1917661939469111373, 1917661939469111373, 1929158203079619056

2026-08-25 23:40:31 -04:00
1 parent cb99a441a2
commit 3beac7ea20
@@ -0,0 +1,54 @@
+---
+title: "@onatt0 The big industry mistake was factoring into thousands of functions in code, instead of just baking all that into a \"protocol\" of data structures in memory."
+type: archive
+source: twitter
+source_url: "https://x.com/NOTimothyLottes/status/1917661939469111373"
+author: "NOTimothyLottes"
+handle: NOTimothyLottes
+post_id: "1917661939469111373"
+date: 2025-04-30
+archived: 2026-08-25
+draft: false
+tags:
+  - archive
+  - twitter
+  - NOTimothyLottes
+description: "@onatt0 The big industry mistake was factoring into thousands of functions in code, instead of just baking all that into a \"protocol\" of data structures in memory."
+in_reply_to: "1917659010024587634"
+---
+
+## Source
+
+- URL: https://x.com/NOTimothyLottes/status/1917661939469111373
+- Author: NOTimothyLottes (@NOTimothyLottes)
+- Posted: 2025-04-30 19:26:49
+
+## Thread
+
+**1/** **@NOTimothyLottes** ^1917661939469111373
+
+**@onatt0**
+
+The big industry mistake was factoring into thousands of functions in code, instead of just baking all that into a "protocol" of data structures in memory. Like OOP member functions to load or mutate one variable = vomit.
+
+**2/** **@NOTimothyLottes** ^1917662460716212239
+
+**@onatt0**
+
+I do all my custom CPU side stuff more like treating the register file like a "memory" of which the contents are aliased to different shared structures for different purposes across time
+
+**3/** **@NOTimothyLottes** ^1917662813817835551
+
+**@onatt0**
+
+So the register file is more like an aliased global namespace. And "functions" are free of arguments and free of returns. This way of working with the HW is way better and easier than the 'C' model.
+
+**4/** **@NOTimothyLottes** ^1917663219289559511
+
+**@onatt0**
+
+In the few cases where you need to reuse small code patterns, those end up a compile time macros that inline to different registers, larger patterns are already better factored to data
+
+## Related
+
+- [[archive/threads/onatt0/2025-04-30-holy-truthnuke-and-people-think-c-is-the-optimal-state-of]]
@@ -0,0 +1,44 @@
+---
+title: "@emhyr7_ @onatt0 Things could get really interesting if a company like AMD supported transparent data compression on a last level cache which is shared CPU/GPU."
+type: archive
+source: twitter
+source_url: "https://x.com/NOTimothyLottes/status/1917659515773690071"
+author: "NOTimothyLottes"
+handle: NOTimothyLottes
+post_id: "1917659515773690071"
+date: 2025-04-30
+archived: 2026-08-25
+draft: false
+tags:
+  - archive
+  - twitter
+  - NOTimothyLottes
+description: "@emhyr7_ @onatt0 Things could get really interesting if a company like AMD supported transparent data compression on a last level cache which is shared CPU/GPU."
+in_reply_to: "1917658426072617305"
+---
+
+## Source
+
+- URL: https://x.com/NOTimothyLottes/status/1917659515773690071
+- Author: NOTimothyLottes (@NOTimothyLottes)
+- Posted: 2025-04-30 19:17:11
+
+> @NOTimothyLottes @onatt0 i had the same idea (except for compression) but got frustrated at how tedious it is to specify memory layouts in C. in assembly, it would be much easier: just a bunch of `dd`/`dq`.
+
+## Thread
+
+**1/** **@NOTimothyLottes** ^1917659515773690071
+
+**@emhyr7_** **@onatt0**
+
+Things could get really interesting if a company like AMD supported transparent data compression on a last level cache which is shared CPU/GPU. Something like that could likely get argument compression, saving bandwidth.
