Tonight's Screenshot: LOS and Sensors and lit Science Bot.

[x4000]

Hey all, just a quick screen cap I thought you'd enjoy.

This one has Blue's new science bot lighting and animations up there at the top of the screen, and I think she outdid herself.  You can also see a couple of the new weapon icons over there on the left; Ray did a great job on those.

There's also a view of our new "shadowcasting" implementation which lets us do Line Of Sight that is more permissive and less restrictive.  IE, it behaves more like you'd expect.

We also changed things up so that your line of sight and your sensor range are two different things.  Well, they always were, but we changed how they are graphically represented and what that means for who can attack who when.  Enemies can only fire on you when they are in your line of sight now, as with a lot of roguelikes, and they are immediately alerted at that time, too.  But your sensor range extends beyond that, and you can then see the enemies in a sort of glowing yellow diagram form.  If those enemies are alerted (such as by you seeing them and then running away, or by your blowing your whistle), then you'll see those little yellow schematics turn and come for you, then emerge into full color when they hit your LOS.  It's a neat effect.

The LOS rule changes came about mainly because of the addition of the "flavor objects," a number of which you can see scattered around.  These objects provide destructible cover as well as blocking line of sight and in general making the level look more visually interesting.  The problem was, with our old restrictive LOS rules (that were, worse, invisible unless you held down shift to see your targeting overlays), it got pretty confusing.  "Why can't I attack him!?"  Etc.  It wasn't a dire problem, but it was a midlevel annoyance or more.  This new view is more attractive, makes more sense, and avoids all those issues completely.  I'm really happy with it!

Props to Cath on the flavor object art, by the way.

[mrhanman]

The artwork seems to improve with each new image you share with us!

About the LOS - does it immediately widen at the end of the single tile tunnel to 3 tiles?  shouldn't it stay one tile for a few tiles out of the tunnel, and then gradually widen to 3, maybe continue to widen the further it goes?  Or am I looking at it wrong?

[x4000]

The artwork seems to improve with each new image you share with us!

Thanks!

About the LOS - does it immediately widen at the end of the single tile tunnel to 3 tiles?  shouldn't it stay one tile for a few tiles out of the tunnel, and then gradually widen to 3, maybe continue to widen the further it goes?  Or am I looking at it wrong?

Not my algorithm at all.  This is a generalized algorithm called shadowcasting, implemented by a very respectable fellow here: http://blogs.msdn.com/b/ericlippert/archive/2011/12/12/shadowcasting-in-c-part-one.aspx  There is a little silverlight game showing it that you can use in the browser.

There is also more generalized information about multiple permissive los algorithms here: http://roguebasin.roguelikedevelopment.org/index.php?title=Permissive_Field_of_View

TLDR: Basically, due to the coarseness of the grid, you will see artifacts.  Various other algorithms can have more or fewer artifacts, but none are perfect and some are way more computationally expensive to run.  We've chosen one that has relatively few artifacts and is pretty darn fast.  And which had working code in C# that we could use.

[mrhanman]

Oh, now I see (I think)!  The view into the room widens as you approach the exit of the tunnel.  Based on Mr. Lippert's blog post, beyond the limit of the screenshot, the view should widen, yes?

Will there be the ability to zoom in and out on the board, or is the viewport size fixed?

[Aklyon]

And which had working code in C# that we could use.

The most important of parts for people who make updates as fast as you guys do, I'm guessing.

[x4000]

Oh, now I see (I think)!  The view into the room widens as you approach the exit of the tunnel.  Based on Mr. Lippert's blog post, beyond the limit of the screenshot, the view should widen, yes?

Oh yes, absolutely.

Will there be the ability to zoom in and out on the board, or is the viewport size fixed?

It works exactly like the viewport in Skyward, since it actually is the viewport from skyward.  Which was basically the viewport from AI War.  So not only can you zoom, you can grab with the middle mouse button, pan with the arrow keys or with edge scrolling, etc.  Only when you move does it snap your view back to your exo.  I'm actually constantly zooming in and out because all the fighting takes place at range, unlike most roguelikes.

And which had working code in C# that we could use.

The most important of parts for people who make updates as fast as you guys do, I'm guessing.

Not always, but it is certainly nice.  Lots of the algorithms in our games were either completely original, or in the case of well-studied stuff like roguelikes, implmented from pseudocode or just approach descriptions.  But this one in particular was sufficiently complicated that, yes, the preexisting code was a huge factor.  It was lucky that it was already the approach I wanted to take anyway, based on its characteristics.

[Tridus]

Science bot looks great.

Now what does it do? 

[Misery]

Science bot looks great.

Now what does it do? 

Blows things up with SCIENCE!!!! I'm sure.   What the heck else could science be for?


Well, this is looking ever more impressive.   You guys are doing the alpha a bit after PAX is done, yes?  Or am I making that up?   Either way, I look forward to it.

[x4000]

Science bot looks great.

