00:16 <+bridge> [ddnet] <redix> @fokkonaut there are 2 (possible) reasons for your map problem:
00:16 <+bridge> [ddnet] <redix> 1) in ddnet `GetDataSize` returns the actual size of the data, while in vanilla it's the (compressed) data size in the file
00:16 <+bridge> [ddnet] <redix> 2) 0.7 has tile-skip for better map compression. when the map is loaded, the tile layers are unpacked and the data in the datafile is replaced, but the size is not updated
00:18 <+bridge> [ddnet] <redix> https://github.com/teeworlds/teeworlds/commit/2503dfdaec272720d3cbac407407649b21f1f2e6
00:22 <+bridge> [ddnet] <redix> just remove the check. if the the map is broken, the client will probably already crash when unpacking the layer
00:27 <+bridge> [ddnet] <heinrich5991> don't remove the check
00:28 <+bridge> [ddnet] <heinrich5991> this can lead to a crash on a malformed entities layer, e.g. from another mod, when you switch to entities
01:05 <+bridge> [ddnet] <redix> for tilemap layers with version <= 3 keep the check but use `GetDataSize` from ddnet
01:05 <+bridge> [ddnet] <redix> for version 4 i don't really see a reason for keeping it
01:12 <+bridge> [ddnet] <heinrich5991> @redix to not read out of bounds
01:14 <+bridge> [ddnet] <redix> when unpacking the tile skip stuff it allocates new memory with the exact size anyway
01:15 <+bridge> [ddnet] <redix> but it might be a good idea to add a bounds check there =\
01:15 <+bridge> [ddnet] <heinrich5991> I see. that sounds like a reason 🙂
01:15 <+bridge> [ddnet] <heinrich5991> oh, is it an out of bounds write? 😦
01:15 <+bridge> [ddnet] <redix> nope just read
01:15 <+bridge> [ddnet] <heinrich5991> how does the unpacking work?
01:16 <+bridge> [ddnet] <heinrich5991> I wrote that. I should remember
01:16 <+bridge> [ddnet] <heinrich5991> ah no, I didn't
01:16 <+bridge> [ddnet] <redix> oy wrote the unpacking. linked the commit above
01:16 <+bridge> [ddnet] <heinrich5991> I only wrote the write code. phew
01:16 <+bridge> [ddnet] <heinrich5991> ah, that doesn't look exploitable
01:16 <+bridge> [ddnet] <heinrich5991> good
02:33 <+bridge> [ddnet] <louis> https://cdn.discordapp.com/attachments/293493549758939136/686748466285641759/unknown.png
02:38 <+bridge> [ddnet] <Oblikumquat> louis u little shit
02:38 <+bridge> [ddnet] <Pepe> :kek:
05:47 <+bridge> [ddnet] <fokkonaut> @heinrich5991 @redix I fixed GetDataSize in my client already, can one of you look into this again and make a pr?
09:40 <+bridge> [ddnet] <Learath2> Ah the demo bug, I definitely know which demo bug he is talking about
10:23 <+bridge> [ddnet] <fokkonaut> @Learath2 can you commit your command stuff you did in the branch to master?
10:23 <+bridge> [ddnet] <fokkonaut> would be cool
10:23 <+bridge> [ddnet] <fokkonaut> (ddnet7)
10:23 <+bridge> [ddnet] <Learath2> It 's not a great idea without the patch in master
10:23 <+bridge> [ddnet] <Learath2> I mean teeworlds/teeworlds master
10:24 <+bridge> [ddnet] <fokkonaut> it is fixed in master
10:24 <+bridge> [ddnet] <fokkonaut> Just no release yet
10:25 <+bridge> [ddnet] <fokkonaut> or do you mean the pr
10:25 <+bridge> [ddnet] <fokkonaut> you did
10:25 <+bridge> [ddnet] <fokkonaut> https://github.com/teeworlds/teeworlds/pull/2479
10:27 <+bridge> [ddnet] <fokkonaut> Does this pr + the commits from you in ddnet7 work together?
