- User Since
- Apr 5 2019, 7:29 PM (163 w, 5 d)
Mon, May 23
Sun, May 22
Works as intended.
Fri, May 20
This looks good, I will test later.
Reads and greps correct.
Not renaming is the safest here.
Thu, May 19
I guess they wanted to accept. :)
@marder can commandeer back?
Change is good.
Mon, May 16
Sun, May 15
Sat, May 14
Not useful for PetraAI.
Fri, May 13
Thu, May 12
One thing, I think this lacks is recovering, we stay in the emergency state forever, even if we'd win.
Something like if we're back to ~75% of the previous population the emergency is lifted?
Wed, May 11
[16:38:28] <elexis> try this [16:38:28] <elexis> 1. revert 26867 [16:38:28] <elexis> 2. add an argument GetTemplate to loadCivFiles and replace Engine.GetTemplate by GetTemplate [16:38:28] <elexis> 3. pass Engine.GetTemplate in all calls, except the one call in the simulation scripts (survival of the fittest triggers) [16:38:33] <elexis> minus the question [16:39:10] <elexis> 4. in survival of the fittest you pass Engine.QueryInterface(SYSTEM_ENTITY, IID_TemplateManager).GetTemplate [16:40:09] <elexis> alternatively you could also delete the entire TemplateManager component [16:41:05] <elexis> also notice that still means nonvisual autostart needs a TemplateLoader, so the better approach is actually using if (Engine.GetTemplate) around that const template = Engine.GetTemplate("special/players/" + data.Code); block [16:41:06] <Freagarach> Well, it validates templates. [16:41:30] <elexis> Engine.GetTemplate doesnt? [16:41:38] <Freagarach> Nay. [16:42:01] <elexis> Gamesettings code doesnt use the civ data you inserted at all [16:42:11] <elexis> from what I see survival of the fittest doesnt either [16:42:26] <elexis> so these functions could just skip that Engine.GetTemplate call if Engine.GetTemplate isnt defined [16:42:44] <elexis> also it seems that this data should not used by simulation or gamesettings (nonvisual) in any case [16:43:20] <elexis> so thats a oneline JS fix instead of making a second template loader hidden inside a static local [16:43:51] <elexis> and he is going to introduce another local static template loader for the nonvisual autostart patch that is not needed [16:44:44] <elexis> so those are quick fixes, I dont know if its good to have that JS function combine the data of two files, or if it wouldnt be better to have it separated into nonvisual and visual data in two files. [16:45:04] <elexis> (and loaded by two different calls) [16:45:47] <elexis> (also dont know if it wouldnt be reasonable to just drop JSON at all instead of having XML here then sprinkling some JSON over there and switch back and forth all the time)
Tests can be expanded (by testing _something_ between 0 and Math.PI) but are okay for now.
X11 functions, including screenshots. :)
git diff --no-prefix -U9999 according to The Internet.
Also, please set the repository correctly when manually uploading the patch. That way our linter can tell you part of our coding conventions. :)
Tue, May 10
First of all, thanks for working on this. :)
(Sorry for spamming the diff. It was less work than expected.)
I'll try and rewrite the tests.
Mon, May 9
With the inline and D4605, I get no OOS with the rejointest.
The patch reads correct from a functional view.
The pass class name _was_ serialised? And the reason it worked was probably because the pass class name didn't change ove the course of a match. ^^
If you could fix/merge the convenience function please, then I can move that from this diff and go back to JS. ^^
Sun, May 8
Nope. Do notice the message was already not sent anymore. I only remove the declaration and the listener in cmpTrainer.
Sat, May 7
Purge obsolete message.
See discussion on IRC yesterday.
Fri, May 6
Looks good (in more ways than one :) ).
You'll need to update cmpUnitMotionFlying to also update the clearance in cmpObstruction.
I would like to see the ITA of a unit as a first fix. Than we can later discuss/establish whether the value is necessary at all.
No clue why, and D4593 seem okay.