Now what does it do? 

Blows things up with SCIENCE!!!! I'm sure.   What the heck else could science be for?

It can fight kind of okay, but is the weakest fighter.  It has a lot of slots for specialized stuff like hacking, viruses, lock picking, and so on though.  So it basically lets you have a wide array of abilities at a middling level, or a few incredibly-adept abilities, at your discretion.

Well, this is looking ever more impressive.   You guys are doing the alpha a bit after PAX is done, yes?  Or am I making that up?   Either way, I look forward to it.

Yep!  Probably within 2 weeks anyway for the private alpha.

[Teal_Blue]

I hafta go back and read all the posts, but just wanted to say the art work on the shot you posted was FANTASTIC!!!!!!     Really, really nice!!  (all the nice details, the nodes on the walls, the different sizes of pipes in the pic, the grid of the walkways, the shadows on the overheads, the brightness on the lit side of the light door,   The piping, and the other bots, deactivated? laying on the ground and the new view, the Sensor view? Really, really nice!! 

(Just a quick thought,  Maybe sometimes the bot can be on top of the walkway, like it is here and the sensor shows the bad bot on the same level, but maybe 'sometimes'? You could have our bot under neath? On the walkway below? And the bad guy on a different level? So they see you sometimes, but have to run to some elevator somewhere to come down to the level to get at you as maybe the grid flooring ricochets most of the bullets around trying to shoot through them to get at you.)     
 
Really looking forward to this alot, maybe i can get in on the alpha test and then purchase when you go 1.0? I'm not exactly great with analytical feedback and sometimes my opinions are a bit obtuse.    But i don't want a free game, just a chance to play it with all the rest of you and then pay for it when it goes 'gold', is that what they call it?    Gold?

Some really nice games lately, I mean always was, just... well... i think you know what i mean, really really cool games. Glad they're doing well, hoping Bionic will too! 

Go knock their socks off guys!!!!     haha


p.s2.  Sorry, but just thought that maybe the trainers, you know like Meg and the others? Maybe have like, I don't know... maybe 4 or 5 or 6 of them? Well, maybe they can bring different strengths or approaches to the teams they support, you know? Like i don't know, leaning more on stealth for one team at a cost of firepower, or another team with more firepower up front on the bots, but maybe they are more fragile, less armor? That kind of stuff, just a crazy idea, and you probably got this covered already, but it just came to me when i was checking out your screenshot and imagined my bots creeping down the dark lit... corridor, the squeak of metal somewhere in the distance that the bots ignore, but the 'operator'? Like Meg back at station listening in hears something she thinks is... Then screams through the microphone as she realizes its a heavy caliber cannon loading sound that is echoing off the metal walls where the bots are moving... The Lead bot just gets her audio, when the whole front of the hallway EXPLODES!!! Into fragmented metal sharp debris and smoke and a concussion wave that picks up one of the bots and throws them against the wall! Damaging his right rifle arm, leaving him only with a grenade arm on the left to fill in... 

The thing with the Trainers, you know? I really think that could be cool, not because of me, but because it makes the bots different from loadout to loadout depending on a mission type, but also because of players styles of play, some are more aggressive and rush in kinds of people, others are slow paced and lets go in cautious and take it easy and see what happens.    At least that is my opinion. 

-Teal

[nas1m]

Well, this is looking ever more impressive.   You guys are doing the alpha a bit after PAX is done, yes?  Or am I making that up?   Either way, I look forward to it.

Oh man. PAX is when? I want to play this - desperately .

[Teal_Blue]

Ooops, one more thing!!    What if the load out picture that shows your bots and their exquipment and slots and stuff also showed the operator, or the trainer? Like Meg? So the player would have kind of a visual in game of the trainer they chose, i mean what if i always chose Meg, because she is like cautious and stealthy and that is the way i play, but one time, or maybe a few times i choose Mac, because he's old and got the knowledge about the tunnels in the mission better than the others (his range on the sensors is maybe a third more than the other trainers)  

And one time i use Randy, because even though Randy is rather mouthy and rude alot, he is up front and in your face and guns blazing when he is on a mission and i really needed that to get through that last mission over on the east side there were so many bad bots in that place i thought i was gonna freak!! 

Anyway, Trainers and visuals on the menu on the side and different tactics for each of them in their loadouts for the bot teams they put together.   

Am hoping this is more than just an idle wish kind of thing, but go get 'em, hope the game does super well!

-Teal

[x4000]

Whew, lots of suggestions there -- bear in mind those are really hard to implement for one reason or another, or don't match the general design.  But thanks!

[Teal_Blue]

Just a visual of what i mean by the trainers, and even if your not going to do like the trainer specialties, well... i wish you could, but its not my call. Anyway, thanks for listening. 

-Teal

[x4000]

Cheers; the exo driver you choose doesn't affect anything except a single special ability that is usually non-combat related and definitely always passive.  We kind of need that extra slot for abilities that you may equip!