10:35 <+bridge> [ddnet] <Learath2> Yep, they belong together, otherwise the pr in ddnet7 sends argument format in the ddnet way and the client craps out
10:36 <+bridge> [ddnet] <fokkonaut> Ah yea, but there is some other commit in F-DDrace that fixes commands with latest master
10:36 <+bridge> [ddnet] <fokkonaut> i dont know which right now, but would like to add that to ddnet7
10:36 <+bridge> [ddnet] <fokkonaut> because F-Client users are not able to use commands without pressing esc first on dddnet7, but on FDdrace it works
10:38 <+bridge> [ddnet] <onby> u know the name f-client is already taken..
10:38 <+bridge> [ddnet] <onby> for a variety of reasons u should not name ur client after this well known bot client
10:38 <+bridge> [ddnet] <fokkonaut> this is for 0.7, for new players, it doesnt matter imo
10:43 <+bridge> [ddnet] <fokkonaut> @Learath2 can you merge all the commands.h commits to ddnrt7?
10:43 <+bridge> [ddnet] <fokkonaut> they are missing and required for master client to use commands
10:44 <+bridge> [ddnet] <Soreu> name it fokkolient instead :p
10:44 <+bridge> [ddnet] <fokkonaut> Mh, nah
10:44 <+bridge> [ddnet] <fokkonaut> I dont think its a big problem tbh
10:44 <+bridge> [ddnet] <fokkonaut> Its 0.6 <-> 0.7
10:45 <+bridge> [ddnet] <Soreu> @fokkonaut @ChillerDragon btw, the template is done in around 50%, so should be able to finish it today
10:45 <+bridge> [ddnet] <fokkonaut> Nice
10:45 <+bridge> [ddnet] <onby> if u expect all 0.6 players to stay in 0.6 forever sure
10:47 <+bridge> [ddnet] <fokkonaut> I dont care
10:48 <+bridge> [ddnet] <onby> kiddy answer but ok
10:50 <+bridge> [ddnet] <fokkonaut> @noby I dont care about 0.6 players and whether they move to 0.7. I just want to play in 0.7, develop my client, ddnet7 a bit and F-DDrace, to provide something for those who play 0.7
10:50 <+bridge> [ddnet] <onby> thats cool
10:50 <+bridge> [ddnet] <fokkonaut> Also, I personally like to go with vanilla
10:50 <+bridge> [ddnet] <onby> im not trying to stop u from doing that lol
10:51 <+bridge> [ddnet] <onby> i just think u might want to change the name
10:52 <+bridge> [ddnet] <Brokecdx-> first fix skins, they look utterly trash
10:52 <+bridge> [ddnet] <fokkonaut> I know there was this bot client before, but its not that I have a F-Client for 0.6, where confusion can occur, its 0.7 and most 0.7 playery never played 0.6, so I dont know where the problem is. Also F-Client by Fallen is not that spread
10:53 <+bridge> [ddnet] <onby> well if u dont care if ppl get confused i guess it doesnt matter
10:54 <+bridge> [ddnet] <onby> also i dont think thats true, plenty of players have moved from 0.6 to 0.7
12:02 <+bridge> [ddnet] <Learath2> @fokkonaut what do you mean by commands.h commits?
