![[de]](/mwf/flags/de.png "Germany")
Two ideas to making the grappling hook less cumbersome:
-just speeding up the whole process, the hook shoot, the pull, and the climb
-if you touch material and climb or hangle, and let go again you fall down immediately (while still on your rope), so if you unintentionally touch a small material chunk, you might end up falling down again.
-A tutorial for using /practicing with the windbag would be nice.
-If you are in viewer mode, you cannot zoom with the mouse, just with F5/F6.
-just speeding up the whole process, the hook shoot, the pull, and the climb
-if you touch material and climb or hangle, and let go again you fall down immediately (while still on your rope), so if you unintentionally touch a small material chunk, you might end up falling down again.
-A tutorial for using /practicing with the windbag would be nice.
-If you are in viewer mode, you cannot zoom with the mouse, just with F5/F6.
> -A tutorial for using /practicing with the windbag would be nice.
The windbag serves as an easter egg in tutorial 7 without explanation. I could add an NPC which explains it. Would that be enough?
>Would that be enough?
Honestly no, the windbag is quite tricky to learn. Actually when I started playing OpenClonk I didn't know what to do with it until I saw other players use it online.
And after that it really took some time to master all the jumps, so a simple level where you have to jump around, while each Jump increases in difficulty slightly would be very helpful to newbies, and in fact to me too :D
Actually yesterday when I played, I still died because of timing those jumps wrong, so I felt like I really could use such a level :((
And don't get me started on deflecting objects like rockets to attack, like K-Pone uses them, though that sorta stuff could probably go into the Tower.
![[at]](/mwf/flags/at.png "Austria")
>And don't get me started on deflecting objects like rockets to attack
How. is. that. possible.
You can use the windbag to blow the rockets. That can be used to either boost a rocket's speed or even manipulate it's flying path. You see this in Hot Ice a lot. (At least when playing with Kanibal, DasWipf, Me, or anyone else who uses Boompacks quite often)
This is actually what I like about OpenClonk, instead of magic combo's in CR, you can now combine some of the tools to make them more powerful. But this really takes some skill and especially practise. Same as in CR where for the average player an arrow pack is something you use for the bow but for the pro a chance to cross a wide abyss.
To be honest, a scenario like that sounds very boring. Even for a tutorial. Even Boomshire isn't that exciting and at least it's a two player scenario.But it's probably doable in the editor in 10 minutes or less. A few spots of material, spaced farer and farer apart, a spawn point and a windbag. Off you go. Train as much as you want.
Don't get me wrong, I am not against any useful tutorials to master the game. But we shouldn't have endless scenarios, explaining and training every object and every application in every situation. No one would want to play that, no one would want to create and maintain these. The windbag isn't that important an object. Yes, Hot Ice is played a lot. Yes, using the windbag is essential for that scenario. But apart from that, there are countless scenarios in which you don't ever need a windbag.
Also, no scenario will ever teach you to be as good as K-Pone or similar. He plays OC for years(?), oftentimes Hot Ice variants, probably at least once a week. No scenario would get you to that level of skill. Only practice will.
It's probably fine to have one spot in a tutorial, where the key application of the windbag (shooting at the ground when jumping to jump higher) is explained and applied once.
I figured that you can jump with it, but not how. Pointing at the floor did not work as intended, or had too little effect (this was a while ago in November). Some basic instructions how to use it would be great.
You shoot it at the floor immediately after you start the jump. It takes some practice to master.
![[jp]](/mwf/flags/jp.png "Japan")
As of now it is hard too to use... when even a veteran player like ala can't get the timing right.
The thing in hot ice referred to here is easy to time:
* Have windbag selected.
* Carry a boompack.
* Aim mouse cursor at target.
* Left mouse click to launch boompack.
* [W] to jump of boompack.
* Left mouse click to use windbag.
Just make sure to perform the last three steps as fast as possible and there is no need to adjust the aiming in between. It is really just pressing [LMB][W][LMB] as fast as possible.
* Have windbag selected.
* Carry a boompack.
* Aim mouse cursor at target.
* Left mouse click to launch boompack.
* [W] to jump of boompack.
* Left mouse click to use windbag.
Just make sure to perform the last three steps as fast as possible and there is no need to adjust the aiming in between. It is really just pressing [LMB][W][LMB] as fast as possible.
Dunno. I don't find it hard at all… One point I could imagine though is that it would result in something more like the Hazard agility. I liked that. I guess we should try how it feels.
Molten Monarch and Thunderous Skies:
-They are both KoTH, and this game mode doesn't work with Teams well.
