--- Log opened Tue Mar 19 00:00:17 2019 --- Day changed Tue Mar 19 2019 20190319 00:00:17< gfgtdf> for exampel when recruiting a unit it called once for each trait, 20 times for anem generation, and optionally one time for gender 20190319 00:00:39-!- celmin|away is now known as celticminstrel 20190319 00:00:45< gfgtdf> so a normal unit with 2 traits, a name and a random gender will result in 23 rng invokations 20190319 00:00:47< celticminstrel> What exactly is this discussion about? 20190319 00:01:39< gfgtdf> the 'save random seed' feature and how recruiting a unit before an attack gives a differnt attack result. 20190319 00:01:52< celticminstrel> Ah. 20190319 00:02:07< celticminstrel> I think I would say, not a bug? If you take different actions you get different results. 20190319 00:02:28< gfgtdf> yes. 20190319 00:04:37<+wesdiscordbot> But I can't influence what the ai recruits and his recruit choices are separate from the 'save random seed' feature. So me, the player, not doing anyting different can get me different results. 20190319 00:05:02< celticminstrel> I see. That part might be a bug, I suppose. 20190319 00:05:26< celticminstrel> Not sure. 20190319 00:05:37< celticminstrel> I guess the AI doesn't use the synced RNG or something? 20190319 00:06:00< celticminstrel> And FTR, I'd say it probably shouldn't. 20190319 00:06:16<+wesdiscordbot> If I understood it correctly, it does, but not for all the thing it does. 20190319 00:06:28< celticminstrel> I see. 20190319 00:06:33< gfgtdf> the ai does nto use the synced rng. 20190319 00:06:46< celticminstrel> But perhaps the AI's seed could be saved too. 20190319 00:06:49<+wesdiscordbot> Not even for the gender check? 20190319 00:06:54< gfgtdf> but action (attack) invoked by ai use the synved rng just leke players actions 20190319 00:07:29< celticminstrel> Unit generation is not part of the AI. 20190319 00:07:37< celticminstrel> So it would use synced RNG. 20190319 00:08:06<+wesdiscordbot> Ah. It invokes the synced RNG for some things. But it doesn't use it. (Did I get it right this time?) 20190319 00:08:18< celticminstrel> I don't think that's very accurate either. 20190319 00:08:27< celticminstrel> I would say that the AI itself does not use the synced RNG at all. 20190319 00:08:43< celticminstrel> But actions it can invoke, including attacks and recruits, may use the synced RNG. 20190319 00:08:50<+wesdiscordbot> Yes. 20190319 00:08:56< celticminstrel> So AI decisions do not use the synced RNG. 20190319 00:09:01< celticminstrel> But the actions might use it. 20190319 00:09:09<+wesdiscordbot> *do 20190319 00:09:17< celticminstrel> ??? 20190319 00:09:18<+wesdiscordbot> Most of them. 20190319 00:09:28<+wesdiscordbot> Most of the actions do use the synced RNG. 20190319 00:09:35< celticminstrel> I say "might" because not all of them have a random component. 20190319 00:09:41< celticminstrel> For example, recalling a unit. 20190319 00:09:50< celticminstrel> Or moving a unit. 20190319 00:10:22< celticminstrel> Those actions that have a random component use synced RNG, yes. 20190319 00:10:58-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection] 20190319 00:12:31<+wesdiscordbot> True. The problem can arise when the ai makes ('unsynced') decisions that result into actions that use the synced RNG. E.g. recruiting or not recruiting a unit with two possible genders, resulting in different combat results afterwards. 20190319 00:14:47< gfgtdf> @Konrad I suggest you can make a featurerequest about making 'save random seed' save the 'random state' of the ai aswell. 20190319 00:16:58< gfgtdf> implementign that woudl require seperating the ai rng form the rest of the unsnced rng, (that is used for drawing etc) but it possible. (but it woudl imply some other things in paricular the lua ai coudl no longer use math.random) 20190319 00:17:52< irker584> wesnoth: gfgtdf wesnoth:master 45c2a5aa0e0d / src/mt_rng.cpp: fix mt_rng::rotate_random https://github.com/wesnoth/wesnoth/commit/45c2a5aa0e0db342c57779755f2f22425df6b2b4 20190319 00:19:31< gfgtdf> didn't we recently have less than 900 open issues or was that just some weird dream i had? 20190319 00:21:53<+wesdiscordbot> I remember the old days when reaching 200 was a reason to panic 20190319 00:22:15<+wesdiscordbot> (and by panic I mean going on bug squashing sprees to bring it down to 100 or less) 20190319 00:24:45<+wesdiscordbot> I guess I'll do that tomorrow. (It just disturbs me that this breaks with the premise of the 'save random seed' feature.) 20190319 00:27:28<+wesdiscordbot> Well, I understand that it's seen as breakage from the players perspective, but on the other side it's not really a feature worthy for huge background changes for it 20190319 00:30:30-!- irker584 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Ping timeout: 246 seconds] 20190319 00:31:59<+wesdiscordbot> But rewording the feature would be easy, right? I mean, if you don't make reality accommodate the described feature, why not make the description of the feature accommodate reality? 20190319 00:34:10<+wesdiscordbot> (Either the feature is described wrongly or the feature has a bug. :D) 20190319 00:36:01<+wesdiscordbot> Technically it could also be both, but at least I'm sure that I can exclude that it is neither. 20190319 00:38:44<+wesdiscordbot> Maybe you find a better description to the checkbox in the campaign menu. 20190319 00:38:48<+wesdiscordbot> Or the mouseover there 20190319 00:41:20<+wesdiscordbot> The mouseover it would be. 20190319 00:47:17-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer] 20190319 00:53:21-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20190319 00:58:03-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer] 20190319 01:00:51-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20190319 01:10:42-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20190319 01:26:10-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20190319 02:24:44-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.] 20190319 02:25:33-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 02:25:49-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds] 20190319 02:25:51-!