What is this Swarming wherof you speak?

G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Not having played for a bit, I notice people saying "Swarming doesn't work
well under new host with new Small Weapons and LWs are now favored" or
something like than.

I have run a few tests and don't see much difference compared with really
old (1-2 year ago, not 2 month ago) host versions.

By "Swarming" do you mean large numbers of small/cheap ships?
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

It's not the swarming itself (BTW: swarming means the use of a lot identical
cheap tiny little ships)
Under new host the times between rendezvous between the swarm and targets is
longer than earlier which is less efective as swarms will diffuse after a
while.

GFM GToeroe




"Alfred the Great" <someone@microsoft.com> schrieb im Newsbeitrag
news:S5hqe.2200$Z44.767@newssvr13.news.prodigy.com...
> Not having played for a bit, I notice people saying "Swarming doesn't work
> well under new host with new Small Weapons and LWs are now favored" or
> something like than.
>
> I have run a few tests and don't see much difference compared with really
> old (1-2 year ago, not 2 month ago) host versions.
>
> By "Swarming" do you mean large numbers of small/cheap ships?
>
>
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

We are talking about in the VCR or in the game?

I ran a couple of tests (old ones) with Diplomat in the VCR

Cubes vs Inamoratae
Cubes vs Cavalry Gunships (that has got to be a "tiny ship"

Both run with fleets costing equal amounts (so there are lots more of the
smaller cheaper ships).

The smaller ships (as before) clobber the fewer larger ships. Details on
request (exotic tech and such - exotics disproportionately help small ships.
Adding 100 to the shields does more for the little ships for example)

Or do you mean groups of small ships, all on auto-intercept getting into VCR
battles on different movement ticks?



"GFM GToeroe" <gfm@gtoeroe.de> wrote in message
news:d8c9dp$dpb$02$1@news.t-online.com...
> It's not the swarming itself (BTW: swarming means the use of a lot
identical
> cheap tiny little ships)
> Under new host the times between rendezvous between the swarm and targets
is
> longer than earlier which is less efective as swarms will diffuse after a
> while.
>
> GFM GToeroe
>
>
>
>
> "Alfred the Great" <someone@microsoft.com> schrieb im Newsbeitrag
> news:S5hqe.2200$Z44.767@newssvr13.news.prodigy.com...
> > Not having played for a bit, I notice people saying "Swarming doesn't
work
> > well under new host with new Small Weapons and LWs are now favored" or
> > something like than.
> >
> > I have run a few tests and don't see much difference compared with
really
> > old (1-2 year ago, not 2 month ago) host versions.
> >
> > By "Swarming" do you mean large numbers of small/cheap ships?
> >
> >
>
>
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

hm, could you tell something about the equippment of the ships and the
numbers.


"Beezle" <someone@microsoft.com> schrieb im Newsbeitrag
news:zOmqe.26743$J12.2022@newssvr14.news.prodigy.com...
> We are talking about in the VCR or in the game?
>
> I ran a couple of tests (old ones) with Diplomat in the VCR
>
> Cubes vs Inamoratae
> Cubes vs Cavalry Gunships (that has got to be a "tiny ship"
>
> Both run with fleets costing equal amounts (so there are lots more of the
> smaller cheaper ships).
>
> The smaller ships (as before) clobber the fewer larger ships. Details on
> request (exotic tech and such - exotics disproportionately help small
ships.
> Adding 100 to the shields does more for the little ships for example)
>
> Or do you mean groups of small ships, all on auto-intercept getting into
VCR
> battles on different movement ticks?
>
>
>
> "GFM GToeroe" <gfm@gtoeroe.de> wrote in message
> news:d8c9dp$dpb$02$1@news.t-online.com...
> > It's not the swarming itself (BTW: swarming means the use of a lot
> identical
> > cheap tiny little ships)
> > Under new host the times between rendezvous between the swarm and
targets
> is
> > longer than earlier which is less efective as swarms will diffuse after
a
> > while.
> >
> > GFM GToeroe
> >
> >
> >
> >
> > "Alfred the Great" <someone@microsoft.com> schrieb im Newsbeitrag
> > news:S5hqe.2200$Z44.767@newssvr13.news.prodigy.com...
> > > Not having played for a bit, I notice people saying "Swarming doesn't
> work
> > > well under new host with new Small Weapons and LWs are now favored" or
> > > something like than.
> > >
> > > I have run a few tests and don't see much difference compared with
> really
> > > old (1-2 year ago, not 2 month ago) host versions.
> > >
> > > By "Swarming" do you mean large numbers of small/cheap ships?
> > >
> > >
> >
> >
>
>
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