+
+**2/** **@NOTimothyLottes** ^1917659868141400171
+
+**@emhyr7_** **@onatt0**
+
+Also another interesting note, is that a argument scatter model acts like a cache warming for future argument reads because CPU side it's read-modify-write when scatter storing into cachelines.
+
+## Related
+
+- [[archive/threads/NOTimothyLottes/2025-04-30-3-i-got-side-tracked-by-building-a-language-that]]
@@ -0,0 +1,38 @@
+---
+title: "@pankkor @onatt0 I personally go free form, no data stack, no arguments, just branch points and program specific aliased usage of registers (regions of code have their own \"structure\" of register usage)."
+type: archive
+source: twitter
+source_url: "https://x.com/NOTimothyLottes/status/1929158203079619056"
+author: "NOTimothyLottes"
+handle: NOTimothyLottes
+post_id: "1929158203079619056"
+date: 2025-06-01
+archived: 2026-08-25
+draft: false
+tags:
+  - archive
+  - twitter
+  - NOTimothyLottes
+description: "@pankkor @onatt0 I personally go free form, no data stack, no arguments, just branch points and program specific aliased usage of registers (regions of code have their own \"structure\" of register usage)."
+in_reply_to: "1929128689293304192"
+---
+
+## Source
+
+- URL: https://x.com/NOTimothyLottes/status/1929158203079619056
+- Author: NOTimothyLottes (@NOTimothyLottes)
+- Posted: 2025-06-01 12:48:52
+
+> @NOTimothyLottes @onatt0 Interesting. Do you then inline all the calls and patch registers in reliance to tracked "top"? What do you do for calls that you don't want to be inlined?
+
+## Thread
+
+**1/** **@NOTimothyLottes** ^1929158203079619056
+
+**@pankkor** **@onatt0**
+
+I personally go free form, no data stack, no arguments, just branch points and program specific aliased usage of registers (regions of code have their own "structure" of register usage). APIs based on data structures in memory, not calls. Etc.
+
+## Related
+
+- [[archive/threads/NOTimothyLottes/2025-04-30-3-i-got-side-tracked-by-building-a-language-that]]
@@ -52,6 +52,9 @@ tags:
 - [[archive/threads/NOTimothyLottes/2025-04-30-3-i-got-side-tracked-by-building-a-language-that]]
 - [[archive/threads/NOTimothyLottes/2025-04-30-i-laugh-when-people-say-c-is-like-assembly-they]]
 - [[archive/threads/NOTimothyLottes/2025-04-30-it-would-be-interesting-to-size-stuff-so-the]]
+- [[archive/threads/NOTimothyLottes/2025-04-30-the-big-industry-mistake-was-factoring-into]]
+- [[archive/threads/NOTimothyLottes/2025-04-30-things-could-get-really-interesting-if-a-company]]
+- [[archive/threads/NOTimothyLottes/2025-06-01-i-personally-go-free-form-no-data-stack-no]]
 - [[archive/threads/NOTimothyLottes/2025-07-06-so-lets-root-skg-the-core-problem-is-near-zero]]
 - [[archive/threads/NOTimothyLottes/2025-07-08-0-physical-video-games-are-dying-mvgs-take-https]]
 - [[archive/threads/NOTimothyLottes/2025-07-23-nobody-self-includes-with-defines-to-reorder-c]]
@@ -28,6 +28,7 @@ Example of the note shape (single spine, no branches): [[archive/threads/How to
 - [[archive/threads/VPCOMPRESSB]]
 - [[archive/threads/VictorTaelin]]
 - [[archive/threads/cmuratori]]
+- [[archive/threads/hsn8086]]
 - [[archive/threads/kenpex]]
 - [[archive/threads/levelsio]]
 - [[archive/threads/mike_acton]]
@@ -38,6 +38,7 @@ not only you have to keep the whole stack around with replicated state, it limit
 ## Related

 - Spine: [[archive/threads/NOTimothyLottes/2025-04-30-3-i-got-side-tracked-by-building-a-language-that]]
+- [[archive/threads/NOTimothyLottes/2025-04-30-the-big-industry-mistake-was-factoring-into]]

 ## Visible chain (transcribed verbatim from screenshot)