Arima's Forum Posts

  • When I previewed over LAN on:

    2x Iphone 4S'

    1x Ipad 3rd gen

    1x Samsung Galaxy S2

    1x Dell Laptop

    I got smooth on laptop, Ipad but naff FPS on Galaxy and both Iphones. These were previewing through 'Chrome' on all devices.

    After compiling with CocoonJS it was SUCH an improvement. Couldn't be happier (I think ;])

    Safari is way faster than chrome on iOS, because as far as I understand it, chrome just uses the web view (which doesn't use nitro which speeds up JavaScript) because apple doesn't allow the technologies needed for a competing browser to be used in apps on the App Store. It's why there's no Firefox for iOS.

  • The reason I'm hesitant to use something like RPG Maker is that all of the games it produces seem to have a very Pokemon-style approach to the genre. There's a town area where the player directs an avatar to move around, sprites for every NPC, etc. I'm effectively eliminating that entire element of the game, since, for mobile, I find that the avatar movement mechanic is clunkier than what the HID allows for, and better paradigms for travel and world-interaction can be explored.

    Don't get me wrong, while it's more difficult, that's why I'm using C2 and not rpg maker myself - rm is just too limited and with C2 I can make my game exactly how I want it.

    hen you say that they have a lot of "interlocking systems" that have to come together, do you mean that in a technical sense (getting objects to instantiate, interact, etc, properly from a programming perspective), or in a mechanical sense (make sure that a +1 to attack doesn't stack in a certain way to make characters be able to one hit everything)?

    Well, both, but mainly the technical side from my experience (you still have to program the rules, after all).

    f it's the latter, I've already developed the core system (Attributes, skills, to-hit stats, item balance, etc) to the point where I'm confident in the relative stability of the mathematics. You're not going to have a certain build that is superior to all others in every way. That's where my experience playing, modifying and more recently writing, tabletop (pen and paper) RPGs comes into play.

    I'm not claiming your system isn't as good as you think it is, but you should probably be aware it's very common for people to code something and realize it doesn't play as well as they thought it would before they started - even people experienced in software and game development often encounter this problem, and is why some people like the idea of fluid game design rather than deciding everything about how the game is going to play before any code is written. Just something to be aware of that you might encounter.

    Well is it difficult or not depends on your own take on it. I can tell you from my experience that if you don't approach it as difficult, but rather with attitude that it is simple but might be challenging, then the only thing that stands between you and the finished rpg are how you plan it in advance and how much time you are willing to put in to it. Also on the side note can't wait for your finished game! :)

    Based upon the numerous reports I've read from other people attempting to develop RPGs and the very near unanimous opinions thereof that they simply are very close to the most difficult genre to make (only being beat out by stuff like mmorpgs and insanity like dwarf fortress) as well as my own experience already having tried to make loot pursuit with the attitude that it would be simple and having planned a ton of it in advance only to realize it didn't work how I thought it would once I tried to actually make it, I feel confident in objectively stating that attitude has nothing to do with it, RPGs are simply hard to make. Even with 5-6 years experience with construct (though I certainly don't claim to be the best C2 user, there are people who can easily code circles around me, I'm far better than I was when I started, having almost no idea what I was doing at all), there's just far too much involved in making an rpg to call it a simple task. I certainly understand how it would seem to be simple though, as I once thought that myself.

    Here's one of many, many examples of how things turn out to be more difficult than they seem like they might be. Tapping on a unit to select them on a touchscreen in battle. Seems easy enough, right? On object touched. Except that tapping isn't very accurate compared to a mouse, and as such the user can miss a unit easily, especially if they're small enough, even more so on a small screen like a phone. Ok, so then instead I tried an 'on touch, pick closest unit' event - except there were times where I would try to tap the ui and a unit would be selected. Ok, so I put all clickable elements in to a 'clickable' family and pick the nearest clickable to the touch. Except unit hotspots are at their feet, resulting in inaccurate picking again since people try to aim for the center of the sprite. Ok, so I add invisible clickable objects at the center of each unit to pick the nearest from. Except some ui elements and units are large/small enough that even tapping what seems like close enough to them will still select something else sometimes because of the distance of the hotspots. Ok, so then I divided the screen up into regions and pick the nearest clickable object based upon what region of the screen was tapped on. Great! It worked flawlessly. Until I tried to use two touches at once.

    Etc, etc... This is just one of a ton of examples.

    Also, thanks for the interest in the game. :)

  • Argh, I'm getting that error too, I didn't actually put the node webkit plugin in the project when I tried it. Sorry. Man, I have been messing up the testing process on this repeatedly...

    Anyway, looks like it should be reported as a C2 bug.

    I also agree it would be nice if we could just choose node webkit from a list instead. I tried using the custom setting but it didn't work for some reason.

  • I agree, the 1 pixel lines aren't easily visible enough. I would like this suggestion as well.

  • As others have told you, yes C2 can make rpgs. The reason you haven't seen many is because RPGs are quite complex, and I'm afraid I must disagree with megatronx on this one, they are quite hard to make due to what Aphrodite said - they have a ton of interlocking systems that all need to work smoothly together. They look simple from the outside (I thought one would be simple to make too. Years later, it still boggles my mind how long my rpg has taken - but then again I had almost no game making experience at all when I started on it), but general opinion from those why try to actually make them is that they're one of the most difficult genres around (unless you're using a tool that has a lot of those rpg systems in place already, like rpg maker).