You could try out ie. swarms with PTTs as main weapon, they are still
cheap (and you more importantly do not need to take the DS, but ships
with poor attack ratings), and they will still work. You could use Drac
ships.
And then I did only run a few tests Dracs vs Innamoratas.

KlingonKommand wrote:
> Beezle <someone@microsoft.com> writes
>
> >I ran a couple of tests (old ones) with Diplomat in the VCR
> >
> >Cubes vs Inamoratae
> >Cubes vs Cavalry Gunships (that has got to be a "tiny ship"
> ...
> >The smaller ships (as before) clobber the fewer larger ships. Details on
> >request
>
> Yes details please! A lot of us mere primates are wondering how to do
> that. Are you using Host 193? Our swarm simulations are showing poor
> results for the swarms. The power of small weapon, which they depend on,
> seems to have massively reduced in the last couple of hosts.
> --
> Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Beezle <someone@microsoft.com> writes

>I ran a couple of tests (old ones) with Diplomat in the VCR
>
>Cubes vs Inamoratae
>Cubes vs Cavalry Gunships (that has got to be a "tiny ship"
....
>The smaller ships (as before) clobber the fewer larger ships. Details on
>request

Yes details please! A lot of us mere primates are wondering how to do
that. Are you using Host 193? Our swarm simulations are showing poor
results for the swarms. The power of small weapon, which they depend on,
seems to have massively reduced in the last couple of hosts.
--
Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Let me work on this a little more. All the tests I have done so far have
been with Scavengers as one of the races involved and they have a SW bonus
(double shots IIRC).

I'll try Fed vs Fed this weekend.
"KlingonKommand" <Paul@nurk.fnord> wrote in message
news:Tza16OGmBrqCFwzd@furfur.demon.co.uk...
> Beezle <someone@microsoft.com> writes
>
> >I ran a couple of tests (old ones) with Diplomat in the VCR
> >
> >Cubes vs Inamoratae
> >Cubes vs Cavalry Gunships (that has got to be a "tiny ship"
> ...
> >The smaller ships (as before) clobber the fewer larger ships. Details on
> >request
>
> Yes details please! A lot of us mere primates are wondering how to do
> that. Are you using Host 193? Our swarm simulations are showing poor
> results for the swarms. The power of small weapon, which they depend on,
> seems to have massively reduced in the last couple of hosts.
> --
> Paul Honigmann
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Well, Lokis beat Golems. Lokis low tech (Tech 1 SW, 210 shield, Tech 9
engines) vs Golems with PPCs.

At equal MC per fleet (6 Golems vs 170 Loki) and "some" Exotics (mainly
first level SW, LW, Shield and attack) the Loki's win.

And I gather the Loki's get no benefit from SW exotics with only Lasers,
right?

I gather another controversy here is "are SWs actually different now/is
there finally a reason to go to higher tech SWs". That seems by inspection
to be dependent on who is fighting who.

I'll look at that next.

BTW my prejudice would be that swarming should _not_ be very effective. An
infinite number of small arms shots don't do anything at all to the Yamato,
as it were.

My prejudice would be lots of damage to weapons and shields and PD but not
hull damage. So if indeed small ship swarming had been toned down I am all
for that.


