next »

[Elideb]

Well, hello again.

For the last couple of months we've been way too busy, apparently. We're sorry we forgot posting around here how the game has been progressing. But fret not. Things go forward at a constant (never as fast as you'd want it to) pace, and some great things have happened. We'll delay the official announcements until a bit later, so this post remains focused on the game and we have an excuse to post soon (I hope daily, or as close as sanity allows).

In a general note, both Juan and I have taken part-time, not too lengthy, side jobs to make Heart&Slash's development a little less harsh on our wallets. That'll help us and the company in the future, so we'll have an easier time producing games for some more time.

So, Heart&Slash.

Since the tutorial introduced a little bit of story, we've focused on giving a sense of progression to the experience. We knew that was a weak link in how the game felt, so we decided to tackle it in a couple of ways.

The first one was the story itself. We'll drop clues about what happened here and there, so players can put a part of the puzzle together. The tutorial is an important part, but we'll have dialogues, decorations and sequences for this as well.

Then, we got quests into the game. They will give players something to do, some times regarding the story, others simply giving them something to aim for. These quests will unlock equipment, game modes, characters and the multi-run adventure of Heart & Slash we were talking about since the Kickstarter. The quests for the factory level are pretty much in place, with a few issues still to fix. In the coming weeks we'll integrate quests in level 2 (and make some of the improvements it desperately needs).

And since you can now unlock equipment, we've added, of course, more of it. Mostly new weapons, for the time being, with the next batch bringing in new kinds (grenades, ninja stars  ...). Then we'll implement some new body parts for Heart.
We want all the work our artists dedicated to H&S to be admired, so we've created menus displaying equipment and enemies (quests, too) the player has found. Each has its description, which could be tips or a short history lesson.

No Star Wars were harmed in the development of these blades


Up close, they seem nice bots

And then, the third and final level in the game is close to being finished. We're giving some final touches to the code and basic design. This week will be dedicated to its enemies and adding variation and visual flair. We really hope people like how it looks and plays.

Every step takes as a little closer to the game we want to make and be proud of

[Elideb]

Input, input, input
During the last weekend I dedicated some time to updating Heart&Slash's input system. Many open issues we have with the game are related to it and I'm trying to sort things out in the best way possible.

Since Unity's native input support and configuration capabilities are... underwhelming, we are using InControl as a wrapper around it, with our own input system built on top of it. When we need to, we will be able to drop InControl and replace its layer with any other library without, I hope, affecting the game code considerably.

Recognizing pads
The main issue is, as always, offering support for as many devices in as many setups as possible. Windows, Linux and MacOS all treat gamepads differently (even different versions of each might be inconsistent). Unfortunately, although InControl offers support for several common gamepads in each platform, there are many it does not know how to handle. And, since its gamepad database is part of the source code, users cannot add their own configs.

I got around that by creating my own custom config system, which I'll probably release sometime in the future. The idea is to give players a program which they can use to automatically generate config files. Once those are placed in the appropriate folder (HeartnSlash_Data/Libs/InControl/Assets/Profiles), the game will recognize them right away. There are a couple of configs already in there, for the valiant players.
Once the config system is released, I'd expect players to send us their setups, so we can include them in later builds.

Linux, SDL2 and SDL_GAMECONTROLLERCONFIG
One of our Early Access players, TiZ, told us that in Linux there is a (relatively) recent standard on how to configure and recognize gamepads, through SDL2. I'd rather not have players configure their pads in two different systems, so I'm considering parsing SDL2's config variable as another source of custom configs. Maybe I'll even add support to create the SDL2 config from my device mapper. But that'll have to wait a little, since I haven't found much documentation yet and will have to read the parser's source code to get an idea of all available options.

Mac input
Once those things are solved, we should be almost good to go. However, there is one more hurdle in the way: our Mac builds only recognize input properly if built from a Mac computer. This is an issue that has me quite confused, because I have no idea what could possible be causing this problem. For the time being, I have updated InControl to the latest stable version. I haven't been able to determine if that fixed the issue yet, but I'm not too confident. At least it has added support for a couple extra controllers, including those of PS4 and XboxOne. That's something good, at least.

I'm afraid it could be related to some OS dependent code in InControl, since it compiles the pads for each platform. If that's the case, the fix could prove hard to find. The other option is having to build everything from a Mac, which is an option I'm not too happy with. But, since we still need to pass through a Mac to sign the builds, my opposition is not that hard.