    Here's a video of the old construct classic version of my rpg (don't even start to consider using construct classic over C2 for an rpg. Trust me. There are reasons why I started over in C2 even after years of work - it's vastly superior and more stable than CC). The C2 version looks very similar to this, but the ui and stuff isn't done yet:

    [TUBE]http://www.youtube.com/watch?v=gNRi0xcwW_s[/TUBE]

  • I haven't been able to test this using C2's built in node webkit export because of my antivirus flagging it as a false positive, but here's a way I just tested:

    Download node webkit from it's web page, extract it from the zip file

    Open nw.exe

    Open a C2 game, press preview

    Copy the address from the window that comes up

    Paste it in node webkit's address bar, add ?nw to the end of it, hit enter

    Press preview in C2, and it updates the node webkit window

    Close the browser window that you copied the address from

  • > I guess a mod would have to keep that in check.

    Which means that instead of you having to look and sort the posts, that's a job that would be issued on mods.

    And, as far as I'm concerned (and only talking in my name), that's some time I can't spare/job I'm not willing to take

    I have to agree - not only would that take a lot of doing for us, it's also highly subjective what quality level a work is at and people are likely to get offended if we pick a category that the person disagrees with, not to mention we would have to research or play everything that gets posted to properly determine the category. The amount of work it would require is the same reason that games put on the arcade are auto-approved now - it was just too much for us to deal with to approve them manually.

    I also agree that the 'your creations' section is a bit problematic as it is, though I'm not sure of the best way to do anything about it. Maybe something sort of like the TIGSource forums have might work, having a section for completed games and another for devlogs/in progress ones. Then when someone has finished a game, they can start a new thread in the completed games section?

    Thinking about it, even if users do post in the wrong area and simply we leave them there, having multiple categories might be better anyway. We would get at least some categorization that way, as I suspect most users would place their threads in the right place. Tech demos(or call it experiments instead?)/devlogs/completed games sounds good to me in that case.

  • What would be the difference between sub families and simply putting an object into multiple families as can already be done?

  • Keep in mind scirra very rarely has sales. I think it's been on sale only once or twice since it's creation.

    That's not to say they won't. But not to say they will, either (I don't know and haven't heard any mention of one).

  • C2 uses node webkit for exe, and another game using node webkit was greenlit for steam, so yes, getting a C2 game on steam is possible.

    As for the 1 gb size, to be honest, I'm not sure If C2 is currently suitable for such a thing. I have no idea what node webkit would try to do memory wise, but I think what might happen is node webkit would try to load everything into ram at the start at once, resulting in long load/preview times at the start and probably using more memory than a lot of computers have (1gb of assets is way larger when decompressed), even if layout by layout loading keeps the VRAM usage low by only using some of those assets at a time. Then again, it might handle it no problem. I really have no idea. The only way to find out is to try it yourself, as I'm not sure anyone has tried such a thing yet.

    If it doesnt work well, you could use the node webkit plugin to load graphics from disk when you need them, but this is a cumbersome process if you're talking about loading animations, as it is not automated at all and you have to code it all yourself. It works fine for background images though.

    The only other idea I have is to break it up into multiple 'pages' - when going from one section of the game to another, using the browser plugin to browse to another page. That could also be cumbersome to do, though, and could also be pointless if node webkit can actually handle it automatically.

    If you wanted to test it, you could make a tiny slice of the game without the game logic - by that, I mean import an approximate amount of graphics into the free version (not the same image over and over, C2 has deduplication methods so it wouldn't be an accurate test), but not all into one layout - 1 gb of compressed graphics will be way more than 1 gb when it gets decompressed into VRAM, and would be more than most, if not almost all graphics cards could handle at once. So, have multiple 'objects' layouts. Then run it in chrome and watch the memory use to see what happens.

    Actually, come to think of it, you might also want to ask on the node webkit message board. They might know what would happen.

  • I don't think there's any way for C2 to do that as it is, but you might be able to write a plugin for it. Node webkit supports a lot of features that C2 doesn't support. Check the list of features on its web page to see if it's even possible.

  • Could it be user definable with actions like it was in CC? In a few places I'm doing tricky stuff in one event in various parts of an event branch, and having to arrange things into a group just for profiling might not be practical as it would also profile the later parts of that branch as well as the need to rearrange code a bit to create a group just for it. Also, setting it to use groups wouldn't stop someone from profiling one event, anyway (one event can potentially be pretty intensive by itself). Perhaps it can just mention the limitations of the profiler, similarly to mentioning the limitations of how every 0.001 seconds won't actually run faster than the framerate?

  • Great, thanks, I'm glad I asked first, and that Roger is going to implement what you asked for in 0.7.

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads
  • Nice, thanks Ashley!

  • I know this has been brought up before, but now that we have file operations in the node webkit plugin, it's more relevant now. For those of us wanting to make a game that uses node WebKit's file io, it's very inconvenient to have to export every time we want to preview. I thought I had an idea how to work around the problem, by exporting a node webkit 'game' that was set to do one thing, go to the preview URL on the start of layout.

    At first, I thought that worked great - previewing that way in node webkit works just like chrome, when pressing preview in c2 updates the node webkit window. But then I realized that the node webkit code wasn't working - it wasn't reading or writing files to disk. I looked at the preview source and it seems like the node webkit plugin is a disabled in preview.

    Is there anything that could be done about this? Perhaps something simple like adding node webkit to the list of browsers to preview with? Or a way to make the method I described above work? It would make things much, much easier to be able to preview with one click rather than have to export every time we want to test our code.