"KlingonKommand" <Paul@nurk.fnord> wrote in message
news:Tza16OGmBrqCFwzd@furfur.demon.co.uk...
> Beezle <someone@microsoft.com> writes
>
> >I ran a couple of tests (old ones) with Diplomat in the VCR
> >
> >Cubes vs Inamoratae
> >Cubes vs Cavalry Gunships (that has got to be a "tiny ship"
> ...
> >The smaller ships (as before) clobber the fewer larger ships. Details on
> >request
>
> Yes details please! A lot of us mere primates are wondering how to do
> that. Are you using Host 193? Our swarm simulations are showing poor
> results for the swarms. The power of small weapon, which they depend on,
> seems to have massively reduced in the last couple of hosts.
> --
> Paul Honigmann
 

Sparrow

Distinguished
Mar 16, 2002
239
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Swarms should have an advantage in combat, as they have disadvantages
everywhere else. Some races simply have no biggies and swarming is
their only chance!

Just my 0.02$
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Yes. But first let us fix the current situation. And then add new things.

I guess if only would let the ships try to fulfill their given orders in a
more observable manner then we would reach a stage where the most would say
"Ah, now it is good. Ah, with this I can live!"

Example: Right now one would expect that ships tend to intercept more
agressively if set to "attack most dangerous". Right now they do it, but as
they try to do it one a more or less circumpolar course they do it very
ineffictively. You can test it if you send a number of Cavalries against
several Cubes but only one with LWs. Then you see that they always move
where they expect the LW-Cube to go to. Unfortunately they do it not on a
straight line. And here the bad begins. Too much time it takes to meet the
target a second time. Usually they get scattered and killed one after the
other.
If only the attack pattern for the case where a attacker ship has set
another enemie ship a top target would be changed a lot would be achieved.
And it could be made in a way simple enough for a JSC(me)::

Sub AcceleratedStraightLinePursuit(ByRef ChaserX, ChaserY, ChaserV as
Double, ByVal PreyX, PreyY, ChaserAcceleration, TimeStep, ChaserMaxSpeed,
ChaserStandOff As Double)
Dim Distance, dx, dy As Double
dx = PreyX - ChaserX: dy = PreyY - ChaserY: Distance = Sqr(dx * dx + dy
* dy)
If Distance > ChaserStandOff Then
ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
ChaserMaxSpeed, ChaserAcceleration, 0) * TimeStep + ChaserV
If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
Else
ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
ChaserMaxSpeed, -ChaserAcceleration, 0) * TimeStep + ChaserV
If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
End If
End If
If Distance>1 Then
ChaserX = ChaserX + dx * ChaserV * TimeStep / Distance
ChaserY = ChaserY + dy * ChaserV * TimeStep / Distance
End If
End Sub

This Sub could be used if an object at (ChaserX, ChaserY) is declared as
chaser for an object at (PreyX, PreyY) for the next movement tick. The
chaser then tries to move at current speed
ChaserV accelerated with ChaserAcceleration towards the prey on a straight
line between them. The acceleration is allowed to bring the speed up to +/-
MaxSpeed. If the chaser reaches the standoff range then the acceleration
will slow down (or even reverse) the movement until he leaves the standoff
range.

If could be used for the cases "attack dangerous", "closest first" and
"attack soft" as then always the code searches for suited objects and fills
them in the top ten list of attack tagets.
It could also be used for "flee": Take the most dangerous and calculate the
points with the most distance in vcr zone (easy to make) and set this point
as prey coordinates with no (or low) standoff. It could also be used to
temporally peel off the prey if the LWs or the missiles have to be
recharged.

The fact that there is a HG on board or the ships has experience could be
used to distribute times where it is allowed to use the above routine. One
also could use the system damage in addtion.

Ships which are idle (or forced to be idle) of orders could use the current
"circumpolar" logic.

ChaserAcceleration and ChaserMaxSpeed could be used due to engine damage,
(total) mass, (total) evasive bonus, max hull speed and combat quickness
and due to several special (wanted) race modifications.

GFM GToeroe

"Sparrow" <e.kueper@gmx.net> schrieb im Newsbeitrag
news:1118666162.321167.73210@o13g2000cwo.googlegroups.com...
> Swarms should have an advantage in combat, as they have disadvantages
> everywhere else. Some races simply have no biggies and swarming is
> their only chance!
>
> Just my 0.02$
>
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

So that is why the ships eventually go into a circle? Because they are
constantly heading for the point where the target was (a "lag pursuit" in
Air Combat Terms, rather than an intercept)?