- wedge010 is now known as wedge009 20190319 02:30:03-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 245 seconds] 20190319 02:38:26-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 02:39:26< wedge009> gfgtdf: We did, but it keeps growing so fast I can't keep track of them any more. I don't know how many of them may be duplicates or related or no longer relevant. 20190319 02:41:56-!- irker720 [~irker@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev 20190319 02:41:56< irker720> wesnoth/wesnoth:master gfgtdf bc796fd224 add [modify_unit][set_variable] AppVeyor: All builds passed 20190319 02:41:59-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 02:43:42<+wesdiscordbot> lots of old things related to the UI and such that might be able to be cleaned up, but I don't know which are done and which aren't 20190319 02:43:47-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 245 seconds] 20190319 02:43:57<+wesdiscordbot> did clean up campaign label down to 6 tho 20190319 02:46:38-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 272 seconds] 20190319 03:23:11-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 03:49:50< irker720> wesnoth/wesnoth:master gfgtdf 5192424b85 add [modify_unit][set_variable] AppVeyor: All builds passed 20190319 04:01:17-!- gfgtdf [~Daniel@x5f745f02.dyn.telefonica.de] has quit [Ping timeout: 245 seconds] 20190319 04:52:30-!- celticminstrel is now known as celmin|sleep 20190319 05:09:55< irker720> wesnoth/wesnoth:1.14 Lennard 6216f82df5 style and spelling fixes, especially mis AppVeyor: All builds passed 20190319 05:39:26-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 05:40:43-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 246 seconds] 20190319 05:43:31-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 246 seconds] 20190319 06:28:12-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 07:14:50< irker720> wesnoth/wesnoth:master gfgtdf 45c2a5aa0e fix mt_rng::rotate_random AppVeyor: All builds passed 20190319 07:26:08-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20190319 08:39:40-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Read error: Connection reset by peer] 20190319 08:40:07-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 08:54:59-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 08:56:18-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 245 seconds] 20190319 08:56:18-!- wedge010 is now known as wedge009 20190319 09:00:27-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 246 seconds] 20190319 09:17:29-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 09:50:50-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20190319 10:15:14-!- irker720 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20190319 10:21:16-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Read error: Connection reset by peer] 20190319 10:24:46-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds] 20190319 10:24:58-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 10:44:10-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 272 seconds] 20190319 10:45:54-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 11:57:36-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 11:59:14-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds] 20190319 11:59:15-!- wedge010 is now known as wedge009 20190319 12:08:08-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 12:09:48-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds] 20190319 12:09:49-!- wedge010 is now known as wedge009 20190319 12:31:28<+wesdiscordbot> @Konrad2 does this happen as well with the experimental AI. It's smarter with recruiting, it may be that it recruits always the best. 20190319 12:50:56< celmin|sleep> But what if there's multiple bests. 20190319 12:51:01-!- celmin|sleep is now known as celmin|away 20190319 14:33:03-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Read error: Connection reset by peer] 20190319 14:33:29-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 14:37:14-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 14:37:50-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 250 seconds] 20190319 14:41:17-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 245 seconds] 20190319 14:48:29-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 14:56:06-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 14:56:55-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds] 20190319 14:56:56-!- wedge010 is now known as wedge009 20190319 15:03:37-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 255 seconds] 20190319 15:04:41<+wesdiscordbot> From my experience Exp Ai usually recruits the best in its opinion composition of units. And becouse of that it sometimes can be abused by killing weak support units and it will still try to rebuild them over and over again. 20190319 15:20:42-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 15:27:59-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 15:29:13-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 245 seconds] 20190319 15:29:13-!- wedge010 is now known as wedge009 20190319 15:35:36-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 246 seconds] 20190319 15:38:17-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 15:45:32-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 15:46:15-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 268 seconds] 20190319 15:46:16-!- wedge010 is now known as wedge009 20190319 16:31:53-!- wedge010 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has joined #wesnoth-dev 20190319 16:32:33-!- wedge009 [~Thunderbi@60-241-236-92.static.tpgi.com.au] has quit [Ping timeout: 258 seconds] 20190319 16:32:33-!- wedge010 is now known as wedge009 20190319 18:17:00<+wesdiscordbot> https://steamcommunity.com/app/599390/discussions/0/3398435622571674878/ Anyone any idea? 20190319 18:19:12<+wesdiscordbot> https://github.com/wesnoth/wesnoth/issues/2999 20190319 18:20:53-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 245 seconds] 20190319 18:25:36<+wesdiscordbot> Thanks. 20190319 18:41:03-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev 20190319 21:42:36-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20190319 22:50:05-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection] --- Log closed Wed Mar 20 00:00:32 2019