and has to defeat all of them in order to capture the
hmmm that's indeed not the ideal way either...

that's why the idea to separate surface weapons from point weapons would
add a new dimension to the bombers :D

are there allready separate hit chances when combatting inside a city?
in a city infantry should be somehow superior to armors, just an
example... 
if not maybe that could be an idea?

the other thing is, that basicly all units fight as lone units, there
doesn't seem to be something like a combat group concept. (even the
    fleet concept is weird and doesn't work very well)

e.g. i wanted to test the russian setup with infantry piggy packing
armors: not feasible, the infantry is wiped out even when they are
supposed to hide behind the tanks....

the fire attractor that represent the armors isn't taken into account :D

other thing that wondered me:
i came over with mighty 12 armors and run into a cell with 1 armor, that
armor took me down 7 of mine...... and this too is a direct consequence
of the fact that all units fight alone

-- 
ciao bboett
==============================================================
bboett@adlp.org
http://inforezo.u-strasbg.fr/~bboett
===============================================================

Re: Occupant Combat, Redux

[Eric McDonald]

On Wed, 2 Jul 2003, Bruno Boettcher wrote:

> are there allready separate hit chances when combatting inside a city?
> in a city infantry should be somehow superior to armors, just an
> example...
> if not maybe that could be an idea?

The existing occupant-combat table partially addresses this, I think.

Re: Occupant Combat, Redux

[Eric McDonald]

On Wed, 2 Jul 2003, Bruno Boettcher wrote:

> > > work. If we change this scheme we would also have to
> > > change every game that
> > > uses the standard combat model. Which I doubt would
> > > be a good idea right
> > > now.
> hmmm but i think that if we manage to move to something more consistent
> and logical it would be a necessary step to take IMHO sooner than later
> ;)

I'm sure that there will be other releases after 7.5, in which this can be
addressed. To address this, there are several things to be considered, IMO:
(1) Do we continue to support what would end up becoming a "legacy" combat
model and have more complex code as a result? And thus run the risk of
interesting bugs creeping in?
(2) Or do we completely change over to a new combat model (after agreeing upon
our expectations for it), and then go through each and every game to see if the
new model makes sense in that context, and rectify it if not?

Either way, the resulting activity could significantly delay the 7.5 release.
As it stands now, pre-7.5 already has a number of significant improvements over
7.4.1 that enable it to stand on its own as an important milestone.

From a user perspective, I am as eager as you are for new features. But, as a
developer, I feel that code cleanup, interface updates, and bug fixes should
end out this cycle, not a new major feature. Just my opinion, of course.

Re: Occupant Combat, Redux

[Jim Kingdon]

> Otherwise, I think it's a real problem that when units attack a place,
> they get free attacks on all the occupants in that place.

It makes attacking a transport (including, for example, a city in the
standard game) a fairly powerful move, yes.

As a defender, this means one is well served to distribute one's
occupant unit among a greater number of transports.

With respect to game balance and such, I'm not sure whether this is a
good thing or not, but it certainly is something which has affected my
strategies and tactics quite a bit.

> I know, I know, but I promise I'll pick up Lisp for Dummies next week
> and see how much I can contribute to the actual code.

On the whole, the Language Syntax section of the xconq manual
(refman.texi) tells you everything you need to know about lisp.
Unfortunately, it throws in a bit of lisp terminology here and there
and could use more examples (both things would be nice to fix, given
the number of people who know lisp isn't so great these days).  But
you don't really need to know lisp - the overlap between xconq and
lisp (chiefly syntax) is smaller than superficial impressions might
imply.

I suppose if we really want to address this perceived barrier, we
could switch to XML.  Although it isn't clear whether xconq would fit
XML as well as other uses of XML, which are more document-like (or
record/field oriented, which fits into the document model pretty
easily).

Re: Occupant Combat, Redux

[Bruno Boettcher]

On Wed, Jul 02, 2003 at 02:36:47PM -0400, Jim Kingdon wrote:
> It makes attacking a transport (including, for example, a city in the
> standard game) a fairly powerful move, yes.
exactly

> With respect to game balance and such, I'm not sure whether this is a
> good thing or not, but it certainly is something which has affected my
> strategies and tactics quite a bit.
absolutely
making often not very realistic tactics....

