elexis

Wrong chinese translation, perhaps not the only one. Starting rP21907. Could be detected by D1674. You can help fixing and translating on https://www.transifex.com/wildfire-games/0ad/ if you want.

Actually it looks like I made a rebase error. Perhaps I was the one who didn't delete thier cache file? (-> D2273)

Tried deleting the cache? Unrevisioned files? It really sounds like running like old gui/ JS (in particular following rP22676).

It works for me, any mods active? Or leftover mainmenu.xml? I suppose windows? I suppose that is the outdated menu.js copy. The commits after July 27th are missing. Your only change is this: --- /tmp/menu.js_orig 2019-09-07 01:04:59.261486333 +0200 +++ /tmp/menu.js_de 2019-09-07 01:03:29.124822542 +0200 @@ -849,7 +849,9 @@ }); }; - barterButton.Buy.hidden = isSelected; + if (g_BarterSell == "coin"){ + barterButton.Buy.hidden = isSelected;} + else {barterButton.Buy.hidden = resourceCode != "coin";} barterButton.Buy.enabled = controlsPlayer(player); barterButton.Sell.hidden = false; selectionIcon.hidden = !isSelected; Perhaps to keep track of your chances, you might want to store the diffs along the mod? i.e. the above in menu.js.diff, and then when theres a new menu.js in svn, you can apply the patch again rather than applying the svn commit to your copy.

What was the expected behavior? Perhaps one could also update the paths during mousemove events (essentially doing the current mouserelease processing on mousemove) (but that sounds laggy).

Check the most recent autobuild.

That's what we always wanted. The problem with that is that you have uneven grounds. Consider the passage between two steep mountains as a worst case. There is a 32*32 texture repeated over the radius, one can do more with it than just a blank line. The feature was introduced in D238, somewhere should be some more alternative textures. I suppose if the visualization is very small in diameter and depending on visualization it wouldnt appear that badly if its a 2d disk, but it wont allow displaying the actual aura range until adding that code as a feature (break down the texture into many smaller textures over the area, just like is done with the ring). I suppose thats not impossible to code and probably not impossible for performance either (not so many auras to be displayed I guess). (For the record, the feature is also used for attack rangeof buildings and heal range. Female inspiration aura too.)

You mean the units should start moving before the mouseup event? The idea of the feature is that you can draw an arbitrarily shaped line and the units will spread evenly from start to finish. It means that it has to know start and end before it can do the computation.

rP22854 rewrote the main menu code. Mods that wnat to specify their name to be displayed in the main menu can easily do so by adding a new file instead of rewriting mainmenu.xml. @wowgetoffyourcellphone Your mainmenu.xml contained this diff to svn: and I would recommend to delete this file and put this file under pregame/ProjectInformation_DelendaEst.js: g_ProjectInformation.productDescription.caption = setStringTags(translate("Delenda Est v0.0.03"), { "font": "sans-bold-16" }) + "\n\n" + translate("Notice: This mod is under development and many features have not been added yet."); g_MainMenuItems[4].caption = translate("Atlas Scenario Editor"); g_CommunityButtons[0].tooltip = translate("Click to open https://0ad.mod.io/delenda-est in your web browser."); g_CommunityButtons[0].onPress = () => { openURL("https://0ad.mod.io/delenda-est"); }; The mods that have more weblinks or information to add, see ProjectInformation.js:

Someone who leaves a rated 1v1 without resigning will be warned or banned for more or less time from the lobby. The problem is not what to do with the information that someone left a game, but how to obtain it. If I mention it in chat that you are left a rated game without providing any proof, you get people accusing each other incorrectly resulting in wrong bans. The real problem lies in detecting when someone left. If you trust the host, then the host can send the same packet and also cause others to be banned under his wish. For example banning the opponent from within the game or by dropping the IP address communication, thats the same as the user becoming disconnected. So the only solution that we have discovered in the past years is to have WFG host the game, then WFG can detect who disconnects, or replays being uploaded automatically and having a machine comparing. But the latter sounds like a workaround. So implement that, one first needs to implement dedicated hosting (i.e hosting without rendering the game), which means having to rewrite the gamesetup to allow players to setup the game (instead of the host). So its easily several months work if one wants to do it this way. And the replay-comparison isn't really making it easier except for the gamesetup part, because a dedicated server that receives the simulation orders and compares them is almost the same as the dedicated host. And we want a dedicated host software anyhow. And we want players to be able to setup the game also anyhow (see success of the autociv mod, which has existed in other mods before as well). It will come, but it takes some time, and especially dedication to not work on any other projects for the time it takes to implement that.

Looks like bush_medit_me.xml, there is a second bush with a similar name and the same issue.

refs #4119 better would be not crashing instead of deleting corrupted cache files in the menu (after it possibly already crashed), and deleting the invalid file automatically.

For reference, according to the according PM conversation (9954): So it sounds like a mobile connectivity issue, dunno.

You compiled 0ad but not the glooxwrapper library, fixed by the last glooxwrapper autobuild.

On the positive side of the cursed wolf being a domestic animal, the OP slaughter attack being used is that it makes the map so easy that people dont complain about it being too hard. That diff to have these regular wolves controllable was committed with the first revision of the map but was reverted directly as it was seen that animals should not receive orders because they are animals, unless they are domestic animals, then they should. I think it would be better for mod freedom if there was a "Controllable" boolean property for UnitAI to make it more abstract rather than hardcoded one way or the other.

Which OS are you guys using? From the lobby server logs, it seems like it's failing at the TLS authentication stage. (https://en.wikipedia.org/wiki/Salted_Challenge_Response_Authentication_Mechanism)

Local firewall? Router or ISP censoring wfg? Tried disabling TLS encryption (in case you trust your ISP and arent on a public wifi)? (Perhaps its an outdated cert?) In case you get it working, please tell us what the issue was. We could also look at the server logs.

I thought too about adding a warn("add a civ req to that pair tech"); warning in that place, but it seems the wrong place to check for that in the GUI and only upon capturing with an enemy. And I don't consider a GUI bug about iterator warnings to be a feature. Should be checked when the stuff is loaded.

Since (not only because) you pinged me, I checked. The error can be reproduced by running with -mod=public -mod=delenda_est -autostart="random/new_rms_test" -autostart-seed=-1 -autostart-players=2 -autostart-civ=1:brit -autostart-civ=2:imp -autostart-player=1 alt+d reveal map, select imp CC, "wololo" cheat. It seems to be a classic, as it appeared in rP18653 also. The error happens because GuiInterfaceCall("CheckTechnologyRequirements") says that requirements are not met (accurately, since the two pair techs have the civ requriement unmet), but then it wants to display template.reqs in the tooltip, expecting that to be an array. But template.reqs = DeriveTechnologyRequirements(...) = InterpretTechRequirements(...) = false if the civ requirement. So after a less superficial look the superficial solution (code.wildfiregames.com/D2171) still seems to be the right one, to add a "if (reqs)" just before the "for (let req of reqs)" in selection_panels.js. This way if someone captures a structure that has a pair tech where the parent has no civ requirement but the child techs do, then it will show these buttons grayed out instead of erroring out (and without a "you need to change your civ to research this" tooltip). As suggested by fatherbushido, the civ requirement should also be added to the paired tech, so that the icon disappears completely to the capturing players of other civs. Also I got some other errors in DE in case you are interested:

Thanks for the report. It seems to be a VS2015 bug (as in 0ad code being correct according to the specs), see rP22604#35842.

I plead guilty, it's fixed, thanks for the report!

I guess you can try? I don't see how the error can occur if not outdated files.

You created this forum acount because your previous one was banned?

The mentioned functions are new in a24 as of rP22284, please do a complete clean build.

Queens of the Stone Age with Dave Grohl Glastonbury Festival 2002 Best lineup!