In case you were around during Tryst'99 and were scratching your head as to what idiot came up with such a problem, look no further... apparently the problem text was taken almost "as-is" from this mail, smileys et al...:) ---------------------------------------------------------------------- From debug@poboxes.com Wed Dec 22 11:53:34 1999 Date: Fri, 22 Jan 1999 19:01:45 +0000 From: Deepak Bandyopadhyay Subject: TRYST Computer Art Design problem This just an idea that struck me, be sure to check with Adobe (submit it as a "proposal") or compare it with Alok's / other ideas. It is not final & I don't know if it is feasible. And if you select it & there are no entries I am not liable. So there :) Title: Gear Train Simulator Difficulty : Moderate to hard Software allowed: any PROGRAMMING language with graphics capability such as C/C++/Java/PerlTK/Visual Basic etc, on a MS-DOS, MS-Windows 95, NT or Linux platform (not SGI or any other!!!). A graphical toolkit or plugin for the language (eg. OpenGL, Java 3D API, using VRML2.0 with Java :) is allowed; however, specialized graphic design softwares (eg. AutoCAD) that may create this and export it as a standalone executable are not allowed. Write a gear train simulator. Create a GUI or a scripting language (GUI has more credit) to specify the types, axes and positions of the gears. Gears are basically of 2 types - normal toothed gears that transmit motion in the same plane, and bevel gears that can transmit motion into a perpendicular plane. All gears in the machine will have their axis of rotation as one of the 3 orthogonal axes. Gears can be placed concentrically using shafts. Conveyor belts can also be specified as another means of transmitting motion between gears that are not physically connected. The sim has to allow the user to input gears, shafts and conveyor belts (and specify the gears that are actually connected to a motor and hence start the motion), and then SHOW the mechanism working with the gears moving in real-time (and maybe calculate some statistics as well to prove that the simulation is accurate). Assume ideal gears with no wear & tear and minimal friction. You can render the gears as fixed width or infinitesimally thin disks (CHECK with the Mech guys as to whether such gears actually work!) Show at least one good 3D view of the whole assembly (by good I mean it should not be a horizontal or vertical projection that reduces some of the gears to lines. If some gears at the back are hidden, that's OK). Multiple or user-defined viewpoints will be given extra credit. PLEASE get someone to sit down & rephrase this draft if you are considering it. And PLEASE consult someone from Mech who can tell you all the possibilities for transmission mechanism design. But do simplify it so that they only have to simulate 3-4 different types of gears. Maybe this is not an appropriate problem for TRYST, but for a MACH type of event, I don't know. I thought it emphasizes the "engineering" nature of IITD's education.