12:02 <+bridge> [ddnet] <Learath2> my pr should be adding those aswell
12:03 <+bridge> [ddnet] <fokkonaut> no, commands.h stuff is in vanilla master already, but not in ddnet7
12:04 <+bridge> [ddnet] <Learath2> but it should be a part of my pr in ddnet7
12:04 <+bridge> [ddnet] <Learath2> well not pr, branch
12:04 <+bridge> [ddnet] <fokkonaut> ah
12:05 <+bridge> [ddnet] <fokkonaut> but can you add them to master
12:06 <+bridge> [ddnet] <fokkonaut> all of them
12:06 <+bridge> [ddnet] <Learath2> I can get the entire thing in master, but then vanilla master wouldn't be able to see the commands properly
12:07 <+bridge> [ddnet] <fokkonaut> ? ofc
12:07 <+bridge> [ddnet] <fokkonaut> my client is Vanilla master and on fddrace it works
12:07 <+bridge> [ddnet] <fokkonaut> fddr has those
12:08 <+bridge> [ddnet] <Learath2> It would show all the commands, but the argument lists would break the console
12:08 <+bridge> [ddnet] <Learath2> so you shouldn't be able to use the commands registered by ddnet
12:09 <+bridge> [ddnet] <fokkonaut> ahh, yea
12:09 <+bridge> [ddnet] <fokkonaut> oy has to get active xd
12:27 <+bridge> [ddnet] <fokkonaut> well no
12:43 <+bridge> [ddnet] <heinrich5991> @fokkonaut no, I don't want to look into this. if you tell me `Size` I could maybe make an educated guess
12:45 <+bridge> [ddnet] <heinrich5991> nice that you're doing ddnet on 0.7 🙂
12:46 <+bridge> [ddnet] <jao> 🙂
12:49 <+bridge> [ddnet] <fokkonaut> @heinrich5991 :)
13:22 <+bridge> [ddnet] <Learath2> @fokkonaut I'll look into it for the reasonable price of 10 euros per hour and it's free if it takes less then an hour
13:27 <+bridge> [ddnet] <fokkonaut> xd
14:00 <+bridge> [ddnet] <Marius> > @fokkonaut I'll look into it for the reasonable price of 10 euros per hour and it's free if it takes less then an hour
14:00 <+bridge> [ddnet] <Marius> @Learath2 
14:00 <+bridge> [ddnet] <Marius> Then you will wait 10h free 100€ :troll:
14:03 <+bridge> [ddnet] <heinrich5991> oh no, you just found an exploit in every employment contract
14:07 <+bridge> [ddnet] <Learath2> Most hourly people I know actually have a minimum billable hours clause in their contract, so they bill you atleast X hours even if it takes less
15:04 <+bridge> [ddnet] <redix> @fokkonaut just add `|| pTMap->m_Version == 4` to the check. For tele, speedup, ... layers just leave it as it is. They cannot use tile skip anyway
15:04 <+bridge> [ddnet] <fokkonaut> ?
15:04 <+bridge> [ddnet] <fokkonaut> none of the layers renders
15:05 <+bridge> [ddnet] <redix> Wasn't it just on vanilla maps?
15:06 <+bridge> [ddnet] <fokkonaut> no, also on an aip gores, but not sure whether that was saved with zillywoods client
15:07 <+bridge> [ddnet] <fokkonaut> or vanilla
15:07 <+bridge> [ddnet] <fokkonaut> i will try whether zillywoods breaks this, or vanilla client
15:07 <+bridge> [ddnet] <fokkonaut> if yes, then i will tell chillerdragon to stop support his editor (he made an experimental editor for 0.7 in his client, uses some of ddnet layers)
15:14 <+bridge> [ddnet] <fokkonaut> ok so the bug happens with maps that are saved with vanilla clients
15:14 <+bridge> [ddnet] <fokkonaut> @redix
15:16 <+bridge> [ddnet] <redix> Yeah, because they use version 4 layers with tile skip
15:16 <+bridge> [ddnet] <redix> But these maps don't contain custom layers like tele and speedup
15:18 <+bridge> [ddnet] <fokkonaut> ah, 0.6 doesnt use tile skip?
15:19 <+bridge> [ddnet] <redix> Nope
15:20 <+bridge> [ddnet] <fokkonaut> okay, adding check for version >= 4 works :) thanks
15:20 <+bridge> [ddnet] <redix> 👍
15:20 <+bridge> [ddnet] <fokkonaut> So i dont need to add it for the other layers i guess? but would make sense in terms of keeping it the same
15:20 <+bridge> [ddnet] <fokkonaut> or if i start making a map editor
15:20 <+bridge> [ddnet] <fokkonaut> like, adding those layers
15:22 <+bridge> [ddnet] <redix> Well yes, you could add tile skip for the extended data. But this also requires serverside support. So the maps won't work on the current ddnet7 version
15:23 <+bridge> [ddnet] <fokkonaut> ok
15:23 <+bridge> [ddnet] <fokkonaut> thank you :)
15:23 <+bridge> [ddnet] <fokkonaut> i wont add it then, but would i need to do something in order to be able to save maps from my client for ddnet7?
