Half Life 2: Destructive Interference

[Welsh Mullet]

Right. *cough* I am making a mod. This mod has a page on the valve wiki http://developer.valvesoftware.com/wiki/Half_Life_2:_Destructive_Interference so most of the info about it so far is there.

I have a few questions i need answering, and the talk page is not the best way by the looks of it.

1. If i wish to include single and multiplayer, which base should i start from, or should i start from scratch?
2. If i start with a multiplayer base, will the player's character model show in game, and yet the client still use a view model? How can i do this in the single player base?
3. Anyone seen any British based mods/maps? (I have seen the currently in progress 28 Greggs later).

[Gradius]

The multiplayer and single player code bases are seperately, so you'd have to modify the multiplayer base to include single player functionality or make two different mods for the single player and multiplayer portions.

Oh a subjective note: I really think you're biting way more off then you can chew. It doesnt sound at all like you're experienced with coding (atleast with Source), yet you want to implement. I'd really strip off vague and not really gameplay influencing changes like "physics water" and "improved mirrors". There's no need for them. If you really want to make this happen I'd advise you make the multiplayer portion first and whittle down the mechanics you want and then after you have that down make the single player (I suppose you could also do it the other way around, either way don't make them both at the same time, it isn't a great idea).

[Noodles]

I very much like the idea of the mod, but you need to take a step back, as like Gradius said you do appear to be biting off more than you can chew, especially if you don't have an awesome kick-ass coder.

Personally I'd suggest Single Player first, as the community generally swarm single player mods with /me wants, while generally mp mods are 'oh yet another'. It'll allow you to develop your story and environment as well.

Start off by getting your story and environment sorted out, and detailed. If you can, get a concept artist. Then you can start on the mapping of the storyline. Work out exactly what entities you'll want your coder to create and how they'll work, probably a good idea to start with some custom weapons, then maybe new NPCs.

I'd really suggest leaving some of that stuff behind, like the mirrors and water, not even sure if the water is possible in Source and the mirrors sound like they'd require a lot of work and wouldn't be very optimal on low-end systems.

Good luck, I've wanted to do a UK based HL2 mod myself, though I'd base it in the country not London, thinking about it London would be raped by the Combine, if there was any organised resistance they'd move out to the country where they have greater security. (I was going to use Longleat forest and the overgrown remains of Center Parcs as a resistance base myself).

[Welsh Mullet]

Well i don't plan on having this finished for a few years yet, as I was kinda planning on using to practice coding some stuff as i do my degree. So by the end of 3/4 years I should hopefully be a kick-ass coder :3
Most of the things I've read on doing co-op etc, use the multiplayer base and turn max players to 1, but this seems a really lazy and dirty way to do it.
And as for the London not countryside thing, I was only thinking of starting in London. maybe even only for a level or 2, or the climax level. The other 2 settings I'm now thinking of using are a disused nuclear bunker in England, or Dover, making it some kind of either transport centre to the mainland or an air-works plant thing (like in the hl2 beta).
And yes, some of the things may be ridiculous to code, but I was going to at least have a look into them. Valve may even fix them first in the Portal 2 engine or the Ep3 one. And I did consider the low end system problem, but since valve is now dropping these systems too (no more dx8) I thought it would be a small problem.

[Major Banter]

You're better to work from a sound base than jumping in the deep end.

Choose a single area to concentrate on; SP or MP, and work on in-game code.

As you get more experienced, implementing MP becomes much easier, and more importantly smoother - you'll know how to handle bugs.

[Shrinker]

Regarding your Wiki user page: A jump-start into C++ by making a mod?

It took me 8 years and countless projects to learn how to use that language decently. Read this and then reconsider whether you really want to make a mod, or play around with something simpler first.

[Terr]

While I think having a project is an excellent way to self-teach yourself something, a big unique mod is over-planning. All you're doing it setting yourself up to feel bad when you complete 25% of it and lose interest/time. Instead, what if you could do the same amount of work and being able to happily list a bunch of discrete accomplishments?

My suggestion is to break each item you've listed there into separate projects, pick one you can do as a self-contained unit, and do that. For example, you might make a new player model and put it up for download, or post a code snippet that alters reflections.

[zombie@computer]

learning c++ from the sourcesdk is a bad, bad, extremely bad thing. First learn at least the basics, otherwise you will be learning valvec++ instead of c++. You wont be making any friends with that.

[Welsh Mullet]

Well yea, i need to learn C++. My only experience is with VB.net so far, but that's OO so it's a start. C++ is one of the major topics of my first year of study, and the SDK is a practical use of C++ I can look into.
But a mod seems as good a project as any to learn the skills of the industry, and i don't exactly know what kind of job i want to aim for in 4 years yet.
But i have some good ideas for the story (creative and story writing being my only strong point in English) And i'm going to start storyboarding them soon, so i'll probably get them up. However, any comments on things people would like to see once i get started will be taken into account. I know i'm jumping in deep, but my course of study is deeper. I'm using this mod (and maybe this forum) as a kind of floatation device.

[Terr]

learning c++ from the sourcesdk is a bad, bad, extremely bad thing.

Yeah, I had hoped to do that for some of the gameinfo.txt/initialization code that mods go through to make a DLL, but ended up reimplementing stuff directly in Java instead.

I think the other problem has to do with how there are areas which seem to be deliberately closed-source and undocumented, leading to extra pain.

[Shrinker]

WM: Are you even reading our criticism? You are grossly underestimating the effort and frustration involved in programming a mod for HL2 with C++. That task is definitely not suited to teach yourself the language.

[Welsh Mullet]

But i'm not going to learn for the source code. What i'm saying is what i learn in C++ in University i can attempt to put into practice in this mod. Over 4-5 years.

[Terr]

I still think you ought to try breaking down the kinds of stuff you want to learn into smaller discrete pieces. The problem with a single big mod is that you don't really get any payoff until >90% of the work is done. Better to do a bunch of small projects with the option to re-use the code at a later time. This also means you don't have one aspect with "lingering bugs" blocking you from working on another part, like a codebase which is unplayable due to a fancy 3D graphics experiment that needs to be fixed before you can test AI in it.

[Welsh Mullet]

I might just try and work on the character customisation system i had thought up to start with. From my idea, it seems like it will be similar to hats in TF2, and really only requires the transmission of a couple of bytes for each player. So i might start with that :3

[Welsh Mullet]

Right, don't know if i should start another thread for help on this, but here is the basic idea of my customisation system.

You spilt the players body into more than one model, all attached to a "skeleton" of bones (so you can animate etc.)

So far the body parts are:
Head
--------Face, back, neck, hat

Torso
-------attachment 1,attachment 2,attachment 3

Arms
-------attachment 1,attachment 2,attachment 3

Legs
-------attachment 1,attachment 2,attachment 3

Each part model can also be told to keep it's colour, or tint itself (ala G-mod) with either the players chosen primary colour or their chosen secondary colour.

These colour are selected from a table of 16 colours, meaning only 4 bits are used for each colour.

This colour choice can be over ridden by the server if the game being played is a team game, either changing just the primary colour to the team colour (red, blue, green or yellow) or primary and secondary colours.

All this data will just be tagged onto the client data that is already sent to the server when you join a game.