Add a map check before loading in multiplayer

[akagemegane]

Because I am at the mercy of NukemNet and its convoluted GUI just to actively play the multiplayer at all combined with nobody fixing nor improving things while I'm trying to go through the trouble making my own so I can never use that launcher again, I might as well suggest this.

Perhaps add a check to prevent players from getting booted out of the game when starting a usermap because they don't have the map installed. Before somebody says that can be best managed using NukemNet, it is rough at handling maps to the point that I witnessed multiple players having to ALT+TAB out of the game just to download the maps manually. I know this sounds like I'm ranting but I wish somebody took NukemNet's code and rewrote the GUI in wxWidgets or something.

[akagemegane]

It's also more telling when I ended up using playit.gg (a reverse tunnel) to play NotBlood and playing usermaps is roughly the same way as it is in NukemNet.

[tmyqlfpir]

Hi thanks for the issue.

Is the author of NukemNet gone?

[akagemegane]

Hi thanks for the issue.

Is the author of NukemNet gone?

No problem.

As for the author, not exactly, the last time I communicated with him was on February 21st.

I think the thing here is that it's more that he rarely contributes to the project. The last major update was on October 22, 2025 (four months ago) but it took almost a year to get the major update done. I believe the main issue with NukemNet was that it was initially developed in a strange way that it screams for a refactor/overhaul.

While I do understand it is considered to be in a alpha stage, the structure behind it was somewhat questionable. When I went to suggest adding in checkmarks to enable game settings like frag limits and the like, the author implemented it but in a way that makes hosting a match more intimidating and somewhat discouraging. I feel like adding in tabs would make it more clean. But that's the thing, he didn't design the program with a convenient UI in mind but more so functionality. My experiences coming from previous network guiders like YANG, Meltdown, and even Classic Gaming Arena (Which has Blood support but it is DOS only due to the project's scope) a network guider's purpose is to get matches going easily, not convoluted and somewhat confusing to use. (Only hurdle with launchers like Meltdown and YANG was the port forwarding requirements to advertise a game but thats beside the point.)

Outside of the GUI, NukemNet has bundled a whole bunch of stuff for support of other types of games. While ipxwrapper makes sense for DirectPlay titles, it makes less sense to bundle a combination of ipxbox and a contemporary instance of DOSBox just to start an IPX server (if ipxbox didn't work) when NukemNet could just integrate the IPX tunneling code to it's own servers and their relay server software. That way it would be more clean and less prone to potential problems.

Meanwhile CGA (with the scope of being a launcher only meant for DOSBox titles) only bundles necessary fixes and other things for certain games. Like buildmfx for BUILD Engine titles and WAD patching tools to assist playing Doom II with DEHACKED patches with ease due to the fact CGA had a mod database (which allows users to upload custom content).

The big reason why I am a big fan of Classic Gaming Arena's interface is that it was designed in a way that it felt more integrated and somewhat easier to understand.

Not to mention it is easier to help guide people how to use it and understand what does what. NukemNet on the other hand, you have to basically figure out how Relay and STUN work which can be a handful if you are new and just wanting to quickly play a game or two with a friend. I had witnessed about three to four people especially in recent times feeling discouraged from using NukemNet due to the GUI and most recently had to play NotBlood manually with the help of playit.gg.

Sadly there isn't really any other viable alternatives and is part of the reasons why I suggested these features which were later turned down like map downloading, lobby browsers, port forwarding workarounds, etc. Had Meltdown not got labeled as malware because the developer was part of some stupid Duke Nukem 3D dick swinging drama, I would still be playing on there right now. Even Netplayseeker had potential (its like YANG but with more features) but was unfortunate that the development stopped since they believe they didn't need to continue when NukemNet exists.

But unfortunately this is the sad reality. We are simply stuck with a launcher that desperately needs a overhaul with other network guiders being literal ghost towns or are just avoidable due to petty drama and being controversial. It's probably gotten to a point that despite my "believe when I see it" mentality with Refreshed Supply, I am hoping the major update and onward is enough to be serviceable so I can just play on that completely (even leveraging the LAN functionality) instead of NukemNet (atleast until it gets an overhaul or somebody else makes a better network guider with NotBlood support but I partially doubt it)

For now I am having to tolerate it to get some multiplayer games going at all.

[tmyqlfpir]

While I am interested to help, there is a several hurdles.

1. Adding map uploading support will not be feature complete, specifically if certain maps use xmapedit features and require the INI and external files to also be shared - that poses too much of a security risk compared to sharing a single file
2. Additionally, maps themselves have zero sanity checks and a map could be made to crash and exploit users. Very unlikely but still a possibility - at most I could shove map downloading behind a opt in toggle
3. UPnP IGD is a possibility, but it isn't something as reliable or as easy as using a proper external server (via STUN) to bridge and connect players
4. Time. My time is shared doing paid bounties for other open source software and studying. I've been maintaining NotBlood for over 5 years now doing almost daily updates until 2026, and I don't have the same sense of time to put into refactoring the codebase, so I am trying to do requests that I know realistically possible with my limited time vs requests that would be better suited done upstream

I could maybe look into point 3, but I don't know the requirements or effort it'd take.

I'll investigate that this month, and maybe if I can get that to work I will try another shot at map uploading, although without xmapedit map support it seems like not really worth the time and security risk.

[akagemegane]

While I am interested to help, there is a several hurdles.

