Comments - Hanfling

Thats actually not true. Smirftsch do gives out headers on request, they are just not publically open, so you plain have to ask him.

Theoretically speaking, this requires the same data/graph as it would do randomly generate what happened, and I'm still somehwere at missiong 2 stuck or 3 stuck with random generation code.

The question would be how one could represent such a thing, maybe some graph description language and generate some ui or random code from there.

Yes, downloads are currently always found there, anouncements are atm done over my patreon page: Patreon.com or on dxhq discord.

I should some day post some status update for moddb, but short story the last years was sth. like: I kinda discontinued working on HX, but wanted to do coop later as part of another project, but gave out some private builds, and things kinda escalated from there slightly, so I started fixing up some bugs and I'm continuing development on it.

\heartbeat\7791\gamename\hx

Epic actually included the source code for the game launcher in the unrealpubsrc224v and UnrealGoldPubSrc source releases for modding.

So there is no need for starting to copy the original executable as one can just compile an own version based on source Epic gave out specific for this purpose.

Also, it might be different for the Unreal Steam release, but otherwise the games won't work or worked correctly with the original executable anymore for various reasons.

Freezing rain.

Though there are no plans currently on having coop on Revision, several things (especially rewrite of parts of the code) I did/will do for HX are -- though sometimes for a bit different reasons -- interesting for Revision, and thus they are likely to end up in the same or similar form in Revision anyway. This includes making the game in overall easier to modify by using mutators, gameinfos etc. (instead of modifiying the DeusEx.u file which nearly all DX Mods do -- but at least not HX), a rewritten ingame windowing toolkit, rewritten rendering, etc.).

Not dead. :)

I'm working for some time on an OpenGLDrv for Unreal Engine 1, however I just ran a quick & dirty test for using some sort of cartoon like effect in Nerf Arena Blast: Coding.hanfling.de and imho such an approach could work fairly well for Hatchet, so contact me if you have some interest in it.

I neither have a perma death in nor does the player gets random loot.

In my opinion one common misconception is that singleplayer gameplay translates as is to coop gameplay. Coop gameplay is always at least a bit different compared to singleplayer and what works for singleplayer does not need to work particular well for multiplayer and the other way around.

I'm aware of the fact that try-fail-reload is a core component of singleplayer Deus Ex gameplay. But in multiplayer this easily ends up beeing pretty annyoing. Especially as coop gameplay so far seems to end up faster and more action orientated. This is nothing choosen by design, it plain ends up easily with two players that way. So making often a break for reloading basically does not fit to how the gameplay gets. In fact I really do enjoy the lack of the try and reload gameplay component as this yields a permanent game flow. And basically every time I play Deus Ex in singleplayer I end up doing the exact the same things as I did the previous time I played it, so I end up getting the same game every time I play it. In HX so far it was always different and I did like the fact that the game unfolds differently. Keep in mind that Deus Ex was designed that you can basically do anything without making it impossible to progress (there are a few exceptions though).

And I know how to implement to save a game. I save the map every time you leave, so you can revisit it later. However there are a few caveeats regarding saving a game in multiplayer. Every time you save you break the network, so you can only save during map transitions. So for saving a game one would need to at least reload the map, maybe even need to load another map in between. The next question would be how to deal with the player positions. Easiest solution would be to reset the player positions to the start of the map. As Unreal Engine is an open multiplayer system where you can leave and enter the game at any time, how to deal with additional players if you later reload the savegame? What happens when a player reenters the game 10 minutes later, and that part of the map got inaccessible? (and maybe even imposible to leave?) This can yield up tons of problems regarding how the maps and game was designed and the number of possible issues preventing you from progressing in the game are not containable. So in the best case for a free save system you end up getting an annoying one.

