Dear All,

It almost goes without saying but because what I'm about to say might sound a bit harsh I would like to make it totally clear that I am a huge, huge fan of LT, Josh, the dev logs - in fact every aspect.

However...

It seems to me that 2015 is racing up pretty quickly and Josh is being distracted by things that risk further delays to the release date - a few examples:

1. AI Testbed - I've watched Josh trying to shoot the bad guys and he can't - they seem good enough already!
2. LTSL - I'm no expert but I suspect he has taken it as far as he needs to for the TIME BEING
3. UI tinkering and polish

I've read pretty much all the dev logs and I hope I've not missed anything but there are still some major, major items that have barely been mentioned:

1. Colonies
2. Proper functionality for trading
3. Characters - e.g. police/diplomats/pirates
4. Missions
5. Battle scenarios
6. Space station/science vessel functionality
7. Law and order - has any consideration been given to what happens if you do something bad - e.g. killing civilians/smuggling etc.
8. Asteroid popping

I appreciate Josh tends to show specific things in the video updates and it may well be that the 8 points above are extremely well developed already. However if they are not there has to be a conscious effort by Josh to move away from the weird and wonderful ethereal projects such as AI that some people dedicate YEARS researching and developing and instead focus on nitty-gritty tangible content and gameplay.

I've worked on projects where you are coding for months and months and when the end gets near it can be difficult to maintain focus and get things over the finish line and that is when you need someone to keep you on track. Does Josh have someone who can project manage him?!

Anyway like I say I hope that doesn't sound too harsh - Josh has such a broad range of talents I can see why it must be tempting for him to adopt this scatter bomb approach to development. It's time to get surgical!

My garbled thoughts about the issue:

I'm as impatient as anyone, but I don't think Josh is wasting time here. As I understand, the framework is the most important thing you can develop, and if you want your game good, you need the engine polished to a mirror finish. His work on scripting and other systems is getting in place to simplify later content generation, which is little time, compared to what the AI and nodal system/LTSL etc take up. Most early access devs fall into a sort of a pit where they have to release something flashy to the hungry public weekly, and can't really work in the invisible, long-term stuff, let along restart the whole thing from the ground up like Josh did once! Have a little faith and you'll see the 'major' game aspects assemble in the last month, and get playtested and polished through beta.

The other point is that this is Josh's baby - a first big study project of his that got great support. He wants it to be his true Freelancer tribute, his dream game, and also learn a lot in the meanwhile. And he wants to share his excitement, his learning route, and his realized dream with us. Was LTP playable? Yes. But it would leave a bitter taste in everyone's mouths as a full game, knowing it's just not living up the potential. No rush, better make the most of it, it's a true one-time opportunity.

With all that said, I'm really anxious and curious to see where the game is going to end up, and simply cannot wait to make some input through beta! About the first project I can -truly- influence though feedback, I feel.

After reading the latest dev log, I get what you're saying.

Don't get me wrong, I trust Josh implicitly to deliver his dream game. I'm sure he has a plan, and an organised todo list that allows him to tackle the big problems first.
And while I'd love to be privy to this information, I understand Josh's need for secrecy and his need to honour his agreement with the EC.

But yes, I worry a bit as well. To me, combat AI seems like the icing on the cake. And I wonder why the focus is on that at this point in time. It's probably part of a big plan, or there's the off chance that Josh is intrigued/distracted by the technical challenge it poses, as your topic title suggests.

For september's update I'm hoping to see more gameplay systems in place, combined with an improved UI. Perhaps this gameplay emerges automatically from the procedural nature of LT. And perhaps the current tier in development requires the gameplay to be built from the top down, instead of from the bottom up.

It seams like you're worried about the lack of content in the game so far, but what you have to realize is that LTSL is the best thing that has happened to LT in terms of speeding up content creation. Similarly the AI is what is going to give life to the universe, and change this from one big simulation into an actual game.

Thursdays dev log addresses a few specific code things people were complaining about LTSL since day one, namely prefix vs infix operator notation. This small change is going to save a lot of brainach for anyone who ever opens a LTSL script.

Josh codes faster then 74 normal developers; I have absolute confidence in there being an awesome playable product jan '15. Whether he is able to give up his baby and let it out into the real world is yet to be seen.

I generally agree. I've always appreciated Josh taking the long view, but the recent "AI testbed" development is the first the purpose of which I haven't grokked immediately (and I've read every dev log). And you bring up a good point -- no footage of LTP has indicated any weakness of the combat AI so far, and I question the need for trying to converge on some ideal when what currently exists is quite workable.