1. Adding map uploading support will not be feature complete, specifically if certain maps use xmapedit features and require the INI and external files to also be shared - that poses too much of a security risk compared to sharing a single file
2. Additionally, maps themselves have zero sanity checks and a map could be made to crash and exploit users. Very unlikely but still a possibility - at most I could shove map downloading behind a opt in toggle
3. UPnP IGD is a possibility, but it isn't something as reliable or as easy as using a proper external server (via STUN) to bridge and connect players
4. Time. My time is shared doing paid bounties for other open source software and studying. I've been maintaining NotBlood for over 5 years now doing almost daily updates until 2026, and I don't have the same sense of time to put into refactoring the codebase, so I am trying to do requests that I know realistically possible with my limited time vs requests that would be better suited done upstream

I could maybe look into point 3, but I don't know the requirements or effort it'd take.

I'll investigate that this month, and maybe if I can get that to work I will try another shot at map uploading, although without xmapedit map support it seems like not really worth the time and security risk.

Sounds good. Here are my responses as well.

1. I can see what you are implying. Based on my experiences, a majority of Bloodbath user maps I've been playing never made use of external files. Initially I used to be playing with .INI files (during the NBlood days) but only because it proved to be a hassle making sure everybody had the exact .MAP installed, let alone exiting it out just to try another map. As for XMAPEDIT features, I've never ran into a map that utilized them as most of the maps I've played were made before Blood started to get source ports. To me, only having individual map downloading support should be good enough since I felt like source port exclusive Bloodbath levels are uncommon and cooperative levels are best played in extracted episodes instead of it being a single .MAP file. Best to keep these limitations noted once implemented.
2. This was something that was brought up to me previously when I suggested map downloading. Disabling map downloading has been a option amongst other multiplayer games at the time. (Especially with games using id Tech 3 and has had a history of exploits) I think it would best to have a toggle or at least a warning message.
3. I recall port forwarding workarounds was something I started looking into since my ISP at the time put me under Carrier-Grade NAT (CGNAT) which cock blocked me from self-hosting. I do have access to two VPS servers (Linux and Windows respectively) to host dedicated servers but NotBlood wouldn't be viable without a headless solution. (The server-dedicated branch could fit the bill and I could try testing it to report what needs to be fixed). I do believe the IRC-based lobby browser could still work alongside external solutions like playit.gg. However it would have to be designed in a way that allows players to manually specify the IP address to advertise. Better yet, allowing the source port to specify an IRC server/channel could be beneficial in case one server goes down. In fact, I would love to start my own IRC server just for advertising NotBlood lobbies. Basically saying if STUN, Relay server or even the dedicated server branch is too much time to implement/improve then I believe bringing back the lobby browser while allowing the host to advertise a lobby using their playit.gg IP and port (manually even) on a specified IRC server would be atleast better than nothing.
4. I do understand your position with time, which is also understanding of why you've turned down some of my suggestions previously. Because it could be too much time having to refactor a lot of code, and this is something I am starting to learn as I get more into programming. Keep me updated when you begin investigating and I am open to test when needed.

[tmyqlfpir]

1. Then in that case I think a single map shared is good enough for 99% of use cases.
2. I have a responsibility to provide software that won't get my users ratted or hacked. There is zero sanity checks for map files and creating something that crashes is easy. If the engine itself was using a client/server model like quake, then the damage possible wouldn't be as severe as any exploit is limited to packet parsing - downloading and reading files opens a can of worms.
3. The IRC server browser backend assumed that the ports shown in the browser were open, so unless the player can get their router and ISP to play nice the server browser was a broken feature. The only time I've had port forwarding work is when an application like ARMA 3 uses UPnP IGD, which could be implemented assuming people have a router that supports it.
4. I have to be choosy with what I focus on, as I am donating what limited time I have. I don't like declining features, it sucks from a user POV because I've been there, but I cannot realistically say yes to every request.

All of this is compounded by how difficult it is to test networking features on a single computer.

So to reiterate:

• Attempt to get UPnP IGD API permission to port forward for routers
• If UPnP IGD API does work, return the IRC based browser and offer customized IRC server to connect to for longevity support
• Investigate and implement map streaming again

[akagemegane]

1. Then in that case I think a single map shared is good enough for 99% of use cases.
2. I have a responsibility to provide software that won't get my users ratted or hacked. There is zero sanity checks for map files and creating something that crashes is easy. If the engine itself was using a client/server model like quake, then the damage possible wouldn't be as severe as any exploit is limited to packet parsing - downloading and reading files opens a can of worms.
3. The IRC server browser backend assumed that the ports shown in the browser were open, so unless the player can get their router and ISP to play nice the server browser was a broken feature. The only time I've had port forwarding work is when an application like ARMA 3 uses UPnP IGD, which could be implemented assuming people have a router that supports it.
4. I have to be choosy with what I focus on, as I am donating what limited time I have. I don't like declining features, it sucks from a user POV because I've been there, but I cannot realistically say yes to every request.

All of this is compounded by how difficult it is to test networking features on a single computer.

So to reiterate:

• Attempt to get UPnP IGD API permission to port forward for routers
• If UPnP IGD API does work, return the IRC based browser and offer customized IRC server to connect to for longevity support
• Investigate and implement map streaming again

Sounds good, I do think having it as a toggle would be best then. This combined with adding a check to see if everybody has the map would be best to avoid having to rehost the game to get somebody back in.

UPnP IGD could work for some people, but not sure how it would handle those under CGNAT. From my experience I couldn't port forward anything unless I used solutions like playit.gg (a reverse proxy that worked since NotBlood was UDP) or a VPS. Though advertising servers onto the lobby system would benefit from having the ability to manually assign the IP address when hosting on a reverse proxy. (Unless one was able to successfully get their VPS to self-host a VPN with port forwarding like WireGuard but that's a pain in the ass to setup in my experience.)

For instance I would want to have people connect to the IP associated from playit.gg (most cases I ping the domains to get the actual IP address) instead of my ISP's public IP, thus the option to assign one would tell the client to connect to that instead.