Not logged inOpenClonk Forum
Forum Home Help Search Watchlist Register Login
Up Topic Development / Developer's Corner / 5.3 release planning
1 2 Previous Next
Post
- - By Clonk-Karl [de] Date 2012-04-21 19:32 Edited 2012-05-13 09:22
I think we should have a first 5.3 release rather soon, with the settlement stuff we have so far. Mid-May might be a good date so the OC contest scenario(s?) can be played with a stable game version instead of requiring people to use the latest development snapshot. Also, we got quite some feedback based on the release version so far just to tell people "thats out of date, use the dev snapshots instead".

What would be definitely needed until then?

* Working rope tower, so Skylands can be solved
* Modeled and textured elevator case would be really nice (uv map done by Newton)
* (Written) tutorial for settlement (done by JCäsar)
* Rework loam so it can be produced using the bucket
* Make the flagpole C4D_StaticBack and add a function to recollect it(?)
* Basement for the elevator
* What else?

I can maybe give a the rope tower a try (what exactly is the problem?) but someone else would need to take care of the elevator case (Newton?)

Do people agree this makes sense? If yes we should soon start a 5.3 branch which contains only stuff we definitely want it to have and bug fixes, but no new features (neither in engine nor in game content).
Reply
Parent - - By Zapper [de] Date 2012-04-21 19:52
I am not sure whether it makes sense.
But first:

>(what exactly is the problem?)


Proper rope physics. (Possibly implemented in the engine)

Why I am not sure: I have a lot of unfinished stuff lying around.
But currently I am working more on the settlement scenario for the contest than on directly OC related stuff. That means: Even if the timing for being able to play those scenarios would be cool, the timing to actually get work done might be not so ideal :)
(Also those scenarios do not run away and could be further improved after the end of the contest!)
Reply
Parent - - By Clonk-Karl [de] Date 2012-04-21 20:06

> Proper rope physics. (Possibly implemented in the engine)


So does that mean there is no way this can be fixed w/o a total rewrite?

> Why I am not sure: I have a lot of unfinished stuff lying around.


What's the problem with that? It can go in anytime, just not for 5.3. At some point we need to make a cut or we will never have a release.
Reply
Parent - - By Zapper [de] Date 2012-04-21 20:16

>So does that mean there is no way this can be fixed w/o a total rewrite?


I admit that I did not have a in-depth look at the code. But there is the possibility that a rewrite would be faster than trying to adjust the current solution.

>What's the problem with that? It can go in anytime, just not for 5.3. At some point we need to make a cut or we will never have a release.


True, that was just my: "but I am busy with the settlement scenario :((" :)
(for example wheat would be cool to have in the first release, since we already have the windmill and all, also I wanted to change the visuals of the power system a bit)
Reply
Parent - - By Randrian [de] Date 2012-04-30 08:07
Well, every now and then I work a bit on the rope physics.
One problem, which has to be soved is to make a proper "interface" between the rope physics particles and "engine" objects. The "forces" of the landscape on an object (e.g. a lorry being pushed by a clonk) are much different then those on the rope particles (just a normal "force" precenting it to enter the landscape). I have had many thoughts about this but didn't find a proper solution to that.
The next problem is that lines can intersect with the landscape. I have already written code to prevent that, but that will slow down the ropes a bit.
You refered to an implementation in the engine. This could be a good idea, cause there is much calculation done in the ropes, which could be faster if it would be implemented in the engine. But then we would need a good interface for the ropes and the objects of the ropesegments.
Reply
Parent - - By boni [at] Date 2012-04-30 08:19
ck just started an engine-implementation of the rope. It's in the rope-branch. :)
Reply
Parent - - By Clonk-Karl [de] Date 2012-04-30 09:00
Randrian, I could well need your help there. If you have time for a short discussion in IRC at some point that would be appreciated :)
Reply
Parent - - By Matthias [de] Date 2012-04-30 11:14
I think this is a pretty interesting topic, do you care to document your approach somewhere (Blog?)?
Reply
Parent - By Clonk-Karl [de] Date 2012-04-30 18:27
At the moment I'd rather spend time on actually working on it. I can write something up once (if) it becomes halfway usable.
Reply
Parent - By Randrian [de] Date 2012-04-30 18:06
Ok, I'll try to pass in IRC when I have time. I hope my implementation of the rope physics is of some help and wasn't completly wasted, if you write a complete new one on your own.
Reply
Parent - By Maikel Date 2012-04-21 19:56 Edited 2012-04-21 21:49
I think this makes sense, depending a little on how much work we get done by that time, the pump and related stuff is also a necessity.