Of course, the whole reason for the AI testbed is probably to test higher-level AI functions, but I somewhat doubt that any such testbed will be able to compensate for the current lack of an actual base game.

First, time is relative. Whether you think 2014 is flying by or dragging on doesn't really make any difference to the development, nor does it really say anything about it. It only matters how much time Josh feels he has, if he thinks he still has plenty of time, than I'd venture to say he does, If he thinks 2014 is flying past him, then I might be a bit more worried.

The AI testbed seems to me to be a rapid prototyping tool, doing for AI content what LTSL did for the code. Him spending a couple days on it doesn't really worry me, as he has already said this is not simply for combat AI, but will be about rapidly creating and testing all aspects of the AI, from combat to trading to politics. If he is still extensively working on it in October, it better be so comprehensive and simple that it's on par with the sims...anything else will seem like a bad use of time

I agree, LTSL seems pretty well developed, and doing too much more work on it save for optimization, seems a bit tangential to getting the game out, though I suspect Josh will be toying with it, making improvements and additional capabilities to it while making it more efficient and elegant for years to come. There will simply be a point, rather soon I suspect that he will have to say, this is good enough for LT 1.0, and only a major obstacle will bring me back here until after release.

I can't really comment on the UI until I see what he has done with it. No idea whether it is worth the time or not.

1. Colonies

Colonies are functionally like stations, I suspect that until the planetary ownership addition comes out after release, they will be pretty much stations with primarily aesthetic differences, perhaps there will be some additional mechanics. I suspect that since the after release updates have no official ETA, he can take his time with them, and considering all the fancy tools he built himself, the planetary ownership update will be quite extensive and make planets really something amazing.

2. Proper functionality for trading

I do believe trading is mostly complete, and it was mainly UI work that needed doing

3. Characters - e.g. police/diplomats/pirates

I dont actually think there will be set characters... In the same way that you can become a barista or a doctor or a diplomat or a soldier, I think the AI will simply take up jobs/missions, and as they focus in on a select few mission types, their "character" will emerge.

4. Missions

Given that the AI can now understand the concept of missions, its a matter of getting the ai to understand the types of labor they can delegate... this may be as simple as adding a few lines to each script which defines a given behavior, which tag it as the kind of thing which can be delegated.

5. Battle scenarios

This seems to be what the AI testbed is for. I dont think josh is creating a handful of scripted scenarios, nor would I want him to, I am hoping he is creating AI which can actually understand any situation they are faced with.

6. Space station/science vessel functionality

Seems fairly functional to me, pop a module into a station/ship's hardpoint and it gains that functionality, there something i'm missing?

7. Law and order - has any consideration been given to what happens if you do something bad - e.g. killing civilians/smuggling etc.

Absolutely agreed, I haven't heard nearly enough on this. We have given suggestions, but it is still way too mysterious for my liking. I am however of the mind that Josh would rather build up the AI as individuals first before setting in stone how they would work together. When you have intelligent AI, putting together the social contract, rules and bylaws, concept of jurisdiction and jurisprudence seem fairly tame. I will only really start to worry if November comes around and we still haven't heard anything on those big social/legal mechanics.

8. Asteroid popping
That could be in there during the beta and be left almost for last, as far as i'm concerned.

I don't see the problem.

1. AI work has barely begun. A testbed for AI (and content balancing!) is incredibly useful.
2. LTSL is going to run LT - pretty much - so calling it "good enough" means stopping LT development.
3. I see very little UI polish being done. Only the tools to build a UI.

*shrug*

edmozley wrote: 1. Colonies
2. Proper functionality for trading
3. Characters - e.g. police/diplomats/pirates
4. Missions
5. Battle scenarios
6. Space station/science vessel functionality
7. Law and order - has any consideration been given to what happens if you do something bad - e.g. killing civilians/smuggling etc.
8. Asteroid popping

1. without any information on what you mean with "he needs to fix colonies" at least i cant follow whats your problem
2. is already in, but that needs some UI poloishing, which you just forbid...
3. there are no characters, AI's do what they think is most profitable for them.
   If they get traders or pirates depends on initial conditions
4. are already implemented, just need a handful on AI script lines, like mysticia said, and some UI polish
5. another case of not enough information for me
6. already implemented, needs UI (but i've seen no construction using drones for now...)
7. gets reflected in your faction stances, if you attack someone, they like you less and act accordingly
8. pretty secondary problem imo...

