Lighthouse waypoint project
 

I was wandering about the net and came across a chinese website discussing fritzbot where the beta version of his map was mentionned. Well that got me to trying it out and I liked the textures so much I decided to waypoint the final version a couple of days ago. So far so good but I am having problems with a fake func_construct I have added. The allied engrs twist and point in the wrong direction. The added script creates are;
The z axis is about 60 above the floor of 384 (and I have tried lower z axis values) and the center is to one side of the door it controls (with a small clip protrusion from the wall when built). The bots dynamite ok but fail to construct (which I can do easily). Do I need to shift the center, or further reduce the mins and maxes?

BTW I will eventually want to add a visible brush and model (the existing door) for the blown case.

Re: Lighthouse waypoint project
 

top map!

have you played Bridges?
amazing map... one of the best ever imo!

Re: Lighthouse waypoint project
 

I tried bridges for a bobot 1.03 waypoint only (which was not optimal). Its size was sometimes a problem on my old machine. Anyway i believe the bindlestiff has bridges on his to do list.

btw just posted 0.01 (a couple of activate/deactivates were bugged, other tweaks)

Re: Lighthouse waypoint project
 

I've been working on bridges_final now for a week. Since it's a big map, it may take a few more weeks to complete. This is the one called "__bridges__" with two underscores before and after the map name, constructed by mrfin of the UK. :balloon:

Re: Lighthouse waypoint project
 

Kept at it and just posted 0.02 which has the first fix for the fake constructable. Rope tests still to do, but this fix should improve the play.

Re: Lighthouse waypoint project
 

Here's some initial thoughts and observations.

The Allies can get "stuck" on a little piece of wall that juts on the outside-side of the cracked wall.

The Allies do a terrible job of defense if an Axis sneaks in early. I waited by a door for someone to open it, then rushed in and systematically blew up all three objectives. Their entire team was focused on guarding the main and side doors.

There's a lot of jumping bots, which means there's a lot of bots jumping waaaay over the edge.

The ladder to the main door is a tough one for the axis. They want to jump away from danger if they get nicked, which means they fall to their deaths after being hit once.

I think more effort should be put into blowing up the cracked wall by the Axis. They need multiple entryways, or it's a slaughterhouse.

The Allies seem to have consistent problems with getting through doorways, especially the swinging one by the double gate and the doors in the house they spawn in. This may well be a function of the map itself.

--

Neat map - it's been a lot of fun playing so far!

Re: Lighthouse waypoint project
 

Thanks for the feedback.
in reply:
(1) I will be on the lookout for that allies sticking point.

(2) What you describe is a common problem in Fritzbot. When you sneak in you dynamite objectives that the bots do not view as active yet. In 2hide and Ammodepot I tried adding a defuse only action that was active from the start of the map, but the bots are still not brilliant at reacting to it if at all. What I can do is add roams to try to make the engineers periodically check on things but it robs bots from the defender's front line.

(3) Alas Bot behavior at high cliffs etc. when under fire is just plain kamikaze. They ignore waypoints when fighting and fake brushes feel, well FAKE. That is why I said on my web page that this map may never go past beta. It is also why I scripted in the ~sea level rear dynamite entry.

(4) I might be able to prevent them moving sideways or backwards with fake brushes but I can't prevent them falling on the main entrance ladder.

(5) In work not posted I tried moving the bots to focus on the cracked wall but both the bots climbing and defending end up jumping to their deaths. I may just have to allow the wall to be blown and make the allies defend from the inside.

(6) The doors may need a little tweaking and I could route a few bots to the second door of the spawn, but the door by the gate will still be a bottleneck (I could also route maybe a CVop through the ~sea level rear door but the length of the path disadvantages the allies).

Unfortunately I stopped working on this map because my changes started making things worse. I spent a day adding a model for the construction materials at the rear door and tweaking the clips and then misc stuff including the cracked wall. But then in testing some of those actions would become contrary to the logic in the Aiscript and I could not figure out why. So I will have to diff every thing, revert and the add them back in a bit at a time. Time consuming so its been pushed down on my list Sorry.

Re: Lighthouse waypoint project
 

OK took a look again and corrected some stuff, likely fixing those mentioned bugs (to be verified). The problem now seems to be priority related (like glider). The constructs being prioritized leads the axis engr bots to spend too little time on blowing the main gate and objectives and too much time on the cliff ropes. However I am experimenting on a way to change priorities using 2 construct actions targeted at the same entity E.G.;

So far it looks like it works albeit with a predictable delay. Still have to test some things to be sure. It did not seem to work properly when I used action 0 (the minor construct) for the action test and/or not delay the disable of the major construct #104 but I should retest jic.

I guess I will have to rethink the strategy now as Axis does not win because when the wall gets blown before the gate they don't attack the gate with the force needed to blow it. I may have to make the strategy more sequential.

- - -
2) Defuse only actions are useless alas, they are ignored in all testing I have done. The only way I guess would be to script a fake entity to be triggered on the dyno-planted to activate a normal plant-defuse action for 30 seconds. The bot engineers on the team of the human who planted would likely zombie for that time but at least the dyno could be defused (assuming the bots did not wait for a second plant).

4) Fake brushes at the main ladder do reduce the falling damage because you only fall down one of the two ladder distances.

6) I still do need to reduce the scrum at doors that slows down passage. Routes and better strategy should help.

I will post a WIP when axis start winning at least 20% or better

Re: Lighthouse waypoint project
 

Mrfin released a Lighthouse 2 - link - 13MB

I haven't playtested this one yet. The first version was excellent imo.
I just thought you may be interested in having a look :)

Re: Lighthouse waypoint project
 

Yes the first version was excellent but I felt the second one looked more like committee work (stuffed) when I ran through it some months ago. And since bots aren't at their best on cliff like edges nor on the lighthouse I saw no point porting my original Lighthouse waypoints (which will never be release quality anyway). And since IMO you could play the map with 3 on 3 humans, bots are kind of optional.