peter_spy

Yeah, a good animator or character artist is very hard to find

They do look like crap anyway, just because of how much time have passed. But that's okay, you did what you could, with knowledge and tools that were available to you back then. Does that mean though, that any efforts to bring better assets to the table should be discouraged?

That's really awesome news. It will be interesting to test that

If I see correctly, fog lights are still treated as omni lights, and they mess up the lighting preview. The workaround in previous versions was to put such light on a separate layer and make it invisible. It doesn't seem to work on my end anymore with 3.0.0

Have you tested it performance-wise? Last time I checked, heat haze was getting more resource hungry the more pixels in player viewport it used. So sticking your nose to the glass and looking at another big heat haze surface, like a body of water or something, could be really expensive.

It's nice to see that gamedev is finally adopting common software development practices, adding this one to my fav GDC vids list

First off, at the creation stage it's just "let's create great art", or "let's think of ways to express ourselves and communicate our ideas". GDC presentations are a byproduct of that work. Then, at some point, the PR department steps in and talks about high sales. But historically, games in the Thief era had much more leeway in that regard. So framing it like it was a modern AAA title is misleading. The question of hobby vs. professional framed the above is something that I've seen several times already, and TTLG and here would be the only places in the world that I've seen people frame it like that. There is no hard line between hobbyist and a pro, as your post implies. People do pick up creative hobbies because of pros in the first place. Like you pick up an instrument, because you heard something recorded by pros. And you imitate them, then you try to find your own style, etc. You learn whatever your curiosity and taste leads you to, and as much as you are willing to spend your free time on this. But no one says: "ooh, I'm not learning that, this is for the pros only". "I'm not learning it right now, because I have to work on my technique first" – sure. In the end, the difference between hobbyists and pros is mostly that the latter earn money for it. Not how far someone goes in their learning or skills. That is perhaps one thing we can agree on. Also don't want to continue this, because to me it seems like a discussion for sake of discussion, with walls of text, and not in order to get somewhere interesting. @Dragofer you forgot to move this last post from the conversation to the gear thread.

That people who came up with the principles I'm talking about got to these conclusions while designing their successful games, Thief included, which TDM is based on. Using this knowledge is useful and leads to creating better maps, whether in Thief series or similar games, so again, TDM included. So not knowing or not being interested in what made the game TDM is based on great is nothing to be proud of. And no, referring to knowledge or experts in subject-matter is not running out of arguments, unless you're this type of a person who questions everything, and has your own "philosophy" on everything. Then I'm definitely outta here Not really, you even bolded out that part: "creating situations in which not using your gear is one of the worse possibilities". And that's a purely theoretical situation, it won't hold up during the execution. The rest, i.e. "stealth games are dead" is just a defensive distraction and watering down the discussion; next up will be "what is a stealth game", "what is reality anyway" etc. Again, not my kind of jam

1. Yes, most interactions with the environment mean noise, which can lead to failure. If you want to incentivize using tools and experimenting with guards and environment, why did you choose Thief gameplay and implemented stealth score? 2. Not really, if you balance it out. You can create more paths, e.g. a shorter one, but lit by a lantern that goes out for a while, every now and then. Or another shorter path behind guards back, but with a loud surface. This way player has a risk-reward kind of choice (get somewhere faster, but have to creep carefully behind AIs back, or time their move through a temporarily unlit area). Again, it's easier and faster to discuss these things when you know the contents of that presentation.

Well, there is no rulebook, but there are e.g. ex-LGS devs presentations floating on the web, this was basically a quote from Randy Smith's presentation called "Level building and stealth gameplay", but there are also others. Since these are done by professional developers, and confirmed by many successful games under their belt, then yeah, their expertise and credibility is much much stronger than of anyone here, sorry You can see this approach working in their games, I can see it working in my WIPs as well. And, you know, I'd rather get the fundamentals first, and maybe then try to play the rebel and make up my own rules. Like in any other discipline, really. I'd suggest reading that presentation first, because you're kinda discussing its points taken out of context, which doesn't make much sense.

