## Search found 46 matches

February 6th, 2009, 10:23 am
Forum: Coder’s Corner
Topic: pseudo-random generator + improvement idea
Replies: 47
Views: 19783

### Re: pseudo-random generator + improvement idea

@ Glass Pearl Player: Sorry to disagree, but I have used exactly the Wesnoth's RNG code during my test (over 1.000.000 of time), and in the test I displayed here, it was exactly an unit with 4 strike at 70% chance to hit. So no, there isn't more chance to miss all the 4 strikes, in my test, it happe...
February 2nd, 2009, 10:42 am
Forum: Coder’s Corner
Topic: pseudo-random generator + improvement idea
Replies: 47
Views: 19783

### Re: pseudo-random generator + improvement idea

So assuming the 17th bit has a longer period than the 2nd one, it may indeed hide the issue. The longer the period is, the better it is ;) If the period is longer, there is more randomness. True randomness have no period (infinite period if you prefer). I have written a little program to test the R...
January 31st, 2009, 8:05 pm
Forum: Coder’s Corner
Topic: pseudo-random generator + improvement idea
Replies: 47
Views: 19783

### Re: pseudo-random generator + improvement idea

So I wonder why conceptor of Lehmer pseudo-random generator actually recommand to not use directly modulo to scale numbers... :roll: Few years ago, I have made some test to simulate a 4d6 (4 dices of 6 faces), using 1+rand()%6, 1+(rand()>>8)%6, 1+(rand()>>16)%6. By the way 1+rand()%6 was the worst. ...
January 31st, 2009, 2:58 pm
Forum: Coder’s Corner
Topic: pseudo-random generator + improvement idea
Replies: 47
Views: 19783

### Re: pseudo-random generator + improvement idea

Are you sure it really happens more? The "statistics" part of the save files is there for a reason, so that you can check that it is not just your feeling. Unfortunately checking wesnoth statistic doesn't help. Imagine I attack with 2 units that have 50% chance to hit and 4 attacks. In the first ca...
January 31st, 2009, 12:15 pm
Forum: Coder’s Corner
Topic: pseudo-random generator + improvement idea
Replies: 47
Views: 19783

### pseudo-random generator + improvement idea

Hi! I know you don't like random generator thread, however, my purpose is not only complaining again it: I found why it is not really good random, and I have some idea about how to improve it. So please, don't lock it too early... But now, let's start to complain ;) Like many people, I have experien...
May 15th, 2008, 3:21 pm
Forum: Ideas
Topic: New trait for dwarves : fast-gunner
Replies: 19
Views: 4209

### Re: New trait for dwarves : fast-gunner

BTW they shoot first, THEN reload.
Nobody come into battle with an empty musket, then load it in the heat of battle

About technologies, dwarves always have better technology than human. So I'm sure they could have good gun, like musket.
May 14th, 2008, 2:19 pm
Forum: Ideas
Topic: New trait for dwarves : fast-gunner
Replies: 19
Views: 4209

### Re: New trait for dwarves : fast-gunner

Oups, I forgot to mention this would apply only to the ranged attack... So a nearly dead dwarf with "fast-gunner" trait will be simply killed with a melee attack. And IMO it has always been a bad idea to finish those dwarves with a ranged attack, with or without that trait :) And for the balance iss...
May 14th, 2008, 1:14 pm
Forum: Ideas
Topic: New trait for dwarves : fast-gunner
Replies: 19
Views: 4209

### New trait for dwarves : fast-gunner

Hi ! It's been a long time that I think about it... Give firststrike to all the dwarvish thunderer tree. Why ? Because shooting something with a gun is more fast than using a bow, and more fast than casting a spell. But finally, I think it would be nicer if this wasn't an ability that all dwarvish t...
May 14th, 2008, 12:16 pm
Forum: WML Workshop
Topic: Trouble with map made with 1.3.x on 1.4.x version
Replies: 5
Views: 1226

### Re: Trouble with map made with 1.3.x on 1.4.x version

Thanks a lot, but I have already fixed it for all my map.
But I'll think about it for the next time I update wesnoth
May 13th, 2008, 10:38 am
Forum: WML Workshop
Topic: Trouble with map made with 1.3.x on 1.4.x version
Replies: 5
Views: 1226

### Re: Trouble with map made with 1.3.x on 1.4.x version

Hi ! Sorry, it's not related to word wrap... Anyway, I have found a way to fix it : open the map with wesnoth_editor, and resize the map with +2 width and height, and an offset of -1 for x and y. For example, I have got a map with a size of 40x40 in the former 1.3.x, but in 1.4.x its size is 38x38 w...
May 13th, 2008, 7:34 am
Forum: WML Workshop
Topic: Trouble with map made with 1.3.x on 1.4.x version
Replies: 5
Views: 1226

### Trouble with map made with 1.3.x on 1.4.x version

Hi ! Long time I haven't play wesnoth and work on my campaign ! I have got a little trouble with all my map... I have made them with the editor from the svn 1.3.2 version, and now, I have downloaded the last stable release (1.4.2) and all my map don't work anymore... First when I have tried to launc...
October 8th, 2007, 11:14 am
Forum: WML Workshop
Topic: number of movement can't be under 4 hex ???
Replies: 4
Views: 1316
okey, thanks
October 4th, 2007, 5:30 pm
Forum: Ideas
Topic: allowing changing recall_cost in a [side] tag
Replies: 5
Views: 1127
It's possible to set a recall event so the player will be refunded...
yes but at least 20 gold piece is needed
October 1st, 2007, 2:49 am
Forum: Ideas
Topic: allowing changing recall_cost in a [side] tag
Replies: 5
Views: 1127
Yes, in my case, some scenario follow immediatly another, and it is more logical (and more playable) that for this scenarii, recall would be totally free... I can do a massive auto-recall in WML, but the player don't nedd some kind of unit... Another problem : what if one scenario allow a unit to re...
September 30th, 2007, 4:40 am
Forum: Ideas
Topic: allowing changing recall_cost in a [side] tag
Replies: 5
Views: 1127

### allowing changing recall_cost in a [side] tag

It would be nice, if we can overide the attribut "recall_cost" in the game.cfg (in the [game_config] tag), especially in a [side] tag...
Some attribut are already there like "village_income"... what about extending it to "recall_cost" ?