[r.kachowski]

awesome news about mac! 

i've been going nuts trying to get this to run on my macbook. the "configure gamepad" option is just never available to me no matter what i do

on a tangentially related note, what is the news with the kickstarter backer designed weapons? 

[Elideb]

So, last week we added a considerable amount of new equipment, including throw weapons: ninja stars, grenades, mines, etc. That adds a considerable amount of variety to secondary weapons, which were previously limited to guns and shields.

This week will release the first version of level 3, then new body parts for Heart. There are not many of these left for us to reach what we consider "content complete": some extra weapons and body sets, a couple of enemies and the game endings. Afterwards, there's a lot of tweaking and optimizing left, translations, more tweaking, etc., of course, but things seem to be coming to a close.

Slashing through bugs is now a high priority!

From now on, we'll be dedicating at least a day every week to squash bugs and balance stuff: fix collisions, damages, difficulty, input...
One point we'll have to consider for sure is making the levels a little shorter. The latest build took one of our Early Access players 1 hour and 20 minutes to complete. That's 20 more minutes than the final gameplay we consider desirable, for two thirds of the complete game. We'll have to analyze gameplay data in detail and see what parts of the game take longer to complete, or if that was an atypically long run (got lost often, visited every room, leves where longer than usual).

We'll also have to improve the loading of the game. Making it load asynchronously is easy (already done in one of our branches), but getting everything to wait until the rest is properly initialized is a different story. Let's see how long this takes to implement and properly test.

[Elideb]

awesome news about mac! 

i've been going nuts trying to get this to run on my macbook. the "configure gamepad" option is just never available to me no matter what i do

on a tangentially related note, what is the news with the kickstarter backer designed weapons? 

Hey! All kickstarter weapons but two are done, and one more is still not in the game. We have the guitar, Zabre, BeatBox... We also have the three Shinken katanas in the game, but these are special, because they are part of a Kickstarter character pledge, which will only have those three swords available. You will be able to play with them with other characters, however.

Regarding the input in Mac, we still need to do some more testing. Not all issues are fixed, but we hope things are better now.

[Elideb]

While Juan was completing the next batch of equipment (B-DSS and FRANKY sets, from the factory bosses, scavenger and police from the city and sewers), I have finally merge into the main development branch the space lift level. There's still a lot of work left to do in it, but having the option in the menu feels good 

Now that I see it, we should increase the font size in there...

We are a little closer to the end!

[Elideb]

Today was build generation day, so there was not much time for actually moving things forward.

We agreed some time ago that making parts return a little more health would be positive. And, also, that the scrapping button was too small and uninformative. So, while Juan generated the builds and all that jazz, I got into this.

For the rest of the post, take into account that Heart starts with 6 full hearts, and can get to 14 with its upgrades.

When scrapping a piece of equipment, you used to recover half a heart, plus another half per XP bucket spent on upgrades. That'd be from 0.5 to 2.5 hearts for most parts. Or from 0.5 to 5 hearts for the ones with really expensive upgrades. The low end of those numbers is really low, so we've changed the base value to one full heart, for non-upgraded weapons. More powerful weapons might recover 1.5 or 2 hearts with no upgrades.

That alone makes scrapping equipment more attractive. But we've decided to give the scrapping mechanism a little more importance by adding an extra upgrade to improve it. Now, Heart has a fourth upgrade which multiplies the health recovered by 1.5 (1.4 in the case of the secret character)¹.

Let's make some calculations, to see how the changes look  :

Hearts recovered if scrapping per weapon (full upgrade cost in parenthesis):

	Steel katana (3)	American shield (6)	Electric guitar (9)
Old base	0.5	0.5	0.5
Old full	2	3.5	5
New base	1	1.5	2
New full	2.5	4.5	6.5
New base+upgrade	1.5	2	3
New full+upgrade	3.5	6.5	9.5

That's certainly a considerable extra amount of hearts. It might seem unbalanced, but since there are few other ways to regain health in the game, and experience is not that easy to accrue, getting to the highest numbers takes quite a while. And then you have to destroy a fully upgraded weapon.

So, in the end, we think things balance out. As you progress, you'll find better weapons. Then, the question will be: should I upgrade the old weapon and scrap it to recover health? Should I scrap the new weapon right away, and keep the old one upgraded?

