Lol! Nice! It's rough turned .100" over and has to rest for a day to let internal stress relax. The long diameter will end up straight with not more than .005" taper. That's actually the whole tolerance, so less.
MXS Twitter
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
jlv
- Site Admin
- Posts: 15473
- Joined: Fri Nov 02, 2007 5:39 am
- Team: No Frills Racing
- Contact:
Re: MXS Twitter
How much does it warp after it relaxes?
Josh Vanderhoof
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
Sometimes, it's more than .100" runout. That exceeds the stock amout, so stopping and shimming under the opposing jaw will keep it in line. Luckily it's not too often that it gets that bad. Chinese steel is not consistent. It's sucks we've gotten to this place of depending on them.
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
Been trying to get in the forum all day and just now got on finally. Complete bot takedown! I tried from several different platforms too.
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
Here's a good example...
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
jlv
- Site Admin
- Posts: 15473
- Joined: Fri Nov 02, 2007 5:39 am
- Team: No Frills Racing
- Contact:
Re: MXS Twitter
That's crazy that it warps that much! What exactly am I looking at there? At first I thought the left side had relaxed and the right was freshly cut, but then when you pan across it goes all wobbly again.
Josh Vanderhoof
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
The rusty part is the short, "gearbox" end and there's a flange that separates it from the "chamber" end I'm roughing out. You see 2 or 3 previous passes spinning eccentric and the smallest pass is the truest. The rest of the turned part is running out because with each pass, new stress is relieved. After roughing down to .100" over, it rests for a day, then 3 finish passes to get taper down and relationships dialed in so all are in tolerance by final pass. It'll get threads on one end, the chamber end gets hexed to drive the cutters and they'll spline the gbox end before stacking cutters on it and bearings put on.
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
Pong game recompiles its own source code every frame — winning entry at IOCCC29 was generated by a custom compiler
NewsBy Luke JamesPublished June 29, 2026
The IOCCC29 winner writes each frame as fresh C source, compiles it, then runs it.
Jonah Uellenberg won the Ping Pong Prize at the 29th International Obfuscated C Code Contest earlier this month, with a version of Pong that recompiles its own source code on every frame. Each run of the program prints the C source for the next frame, with the current display and game state embedded inside it, so compiling and running that output advances the game by a single step. Somewhat ironically, Uellenberg didn’t write the obfuscated code himself; instead writing clean, commented source in a programming language he built called Insert, whose compiler generated the scrambled C that took the award.
In his entry notes, Uellenberg writes that the winning code “isn't really obfuscated. It never was,” stressing that it’s compiler output rather than anything mangled by hand. The Insert compiler targets C and runs the size-optimization passes that a production compiler would, such as liveness analysis to strip dead code, constant folding, and a variable-merging pass that reuses one variable slot for many values. Repeated token sequences are then packed into C preprocessor defines to shave bytes. By his account, optimization and obfuscation tend to collapse into the same output, which is why the result reads as a wall of single-character identifiers and a data array that mixes constant and mutable values.
A quine is a program that prints its own source code when you run it. Uellenberg's version prints a modified copy of itself: the source for the next frame, with the updated game state baked in. Run that, and it prints the frame after, and so on.
The player moves the right paddle by passing “w” or “e” as a command-line argument, competes against a computer opponent, and runs a bundled script that automates the compile-and-run cycle so the match plays in real time. Beating the computer switches the program to a second game, a side-scrolling jumper modeled on the Chrome dinosaur runner, and clearing that switches it back.
IOCCC29 is the second contest since organizers ended a four-year hiatus running from 2020 to 2024. 2025’s winners were presented via a YouTube livestream on June 6th before the source went to the contest's GitHub repository.
Yusuke Endoh, whose self-modifying quines Uellenberg credits as the reason he started writing them, took three of this year’s awards. Run since 1984 by Landon Curt Noll and published under a Creative Commons license, the contest drew a judges’ note on the Pong entry asking entrants to resist building “Quine Doom” for next year.
The entry sits alongside other recent feats that wring full systems out of unlikely materials, including an x86 CPU emulator written entirely in CSS and a playable build of Doom rendered in pure SQL. The judges plan to open IOCCC30 toward the end of 2026.