However I percieve Deus Ex in coop different then singleplayer. It is best described as a mix of a coop like in Unreal, Rune, Quake II with the with the ability to mess with tons of stuff as in Deus Ex singleplayer -- well except that you are not alone messing with stuff anymore. I don't perceive it as annoying that we can't retry if we didn't get something right, but other basic things: If you start later on in the game, you never had the NanoKeys for Unatco or never visited Smuggler. What I did like in turn was that we were not overequiped when starting at a later mission, but instead need to figure out first how to get things done with what we just found, which impose really nice little challenges instead of whacking everything up with the dragon tooth, which actually kills the gameplay after Hong Kong. Without the Dragon Tooth playing the later missions is plain more interessting and fitting more the gameplay of the first half of the game.

However regardless of a resume or a save system in coop you can just load any mission(/map) you would like to play and making the start there as pleasing as possible is important.

However I agree that it would be nice to continue playing with the progress you made the last time you played with a fried, thats why I already wrote in the comments of the news post that I will add a resume a mission start feature which would at least address this issue.

But in any case. If something plain does not work out I will change it. But for now my guts tell me that my way to go will do work out better. But, it's frustrating that some people basically just straight ahead assume that how I do things will break the game ultimately, without ever having played HX ignoring the fact that one could just do a mod for HX changing a few aspects to match their opinion how it should be done.

However I guess it is best to continue the discussion about how this should be done after a first release of HX as this will give you the experience about how Deus Ex feels in coop.

I doubt that I'll be adding a vending machine. :D

Well basically Deus Ex uses just a few different meshes for all the characters. And for some of them, like the NSF Trooper player animations do exist. However as the NSF Trooper mesh is used by a couple of characters so those should be easy to integrate without any changes needed to animations, etc. So it breaks down for just making a player class for each of the characters one could get that way. And thats nothing more as just setting some properties for the class.

No, HX will never work with Shifter Mod. It basically bypasses all Deus Ex classes which shifter modifies and by the way Shifter was done as a modification of base game files there is no easy way of porting Shifter over and one would basically need to rewrite shifter from scratch.

However as for the extra character models, I already noticed that there is a great deal of shared meshes which would enable the use for many of the meshes for player characters and in fact it would be no great deal to add them to HX. I even consider adding the player classes for conveniance, to allow mods for HX to make use of them (basically just allow them in the Gameinfo) or making the allowed character models just a setting in the ini files.

However, my descission for core HX was to just use the JC model with it's five skins, as this is less irritating and fits better.

But in any case, it would be no big deal to add extra character models for everyone wanting to use them.

Please don't shoot the bartender.

Without the randomisation when you start at a later mission point you would only get the default (e.g. nothing happened before), which is pretty annoying because you have no nanokeys, each time Anna is alive, etc. and doesn't feel particular right.

However as most people seems to be uncomfortable about a lacking save feature -- which is not a feature coop games usually offer -- I'll reconsider adding a resume at mission start and some (unsupported) commands for saving/loading the game. However, keep in mind that saving in mp really sucks. For network reasons you can just do this during at map change, so saving the map would always involve a map change and as player count can vary, etc. you would be put to spawnpoints at map start/where you entered the map each time you save the map.

We simply have more women, drugs and sportcars then Romero.

"When it's done."

However, it'll not be in the next weeks, but it'll neither take a year. One (soft) deadline I have set for myself is to release within half a year after Revision release.

A rough (incomplete) overview of what I want to finish before first release is:
- Rewrite Datalink playback system.
- Finish the previous missions randomization feature.
- Fix Rebreather.
- Some (sparse) additional items at the start of certain missions, to make it more conveniant to start there (You won't believe how glad you can be about a combat knife!)
- Lots of minor improvements/fixes.
- Rework of weapons effects system (including fixing the spider bot attack arcs).
- Further network optimisations.
- Make notes persist after map changes.
- Fix chopper is sometimes invisible at the end of some maps/missions (however it reappears when revisiting the map, but still annoying)
- Fixing issues for multiple players (like the missing elevator buttons in Oceanlab were before).
- Code cleanups.
- Fixing smoke for flares and incestburners.
- Implementing a (partially) heuristic approach for visibility determination for replication (should fix some remaining issues with decorations not showing up which are partially occluded by the map).
- Clients can't access main window, after server was shutdown.
- Added distance checks for display of conversations other players have (It's a bit weired if they do show up when the other player is on the other end of the map).
- Better handling if players leaves the game when he is still dead, during conversations, etc.
- Doing sth. about game no immediatly ends for listen servers local client (using the versalife backentrence is basically jumping into death)
- etc.