Hmm, to do an intercept you need to know where the target is going, and if
you are both trying to intercept each other, or trying to intercept someone
who is trying to intercept a third ship or run from you it gets more
complicated.

The simplest would be having the ship not move each tick to where the target
ship currently is, but to move to the point where it will intercept in least
time if the target maintains current course ("lead pursuit" in air combat
terms) (and for fleeing ships, move away from the current projected course
of some pursuer, perhaps the closest).

But the map is small so they will still tend to run around the edges. How
much difference would plotting lead pursuit instead of lag pursuit actually
make?

"GFM GToeroe" <gfm@gtoeroe.de> wrote in message
news:d8k7ce$31u$03$1@news.t-online.com...

> Example: Right now one would expect that ships tend to intercept more
> agressively if set to "attack most dangerous". Right now they do it, but
as
> they try to do it one a more or less circumpolar course they do it very
> ineffictively. You can test it if you send a number of Cavalries against
> several Cubes but only one with LWs. Then you see that they always move
> where they expect the LW-Cube to go to. Unfortunately they do it not on a
> straight line. And here the bad begins. Too much time it takes to meet the
> target a second time. Usually they get scattered and killed one after the
> other.
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Define what function (or method) you mean with IIf in the algorithm, it
is enough to say what you want to get as a return value (not how it is
computed), afterwards I will comment on this (I will take in mind that
it was written by a JSC).

GFM GToeroe wrote:
> Yes. But first let us fix the current situation. And then add new things.
>
> I guess if only would let the ships try to fulfill their given orders in a
> more observable manner then we would reach a stage where the most would say
> "Ah, now it is good. Ah, with this I can live!"
>
> Example: Right now one would expect that ships tend to intercept more
> agressively if set to "attack most dangerous". Right now they do it, but as
> they try to do it one a more or less circumpolar course they do it very
> ineffictively. You can test it if you send a number of Cavalries against
> several Cubes but only one with LWs. Then you see that they always move
> where they expect the LW-Cube to go to. Unfortunately they do it not on a
> straight line. And here the bad begins. Too much time it takes to meet the
> target a second time. Usually they get scattered and killed one after the
> other.
> If only the attack pattern for the case where a attacker ship has set
> another enemie ship a top target would be changed a lot would be achieved.
> And it could be made in a way simple enough for a JSC(me)::
>
> Sub AcceleratedStraightLinePursuit(ByRef ChaserX, ChaserY, ChaserV as
> Double, ByVal PreyX, PreyY, ChaserAcceleration, TimeStep, ChaserMaxSpeed,
> ChaserStandOff As Double)
> Dim Distance, dx, dy As Double
> dx = PreyX - ChaserX: dy = PreyY - ChaserY: Distance = Sqr(dx * dx + dy
> * dy)
> If Distance > ChaserStandOff Then
> ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
> ChaserMaxSpeed, ChaserAcceleration, 0) * TimeStep + ChaserV
> If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> Else
> ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
> ChaserMaxSpeed, -ChaserAcceleration, 0) * TimeStep + ChaserV
> If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> End If
> End If
> If Distance>1 Then
> ChaserX = ChaserX + dx * ChaserV * TimeStep / Distance
> ChaserY = ChaserY + dy * ChaserV * TimeStep / Distance
> End If
> End Sub
>
> This Sub could be used if an object at (ChaserX, ChaserY) is declared as
> chaser for an object at (PreyX, PreyY) for the next movement tick. The
> chaser then tries to move at current speed
> ChaserV accelerated with ChaserAcceleration towards the prey on a straight
> line between them. The acceleration is allowed to bring the speed up to +/-
> MaxSpeed. If the chaser reaches the standoff range then the acceleration
> will slow down (or even reverse) the movement until he leaves the standoff
> range.
>
> If could be used for the cases "attack dangerous", "closest first" and
> "attack soft" as then always the code searches for suited objects and fills
> them in the top ten list of attack tagets.
> It could also be used for "flee": Take the most dangerous and calculate the
> points with the most distance in vcr zone (easy to make) and set this point
> as prey coordinates with no (or low) standoff. It could also be used to
> temporally peel off the prey if the LWs or the missiles have to be
> recharged.
>
> The fact that there is a HG on board or the ships has experience could be
> used to distribute times where it is allowed to use the above routine. One
> also could use the system damage in addtion.
>
> Ships which are idle (or forced to be idle) of orders could use the current
> "circumpolar" logic.
>
> ChaserAcceleration and ChaserMaxSpeed could be used due to engine damage,
> (total) mass, (total) evasive bonus, max hull speed and combat quickness
> and due to several special (wanted) race modifications.
>
> GFM GToeroe
>
> "Sparrow" <e.kueper@gmx.net> schrieb im Newsbeitrag
> news:1118666162.321167.73210@o13g2000cwo.googlegroups.com...
> > Swarms should have an advantage in combat, as they have disadvantages
> > everywhere else. Some races simply have no biggies and swarming is
> > their only chance!
> >
> > Just my 0.02$
> >
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