Release date is also not 1.1.2015 but "early 2015" which is somewhere up to 31.3.2015, so josh has 6 further months of work without moving again.


Wording concerns is fine, saying "josh has to do what i say or he will fail"... less

Beta is due before the end of the year. Keep that in mind.

However, I think at the very least we ought to keep an open mind through September. Especially with this AI work.

I have to admit I am starting to wonder the same notion as the OP. Not because I doubt Josh or the current progress made. More so because LTSL seemed to be a sidestep that was taken to better equip Josh for content creation. Yet he's still working on it. Still polishing it. I'd would go as far as to say he's almost gotten sidetracked with his focus on LTSL. Now that being said, LTSL will help him considerably in doing everything he wants to do. However, I do believe the testbed functionality IS straight up sidetracking. Unless his AI was in serious need of improvement, which I cannot imagine being the case. I'm really not sure what the point of it is other than to generate healthy statistics to further enhance/improve various aspects of the game. Which is great, don't get me wrong. Imo though, you shouldn't start building a glorified statistics generator for a game that isn't even finished yet. That should be a separate thing that you do to enhance. Not to build.

Now maybe the AI he's designed thus far is in need of some sprucing up, I don't know. I am starting to grow slightly concerned that his philosophical approach to the project is potentially getting in the way of physical results. I really hope he is focused and he doesn't get too sidetracked. He still has plenty of time to plow through the content creation aspect. Keep in mind also that being a procedural game, most of the content will be procedural and functional. So I'm really hoping his testbed focus is basically him running the various elements of the game through testing to see if they are up to the standard he wants. Which would then possibly mean that the game is almost ready for beta. We can hope!

TGS wrote:I do believe the testbed functionality IS straight up sidetracking. Unless his AI was in serious need of improvement, which I cannot imagine being the case. I'm really not sure what the point of it is other than to generate healthy statistics to further enhance/improve various aspects of the game.

Umm, the AI IS the game. Without the AI LT would just be a pretty system simulator. So I do believe the AI needs all the work being put into it. At the end of the day, Josh is the best judge of whether the AI needs improvement or not. Remember that you'll be playing this game for hundreds of hours. The AI needs sufficient complexity and variety to not make those hours boring.

As for the combat, not being able to destroy enemies is not necessarily a sign of AI proficiency, it can be due to anything ranging from overpowered enemies, missing UI elements to the controls needing polish.

Multiple people have already suggested the importance of LTSL, so I'll move past it.

UI tinkering and polish- One of the reasons X Rebirth was heavily criticized was because of it's confusing UI. The UI is your window into the game, and for a game as complex and varied as LT, a bad UI could ruin a fantastic game.

Cornflakes has already responded to the other perceived problems. As for asteroid popping, it is a solved issue. It just wasn't implemented in the previous update. The imposter tech required for the solution is already implemented.

I think the major reason for posts of this nature is that instead of posting crowd-pleasing devlogs and working on features to show-off in the updates Josh has started digging into the necessary technical work required to complete the game and it is the lack of understanding regarding the work that is making most of us nervous.

Skepticism is healthy though and I think both the community and Josh can benefit from that.

Vartul wrote: Umm, the AI IS the game. Without the AI LT would just be a pretty system simulator. So I do believe the AI needs all the work being put into it. At the end of the day, Josh is the best judge of whether the AI needs improvement or not. Remember that you'll be playing this game for hundreds of hours. The AI needs sufficient complexity and variety to not make those hours boring.

Obviously lol. However we already know that AI does exist in the game. Not even terrible AI. From what I understood with regards to the testbed aspect is that this essentially opens up a means of having dynamic range to the AI. Allowing for a ranking system both for the AI as well as the player based on skill. This is ultimately why I see it as sidetracking. Does the game need a ranking system? Does the AI need improvement to the point of offering a wide range of skill levels? Does this functionality not already exist? See what I mean? AI exists. It isn't even terrible. Thus this is ultimately an enhancement to the AI that already exists. Yes I agree that the AI should be good. But I am simply concerned that such things are side projects that aren't essential. I'll put it in a slightly different way. A few months ago it was looking like Josh was starting to wrap up the "theory" side of the game (E.g design vs implementation) and was about ready to dig into content additions/creation. But he seems to be pulling more theory out of what appears to be no where and going on philosophical tangents. While very interesting and amusing to behold, it's very easy to lose focus whilst doing this. He could end up creating a considerable amount more unnecessary work for himself through these deep thinking sessions.