What I might do before release:
- Refactor Inventory pickup code.
- Rewrite parts of current mutator setup to make it easier to hook in custom classes for easier to support other singleplayer mods/hdtp support and/or gameplay modifications.
- Cooperative notes editing features
- Fix the (vanilla) bug where you get locked in air when using the drone when in air.
- Custom artwork for HX (i already have some sketches).
- Fix mission start text display for clients
- Chat messages starting with ; do not work (like ;D)
- More code cleanups.
- Fixing friendly NPCs become hostile towards you after suicide.
- SpawnPoint packages instead of hardcoded ones.
- Fixing the sometimes odd accuracy display of weapons.

What I want to work on after first release:
- Further work on easier support of 3rd party singleplayer campaigns/mutators and mods for HX.
- Major work on AI.
- Major work/balancing difficulties.
- Maybe moving the replacement Actors for vanilla DeusEx to another package.
- Readding corpse carrying (there are no 3rd person animations for it in)
- Improve 3rd person view on things like the fire extinguisher, etc.
- etc.

I'm afraid there is no easy answer to that question. If I would need to break it down to a single reason, I would answer: gut feeling.

To give a more comprehensive question I need to step a few years back, well about one and a half decades to be precise, when a friend and I just got our hands on the Deus Ex demo and we tried to make it work in mp. Iirc after struggling to renable the console we managed to join the game, but it did crash or at least did not really work. Probably soon after the DeusEx SDK came out I started to try to make the game work in Coop with occasional rewrites from scratch every year or so, mostly focusing on another aspect of the game I had not yet succeded to address in a previous rewrite. I can remember having a fairly large codebase around 2005, but all of the rewrites in the end suffered from not beeing able to save in multiplayer (especially on a dedicated server). Despite that there might have been breaks for two or three years where I didn't try working on a Coop mod for Deus Ex the idea of writing a Coop mod for Deus Ex never let the grasp on me and it somehow writing a proper Deus Ex coop mod became more and more a goal I need to achieve in my lifetime.

So for the current rewrite I started in late 2012, I focused first on getting the saving/loading and map transitions to work and almost cried after I finally got that working (despite that it still took ages to get it working stable), but until recent the main development focus remained on larger or technically difficult features and not on making HX feasonable to play a soon as possible, although I fixed plenty of smaller issues (like Faucets not working) when I got sick of figuring out stuff which did not make sense. It even took over a year until I fixed (the rather simple bug) that prevented NPCs from shooting in MP.

So during that time of development there were rarely builds which would have been reasonable to release, because they wouldn't have been more like a bit of feature demonstration which likely would have probably be fun for a few hours, but not anything which would have been reasonable to play yet.

Recently the situation is different. Builds about 2-3 months ago were like 90% complete (in terms of what I want for a first release), and now they are like 95% complete. Despit that it's just a small fraction, gameplay feels times better now, and I'm sure that the remaining 5% work will still be a significant improvement and although it might be a bit selfish, I want a reasonable first release for HX. Although the situation is slightly more favorable for mods then games, the first release of a game or mod is nearly always the most important release by which it is judged and remembered, and it feels like a waste to release HX after all this long time too early, when just a minor fraction of polish is left which will make a difference.

You can play it with more then two players and I actually put at least four spawnpoints for each map transition in, but my impression is that the singleplayer storyline is best played with two players and that five players would be clearly too much. However the situation might change with custom maps designed for a higher player count.

My impression was more that they used that thing for development (and not considered putting it ingame).

So how did you do it? Load some metaformat and decode it using unrealscript?

I'm afraid to tell you we had to change that to the Dragon Bone and make it out of some kind of rubber compound. It's non leathel now, but will allow you to render most targets unconscious with a single blow.