That scenario is in direct conflict with one of the fundamental stealth design principles, where in every situation you should have at least one "zero-failure path". It's not that it cannot be contested in any way by AI (looked at, crossed by patrolling, etc.), and it might (or even should) be the longest way to go from A to B, but it has to be there. Using noisemakers, water arrows, or killing guards is always an option. But it's also a play style choice some might not like. Even using blackjack as one of reusable tools might not be preferable by a portion of players. Not to mention that you can fail using any of these means, and you'll be punished by swift death, in most cases. The fundamental difference between stealth games and titles like Quake is that in the latter, developers expect you to partially fail, so they scatter around medkits. In stealth games, you don't expect player to fail (typically).

I wonder what practical example of that would be, since most player tools (arrows) are finite, and you always risk player running out of them (whether by accident or not), before arriving to a place where they'll have to use them. Even if you trap the player in a room with variety of arrow types they have to use to get out of it, you always risk a situation where they could make a mistake, and permanently lock themselves in. The only infinitely reusable tools are blackjack, sword, lockpicks, and... lantern?

It seems to be doing that, actually. It seems like AI can go from level 4 or 5 to 0, it just plays the "agitated search" barks until it reaches Alert 0 on level 0, and it never puts away its weapon since. I thought the very slow cooldown for the 1.5-0 range might be there to add a little variation if player manages to alert the same AI second time in a row, in a time between alerts shorter than 3 minutes. Anyway, thanks for digging that up, this is interesting

Thanks! I wonder how this one has gone unnoticed for so long (since 1.7, apparently ), or maybe it hasn't been thoroughly tested in the first place? Anyway, a somewhat related question: was there a specific model you went after for alertness going down? It seems like it's different for some alert levels, e.g. it goes down very fast for "suspicious" and "observant", while it's very slow for "idle" (0-1.5): around 0.01 units/s, which seems like a very long cooldown.

IMO it's overthinking. When I have a toggle function, I press a key once, I don't hold it, so it works correctly. I only noticed it because this thread pointed it out. To me this change is fairly irrelevant. And given how much it breaks, I'd think it's not worth the effort.

When you add absence_noticeability 1 to a loot item in DR, and you steal it in-game, the AI alert level immediately jumps up to 18.1, which is "agitated searching". If you add the absence_alert spawnarg, it does nothing, regardless of its value. Not sure what was the overall intention here, but the description of the absence_alert is incorrect then:

Fantastic work Epi, as always!

The problem is, that leads to incessant clickfest, because you might not remember which category given keybind belongs to. A clear-looking scrollbar will always indicate that there are more options available. Moving keybinds to the center of the screen and giving them alternate column for secondary bind could also be an improvement.

By the way, that stuff with so many Settings' subsections has to go, if we want to talk about good UX. One continuous scrolling list per category, and there options can be grouped by subcategories. The order of main categories needs to be changed too, from most to least often changed settings (not alphabetical or any other).

Arrow crystals are a very Thief 3 idea. Even in T3Ed they were called crystals, and the main point was that those were more like a manifestation of elements, and Garrett could make use of them (perhaps due to his ex-keeper training?). Adding shafts and some kind of crafting wouldn't give quality time to players, more like a ubisoft-like busywork.

Looks fun!

Arkane games' menus are usually a kind of UI experiments, and they're usually bad in terms of user experience (Dishonored 2 and Deathloop, I'm looking at you). And while typically boring and generic, games from companies like Ubisoft or Square Enix have a lot of research going into UX and accessibility, so all the AssCreeds, Far Crys and Tomb Raiders might be worth analyzing in that regard.

It sure does, as everything from button placement and spacing to options order and layout feels clunky and unintuitive, in comparison to other games. The best approach would be to have someone with UX and interface design background take a look at all options and design new user flows and layout. The second best option would be to take a look at a few modern games (because at least UI design and accessibility is what they do very well) and try to establish common rules they use, and then apply them while redesigning TDM menus.