tulamide's Forum Posts

  • ups <img src="smileys/smiley9.gif" border="0" align="middle" />

    <img src="smileys/smiley36.gif" border="0" align="middle" />

  • This is driving me crazy, because nothing I try changes anything, and I don't know why.

    On my last pc, some month ago, I worked with a HD2600XT and Firefox 4. Ghost shooter demo: 10 fps.

    FishIE Tank: 33 fps

    On my new pc I had only an on-board-graphic chip (NVIDIA 7025) with the newest driver and Firefox 6.

    Ghost shooter demo: 10 fps.

    FishIE Tank: 29 fps

    A friend of mine was so kind to give me his old card, a 9400GS. Still Firefox 6 and newest driver.

    Ghost shooter demo: 10 fps.

    FishIE Tank: 32 fps

    Scott Porter's Canvas Performance Test: 48 fps

    It doesn't help if people tell that they are able to run at 60 fps, because if their computer is powerful enough, it might reach that value without hardware support. And I can't find any message, if Firefox is using the hardware or not.

    1) Why does it appear to have no hardware support, although it should?

    2) Why is the ghost shooter demo so slow, while a HTML5 3D Canvas Test results in 48 fps?

    3) Why does it seem, that no matter the cpu and gpu, I always get approx. the same rates?

    Test pages used:

    FishIE Tank

    Scott Porter's Canvas Performance Test

    EDIT: I also tested the on-board-chip and the current card with a project in CC. On-board chip 400 fps, new card 1400 fps

  • Try Construct 3

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

    Try Now Construct 3 users don't see these ads

  • What's a MIDI?

    Musical Instrument Digital Interface

    The files created with the extensions .mid or .midi (and some others, like karaoke files) contain instructions for the interface (mostly a software synthesizer nowadays) to generate music playing. It contains no audio, and is therefore a very small file format. Imagine mod files without all those samples ;)

    11clock

    As shinkan said.

  • for not playing Portal anymore. It is offered for free until 20th September on Steam.

    http://www.eurogamer.net/articles/2011-09-16-portal-free-on-steam-until-20th-sept

  • I don't know much about audiere, but it sounds like digital distortion. This happens if there are too many sounds playing at a too high volume level. Also, using 300 sounds at once is a huge amount. This could swamp the engine. You should try to decrease the number of sounds playing at the same time and set the volume of those sounds to a significantly lower level.

  • It is most likely that the group, that you found responsible for the framerate drop, contains loops that run longer than you intended, or create more objects than you intended. Did you have a look at the debugger to see how many objects actually are created during those events? And, if you change, for testing purposes, the loops to always create a fixed amount of sprites, instead of collision checking, does it also drop the framerate?

  • <img src="smileys/smiley4.gif" border="0" align="middle" />

    I wouldn't be ashamed. You managed to solve this issue. That's all that counts <img src="smileys/smiley2.gif" border="0" align="middle" />

    (And to make you feel better: I downloaded the cap yesterday and did not stumble upon that. I was about to blame the wait actions. What a relief that I don't answer anything unless I'm at least 50% sure about a possible solution <img src="smileys/smiley41.gif" border="0" align="middle" /> )

  • As Jayjay said, CC will handle this just fine. If you will have larger arrays or are just curious to see other implementations of sort algorithms, you may also want to have a look at Verve!, that presents a divide & conquer algorithm, which always needs the same amount of time to find the right position, no matter if that position will be, for example, the first item in the list or the 10000th. But for a simple 10 place hiscore you won't need it.

  • Well, contrary to what tulamide says, I don't know how how to do it, but I just learned from a few internet serches and about an hour of tinkering.So, despite of what you said, in the end you proved me right <img src="smileys/smiley4.gif" border="0" align="middle" />

    I just can't think of anything in programming that you couldn't solve with such an ease and clear, comprehensive concept. It is pure admiration. For example, I too started an internet search but I didn't come up with anything but pygame. If anything fails, you are reliable. That was needed to be said <img src="smileys/smiley1.gif" border="0" align="middle" />

  • Yeah, it's an old thread.

    But this update is worth mentioning, I think:

    On September, 9th, a german court "D�sseldorfer Landgericht" confirmed Apple's preliminary injunction, whereupon Samsung's Galaxy Tab 10.1 is too close to the iPad. Samsung was forbidden to sell the Galaxy Tab in germany and even more the court prohibits Samsung to offer any Tablet-PCs with certain characteristics (rectangular shape, rounded edges, flat clear surface, curved back) europe-wide.

    There's no english version of the news, I'm afraid. But I summarized it as conscientiously as possible.

  • There's no such action. It was an indicator that your actions to drop something would have to be inserted there. The point was to show a way to create a random rate by using "every x milliseconds" in combination with the expression random() <img src="smileys/smiley2.gif" border="0" align="middle" />

  • Thank you.

    I'm glad it helps <img src="smileys/smiley1.gif" border="0" align="middle" />

  • No they don't list just any. They query the system, which reports any resolutions, that the attached monitor reports.

    There is no built-in support for that query, or I haven't found it. I'm sure, ROJO knows how to access these informations directly from the system using Python ;)

    However, using Python you can also get such a list if you install PyGame. The function "pygame.display.list_modes" does exactly what you want.

    It would be used like so:

    import pygame.display

    pygame.display.init()

    modelist = pygame.display.list_modes(32)

    32 indicates that you are looking for resolutions with 32bit color. The result is a list with tuples:

    [(640, 480), (1024, 768), (1440, 900), etc...]

    [EDIT] Link to the pygame function description: http://www.pygame.org/docs/ref/display.html#pygame.display.list_modes

  • >tulamide

    Just curious, any reason why you used 'move at angle' for the player instead of a behavior?No particular reason. I'm just used to do it this way if I'm in a hurry. The cap was a quick&dirty product :)

    Now this, you solved if I were to have another object blocking the vision but the problem is that I need to know:

    Is any "Wall&EnemySprite" blocking the line of sight to any other "wall&EnemySprite"?

    That is:

    What "Wall&EnemySprite"s can the "PlayerSprite" see, and what "Wall&EnemySprite is blocked by other "Wall&EnemySprite"s?.

    I don't know how well I explained it, but I really only have 2 "Active objects" in the game; a PlayerSprite and a sprite representing enemies AND obstacles. (And I need it that way, as walls need to change to enemies and vise verse in my game)This is what I mean by "think too complicated".

    For every game there comes a time when you have to optimize your code/thoughts. Even the biggest game developers wouldn't be able to constantly check roughly 1 million times per tick without dramatically slowing down their game. So it is about tricking. Like in 3D games high poly models are exchanged by low poly ones when they are displayed far away.

    1) You don't need to test all instances. Limit the tests to the range that your player sprite will "see".

    2) Don't loop 360 times to get a full circle. Instead create one circle sprite and set it to the radius you want the player to be able to see. Select all blocks that overlap that sprite (Or use no object at all and select those blocks with math only). Now loop through those blocks and check with sample points every blocksize against the angle to the player sprite if there are other blocks in the way. If so, set it to invisible, else set it to visible.

    This should speed up a lot. I didn't have a look at JayJay's cap. Maybe it is exactly doing so. If not, I might do a cap, if I find the time.

  • Damn, I worked on a cap to play around with PerlinBias to find the right values. Was too late :(

    Anyway. I am posting it, if you don't mind, in case someone else might need it:

    http://www.mediafire.com/file/d7iwraufey7v6nw/PerlinBias.cap