Jonas contacted me in late 2008 asking if they could use my launcher for TNM because it fixed the issues caused by the advent of speedstepping and multicore hardware, so I sent him back a version U used for HX during that time, which also used an own User.ini, Default.ini and DefUser.ini set, which apparently ended up in first TNM release. :)

Thanks for bringing up the NanoSword point, I have not thought about that yet. Do you know if it shows up at some later points? I just knew about a WIB in 06_HongKong_MJ12Lab and MaggieChow in 06_HongKong_Storage, so one can assume that for 2 Player both could have a NanoSword at that point. One matching location to put another one in later in the game would be 14_OceanLab_UC. However, after having the NanoSword the game tends to degenerate to smashing everything with the NanoSword, which is not that great in particular. An good idea might be to make later NanoSwords no to easy to get like giving them to MIB/WIB.

You could have saved you the screenshoting by just exporting the DeusExText package (there is a tool for it around). And indeed you need to add add some ServerPackages or things go weired regarding FNames. But you won't need to add the audio only files.

Currently i have:
ServerPackages=DeusExConversations
ServerPackages=DeusExText
ServerPackages=DeusExConText
ServerPackages=DeusEx
ServerPackages=DeusExConAudioAIBarks
ServerPackages=HX

Maybe these snippets are of some use for you regarding DeusExTextParser and Conversation handling:
Nopaste.waigie.de

Your major misconception is that singleplayer concepts translates exactly to multiplayer, but Deus Ex plays, like each game, way different in coop. As usual the focus shifts away from story to a more aggressive/offsensive gameplay. It's basically the same happening in Unreal, where you barely used the Translater in coop, while you could spent quite some time using it in singleplayer. Also players tend to play more offensive/aggressive compared to singleplayer.

So it's a noticable different feeling comparted to singleplayer, but it does feel great as it is, so why shouldn't I focus on things which makes it great in coop? Also there is no such thing as a permadeath -- players respawn like it's used to be in multiplayer coop games so no permadeath can ever happen (despite that Deus Ex in coop is probably not even half of the 25 hours as you stated, and there is one only larger mission - Hong Kong). And personally I never felt it as any kind of penalty to start the coop at some mission start. I select my prefered augs and aug ups, setup the server and go ahead. After a few minutes of playing I have again tons of equipment.

Thanks for the offer, but your GMDX singleplayer only code is of little use to me.

Maybe another idea would be to influence the chance how the outcomes are random generated by what the players did the last times they played it. E.g. if you kept Paul alive, the next time you start a new game at a later mission Paul would have a higher chance to be alive.

As you rarely load a savegame which is not the latest one, the Splinter Cell system in the ends breaks down to being the nearly the same as an unlimited system.

The progress you make would still be your own for the parts you played, but will be random for the parts you didn't play.

But I agree that it would be nice to have the progress one made earlier, but beeing able to reload would break what works great in HX: You have to deal with what happened.

Even right now it works great to start the game at some later missions while the only thing beeing slightly annoying is that you have no NanoKeys at the start.

As for TextParser, it has been a long time and i can't remember having serious troubles with it.
For conversations, i use a serverside playback, for datalinks a clientside along a serverside datalink dryrunner handling the logic. However one of the clientside issues was that the DeusExLevelInfo got removed and thus the clients lost the ability to check which ConversationPackage to use. Binding conversations does work in multiplayer and is not strictly tied to having the DeusExLevelInfo around. However I moved most of the converation code off Player to GameInfo and use replicated functions for displaying next conversation screen or bouncing back a selection choice to the server. One thing worth to mention is that just the BindName matters when binding conversations to an Actor. Thus my GameInfo has JCDenton as BindName. Another thing which is kinda annoying that you might get the wrong player for the events in ConPlay and need to account for it. For lipsynch you basically need to know that lipsynch is done whenever the pawn is playing a sound and bIsTalking is true. Also you need to add some remap system to be able to stop the conversation sounds on clients when you skip over sections, etc. So much for the basics. Technically the conversation system was one of the rather easy parts.

Let it spill over into the schools and churches, let the bodies
pile up in the streets. In the end, they'll beg us to release it.

Actually caused by me because I had two instances of the game running at the same time, so they playback the sound with a small delay which sounds like reverb.