This is your error:
Mon, May 2
#1 -> This leads to weird code, because:
- Maps don't care about non-JSON data
I'm not sure what you mean here, but grep -irn maps/ -e 'gettemplate' results in quite some hits. ^^
- GUI cares about JSON data and the 'Identity' part of the XML data, but only that part.
Also civ/team bonuses, available templates etc.?
- Sim mostly doesn't care about JSON data.
Working to ensure that. ;P
Back to non-generic solution.
Yeah, just curious to what the symptoms were. ^^
Sun, May 1
Basically back to https://code.wildfiregames.com/D4486?id=19699. ^^
What are the symptoms of this mistake?
Sat, Apr 30
[07:41:45] <elexis> does that patch impact mods writing to custom config files? I dont know if that is currently well supported
Fri, Apr 29
Change is good.
Reads correct (apart from the inlines).
I haven't tested it.
Thu, Apr 28
I guess a general order ought to be established?
Apr 26 2022
Apr 25 2022
But 0 != 10?
I know there are keyboards out there with more than 10 numbers, to be used specifically for control groups, how to handle one with a 10?
Apr 24 2022
Apr 21 2022
Apr 20 2022
Wed Apr 13 2022 [17:00:53] <elexis> RequirementsHelper.prototype.ChoicesSchema -1 tab whitespace [17:01:10] <elexis> "" could be string.Empty or so [17:01:27] <elexis> also +1tab there [17:01:34] <elexis> return "" + could be return \ I think [17:02:33] <elexis> Object.keys(template).length object construction, would be good to avoid [17:02:51] <Freagarach> There was something about string literals in schema, IIRC? [17:03:31] <elexis> not that I know of, youre constructing a string there not more not less [17:03:35] <elexis> if (requirementType === "All" && !this.AllRequirementsMet(requirement, cmpTechManager)) [17:03:35] <elexis> return false; [17:03:35] <elexis> if (requirementType === "Any" && !this.AnyRequirementsMet(requirement, cmpTechManager)) [17:03:35] <elexis> return false; [17:03:40] <elexis> can be combined using || [17:04:27] <elexis> (same performance as || and && use shortcircuiting) [17:04:28] <Freagarach> I meant using ` for schema strings. [17:04:36] <elexis> ah those, never used that [17:05:20] <Freagarach> https://trac.wildfiregames.com/ticket/3825
[17:56:35] <elexis> compartment mismatch is not a hard problem [17:57:01] <elexis> it means that in C++ a JS object is allocated in one script context, but then used by another script context without having it cloned [17:57:25] <elexis> so it might be possible to locate it by reading and some circumstancial evidence [17:57:42] <elexis> you probably find someone who can build in debug mode [17:58:20] <elexis> https://trac.wildfiregames.com/changeset/24116 [17:58:54] <elexis> https://trac.wildfiregames.com/changeset/15589 [17:59:13] <elexis> https://trac.wildfiregames.com/changeset/15561 [18:05:47] <elexis> also RegisterScriptFunctions_GUI, RegisterScriptFunctions_Simulation, RegisterScriptFunctions_Maps are identical, so can be deduplicated
Nope, it should be kept, else one can't order a single mangonel to move, IIRC.
Apr 19 2022
Do note this is wrong. I've noticed a long time ago that when you attack siege set to passive, interesting things happen to see by those who wait.
Apr 17 2022
Is this the forum thread you meant? https://wildfiregames.com/forum/topic/75691-trader-bonuses-international-vs-landsea-traders/
Saving some progress.
Apr 14 2022
You mean because they won't get a 404? IIRC that is the case with most of our libs. Also, being able to specify the zlib version is right now kind of useless, as there is really merely one option (latest) with the current URL. The patch allows easy changing.
Apr 12 2022
Apr 11 2022
Good that you also have the observer see this. We need to think about:
- Teams obtaining victory/getting defeated.
I guess a review? Since this patch touches C++ and I am bad at that. ^^