And, in my opinion, that's quite an interesting decission for the player to make...


¹ Actually, 1.51 and 1.41, to avoid issues with roundings.

[Elideb]

I used some (not really) free time to also improve the Scrap button. It now indicates how much health will be recovered by scrapping the part. It was too important a fact to not tell players about.

[Elideb]

So... Unity has decided to reimport the whole of Heart&Slash and I find myself with a little unexpected free time. Let's post!

Since our last post, Juan has finished the integration of new equipment parts (doing all their pixel-y sprites for the menus and HUD) and I've dedicated a couple of days to bug fixing. From now on I'll have a lot more free time, so I expect to have one whole day each week for bug fixing and fine tuning of the game.

Now, our game should be showcased this weekend at PAX East, by the hand of our publisher Badland Games (not related to the Badlands games). They'll have the latest demo available and we hope people will enjoy their time playing it there. We'll post pictures as soon as we get them.

Also, we have come to accept that our atlas need some rethinking. Originally, they were separated in menus and HUD atlases. Unfortunately, the equipment icons are displayed in both, so they are duplicated. That is a waste of memory, sure, but I'm more worried about it being a waste of our time and a constant source of worries. It's easy to lose track of which textures were added where, or pick the wrong atlas for a certain element. We'll probably have to create an independent atlas just for the equipment, which won't be a major problem, but is still one more task to take care of. We'll probably use the chance too to improve how we reference sprites in an atlas from our components. Our current approach is a little shady, as it requires having the original texture's name at hand. I shiver everytime I remember that

Also, this weekend we'll participate in a videogame developers meeting called Freakend. Juan will talk about our Kickstarter campaing and I'll probably do a (still alive) post-mortem. If they are made available afterwards, we might share them, depending on how well they come out. I don't know if the presentations will have to be in Spanish, but I'd like at least mine to be in English.

[Thomas Finch]

I like all the progress you guys have made and can't wait for this game! Just a note: The title says "Xbos One".

[Juan Raigada]

Ooops  . Fixed

[Elideb]

Our publisher had a Mac demo of Heart&Slash on display this weekend at PAX East. That's been, by far, the biggest event Heart&Slash has been a part of, which is great, but we couldn't afford to attend 

They used a Mac build of the game, running on a Mac Mini (which are not that Mini, I'd say). Unfortunately, there were several technical issues: the game seemed to suffer from a memory leak which made it crash every hour or so. Also, the last day the machine was accidentally unplugged. Everything seemed to recover nicely, but the controller support was not there anymore, despite it being detected by MacOS. We knew our pad support was brittle in Macs, but that is pretty scary.

On our side, Juan and I attended Freakend, a 2 days game developers meeting, organized by our friend @DiegoBez. We gave a couple of conferences about KickStarting and developing Hearth&Slash, and learnt a lot about optimization in Unity (yeah, there's lot of stuff we did not know) and marketing (which we need in spades, as @david_bonilla heartlessly told us). Also, half of our sound team, Mirella, sung several videogame covers, with our friend Hernán.

So it's been quite the weekend. Personally, I'm still trying to recover my forces.

[Elideb]

A quick update, while a test build is generated. Juan is working on the final boss fight (although some assets are missing), which should be ready for early testing next week. It'll have to be tweaked, as everything, and then we'll add the endings.

Meanwhile, I'm on the bug-fighting squad. I'm trying to improve the performance of the game overall, but specially on the city level. Police lifters and turrets do something weird which produces huge amounts of alpha rendering. However, the problem seems to be a lot more noticeable in builds than in Unity. It's all a bit weird.

I'm also working on making our enemies use as few materials and textures as possible. My future includes lots of atlasing, UV mappings and other niceties.

[Elideb]

Aaaand we have found the bug that caused that dreaded framerate drop!

Our conclusion is that DirectX must have a bug somewhere that makes it take a lot of time to render transparent objects with 0 scaling. This does not happen in OpenGL, so Unity's editor did not display the issue (and neither should Linux or MacOS, but we have to check).

Transparent objects, soon to have 0 scaling! Yuuuhuuuu!

The warning signs in that image were scaled according to a curve so that they'd became of size 0 when the turrets touch the ground. Then we forgot to disable them, so they remained around at scale 0, destroying the game's framerate.