-you can kill your Team mates for kills, and be killed by them to save the half dead king
-the points are not added up
-both rounds are KoTH, but have different specific details of how this works (what the king does, how many points), this is not really clear in game and adds to confusion.
-the kill tracing doesn't work very well, especially if you hurt someone and they jump into the underground or in lava. A side effect of this is that people can kill themselves rather than giving you the kill.
-the maps are rather difficult to navigate
-getting weapons from chests is rather slow
And finally in Cauldron: I couldn't climb the top left ladder, because some pixel blocked the top.
-They are both KoTH, and this game mode doesn't work with Teams well.
-you can kill your Team mates for kills, and be killed by them to save the half dead king
-the points are not added up
-both rounds are KoTH, but have different specific details of how this works (what the king does, how many points), this is not really clear in game and adds to confusion.
-the kill tracing doesn't work very well, especially if you hurt someone and they jump into the underground or in lava. A side effect of this is that people can kill themselves rather than giving you the kill.
-the maps are rather difficult to navigate
-getting weapons from chests is rather slow
And finally in Cauldron: I couldn't climb the top left ladder, because some pixel blocked the top.
Molten Monarch is indeed a completely misconcepted round. Teams should be disabled for KoTH, is this possible in network mode?
I am not really happy with the KoTH goal in general. It just doesn't play as it was intended.Someone once made a round where you gain points while actually IN an area. I think that would be a better KoTH for our rounds.
I like the KotH rounds. It's mostly luck-based, but they play a bit differently and there is no camping.
Hm, I had the impression that most people outside of the developer team don't understand that you are supposed to target the king to get points
Can't we draw the description over the loader while loading the scenario?
> -both rounds are KoTH, but have different specific details of how this works (what the king does, how many points), this is not really clear in game and adds to confusion.
That comes from a time when enthusiastic developers thought that people would actually read the scenario description!
I forgot pluto's suggestion:
It is still hard to tell if the Blunderbuss is loaded or not, so he proposed a background color in the item slot. Like make it slightly red if it is not loaded or something.
This could also apply to the grenade launcher (where it is easier though).
It is still hard to tell if the Blunderbuss is loaded or not, so he proposed a background color in the item slot. Like make it slightly red if it is not loaded or something.
This could also apply to the grenade launcher (where it is easier though).
-Escape the Volcano Extreme: One can climb on the border of the map
-if you switch items while holding the musket, the loading action will continue
Hide out:
-Our door stopped closing
-too long, map too big, too exhausting
-not visible how many flags to capture
FrozenFortress:
-Instant Kill through IceBall
-IceBall uses Fire Sound
-Respawn zu schnell
-not visible how many flags to capture
Cauldron:
-player got stuck
-if you switch items while holding the musket, the loading action will continue
Hide out:
-Our door stopped closing
-too long, map too big, too exhausting
-not visible how many flags to capture
FrozenFortress:
-Instant Kill through IceBall
-IceBall uses Fire Sound
-Respawn zu schnell
-not visible how many flags to capture
Cauldron:
-player got stuck
>-too long, map too big, too exhausting
It isn't big at all. In CR, CMC's Oasis is even bigger and is played with CTF aswell.
> -if you switch items while holding the musket, the loading action will continue
Same goes for the grenade launcher. But I wouldn't see this as a bug. In Hot Ice it is very common to be used to reload while performing windbag-jumps.
> -player got stuck
I think this happens on other maps than cauldron as well.
The random petrification / depetrification of the Lava Dude (played in that Tower Level) is a bit frustrating, you never know when to jump.
This is actually the first time I saw the Lava Dude in action, I think we could make it a bit more valuable, suggestions for change:
-only make it petrify when outside of lava
-in turn, always enable walking on the shell side of the LavaDude
-Now make the lavadude move around much less, maybe only floating on top of the lava and turning.
That way we can use it more Mario-like, to cross lava. The way it is now, furiously moving and randomly petrifying we cannot use it like that.
Also his attack:
-I would make it range further, but with much less damage, and without explosion. This way he would hit more often, but less crucial - and that way probably more fun for settling rounds, when you get somewhere near him.
-------
The Mooq:
Doesn't use the sounds I made for the mooq.
-------
Lava bubbles:
If they were smaller, they would probably look better.
This is actually the first time I saw the Lava Dude in action, I think we could make it a bit more valuable, suggestions for change:
-only make it petrify when outside of lava
-in turn, always enable walking on the shell side of the LavaDude
-Now make the lavadude move around much less, maybe only floating on top of the lava and turning.
That way we can use it more Mario-like, to cross lava. The way it is now, furiously moving and randomly petrifying we cannot use it like that.