The function IIF(A,B,C) returns B if A is True and C if A is False.

Sometimes I use it to make code shorter and more readable but don't know if
it is faster than the normal "IF THEN ELSE". And no it is a difference as
now others can now use the algorithm for own purposes.

And maybe I rise to a JFC...

GFM GToeroe





"Nameless" <unknown_ai@web.de> schrieb im Newsbeitrag
news:1118760684.132327.208280@g44g2000cwa.googlegroups.com...
> Define what function (or method) you mean with IIf in the algorithm, it
> is enough to say what you want to get as a return value (not how it is
> computed), afterwards I will comment on this (I will take in mind that
> it was written by a JSC).
>
> GFM GToeroe wrote:
> > Yes. But first let us fix the current situation. And then add new
things.
> >
> > I guess if only would let the ships try to fulfill their given orders in
a
> > more observable manner then we would reach a stage where the most would
say
> > "Ah, now it is good. Ah, with this I can live!"
> >
> > Example: Right now one would expect that ships tend to intercept more
> > agressively if set to "attack most dangerous". Right now they do it, but
as
> > they try to do it one a more or less circumpolar course they do it very
> > ineffictively. You can test it if you send a number of Cavalries against
> > several Cubes but only one with LWs. Then you see that they always move
> > where they expect the LW-Cube to go to. Unfortunately they do it not on
a
> > straight line. And here the bad begins. Too much time it takes to meet
the
> > target a second time. Usually they get scattered and killed one after
the
> > other.
> > If only the attack pattern for the case where a attacker ship has set
> > another enemie ship a top target would be changed a lot would be
achieved.
> > And it could be made in a way simple enough for a JSC(me)::
> >
> > Sub AcceleratedStraightLinePursuit(ByRef ChaserX, ChaserY, ChaserV as
> > Double, ByVal PreyX, PreyY, ChaserAcceleration, TimeStep,
ChaserMaxSpeed,
> > ChaserStandOff As Double)
> > Dim Distance, dx, dy As Double
> > dx = PreyX - ChaserX: dy = PreyY - ChaserY: Distance = Sqr(dx * dx +
dy
> > * dy)
> > If Distance > ChaserStandOff Then
> > ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
> > ChaserMaxSpeed, ChaserAcceleration, 0) * TimeStep + ChaserV
> > If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> > If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> > Else
> > ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
> > ChaserMaxSpeed, -ChaserAcceleration, 0) * TimeStep + ChaserV
> > If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> > If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> > End If
> > End If
> > If Distance>1 Then
> > ChaserX = ChaserX + dx * ChaserV * TimeStep / Distance
> > ChaserY = ChaserY + dy * ChaserV * TimeStep / Distance
> > End If
> > End Sub
> >
> > This Sub could be used if an object at (ChaserX, ChaserY) is declared as
> > chaser for an object at (PreyX, PreyY) for the next movement tick. The
> > chaser then tries to move at current speed
> > ChaserV accelerated with ChaserAcceleration towards the prey on a
straight
> > line between them. The acceleration is allowed to bring the speed up to
+/-
> > MaxSpeed. If the chaser reaches the standoff range then the acceleration
> > will slow down (or even reverse) the movement until he leaves the
standoff
> > range.
> >
> > If could be used for the cases "attack dangerous", "closest first" and
> > "attack soft" as then always the code searches for suited objects and
fills
> > them in the top ten list of attack tagets.
> > It could also be used for "flee": Take the most dangerous and calculate
the
> > points with the most distance in vcr zone (easy to make) and set this
point
> > as prey coordinates with no (or low) standoff. It could also be used to
> > temporally peel off the prey if the LWs or the missiles have to be
> > recharged.
> >
> > The fact that there is a HG on board or the ships has experience could
be
> > used to distribute times where it is allowed to use the above routine.
One
> > also could use the system damage in addtion.
> >
> > Ships which are idle (or forced to be idle) of orders could use the
current
> > "circumpolar" logic.
> >
> > ChaserAcceleration and ChaserMaxSpeed could be used due to engine
damage,
> > (total) mass, (total) evasive bonus, max hull speed and combat
quickness
> > and due to several special (wanted) race modifications.
> >
> > GFM GToeroe
> >
> > "Sparrow" <e.kueper@gmx.net> schrieb im Newsbeitrag
> > news:1118666162.321167.73210@o13g2000cwo.googlegroups.com...
> > > Swarms should have an advantage in combat, as they have disadvantages
> > > everywhere else. Some races simply have no biggies and swarming is
> > > their only chance!
> > >
> > > Just my 0.02$
> > >
>
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

