2026-07-04 00:23:42 <oak         > Noodling with an idea for catjam. Envisioning writing Forth differently... https://kira.eight45.net/pub/visual-forth.png
2026-07-04 00:40:03 <oak         > I have a hard time keeping the stack in my head AND doing problem-solving, so I thought it'd be nice if the computer could visually track it for me in real-time.
2026-07-04 01:28:28 <cleobuline  > you can add a user root on macos
2026-07-04 01:32:17 <cleobuline  > veltas:  https://labynet.fr/expert/CAM-8.dmg
2026-07-04 01:32:32 <cleobuline  > beta version
2026-07-04 07:06:22 <xentrac     > cleobuli_: the user root on macos is not the true root
2026-07-04 08:19:42 <veltas      > I don't have a mac, I just own Apple
2026-07-04 08:20:22 <iv4nshm4k0v > 
ACTION does not own Apple, but has a couple of apples in his pantry 

2026-07-04 08:20:36 <veltas      > I was tempted to get the $500 Neo and was a little upset I didn't get it before the price went up
2026-07-04 08:21:00 <veltas      > Which I was expecting but there was some controversy in my household about me getting it, probably would have been a waste of money anyway
2026-07-04 08:23:27 <xentrac     > controversy in your household?
2026-07-04 08:23:39 <veltas      > Yes my wife was saying I wouldn't use it or it would just break immediately
2026-07-04 08:23:46 <veltas      > She's probably right on both accounts
2026-07-04 08:24:11 <xentrac     > what was the Neo?
2026-07-04 08:24:31 <veltas      > A cheap macbook they released a few months ago
2026-07-04 08:25:03 <xentrac     > ah
2026-07-04 08:26:28 <veltas      > I promise as a shareholder I will vote for rootable devices if the question is ever presented to the shareholders
2026-07-04 08:26:55 <veltas      > I would guess that a lot of investors would prefer that and see value in it, whether profit or just social value
2026-07-04 08:28:22 <xentrac     > it will never be presented to the shareholders, but you can raise it as an issue
2026-07-04 08:28:44 <veltas      > I don't think my 1 share would let me do that, I suspect you need to have a certain number to be able to do that
2026-07-04 08:28:50 <xentrac     > shareholders generally just vote for the board of directors
2026-07-04 08:28:52 <veltas      > I could be wrong
2026-07-04 08:29:22 <veltas      > I did vote last time and it was more than board questions
2026-07-04 08:29:23 <xentrac     > I think you're entitled to attend the shareholder meeting and maybe even to speak
2026-07-04 08:30:06 <veltas      > Shareholders can bring forward things to vote on, and the board will just comment on whether they agree with the motion or not
2026-07-04 08:32:41 <xentrac     > I suspect *that* might require more than one share
2026-07-04 08:47:56 <veltas      > This was done with the Right to Repair stuff and something ended up being agreed, I don't know how good it was but sounds like there was a little progress
2026-07-04 08:48:04 <veltas      > We just need right to software repair too
2026-07-04 08:50:26 <xentrac     > yes, that's what the free software movement is about.  a friend of mine is visiting the executive director of the FSF this week; I'm interested to hear his report on how she sees the current situation
2026-07-04 08:56:28 <cleobuli_   > xentrac:  what you mean not a real root
2026-07-04 08:58:12 <lofty       > Uh, I'm guessing System Integrity Protection
2026-07-04 09:00:04 <cleobuli_   > it may also be disable
2026-07-04 09:00:48 <iv4nshm4k0v > How "real" could "root" even get on a system with a TPM?  Or Intel ME?
