Gino van den Bergen

Mostly serious game developer

[Nuseir Yassin @nasdaily]

peace

[Massimo @Rainmaker1973]

40 years of Boston Dynamics (1983 - 2023)

[📹 World Data Center, HD: https://t.co/m6VDYgvZXI]

[Gino van den Bergen @dtecta]

@SebAaltonen @olson_dan @ninkibah It's nonsense in a function declaration. It's not nonsense in a function definition.

[Erin Catto @erin_catto]

A physics engine should implement collision code independent of dynamics code. This helps avoid mistakes like using center of mass instead of shape centroid. It also makes it easier to test collision algorithms in isolation.

[Gino van den Bergen @dtecta]

@logicalerror I have to disagree. Abstracting means focussing on what's important. Clearing out all the unimportant stuff makes programs easier to understand. For example, sorting algorithms only need to focus on when one object of a type is less than another and how they are stored in memory.

[Gino van den Bergen @dtecta]

@EricLengyel I guess that for many applications (configuration space, geodesics) you can unambiguously refer to "the closest point" and drop "to the origin". Not very sexy, but concise enough to be repeated many times in you text.

[Gino van den Bergen @dtecta]

@PierreTerdiman @ChristerEricson Compilers might be more lenient with "float(x)", since it mimics a constructor call. This helps in writing function templates that take arbitrary numerical types (float, std::complex<float>), and float proxies such as ErrorTracer<float>. See also https://t.co/okRsKBp6J3

[Gino van den Bergen @dtecta]

Looks like some links to publications I posted here went stale. Please, visit https://t.co/mFYr66Poxw for a list of my publications.

[Gino van den Bergen @dtecta]

@cmuratori @PierreTerdiman I'll take that as a compliment.

[Gino van den Bergen @dtecta]

@tuxedolabs Congratulations!

[Gino van den Bergen @dtecta]

@tuan_kuranes @Official_GDC Glad you could join, and thanks for the nice words and the link!

[Gino van den Bergen @dtecta]

In case you missed it or need a recap, here are the slides to my @Official_GDC talk on conservative mesh decimation: https://t.co/WxOGuJ1gKq #GDC21 #gamedev

[Gino van den Bergen @dtecta]

My @Official_GDC talk on conservative mesh decimation is going live today at 1:20pm PT (22:20 CEST). Come and see! https://t.co/mOXzSrwiqh #GDC21 #GDC2021 #gamedev

[Mark Hamill @MarkHamill]

#FavePhotoFriday 😍

[Gino van den Bergen @dtecta]

Sometimes you need to be conservative. Please, check out my @Official_GDC talk: Math In Game Development Summit: Conservative Mesh Decimation for Collision Detection and Occlusion Culling https://t.co/mOXzSrwiqh #GDC21

[Gino van den Bergen @dtecta]

@PierreTerdiman Didn't Shrödinger solve that one? He probably left gravity out of the equation, so it must be one of those unsolved physics problems.

[Gino van den Bergen @dtecta]

@Reedbeta Yes, I have been guilty of this in the past and to be frank I like to be able to use math expressions in algorithms. Our current math language has taken over 400 years to evolve into a concise an unambiguous notation, and in that sense is hard to beat by any programming language.

[Gino van den Bergen @dtecta]

@wabuilderman Thanks for your interest in the paper! You can pretty much subtract any finite or infinite point set from any other. For Y = {y_0, ..., y_n}, "{x} - Y" denotes "{x - y_0, ..., x - y_n}". Makes sense? Please check an updated version of the manuscript in https://t.co/H7jnUIxOds.

[Gino van den Bergen @dtecta]

@richgel999 There are many things wrong with Modern C++ but templates and variadics aren't. Dijkstra's problem with goto was the complexity of proving correctness. C++11 has many constructs that allow you to lift programming to a more algebraic level and Dijkstra would have loved it.

[Gino van den Bergen @dtecta]

@EliteDangerous Arriving at Selene Jean with 3% hull and unable to request docking due to a broken sensor, I thought it was a good idea to reboot and repair. Gravity thought otherwise.