GFM GToeroe wrote:
>Yes. But first let us fix the current situation. And then add new things.

>I guess if only would let the ships try to fulfill their given orders in a
>more observable manner then we would reach a stage where the most would say
>"Ah, now it is good. Ah, with this I can live!"

To make the manner more observable the vcr client would need to be
changed.
Ie. a few statistics which are safed anyway, would need to be there in
written form, ie.
accessible via the combat report. This includes targetting, and when
the target changes, the distance when a weapon is fired (including
which of them did hit and miss)...

>Example: Right now one would expect that ships tend to intercept more
>agressively if set to "attack most dangerous".

No that is actualy not the expectation, the expactation is that it does
then target
the more dangerous enemy objects first (and here lies a problem which
is that everybody
and every race (and every fleet) has other objects which he/she
considers more dangerous).
And considering the way it is written in the help files the amount of
weapons measure that. Correct apoarch towards this attack option would
be to let each player for each fleet decide what are the most dangerous
enemy objects, if that is via the fleet command or the Ministers does
not matter (or renaming this attack option).

> Right now they do it, but as
>they try to do it one a more or less circumpolar course they do it very
>ineffictively. You can test it if you send a number of Cavalries against
>several Cubes but only one with LWs. Then you see that they always move
>where they expect the LW-Cube to go to. Unfortunately they do it not on a
>straight line. And here the bad begins. Too much time it takes to meet the
>target a second time.
> Usually they get scattered and killed one after the
>other.

The other possibility is that the combat ends prematurely.

And considering one test I had with current host, a ship leaves the
swarm
when it got hit (usually a LW) and then flies in fast circles around
the center of the vcr (about the same radius
as the distance of the swarm was to the center).

>If only the attack pattern for the case where a attacker ship has set
>another enemie ship a top target would be changed a lot would be achieved.
>And it could be made in a way simple enough for a JSC(me)::

>Sub AcceleratedStraightLinePursuit(ByRef ChaserX, ChaserY, ChaserV as
>Double, ByVal PreyX, PreyY, ChaserAcceleration, TimeStep, ChaserMaxSpeed,
>ChaserStandOff As Double)
> Dim Distance, dx, dy As Double
> dx = PreyX - ChaserX: dy = PreyY - ChaserY: Distance = Sqr(dx * dx + dy
>* dy)
> If Distance > ChaserStandOff Then
> ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
>ChaserMaxSpeed, ChaserAcceleration, 0) * TimeStep + ChaserV

Here is a problem, consider ie. the ship (A) targeted by the ship (B)
using this routine is
behind the ship (B) and so far ship B has flown away from ship A, could
happen ie.
through target change. In any case this routine does not take the
current speed vector into
account only its norm.