NewsBy Luke JamesPublished June 29, 2026
The IOCCC29 winner writes each frame as fresh C source, compiles it, then runs it.
Jonah Uellenberg won the Ping Pong Prize at the 29th International Obfuscated C Code Contest earlier this month, with a version of Pong that recompiles its own source code on every frame. Each run of the program prints the C source for the next frame, with the current display and game state embedded inside it, so compiling and running that output advances the game by a single step. Somewhat ironically, Uellenberg didn’t write the obfuscated code himself; instead writing clean, commented source in a programming language he built called Insert, whose compiler generated the scrambled C that took the award.
In his entry notes, Uellenberg writes that the winning code “isn't really obfuscated. It never was,” stressing that it’s compiler output rather than anything mangled by hand. The Insert compiler targets C and runs the size-optimization passes that a production compiler would, such as liveness analysis to strip dead code, constant folding, and a variable-merging pass that reuses one variable slot for many values. Repeated token sequences are then packed into C preprocessor defines to shave bytes. By his account, optimization and obfuscation tend to collapse into the same output, which is why the result reads as a wall of single-character identifiers and a data array that mixes constant and mutable values.
A quine is a program that prints its own source code when you run it. Uellenberg's version prints a modified copy of itself: the source for the next frame, with the updated game state baked in. Run that, and it prints the frame after, and so on.
The player moves the right paddle by passing “w” or “e” as a command-line argument, competes against a computer opponent, and runs a bundled script that automates the compile-and-run cycle so the match plays in real time. Beating the computer switches the program to a second game, a side-scrolling jumper modeled on the Chrome dinosaur runner, and clearing that switches it back.
IOCCC29 is the second contest since organizers ended a four-year hiatus running from 2020 to 2024. 2025’s winners were presented via a YouTube livestream on June 6th before the source went to the contest's GitHub repository.
Yusuke Endoh, whose self-modifying quines Uellenberg credits as the reason he started writing them, took three of this year’s awards. Run since 1984 by Landon Curt Noll and published under a Creative Commons license, the contest drew a judges’ note on the Pong entry asking entrants to resist building “Quine Doom” for next year.
The entry sits alongside other recent feats that wring full systems out of unlikely materials, including an x86 CPU emulator written entirely in CSS and a playable build of Doom rendered in pure SQL. The judges plan to open IOCCC30 toward the end of 2026.
I know, right? Thought you would dig this if you haven't yet.
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
jlv
- Site Admin
- Posts: 15473
- Joined: Fri Nov 02, 2007 5:39 am
- Team: No Frills Racing
- Contact:
Re: MXS Twitter
That pong game sounds brutal!
How do you cut the spines on something that big? You'd use a broach for something small right?
How do you cut the spines on something that big? You'd use a broach for something small right?
Josh Vanderhoof
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
We have several different gear hobs. One is a big CNC unit that has internal spline cutters for ID splines using reciprocal cutting, and older machines that have external, rotary hobs that coordinate linear and rotary motion to cut the external splines. I've never seen a standard broach in our inventory yet. Those can look like a corncob or a single-row saw blade. The tons of torque generated by the finished product is staggering!
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa
-
jlv
- Site Admin
- Posts: 15473
- Joined: Fri Nov 02, 2007 5:39 am
- Team: No Frills Racing
- Contact:
Re: MXS Twitter
It seems like cutting the splines would allow warping as well. Do you have to cut those in stages too?
Josh Vanderhoof
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
Sole Proprietor
jlv@mxsimulator.com
If you email, put "MX Simulator" in the subject to make sure it gets through my spam filter.
-
TeamHavocRacing
- Posts: 8657
- Joined: Thu Nov 19, 2009 5:52 am
- Team: Havoc Racing
- Contact:
Re: MXS Twitter
They take a rough and finish pass. Splining doesn't relieve that much stress. The hexing will cause a bunch more warping if they don't stagger the passes.
jlv wrote:If it weren't for Havoc I'd have been arguing with the 12 year olds by myself.
R.I.P. Sofa