I'm not sure about the rest of you, but personally I'd like to see the game released next year. I hate to use the term feature-creep because I don't believe that is a/the concern at all. However becoming unfocused can seriously reduce the amount of practical time he has to crunch whilst still meeting his target(s). He still has plenty of time so I'm not too worried.

Vartul wrote:As for the combat, not being able to destroy enemies is not necessarily a sign of AI proficiency, it can be due to anything ranging from overpowered enemies, missing UI elements to the controls needing polish.

Again all I can say is that there is already AI. I just worry about the amount of time he might spend on it. That being said, the testbeds can serve functionality on a much larger scope than simply AI so it's all good in that respect.

Vartul wrote: UI tinkering and polish- One of the reasons X Rebirth was heavily criticized was because of it's confusing UI. The UI is your window into the game, and for a game as complex and varied as LT, a bad UI could ruin a fantastic game.

Funny you should mention this game, one of the main reasons X:R sucked so bad was reiteration. They simply kept redoing things they'd already done. They kept wasting time. The first person elements were only tacked on in the last year of development. Of a 7 year project. There was a thread outlining it all on the egosoft forums at one stage. It's easy for developers to never finish something because they are consistently reworking and retooling things and adding insignificant but costly (dev wise) features/content.

Vartul wrote: Cornflakes has already responded to the other perceived problems. As for asteroid popping, it is a solved issue. It just wasn't implemented in the previous update. The imposter tech required for the solution is already implemented.

I actually mentioned this in another thread and on IRC that I believe the game is in a lot better shape than we realize simply because we haven't seen it in it's entirety. So things might be fine, Josh might have all the time in the world to play around with theory and go on philosophical theory binges. We simply do not know. I hope that is the case, or it might come to bite him in the rear later.

Vartul wrote: I think the major reason for posts of this nature is that instead of posting crowd-pleasing devlogs and working on features to show-off in the updates Josh has started digging into the necessary technical work required to complete the game and it is the lack of understanding regarding the work that is making most of us nervous.

Skepticism is healthy though and I think both the community and Josh can benefit from that.

I would agree with this for the most part. At the end of the day though with regards to dev logs. Sometimes he goes on technical oriented posts which some people love, some people hate. Sometimes he gives us a good breakdown of what he's done but there simply isn't a great deal to tell. Then at times he goes on these philosophical fluff ridden logs. Some seem to love them, but it naturally makes others nervous. Personally I get nervous whenever he seems to be going back to theory because theory with regards to game development to me can seem like incoherence. It's the thought process that you use to basically figure out what you wanna do and it is fairly uncommon to see that stuff so late in the process. Unless he isn't that late in the process afterall, which then begs the question... are the current targets realistic? Usually by now the entire game plan would be mapped out and deviations on design would only occur on the small scale unless some significant breakthrough came to one of the team members that somehow improved a random component, feature or direction. Josh seems to have had a large number of these breakthroughs. Such breakthroughs can seriously slow down the entire process.

Good AI is absolutely essential to a single-player game. I think he's okay with working on it, so long as running it doesn't cut into his coding time.

IMO, most the games that you perceive as failed due to feature creep actually failed because of absence of programmer skills / time. Realizing a good idea requires quite a lot of engine foundation work. After that, it should be quite easy to do.

The logic in points 1-8 for the stuff you mention wouldn't be too difficult on a fairly good engine anyhow. Starting now seems *way* excessive. It'll probably be good enough to start fleshing them out beyond some engine test cases in the last 1-2 months. They probably do not need to be anywhere as sophisticated as the engine's code for world generation, UI, AI (economy and fighting behaviour) to be fun, anyways.

Funny you should mention this game, one of the main reasons X:R sucked so bad was reiteration. They simply kept redoing things they'd already done. They kept wasting time. The first person elements were only tacked on in the last year of development. Of a 7 year project. There was a thread outlining it all on the egosoft forums at one stage. It's easy for developers to never finish something because they are consistently reworking and retooling things and adding insignificant but costly (dev wise) features/content.

How much of that seven year period was spent trying to get rebirth to work on consoles? Most of the stuff in that game seems to be slapped on as afterthoughts, and designed more to give the impression that the game was designed purely for PC, which it obviously wasn't. It was a money grab.

They tried to pull a fast one. I don't see Josh doing that. I see Josh making a game because it's been his dream and because he loves these types of games. He is trying to make a fun game, for himself and anyone else who likes these kind of games. His goal isn't to, "appeal to the greater console market."