elexis

You start xpartamupp with --room arena21 but join with the alpha 22 client

It means the WFGBot (XPartaMupp.py) is not running

It appears as if you have been banned for rating fakery.

What matters is having the same IP address / internet connection

You tried to play with a friend and this is the second account you wanted to create? There is a spam protection (1 hour)

Sure, just edit game_speeds.json. But our code uses 100% of a single core and I don't think that even the most recent CPU will help in reaching that number. It helps more to reduce the mapsize, number of AIs and population count.

20x is only the upper limit, the CPU is the bottleneck. When simulating many AIs players on a larger map with greater pop counts, it becomes increasingly slow and often one can reach only 2x gamespeed at most. So setting a greater number in game_speeds.json doesn't have any effect. The non-visual modes (replay and autostart) will run a bit faster as they don't have to render anything. That is mostly useful to debug code or test AI results. I didn't get into touch with the JSON summary output yet.

IMO with alpha 21 walls became well balanced (at least the best balance Ive seen in a number of alphas). They can be built quickly and provide a good defense as long as one can defend vs siege engines. While it is true that they don't keep an attacker out if the area isn't completely walled, they are still useful to stabilize areas in the frontline, for example building a wall around a newly built civic center or mines in that area (so that enemies will be deterred from wall turret arrows and towers don't hit the mining units). Dragging on the lategame is their job, but an enemy that is an actually better position can tear down those walls very quickly with siege engines (1 ram is still faster than 50 swordchampions).

When upgrading an entity, the old one is deleted and a new one is created. So the old one will count as lost (OnOwnershipChanged to -1). That will be initiated from ChangeEntityTemplate from Upgrade.js. As we can see in that function in Transform.js, an EntityRenamed message is sent just before deleting the entity, so the victory condition could listen to that and replace the remembered entityID before it's deleted.

Re temple: (3) Because the building is square and they start shooting once in range. (4) Agree, but the issue is especially that we can't select multiple buildings of allies at once. (5) sure (7) That's controversial, discussion goes back forth. Most recently here #4576. (8) We've noticed th half-second delay for tooltips in D459 too. Agree to remove it, trivial. (12) Agree, the civs are well differentiated, all unique, even all hellenic ones. Apparently this is wanted though, so a new page was proposed once upon a time #3987

Already had issues finding that improvement again, as it's buried on page 132 of a collective thread and the remarks are soon becoming forgotton history, so please create trac tickets for these issues, temple. For the civ dropdown size, I've created D527.

Can we add that check to Attack.js until someone is willing to rewrite BuildingAI? :-(

The current version of the copy of delenda est's file is the same as in the public mod one excluding this cost diff. The file (when it was added to delenda est on february 22nd), was either outdated by a commit which was the reverse of that hunk above (rP19213), or intentionally used the prior version. Now since that commit specifically fixed a delenda est bug, I wonder why delenda est reverts that x) This comment is especially about the changed line https://code.wildfiregames.com/D113?id=402#inline-2220 IMO the only real option you have here is deleting that file to stop having to keep it in sync and tell us if something is broken, we can fix it quickly. I've tested de with the deleted file and it works just fine. If you don't delete it, it will just throw random errors the next time we change it. Furthermore, your ptol_hero_cleopatra_3.json doesn't have a name nor description, thus bugs with the tooltips.

Comparing the delenda est version with the public mod version from that time, the only difference is: @@ -368,7 +368,7 @@ ret.cost = { "time": template.researchTime ? +template.researchTime : 0 }; for (let type of resources.GetCodes()) - ret.cost[type] = +(template.cost && template.cost[type] || 0); + ret.cost[type] = template.cost && template.cost[type] ? +template.cost[type] : 0; ret.tooltip = template.tooltip; ret.requirementsTooltip = template.requirementsTooltip || ""; Since that file is outdated, things like the resource trickle tooltip and aura range parsing are missing here: https://code.wildfiregames.com/source/0ad/history/ps/trunk/binaries/data/mods/public/globalscripts/Templates.js https://github.com/JustusAvramenko/delenda_est/commits/master/globalscripts/Templates.js

This sounds a lot like the Templates.js file being outdated: https://github.com/JustusAvramenko/delenda_est/blob/master/globalscripts/Templates.js Didn't delenda est have an issue with that some weeks ago already? Why is a copy of that around to begin with? If you need a new function, add it to a new file. If you need to change an existing function, tell us and we'll see what we can do to make 0ad work better with mods.

This could be a ranged aura without aura range. I checked delenda est's aura folder, but all ranged ones have a radius too. I'd just add a warn(auraID); to see the name of the aura that it doesn't digest.

This guy got the same issue: #4561. Now if some of you would figure this out, we could close it for good for more than 3 people.

Thanks, I could reproduce it, even on Alpha 21: 1. Enable persist-match-settings 2. Select a skirmish map 3. Disable locked teams & enable last man standing 4. Get back to main menu, enter gamesetup again 5. Notice the more options dialog has both settings enabled and that this warning is thrown when starting the game. It is mostly visible because these two maps have lock-teams = true by default, but equivalent bugs can be possible for those maps that specify lock-teams=false. We can set that lock-teams = false for that new map, but it's no use. The persist-match-settings feature and with it the entire concept (still) of saving settings has to be redesigned. That warning seen above won't hurt the game, since last man standing will still be disabled, so probably no need for immediate action.

I suspect this is a new (old) bug, so thanks for testing! Can you reproduce the bug? Was it a random map? Did you enable last man standing in the previous match? Was it a lobby game? Do you have a replay of that game?

Account registrations were enabled again.

This is a frequently requested feature: #3491, #57. It should be uploaded as a patch file, see SubmittingPatches and GettingStartedProgrammers, so that even if the patch won't be finished, that others might take inspiration from this. It should be done in independent steps (pinging and the resize independent of that, Resizing also needs some thinking on how we can yield the best usability from this).

1+2 -> Yep, wall placement code is odd, leaves gap, can't continue walls if the turret is missing. Will be tough to fix 3 -> Had proposed some patch to make this an actual gate, but the model guys weren't really happy with it, wanted to improve it, and that's where we left last release. What was there actually todo @stanislas69 ? Making sure the towers don't overlap? The entire thing just being too huge for a gate? Only found this screenshot in the irclogs Afaics it historically was part of a wall too

The shared ally vision technology technology at the market seems easier (and cheap enough, perhaps making it more expensive might make players reconsider). But the trick seems handy for diplomacy games and neutral players

Fixed next release by putting all elephants, siege engines and ships on a strict diet. See the screenshots https://code.wildfiregames.com/D253#9665 And if the passage is too small (alpha 21), you have to take ships.

Having a trac ticket that links to the forum thread about a topic that is desired by many but not implemented nor documented yet isn't bad. Endless discussions is a thing. There are definitely some missing design decisions, but nothing is impossible to solve. The main problem that most of the new design concepts have is that there is no programmer who is enthusiastic enough about this to finish the design concept, convince the team and write all of the code which can take several months to complete.