> If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> Else
> ChaserV = IIf(ChaserV > -ChaserMaxSpeed And ChaserV <
>ChaserMaxSpeed, -ChaserAcceleration, 0) * TimeStep + ChaserV

Same problem as above.

> If ChaserV > ChaserMaxSpeed Then ChaserV = ChaserMaxSpeed
> If ChaserV < -ChaserMaxSpeed Then ChaserV = -ChaserMaxSpeed
> End If
> End If
> If Distance>1 Then
> ChaserX = ChaserX + dx * ChaserV * TimeStep / Distance
> ChaserY = ChaserY + dy * ChaserV * TimeStep / Distance
> End If
>End Sub

>This Sub could be used if an object at (ChaserX, ChaserY) is declared as
>chaser for an object at (PreyX, PreyY) for the next movement tick. The
>chaser then tries to move at current speed
>ChaserV accelerated with ChaserAcceleration towards the prey on a straight
>line between them. The acceleration is allowed to bring the speed up to +/-
>MaxSpeed. If the chaser reaches the standoff range then the acceleration
>will slow down (or even reverse) the movement until he leaves the standoff
>range.

>If could be used for the cases "attack dangerous", "closest first" and
>"attack soft" as then always the code searches for suited objects and fills
>them in the top ten list of attack tagets.
>It could also be used for "flee": Take the most dangerous and calculate the
>points with the most distance in vcr zone (easy to make) and set this point
>as prey coordinates with no (or low) standoff. It could also be used to
>temporally peel off the prey if the LWs or the missiles have to be
>recharged.

>The fact that there is a HG on board or the ships has experience could be
>used to distribute times where it is allowed to use the above routine. One
>also could use the system damage in addtion.

What about ship skill?

>...
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

GFM GToeroe wrote:
> "Nameless" <unknown_ai@web.de> schrieb im Newsbeitrag
> news:1118932499.790479.49390@g14g2000cwa.googlegroups.com...
>
> > To make the manner more observable the vcr client would need to be
> > changed.
> > Ie. a few statistics which are safed anyway, would need to be there in
> > written form, ie.
> > accessible via the combat report. This includes targetting, and when
> > the target changes, the distance when a weapon is fired (including
> > which of them did hit and miss)...
>
> Yes, this is a good idea.
> But: "...The big trick is to have movement like you suggest and keep the VCR
> files small, if just a small amount more of movement detail is added the
> size of the VCR files balloon..." (Tim)

What has that to do with the above. There would not be stored more data
in the vcr files (as there is now), the info is their in form of the
displayed vcr combat simulation,
it just needs to be extracted (by the vcr routine - client side) and
then displayed in the combat report. IMO only the size of the rst send
via email from the host to the client is important, how big an
extracted (and maybe even temporary) vcr file is does not matter.
It is stored in the vcr files, when ships do change targets, also when
they leave combat (eventhough this part seems to be not working
correctly), also the x,y position of all objects is given in the vcr at
any given tick (and therefore the distance is implicitly given between
two objects on any given tick), it is also possible to hide missed
shoots (in vcr), so it is safed which weapons do hit and miss, it is
displayed in the client at which point weapons are fired, so also no
problem to display it in the combat report.
....
So I see here absolutly no problem.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

"Nameless" <unknown_ai@web.de> schrieb im Newsbeitrag
news:1118932499.790479.49390@g14g2000cwa.googlegroups.com...

> To make the manner more observable the vcr client would need to be
> changed.
> Ie. a few statistics which are safed anyway, would need to be there in
> written form, ie.
> accessible via the combat report. This includes targetting, and when
> the target changes, the distance when a weapon is fired (including
> which of them did hit and miss)...