> I suppose if we really want to address this perceived barrier, we
> could switch to XML.  Although it isn't clear whether xconq would fit
> XML as well as other uses of XML, which are more document-like (or
> record/field oriented, which fits into the document model pretty
> easily).
hmmm i allready had some thought about this, and had begun to amke some
reflexions about a suitable semantic, but a lot of xconq games are
tables, and those are a real pain to reflect in xml, so i abandoned that
idea...

-- 
ciao bboett
==============================================================
bboett@adlp.org
http://inforezo.u-strasbg.fr/~bboett
===============================================================

Re: Occupant Combat, Redux

[Eric McDonald]

On Wed, 2 Jul 2003, Jim Kingdon wrote:

> On the whole, the Language Syntax section of the xconq manual
> (refman.texi) tells you everything you need to know about lisp.
> Unfortunately, it throws in a bit of lisp terminology here and there
> and could use more examples (both things would be nice to fix, given
> the number of people who know lisp isn't so great these days).  But

Hi Jim,

  What sort of modifications to this section did you have in mind? I'll
make them, if you want. I took a quick read through it a little while ago,
and could only think of these potential modifications:

  * Mention the current range of Xconq numbers (that it uses shorts and is
thus capped at 32767). This can be inferred from looking at the paragraph
on dice specs, but could be made explicit.
  * Put emphasis on the sentence that tells what the default values are
assumed to be when they are not explicitly stated in the documentation.
  * Give an example of an atom versus a list containing an atom versus a
list containing multiple atoms.

  Are there others?

  Thanks,
    Eric

Re: Occupant Combat, Redux

[Jim Kingdon]

>   * Give an example of an atom versus a list containing an atom versus a
> list containing multiple atoms.

This would help.

Reword the sentence "GDL resembles Lisp, but instead of defining
functions" to say something like "GDL syntax allows you to include
numbers, strings, lists, and the like.  For those who know lisp, it
uses a similar syntax, but lisp knowledge is not needed to write or
read GDL".

Change "`Dotted pairs'' are not allowed." to "There is nothing similar
to the ``dotted pairs'' of lisp (and if you don't know what that is,
it means you don't need to worry about it).

Explain "Anything that is not a list is an @dfn{atom}." better.  It
isn't strictly true (it doesn't mean comments, for example).  Maybe
"an atom is a symbol, a number, or a string" is closer to the truth.

Reword the rest of this paragraph to nuke the word "forms" (it is
defined, but only later).

Put an example at the start.  Put more examples throughout (especially
in the "Lexical Elements" section and the discussion of
interpolation-list in the next section).

There might be others (a really good reviewer would be someone who was
trying to figure it out for the first time, without any knowledge of
lisp), but those are a good start.

Your other two suggestions also look OK to me.

Re: Occupant Combat, Redux

[Eric McDonald]

On Thu, 10 Jul 2003, Jim Kingdon wrote:

> There might be others (a really good reviewer would be someone who was
> trying to figure it out for the first time, without any knowledge of
> lisp), but those are a good start.

Yes, that was the main reason I was keeping this discussion public, _in the
hopes of getting feedback on the manual from non-Lisp people with very
little programming knowledge overall.

Anyway, I think I have incorporated the bulk of the changes that you and
Hans suggested. After I move and find a new ISP, I may run texi2html on the
results and post them in whatever Web space I am given. That way people
without viewers for .info files or who don't check things out of CVS
can comment on the revisions and suggest additional improvements.

  Thanks,
    Eric

Re: Occupant Combat, Redux

[Hans Ronne]

>  * Mention the current range of Xconq numbers (that it uses shorts and is
>thus capped at 32767). This can be inferred from looking at the paragraph
>on dice specs, but could be made explicit.

Perhaps also mention that 9999, 999 and 99 (rather than 32767) are used to
represent an "unlimited" large number in many games.

>  * Put emphasis on the sentence that tells what the default values are
>assumed to be when they are not explicitly stated in the documentation.

The value -1 has a special meaning in some tables. For example, if
fire-hit-chance is set to -1 for a unit that can fire, the standard
hit-chance table is used instead also for fire actions.

>  * Give an example of an atom versus a list containing an atom versus a
>list containing multiple atoms.
>
>  Are there others?
>
>  Thanks,
>    Eric