slipher commented Dec 7, 2021

I don't get any problem now with binding char:< (and tutorial on). Possibly fixed by d2aaff7 which turned on escaping in more places. To future rea…

slipher commented Dec 7, 2021

You did the TODO so delete it :)

slipher commented Dec 7, 2021

The nth item in the list is the name of the "plus" command for the nth button. So it explains where the buttons are defined.

slipher commented Dec 7, 2021

Just get rid of the move assignment operator and I'd be fine with it

bmorel commented Dec 6, 2021

Ok. Well, I replaced it with a range anyway, which makes more sense (as long as NUM_BUILDABLE is really the last one). Or there's also a specific a…

slipher commented Dec 6, 2021

LGTM except the submodule should be updated

slipher commented Dec 6, 2021

assert not equal

bmorel commented Dec 6, 2021

I'm not even sure what this comment is really meant to explain.

bmorel commented Dec 6, 2021

I'm not saying I understand what IBE thingy is, but I probably grepped and found this line: #define SB_BUILDABLE_TO_IBE(x) ( ( ( x ) & SB_BUILDABL…

bmorel commented Dec 6, 2021

Now that I look anew at it, yes, this would make more sense. Looks like buildableInfo is really a "struct" with both the buildable_t and some flags…

bmorel commented Dec 6, 2021

What this does is silence a warning which is emitted a lot. It does not alter the code's behavior, since what was the standard before C++11 was tha…

slipher commented Dec 6, 2021

Links! Unvanquished/Unvanquished#1617 UnvanquishedAssets/res-buildables_src.dpkdir#7 I suppose it's smart to move it here, so that we can use the n…

slipher commented Dec 6, 2021

Why is it called mask? It's not like a set of bit flags

slipher commented Dec 6, 2021

The assertion should be on buildable not buildableInfo, right? Also ASSERT_NQ

slipher commented Dec 6, 2021

I think I was wrong about the semantics of selector nodes. Actually they go and retry every action from the beginning every frame. So the nth node …

slipher commented Dec 6, 2021

Looks like I was wrong about that extra paks thing. Although the command is enqueued before the end of the previous map, the sgame entirely restart…

slipher commented Dec 6, 2021

I don't understand the intended semantics of the example. if a bot is further than 300, it will always go back if a bot is closer than 100, it wil…

slipher commented Dec 6, 2021

What's wrong with it?

slipher commented Dec 6, 2021

Mention the new location?

slipher commented Dec 6, 2021

Actually BUTTON_WALKING is used directly by pmove to control speed. And there isn't anything anywhere called MOVE_RUN. So better to remove this com…

slipher commented Dec 6, 2021

Well I should say, it will probably use the copy constructor of StringBase. So it won't exactly do a wrong thing, but it certainly won't do a useful …

slipher commented Dec 6, 2021

StringBase has raw pointer members, and does not define a move constructor. So this seems unlikely to do the right thing. In general I don't sugges…

bmorel commented Dec 5, 2021

Note that I never said I'm against the idea of showing complete view of map, I was just pointing at some inherent flaws of them for most unvanquish…

illwieckz commented Dec 5, 2021

About displaying the whole minimap, we can start to generate them in higher resolution. Maybe 4K×4K?

illwieckz commented Dec 5, 2021

The game had mechanism for layered minimaps but it may be broken because I remember hangar28 author had to blend them in one with different color p…

necessarily-equal commented Dec 5, 2021

@Gireen: in your example, if the distance is greater than 300, the bot will go to the overmind, while in mine, the bot will go to the medipad, so i…

Gireen commented Dec 5, 2021

It is at least better than nothing even with the current resolution and only in 2d. If i remember right the minimap can also have multiple layers, …

Gireen commented Dec 5, 2021

it was set in default.cfg for maps so it should have worked. And it also did if one tries to build a message appears that its not possible. So on t…

necessarily-equal commented Dec 5, 2021

One question left to answer is: should we reset the hysteresis for a bot when the action failed or succeeded? I would tend to say yes, but this is …

necessarily-equal commented Dec 5, 2021

(ditto)

necessarily-equal commented Dec 5, 2021

I find these confusing, how about changing the two lines to BTN_TALK = BUTTON_TALK, // = 1 (redefined from daemon), disables actions