Also, for really interesting settlement burning and destructible structures would be nice.
Reply
Parent - - By boni [at] Date 2012-04-21 20:03
Telling me in advance so I keep everything that migth break stuff away from default. ;)
For me, the most important thing is that the newly introduced controls/gui stuff works as intended. Worst case we have to quickly bugfix-release. :I
Reply
Parent - - By Clonk-Karl [de] Date 2012-04-21 20:07

> Telling me in advance so I keep everything that migth break stuff away from default. ;)


You can put as much experimental stuff in default as you want, just not in the 5.3 branch.
Reply
Parent - By boni [at] Date 2012-04-21 21:31
Yeah, I just remembered that releases are their own branches. :)
Reply
Parent - By Clonkonaut [ie] Date 2012-04-21 20:11
Agreed. The Controls branch is just about finished enough.

> I can maybe give a the rope tower a try (what exactly is the problem?)


There are quite a lot problems:
1. Even if there is no corner of whatsoever shape, the objects are simply lifted into the air towards the tower. This is no rope pulling an object across the ground because gravity is killed completely. Same applies if the object is higher than the tower. It's then slowed down to the pulling speed of the tower.
2. If there is a light corner like less than 90°, the rope itself often is fine but the pulling object usually gets stuck. Because the engine behaves poorly with directional forces (speed) and just stops the object. My approach was to 'shake' the object free; whenever it's stuck it gets random impulses. This kinda works but looks silly.
3. If there is a sharp corner, there are a lot of difficulties. First of all, the rope itself often clips into the material which of course results in a major meltdown. When this by chance does happen of course the object gets stuck because the rope does not pull it around the corner. A real rope would get some distance between itself and the wall when it's tightened because the object is somehow pulled around; our rope doesn't. Therefore the object gets stuck.

It's hard to explain this but just check the three different cases and you will clearly see the problems.
Reply
Parent - By Clonkonaut [ie] Date 2012-04-21 23:35
A tutorial (even if written) for the settlement content might be helpful.
Reply
Parent - - By Newton [de] Date 2012-04-22 00:34
Level1? Not so much todo anymore though.
Reply
Parent - - By Clonkonaut [ie] Date 2012-04-22 00:41
But we should better surrender the chemlab...
Reply
Parent - - By Newton [de] Date 2012-04-22 11:24
Why?
Reply
Parent - - By Clonkonaut [ie] Date 2012-04-22 13:02
It doesn't look like someone will do this in the near future :/
Reply
Parent - - By Newton [de] Date 2012-04-22 13:51
What about Matthi?
Reply
Parent - - By Clonkonaut [ie] Date 2012-04-22 13:58
He said he won't make it. Stuck deep in university work or similar. There is no one actually working on this.
Reply
Parent - - By Heyub [us] Date 2012-04-23 14:22
I could make a model, if that is what is needed?  May not be the highest of quality but could do as a place holder... It just sounds fun to model a clonk building, whether or not you use it is entirely up to you, I'm just having fun. ;)
Reply
Parent - By Clonkonaut [de] Date 2012-04-23 18:57
Well, of course you can contribute ;) I'll ask him for the conceptual stuff, so you get some sketches, I think he told me that he has some from Fungiform.

In case you, Matthias, read this, could you attach all information you have on this?
Reply
Parent - By Matthias [de] Date 2012-04-23 21:39 Edited 2012-04-23 21:43
I'll attach the lastest sketches I got from fungiform on this. I hope he doesn't mind. I preferred the one with the huge glass tank, but glass is kinda tricky, and due to a sudden drop in free time, I haven't got around to actually make this model.

Edit: No wait, I put it where it belongs.
Reply
Parent - - By Clonk-Karl [de] Date 2012-04-25 20:50
Can you say what's the status of the elevator case?
Reply
Parent - By Newton [de] Date 2012-04-26 13:54
Didn't further work on it. The newest state is the one in the repository.
Reply
Parent - By Newton [de] Date 2012-05-04 19:11
Made some changes on the elevator case and created the uv map. Anyone's up for texturing it?
Reply
Parent - - By Maikel Date 2012-04-29 10:42
I think we should maybe focus on repairing the old stuff first, 50% of our scenarios are broken due to some engine and control changes that were not tested well enough....
In the future it is maybe useful to start more than one scenario if you push big changes.
Reply
Parent - - By Zapper [de] Date 2012-04-29 11:55
Anything power-related? :|
Reply
Parent - By Maikel Date 2012-04-29 12:03
koth (Goal_Melee->BlaBlup() does nasty stuff now)
Reply
Parent - - By boni [at] Date 2012-04-29 12:22
I suspect that there'll be more things that broke through the Controls-Changes and haven't been found yet. That's the main reason why I wanted to push Controls to default asap. ;)
There'll probably be some errors due to scenarios overloading stuff, and some others because implementations changed. We should probably playtest each scenario at least once.
Reply
Parent - By Clonk-Karl [de] Date 2012-04-29 22:44
Yes, I agree we should schedule a few hours for testing various scenarios before releasing anything.
Reply
Parent - By Maikel Date 2012-05-03 19:23
This needs to be fixed: http://bugs.openclonk.org/view.php?id=638
I am kinda tired of this one by now.
Reply
Parent - By Clonkonaut [ie] Date 2012-05-04 19:06
[!] Rework loam so it can be produced using the bucket