Now, we never scale them below .1 (forced by the script), have adjusted the curve and properly disable the markers when we're finished with them. We've gone from 10-20 FPS with 4 turrets on screen to 25-35 FPS. That's measured in my laptop, with an Intel HD4000 (ultraportable specific card, so very low end).


So that one, which took quite a while to detect, is now fixed. We still need to improve the overall framerate in level 2, but at least it is playable now.

Juan and I have been discussing how to improve meshes and materials in the game. The tasks have been defined and I'll comment them as they progress.  Let's see how much we can improve things!

[Elideb]

This weekend I implemented our pooling system, of which I'll comment something some other time (prefab based, easy to configure), and spent the whole Monday configuring all missile launchers and missiles in the game (but one) to take advantadge of it. Once I've moved all the combat particles too, I'll check how much has memory consumption and performance improved.

Juan is working on the final boss battle, which will have him busy for some time. So far, although it's missing the textures, QuAsSy looks great.

We've also been discussing how to fix issues our sound guys are having. The tools we provided them have some mild-to-severe shortcomings which are slowing their work a little too much. I'll probably get to work on that today or tomorrow, with Juan's constant supervision. It is his system and he knows how it works better than me.

On a positive note, we still get positive comments from people playing the game on Early Access, which feels great.

[Elideb]

The pooling and the final boss battle are going well. I haven't taken measurements yet (I have to configure all floors to use the new system), but things should be smoother now, at least on the CPU/memory side. Let's see.

Now, I'll be dedicating a few days to improve the audio integration pipeline. Our sound team is having problems with how long and hard the task is for each attack, taking a whole day per weapon. This should be time well spent if we can get them to integrate several weapons per day and reduce the time it takes to get their work into the game.

Excuse the lack of images. I don't want to show too much of QuAsSy and captures of pooled particles systems are not exactly amazing.

[karlozalb]

I want a sexy particle pool system screenshot!!!

We've done a pooled particle system for our project too. We realized that our game's performance decreased a lot without it, specially when we were testing it on Unity's editor. I've not tested the x64 version of the editor yet, but I expect to see a reasonable performance improvement (please xD).

Looking forward for that boss battle

[Elideb]

I want a sexy particle pool system screenshot!!!

We've done a pooled particle system for our project too. We realized that our game's performance decreased a lot without it, specially when we were testing it on Unity's editor. I've not tested the x64 version of the editor yet, but I expect to see a reasonable performance improvement (please xD).

Looking forward for that boss battle

Haha! Well, the greatest thing about the new particle pool is that it looks exactly like it did before!!!! And [almost] bug free!

For us, the biggest boon hasn't come from the performance improvements (it's noticeable, but not a dramatic frames per second rise), but from the ease of configuration. We now know exactly which pools are available, so we can just pick, instead of looking through all particles or missiles we used to have. With the pools, all weapons use a few projectiles and particle effects.

[Elideb]

Today we have released one big update, including the final boss battle, one of the endings a couple new enemies, lots of fixes... Things are looking good.

Face off!

We need to integrate the City/Sewers quests, some more weapons/equipment and that's it for game content. From then on, it's bugfixing and giving some final touches (replace the Space Lift minimap, balance combat, integrate cutscenes...). We see the end now, which feels really good after all this time with a huge pile of "pending tasks". Once the game per se is complete, we can look at ports or localization with calm.

[karlozalb]

Today we have released one big update, including the final boss battle, one of the endings a couple new enemies, lots of fixes... Things are looking good.

Face off!

We need to integrate the City/Sewers quests, some more weapons/equipment and that's it for game content. From then on, it's bugfixing and giving some final touches (replace the Space Lift minimap, balance combat, integrate cutscenes...). We see the end now, which feels really good after all this time with a huge pile of "pending tasks". Once the game per se is complete, we can look at ports or localization with calm.

woohoo! the light at the end of the tunnel

By the way, I've remembered a question about pooled particle systems. We have a problem with them, sometimes when we recycle one of them and we put it in another place, we can a see a trace of particles between the last position and the new position for a moment (we stop the particle emmision as soon as it is recycled). Have you ever suffered this problem? How do you avoid it? I've thought about waiting a second before setting the "recycled" flag on, in order to give them some more time to go away, but maybe you have a fancy method

Hope you understand what I'm saying and sorry for the off-topic