15:28 <+bridge> [ddnet] <redix> If you want to support all the custom layers, you ofc need support for them in the editor :D
15:29 <+bridge> [ddnet] <redix> Same for other things like sounds, ...
15:29 <+bridge> [ddnet] <redix> Right now the client will just drop them when you load a nap from ddnet
16:36 <+bridge> [ddnet] <Learath2> I forgot about the layer extensions project 😦
17:28 <+bridge> [ddnet] <heinrich5991> @Learath2 not on vanilla, interesting 😛
17:28 <+bridge> [ddnet] <heinrich5991> @Learath2 how can I declare a function pointer to a function with C calling convention?
18:08 <+bridge> [ddnet] <fokkonaut> hey, in ddnet client the entities background color will be rendered more and more when cl_overlay_entities increases. at 100 ofc, it is fully rendered without transparency. in my client, it is too on 100, but before 100 it is not rendered partly, but the main background gets more and more transparent, until at 99 it is basically white and at 100 again the entities bg color applies. No idea why, looked through everything and cant find why,
18:56 <+bridge> [ddnet] <Learath2> @heinrich5991 do you know the signature of the function you are pointing to?
18:57 <+bridge> [ddnet] <heinrich5991> yes
18:57 <+bridge> [ddnet] <heinrich5991> it seems the only way to do it is to typedef the function pointer type in an `extern "C"` block :/
18:59 <+bridge> [ddnet] <Learath2> oh in C++?
19:00 <+bridge> [ddnet] <Learath2> yep you need the `extern "C"` but you don't need an entire block, maybe that'd make it look a little better
19:02 <+bridge> [ddnet] <Learath2> @heinrich5991 a `std::function` doesn't really care about calling convention, idk what you are doing but maybe you could abuse that
19:02 <+bridge> [ddnet] <heinrich5991> how do you not do a block?
19:02 <+bridge> [ddnet] <Learath2> `std::function` can come with overhead though, not sure if it HAS to have overhead
19:03 <+bridge> [ddnet] <heinrich5991> `extern "C" void (*m_pfnLog)(const char *pLine, void *pUserdata);` didn't work
19:03 <+bridge> [ddnet] <Learath2> `extern "C" typedef int (*fcnptr)(int)` should work
19:04 <+bridge> [ddnet] <Learath2> @heinrich5991 define didn't work?
19:04 <+bridge> [ddnet] <heinrich5991> ah with typedef it works
19:04 <+bridge> [ddnet] <heinrich5991> define?
19:05 <+bridge> [ddnet] <Learath2> As in what happens? Does it assume m_pfnLog is a function it should call with c++ calling conventions?
19:05 <+bridge> [ddnet] <heinrich5991> it gives a syntax error
19:07 <+bridge> [ddnet] <Learath2> Hmm, what happens when you don't have the extern "C" anyway?
19:07 <+bridge> [ddnet] <Learath2> Linker issues?
19:08 <+bridge> [ddnet] <heinrich5991> haven't tried
19:08 <+bridge> [ddnet] <heinrich5991> C/C++ don't necessarily share a calling convention (or do they?)
19:10 <+bridge> [ddnet] <heinrich5991> @Learath2 happen to know that? ^
19:12 <+bridge> [ddnet] <Learath2> I'm not sure how it applies to function pointers, but C++ mangles function names which is the only incompatibility iirc
19:12 <+bridge> [ddnet] <Learath2> I think you are fine without the extern "C"
19:12 <+bridge> [ddnet] <Learath2> Oh and operator overloading, that also creates an incompatibility (still due to naming issues)
19:13 <+bridge> [ddnet] <heinrich5991> but function pointers don't care about names
19:13 <+bridge> [ddnet] <Learath2> yeah, that's why I think you don't need the extern "C"
19:13 <+bridge> [ddnet] <heinrich5991> ok, will do without
22:53 <+bridge> [ddnet] <Ryozuki> ^ useless