Eric Lengyel @ericlengyel.bsky.social · 27/09/2026I'm very grateful that my book is still getting reviews like this two and a half years after publication. www.amazon.com/dp/B0CXY8C72... 1232
Eric Lengyel @ericlengyel.bsky.social · 23/08/2026Thanks for the high praise! It indeed was very fun and satisfying to make it. 011
Reposted by Eric LengyelKirill Osenkov @kirillosenkov.com · 22/08/2026I bought Radical Pie and played with it, and it's delightful It must have been such a blast to work on @disconcision.com @neurocy.bsky.social have you seen this? 191
Eric Lengyel @ericlengyel.bsky.social · 19/08/2026Absolutely. I just heard from Derek Johnson as well. I will include both of you on the email. Thanks. 010
Eric Lengyel @ericlengyel.bsky.social · 19/08/2026Thanks, that would be awesome! There's some kind of state inconsistency bug that's preventing our product from being usable in PPT. Easily reproducible. My gut tells me it would be a quick fix, and embedded Radical Pie equations would become 100% functional. 100
Eric Lengyel @ericlengyel.bsky.social · 16/08/2026I talked to someone at Microsoft support today and identified myself as the creator of Radical Pie. I demonstrated a serious bug in PowerPoint that I wanted to report. He immediately said there was no issue in PowerPoint and that I should contact the developers of Radical Pie.🤦♂️ 3172
Eric Lengyel @ericlengyel.bsky.social · 09/08/2026Position vectors and quaternions are fundamentally different types of quantities, and attempting to represent vectors by quaternions leads to a massive time-wasting trap. People have been trying to jam spacetime positions into quaternions for a century, and the results have always been a janky mess. 100
Eric Lengyel @ericlengyel.bsky.social · 07/08/2026It has been about 10 months since Radical Pie was first released. In that time, 14 free updates have been provided containing new typographic capabilities, new usability features, expanded LaTeX compatibility, and lots of additions to the built-in math font. radicalpie.com 051
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026Yeah, I suppose I'm thinking from a projective / homogeneous coordinates perspective, where W is special. Similar things happen in spacetime as well, where nothing is projective but the time coordinate T is special. It seems like the universe really wants things to come in threes. 100
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026So if you have {1, X, Y, Z, W, YZ, ZX, XY, WX, WY, WZ, WYZ, WZX, WXY, XYZ, XYZW} in 4D, then you'd also have {U, XU, YU, ZU, WU, YZU, ZXU, XYU, WXU, WYU, WZU, WYZU, WZXU, WXYU, XYZU, XYZWU} in 5D, where U is your fifth axis direction. The ten bivector elements give you all your 2D planes. 001
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026The way I handle this is to take everything you have in 4D and add a copy in which each element is multiplied by the fifth dimension on the right. (This is discussed on page 183 in PGA Illuminated.) 100
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026This applies to dual quaternion motion operators as well. A positive displacement in a screw transformation goes in the direction of the {WX, WY, WZ} components of the screw axis (which is a bivector line). 110
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026I prefer the first one because, when a line is constructed as a bivector by multiplying two points P and Q together, the direction from P to Q will show up on the {WX, WY, WZ} basis elements. If you were to use {XW, YW, ZW}, then the direction would be negated. 110
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026As for the planes involving W, you could go either way as long as you're consistent. So it's either {WX, WY, WZ} or {XW, YW, ZW}. 100
Eric Lengyel @ericlengyel.bsky.social · 06/08/2026Hi -- To be consistent with XY and YZ, you'd definitely want to continue using ZX, just as you said for 3D compatibility. That keeps for orientation the same for each of the X, Y, and Z axes, and it gives consistent signs to quaternion components. 110
Eric Lengyel @ericlengyel.bsky.social · 25/07/2026The slides from my talks at the Better Software Conference are now available here: terathon.com/lengyel/#spe... 084
Eric Lengyel @ericlengyel.bsky.social · 24/07/2026I just returned home from an amazing week at the Better Software Conference in Sweden. So many great people there, and very impressive content. 080
Eric Lengyel @ericlengyel.bsky.social · 12/07/2026Apparently, there will be a link for the live Twitch stream on the website: bettersoftwareconference.com 100
Eric Lengyel @ericlengyel.bsky.social · 11/07/2026I'm giving two talks at the Better Software Conference in Sweden next week: 1) Software Design in the Radical Pie Equation Editor 2) Math and Container Libraries in the C4 Game Engine Architecture I believe videos will be available, and I will post slides. 2345
Eric Lengyel @ericlengyel.bsky.social · 10/07/2026Do you use the equation editor built into Microsoft Word? If so, you can put an end to your suffering by upgrading to Radical Pie. radicalpie.com 0277
Eric Lengyel @ericlengyel.bsky.social · 07/07/2026Version 1.12 of the Radical Pie equation editor has been released. This version contains some major new chemistry features that were used to construct the diagrams below. This is once again a free upgrade. 070
Eric Lengyel @ericlengyel.bsky.social · 18/06/2026On the topic of Adobe, I recently reported a bug in Illustrator that appeared in the latest update (version 30.5.1). I got a surprisingly fast response and acknowledgement that they were able to reproduce it. Microsoft should take notes. 020