Yes, this is a good idea.
But: "...The big trick is to have movement like you suggest and keep the VCR
files small, if just a small amount more of movement detail is added the
size of the VCR files balloon..." (Tim)

> Here is a problem, consider ie. the ship (A) targeted by the ship (B)
> using this routine is
> behind the ship (B) and so far ship B has flown away from ship A, could
> happen ie.
> through target change. In any case this routine does not take the
> current speed vector into
> account only its norm.

The idea is to explicitely choose a parameter representation for a path and
use the arc differential to link a speed to a suited change of the
parameter.
It ensures a travelling along the path at the given scalar speed. Look at
vgap4-developer for more details. This is an attempt to have a better
control over the global motion.

GFM GToeroe
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

>
> Yes, this is a good idea.
> But: "...The big trick is to have movement like you suggest and keep the
> VCR
> files small, if just a small amount more of movement detail is added the
> size of the VCR files balloon..." (Tim)
>

Why is it so difficult to keep them small?
Just save the starting conditions and all host has to do is to actually roll
dices x times.

Take the starting conditions and add 1 seed for the random generator and
every vcr should replay exactly like host rolled it (?)

Lordfire
 

nameless

Distinguished
Jun 18, 2002
213
0
18,680
Archived from groups: alt.games.vgaplanets4 (More info?)

Lordfire wrote:
> >
> > Yes, this is a good idea.
> > But: "...The big trick is to have movement like you suggest and keep the
> > VCR
> > files small, if just a small amount more of movement detail is added the
> > size of the VCR files balloon..." (Tim)
> >
>
> Why is it so difficult to keep them small?
> Just save the starting conditions and all host has to do is to actually roll
> dices x times.
>
> Take the starting conditions and add 1 seed for the random generator and
> every vcr should replay exactly like host rolled it (?)

Won't work, since random number generators are system dependent.
So if we would do that, I could watch a totally different vcr (from the
same data!) on a different computer.
Saving all random numbers rolled in the vcr would help, but then again
I am not entirely sure that we would end up with less safed data per
vcr.
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

Yours and Nameless's answer is almost the same what I've answered to Tim.
But an reply by him is still missing.

GFM GToeroe

"Lordfire" <a@b.com> schrieb im Newsbeitrag
news:3he7rgFgl985U1@uni-berlin.de...
> >
> > Yes, this is a good idea.
> > But: "...The big trick is to have movement like you suggest and keep the
> > VCR
> > files small, if just a small amount more of movement detail is added the
> > size of the VCR files balloon..." (Tim)
> >
>
> Why is it so difficult to keep them small?
> Just save the starting conditions and all host has to do is to actually
roll
> dices x times.
>
> Take the starting conditions and add 1 seed for the random generator and
> every vcr should replay exactly like host rolled it (?)
>
> Lordfire
>
>
 
G

Guest

Guest
Archived from groups: alt.games.vgaplanets4 (More info?)

In that case:
Is is so hard write a random number generator by yourself?
I have no idea .. just curious..

Lordfire


"Nameless" <unknown_ai@web.de> wrote in message
news:1118996065.962244.76010@g49g2000cwa.googlegroups.com...
>
>
> Lordfire wrote:
>> >
>> > Yes, this is a good idea.
>> > But: "...The big trick is to have movement like you suggest and keep
>> > the
>> > VCR
>> > files small, if just a small amount more of movement detail is added
>> > the
>> > size of the VCR files balloon..." (Tim)
>> >
>>
>> Why is it so difficult to keep them small?
>> Just save the starting conditions and all host has to do is to actually
>> roll
>> dices x times.
>>
>> Take the starting conditions and add 1 seed for the random generator and
>> every vcr should replay exactly like host rolled it (?)
>
> Won't work, since random number generators are system dependent.
> So if we would do that, I could watch a totally different vcr (from the
> same data!) on a different computer.
> Saving all random numbers rolled in the vcr would help, but then again
> I am not entirely sure that we would end up with less safed data per
> vcr.
>