Gireen commented Dec 5, 2021

Whats the difference to a normal condition? selector { condition ( distanceTo( E_H_MEDISTAT ) > 100 && distanceTo( E_H_MEDISTAT ) < 300 ) { action …

necessarily-equal commented Dec 5, 2021

(nitpick) Typically one adds a comma to help limiting future diffs BTN_RALLY = 14, // also named "taunt" in some places

necessarily-equal commented Dec 5, 2021

This is still partly lowercase

bmorel commented Dec 4, 2021

I had a quick read at the "purpose" of those in engine. To me, only BUTTON_TALK seems to be useful. BUTTON_ANY is never read at all, and BUTTON_WAL…

bmorel commented Dec 4, 2021

The only ones engine actually use are TALK/ANY/WALKING. It could very likely be fixed, I'm not sure. Functions which uses those symbols only writes…

slipher commented Dec 4, 2021

I was thinking, for ones that Daemon doesn't define a specific purpose for, there could be definitions like BUTTON_USR5 etc. which the gamelogic co…

slipher commented Dec 4, 2021

LGTM

bmorel commented Dec 4, 2021

@illwieckz : done.

illwieckz commented Dec 4, 2021

Just replace yellowbuild.png with optimized https://dl.illwieckz.net/j/yellowbuild.png and everything would be fine for me.

slipher commented Dec 4, 2021

BPs are already checked. There's a dollar bill icon when there is not enough. Search BuildableAvailability for source code

illwieckz commented Dec 4, 2021

@slipher I may be wrong, but I have the feeling new stuff are added with camel case (and 99% may be legacy anyway).

illwieckz commented Dec 4, 2021

ℹ️ Tip: If you make a list of issue numbers this way: - #1604 - #1613 (the warning) - #1591 GitHub displays the title issues, this way: #1604 #1613…

slipher commented Dec 3, 2021

I suppose localtime() doesn't work in NaCl

slipher commented Dec 3, 2021

Sometimes buildings disabled with g_disabledBuildables are displayed possible to build for players. The value of the cvar is reset after each gam…

slipher commented Dec 2, 2021

Probably the opening files for reading should be split into two syscalls: One that only looks in paks. One that looks first in game/, then in paks…

slipher commented Dec 1, 2021

Technically speaking, you can load navmeshes from the game/ directory... but only if there aren't any in the paks! FS_Game_FOpenFileByMode should …

slipher commented Dec 1, 2021

I prefer to merge this at the last minute before release Why would that be needed? If it is really needed, then the right thing to do is put it on …

illwieckz commented Dec 1, 2021

I prefer to merge this at the last minute before release, maybe we should create a special tag for that?

slipher commented Dec 1, 2021

Pinging this since the PRs were approved. Won't merge it myself since I don't understand the inter-repo dependencies here.

slipher commented Dec 1, 2021

I don't think there's any reason for this to be fixed-size. It's no worse than the current code though, so I'm not demanding to rewrite.

slipher commented Dec 1, 2021

Yes it is true. Foo a[3]; is no different from Foo a0; Foo a1; Foo a2; in terms of destruction.

slipher commented Dec 1, 2021

So the placement new seems to crash nacl-clang. You can try parentheses instead of curly braces there - I think that worked to stop MSVC crashes in…

slipher commented Dec 1, 2021

LGTM

slipher commented Nov 30, 2021

LGTM

necessarily-equal commented Nov 29, 2021

I've merged the first 3 commits manually so this can be about nitpicking the spectator thing. I still think the current code is fine :)

necessarily-equal commented Nov 29, 2021

well, that'll be for another day

bmorel commented Nov 28, 2021

reminds me I need to do a PR to enable Recast logs. I had the same error, and fixed it by increasing some hard-coded value, as suggested by Recast'…

necessarily-equal commented Nov 27, 2021

Fixed