I don't recommend releasing without this fix. I'll do after the 9th if no one else did but not before.
Reply
Parent - - By Clonkonaut [ie] Date 2012-05-04 19:38
Two others things:
- Make the flagpole C4D_StaticBack and add a function to recollect it
- Something basement-like for the elevator. I have an idea but won't be able to it until the 9th ;)
Reply
Parent - - By Zapper [de] Date 2012-05-04 20:15

>- Make the flagpole C4D_StaticBack and add a function to recollect it


Few open questions there:
Should you be able to recollect a flagpole? How do you builld flagpoles in a way that makes them relatively expensive but affordable through the whole game? (aka, you need a second one fast. But you should not be able to cover the whole landscape in flagpoles after 30min)?
How do you deconstruct/demolish enemies' flagpoles?

All have good pro and con arguments. But I guess for a fast settlement release1.0, we should only think about whether we want to be able to recollect a flagpole. (it's a normal building atm, right? There is not "flagpole" object?)
Reply
Parent - By boni [at] Date 2012-05-04 21:13
A completely different approach: Make flagpoles cheap-ish, but limit the amount buildable per player.
Reply
Parent - By Clonkonaut [ie] Date 2012-05-04 21:16
The quick solution would be Split2Components. But we can leave this out for now. Just make it static back so the flag doesn't move away from its radius.

The question can be answered later when we think about melees and balancing. In another thread then ;)
Reply
Parent - - By Isilkor Date 2012-05-04 23:58

> How do you builld flagpoles in a way that makes them relatively expensive but affordable through the whole game?


Make them more expensive the more the player already has.
Reply
Parent - By Günther [de] Date 2012-06-05 15:10
This sounds like the the most flexible approach.
Reply
Parent - - By PeterW [gb] Date 2012-05-05 14:32
Recollection/Destruction: Grab the flagpole for 10 seconds to un-hoist it?

Production: Isilkor has a point - it would be nice to kind of naturally limit the practical amounts of flags around. A way of not going directly over cost would be some gameplay requirement such as forcing the player to collect all existing flags in one place before being allowed to construct another one.
Reply
Parent - - By Zapper [de] Date 2012-05-05 15:10
I am still in favor of the unproducable flag-item (compare: CR). That would limit the flags per scenario/player naturally (the scenario has to provide X amount of flag-objects), make some room for playing around (hiding flags in chests etc) and not feel weird in the way that the player thinks why the heck he can only build X flags even though he has all the material.

But for this release where you can build an endless amount of cheap flags (at the moment): Do we need to be able to deconstruct flags at all?
Reply
Parent - - By boni [at] Date 2012-05-05 16:54
I"m still against said object because of the "will inevitably land in a lava/acid lake" syndrome.
Reply
Parent - By Caesar [de] Date 2012-05-05 18:02
Make it burn/dissolve there and allow to rebuy it then. :P
Reply
Parent - - By PeterW [gb] Date 2012-05-05 18:58
Yeah, basically my viewpoint, too. But I thought we agreed that this sort of approach would be too limiting for cooperative multi-player? Maybe we just make them really expensive, and make it default to start every settlement scenario with $player_cnt+1 flags or something?
Reply
Parent - By Zapper [de] Date 2012-05-05 19:42

>But I thought we agreed that this sort of approach would be too limiting for cooperative multi-player?


How so?
The the scenario designer wants the players to have independend bases (f.e. if every player is on a different island), he could give every player X flags. If the scenario is more in the direction of Gold Rush (where the players work together), the players would just get X flags in total.
As I see it at the moment, more players do not necessarily need a bigger base (they will not suddenly start constructing every building twice, only because they are more).
Reply
Parent - - By Newton [de] Date 2012-05-10 18:48
Regarding the next release: Is there any chance that the next release will be included in the debian repository? Does someone have contact with pkern?
Reply
Parent - By Clonk-Karl [de] Date 2012-05-10 21:17
Yes, since we are no longer in conflict with their free software guidelines that should be OK. I can get in contact with Philipp easily when we are there.
Reply
Parent - By Clonkonaut [ie] Date 2012-05-13 01:43

> Modeled and textured elevator case would be really nice (uv map done by Newton)
> Rework loam so it can be produced using the bucket


Done.
Reply
Up Topic Development / Developer's Corner / 5.3 release planning
1 2 Previous Next
Post

Powered by mwForum 2.29.7 © 1999-2015 Markus Wichitill