Also his attack:
-I would make it range further, but with much less damage, and without explosion. This way he would hit more often, but less crucial - and that way probably more fun for settling rounds, when you get somewhere near him.
-------
The Mooq:
Doesn't use the sounds I made for the mooq.
-------
Lava bubbles:
If they were smaller, they would probably look better.
By Anonymous ![[us]](/mwf/flags/us.png "United States")
Date 2018-02-20 18:29
> -Now make the lavadude move around much less, maybe only floating on top of the lava and turning.
I think there could be some kind of warning sign one second before it moves. That way, you could plan ahead. We could also add it as a parameter, so the scenario designer can place both aggressive and more passive lava dudes (ideally as a dropdown in the editor).
Can you paste the feedback somewhere where it's not forgotten? I.e. milestone trello board or bugtracker?
Some feedback from me:
- Construction sites could be auto selected in the object interaction menu. It is a bit time consuming to click on "Construction site" each time when adding new building material to a construction site
- The Iron Peak scenario is running really slow on my notebook, probably due to the snow particles. Maybe add a graphic option in the game options that allows the user to limit the number of particles, disable shaders etc.
Keep up the good work!
Edit: Iron Peak works fine now after disabling antialiasing, but smoke particles are still a problem (See attachment)
- Construction sites could be auto selected in the object interaction menu. It is a bit time consuming to click on "Construction site" each time when adding new building material to a construction site
- The Iron Peak scenario is running really slow on my notebook, probably due to the snow particles. Maybe add a graphic option in the game options that allows the user to limit the number of particles, disable shaders etc.
Keep up the good work!
Edit: Iron Peak works fine now after disabling antialiasing, but smoke particles are still a problem (See attachment)
Attachment: Screenshot014.png (204k)
> The Iron Peak scenario is running really slow on my notebook, probably due to the snow particles.
Unfortunately, the lag is purely happening because of the very complicated area control math (the scenario goal). The effects are not that complicated.
This has been an issue since 2014. Unfortunately Maikel insisted on keeping the (theoretically sound) monte carlo approach to the area calculation. I'd still say just using an approximation using rectangles (or even just the distance to the nearest flag) would be fine for that goal.
Looking at the code, I think the monte carlo algorithm isn't actually the issue here. It's slow because it's calling IsFlagpole on all objects ten times each frame. Also the average of all the samples is calculated by always looping over the full array (of 2000 elements), which is not necessary.
True, the performance of the MC approach can probably be further optimized. I am still doubtful that it is noticeably more accurate than just using any approximation that could be calculated every time a flagpole is built.
>every time a flagpole is built
It's a bit unfortunate that we don't have a good way to query newly-created objects. A call from the flagpole into the expansion goal would create a rather ugly coupling that is irrelevant for almost all scenarios.
Quick idea: Maybe we could add a
Find_SinceObjectNumber(int num) that would select all objects with number between [num, EnumerationIndex] (plus some handling for (signed?!) overflow) which (I assume) can be selected efficiently. This would help all periodic FindObjects calls that select a relatively small set of objects.
Oooor, just put the goal and an appendto to the flagpole in the scenario. No need to over-engineer things for now :)
I just realized that FindObjects is actually not used anymore. I was looking at an old commit from when Maikel worked on the issue linked above. So there's no obvious reason why it would be slow...But still, I think "periodic FindObjects" is a fairly common thing that can currently be hard to optimize, so it might be worthwhile to think of a more general solution that doesn't involve #appendtos and manually-organized object arrays.
Could one optimize it by assigning the objects to the same layer as the game goal object, or some other reference object, for things that should not find actual ingame objects but rather "conceptual" objects?
Here's a StackOverflow answer that may help with an exact solution: https://stackoverflow.com/questions/1667310/combined-area-of-overlapping-circles/1667789Looks like a lot of work to get right in code though (handling holes, etc.)
I've tried to implement this. (Not sure I still have it.)
I don't quite remember how, but you run into trouble with precision. You need to decide where circles touch each other and in which order the intersection points form a polygon, or something…
I think the special cases with holes in the centre and such were comparatively trivial.
I don't quite remember how, but you run into trouble with precision. You need to decide where circles touch each other and in which order the intersection points form a polygon, or something…
I think the special cases with holes in the centre and such were comparatively trivial.
What about horizontal sweeping? It just needs a list of flagpole coordinates. It could give a good approximation and is relatively straight forward to implement with some min max and sin cos wizardry. The sweeping process itself could be carried out over multiple frames, to save time.
Tell me again, what's the difference between monte carlo and horizontal sweeping? Only the order of checking the points, right?
Powered by mwForum 2.29.7 © 1999-2015 Markus Wichitill