Eric Lengyel @ericlengyel.bsky.social · 18/06/2026Adobe is now a double-licensee of the Slug Library, having just licensed it for a second software product. Nothing says a customer is happy more than repeat business! 0231
Eric Lengyel @ericlengyel.bsky.social · 03/06/2026I'm going to be speaking at the Better Software Conference in Sweden next month. Details to come. bettersoftwareconference.com 0151
Eric Lengyel @ericlengyel.bsky.social · 28/04/2026Btw, more about relativistic quaternions here: terathon.com/blog/relativ... 000
Eric Lengyel @ericlengyel.bsky.social · 27/04/2026I have posted the slides from my talk at the Virginia Tech geometry seminar last week: terathon.com/pga_relativi... 1141
Eric Lengyel @ericlengyel.bsky.social · 21/04/2026I’m at Virginia Tech again this week to speak in the math career connections series. I’ll be telling students all about what I’ve been doing with my VT math degrees over the past 30 years. 050
Eric Lengyel @ericlengyel.bsky.social · 12/04/2026I posted an unusual entry in my blog that reminisces about my first professional software engineering job before I entered the games industry in the mid-90s. Bespoke Software for Oscillating Bob Viscometry terathon.com/blog/viscome... 0122
Eric Lengyel @ericlengyel.bsky.social · 12/04/2026Porting to the Mac (without MS Word integration) would actually be rather straightforward. It's deploying on the Mac that's difficult. 100
Eric Lengyel @ericlengyel.bsky.social · 11/04/2026Several non-Windows users have now told me that they bought a Windows machine just so they could run Radical Pie. Pretty cool! radicalpie.com 170
Eric Lengyel @ericlengyel.bsky.social · 02/04/2026NASA has a live stream of the Artemis II mission that will run for the entire duration: www.youtube.com/watch?v=m3kR...youtube.comNASA's Artemis II Live Mission Coverage (Official Broadcast)YouTube video by NASA 2122
Eric Lengyel @ericlengyel.bsky.social · 25/03/2026Yes, it is. Just log into your account, and you can download the new version. 110
Eric Lengyel @ericlengyel.bsky.social · 25/03/2026Radical Pie 1.8 was released today, and it includes the ability to parse LaTeX. If you have a big equation already written in LaTeX, then you can just copy/paste the plain text into the editor, and Radical Pie will lay it out visually. radicalpie.com 180
Eric Lengyel @ericlengyel.bsky.social · 25/03/2026Currently working on fur rendering for Chapter 15 in FGED3. Thor, the dog of thunder, is helping with some real-world BRDF references. 0180
Eric Lengyel @ericlengyel.bsky.social · 17/03/2026New blog post: A Decade of Slug This talks about the evolution of the Slug font rendering algorithm, and it includes an exciting announcement: The patent has been dedicated to the public domain. terathon.com/blog/decade-... 1127386
Eric Lengyel @ericlengyel.bsky.social · 09/03/2026The equation editor called Radical Pie that I developed throughout 2025 has been getting a lot of nice feature updates over the four months since its initial release. We're already up to version 1.7. Free trial available on radicalpie.com. 070
Eric Lengyel @ericlengyel.bsky.social · 02/03/2026I've written a new blog post called "Dual Approaches to Projective Geometric Algebra". terathon.com/blog/dual-pg... 0164
Eric Lengyel @ericlengyel.bsky.social · 27/02/2026I wrote a short blog post about the special mathematical / scientific font included with the Radical Pie equation editor. terathon.com/blog/radical... 0130
Eric Lengyel @ericlengyel.bsky.social · 07/02/2026At some point, you neglect infinitesimals. The notion of equality changes from something exact to something that's infinitely close. 030
Eric Lengyel @ericlengyel.bsky.social · 07/02/2026Then 1 − 1 = 1 + (−1) = 000...001 + 111...111 = 000...000.000...001, which is the smallest representable number. Now let n be arbitrarily large, and this smallest number looks like an infinitesimal. If you want to see something neat, sum 2^n terms of Grandi's series like this. 150
Eric Lengyel @ericlengyel.bsky.social · 07/02/2026Imagine an n.n-bit fixed-point number representation where addition is defined as usual for two's complement values, but any carryout on the left of the whole part wraps around to the far right of the fractional part. 240
Eric Lengyel @ericlengyel.bsky.social · 07/02/2026Yes, there are algebras in which a − a is essentially an infinitesimal and not exactly zero. My favorite is this: 170
Eric Lengyel @ericlengyel.bsky.social · 06/02/2026Here, ^★ and _★ are the right and left Hodge duals. The expression b ∨ (a^★) is the definition of the right contraction, sometimes written b ⌊ a. Likewise, (a_★) ∨ b is the left contraction a ⌋ b. 020
Eric Lengyel @ericlengyel.bsky.social · 06/02/2026When a is a vector and b is any multivector, the geometric product a ⟑ b decomposes as a ⟑ b = a ∧ b + b ∨ (a^★), and b ⟑ a decomposes as b ⟑ a = b ∧ a + (a_★) ∨ b. If a and b are both vectors, these both reduce to a ⟑ b = a ∧ b + a ⋅ b. 120
Eric Lengyel @ericlengyel.bsky.social · 29/01/2026Whoa, the SIGGRAPH exhibit floor is tiny this year. This used to be a huge show! What the heck is going on? maps.goeshow.com/acm/siggraph... 362
Eric Lengyel @ericlengyel.bsky.social · 29/01/2026This might be the first year since 1999 that I don't attend GDC. 0150
Eric Lengyel @ericlengyel.bsky.social · 27/01/2026This is a reminder that every diagram from my book Projective Geometric Algebra Illuminated is available on Wikimedia Commons under the CC-BY-4.0 license. They can be freely used in slides, papers, books, Wikipedia articles, etc., with proper attribution. commons.wikimedia.org/wiki/User:Er... 03710