# overgrowth by wolfire too long in development thinking in terms of a 'permanent alpha' approach might help projects like this ie. release teh garme as it is and simply state that it is in a permanent state of continual development / evolution while long gestation periods are often good and necessary for better garmes there's also the strong need to balance the time needed to complete a project with the immediate needs of hungry garmers who want to play it there doesn't seem much in overgrowth's core idea of 'sandboxed fighting rabbits' that could not from eg. simply importing rabbit models into garry's mod and arming them with oversized swords garme developers wolfire state that instead of licensing someone else's engine we've built our own from scratch so that we can custom tailor every feature to support our innovative combat-based garmeplay this too is fair and good; one problem with such an approach however is that 'customization' and 'combat-based garmeplay' are nothing special nowadays and may easily be set up in other engines developers might benefit by asking "what is it about our ideas that requires this unique level of crafting?" what are the most 'immediate' ways to achieve the same or at least 'not noticeably different' effects? matthew weise in ir gdc talk: said ask what is the necessary level of specificity needed to achieve the effect you want?" a near-realtime on-the-fly development style seems to be overtaking the impossibly old "when it's done" approach (despite the false example of "half life 3" / "valve time" valve is in fact a leader in the use of a rapid iterative update development model) there also seems a strong link between the 'embodied immediacy' granted by technologies like the oculus rift and the 'instant garmeplay' made possible by garming concepts like garry's mod; players who enjoy these technologies seem less concerned with modern graphical fidelity and more with the instant theatrical amusement they generate - by the new or unusual garming viewpoints they allow or the particular novel forms of 'suspension of disbelief' they encourage the thing about gmod in particular is that it allows for multiple "garme modes" within the same sandbox. this type of more freeform play or 'playgrounding' is however offset by the rather flat dry look of the old source engine perhaps the time's come to make more use of such 'playground engines' ones which can more easily scale up ir garmes / modes (in terms of programming complexity and graphical fidelity as and when needed) the need to control each and every tiny little detail of your garme before release is often not as interesting as brian eno's approach ie. 'watching seeds grow' - providing garmers with the core ideas and tools they more instantly need to make (play out) your particular set of ideas and expand them in interesting unexpected ways players are far more interesting (in that strange space called "player interaction") than any particular video garme with specific garmeplay types or character models on overgrowth's receiver also arguably far more interesting than overgrowth is wolfire's garme "receiver" // video here a wonderfully somewhat abstract 'garme space' which was developed in 7 days has a 'pressurized density' of ideas a strong look - and is now open source - feels far more exciting than something overly large and complex that needs untold years of developmental cooking (by which time everyone's long forgotten what your initial 'big idea' was all about - eg. "something vague / zombies") idea: combine the increased expressive control of receiver with the time mechanics of superhot // republic of bob