2026-07-04 09:02:37 <xentrac     > iv4nshm4k0v: that's a worrisome question
2026-07-04 09:05:30 <veltas      > It sounds like you can install Linux on macbooks
2026-07-04 09:05:37 <veltas      > So I guess if you care you can install Linux
2026-07-04 09:18:13 <cleobuli_   > it does not have any sense to install  linux on a macbook
2026-07-04 09:18:55 <cleobuli_   > macos is a posix machine
2026-07-04 09:20:25 <lofty       > eh, it has a rather clumsy developer experience in my experience
2026-07-04 09:21:59 <cleobuli_   > Mac OS X is based on the Mach kernel and the BSD implementation of Unix,
2026-07-04 09:22:20 <lofty       > I am aware
2026-07-04 09:22:24 <cleobuli_   > everithing you develop on linux mays run on a macos x
2026-07-04 09:23:19 <cleobuli_   > i develop the forthBot on linux and it run on macos x without any modifications
2026-07-04 09:23:59 <cleobuli_   > ( from sources with gcc )
2026-07-04 09:24:19 <lofty       > https://ck-hack.blogspot.com/2013/09/unnamed-semaphores-and-pososx.html?m=1 <--- that's not always been true
2026-07-04 09:27:43 <lofty       > (also, famously, macOS considers OpenGL deprecated and does not support Vulkan)
2026-07-04 09:30:38 <lofty       > I've generally been in favour of Apple's approach to hardware; lots of relatively small-surface-area accelerators rigorously protected by IOMMUs, rather than the "one CPU which can access the universe" philosophy.
2026-07-04 09:30:56 <lofty       > But macOS in general is not the most fun to work with.
2026-07-04 10:04:45 <cleobuli_   > https://labynet.fr/videos/worms.mp4
2026-07-04 10:04:58 <cleobuli_   > beta 0.2
2026-07-04 10:13:15 <cleobuli_   > macos is showing off
2026-07-04 10:47:56 <cleobuli_   > a good reason to buy a macintosh , is to be able to use the labynet's sofware !
2026-07-04 10:48:56 <cleobuli_   > https://labynet.fr/videos/dentrites.mp4
2026-07-04 10:49:49 <cleobuli_   > i am very near the complete cam-6 machine of the book https://people.csail.mit.edu/nhm/cam-book.pdf
2026-07-04 10:51:14 <cleobuli_   > the forth is near to be completed :)
2026-07-04 10:54:51 <cleobuli_   > https://ibb.co/S46QLBsy
2026-07-04 12:06:29 <cleobuli_   > better with sound !!!! https://labynet.fr/videos/brain-synchro.mp4
2026-07-04 14:19:18 <veltas      > Computer graphics are not very portable, modern OpenGL support was quite bad on Windows for a while, and OpenGL has lots of differences between GPUs and vendors.  If you want to support a range of OS's and machines then you have to abstract as much of it away as possible, and write a bunch of platform-specific shaders, unfortunately.
2026-07-04 14:19:35 <veltas      > Apple say that you can convert shaders to Metal shaders, so maybe some of that work is easier
2026-07-04 14:20:33 <veltas      > It seems to me like if you want a really broad support for new and old devices you want to have a classic OpenGL option, and vulkan option.  There is a third-party vulkan compatibility layer for metal.
2026-07-04 14:50:44 <iv4nshm4k0v > On top of that, nine times out of ten, to use a GPU at all, you need proprietary software components on your OS - even if it's something that runs on the GPU itself (so-called "GPU firmware.")  Personally, I've decided that if something requires OpenGL, I don't need it.
2026-07-04 15:08:02 <veltas      > There are okayish CPU GL implementations, I'm interested in playing with at some point
2026-07-04 15:13:51 <iv4nshm4k0v > Not that I'm an expert in such things, but given that unaccelerated graphics existed for decades before GPU became commonplace, I'd expect there to be APIs for the same.  Such as X Window, for example.  What advantage might OpenGL have over such APIs, when GPUs are not in use?
2026-07-04 15:14:38 <veltas      > I suspect the X Window stuff actually is accelerated and faster even than OpenGL for certain things
2026-07-04 15:15:11 <veltas      > I know crc for instance had serious trouble getting SDL to run right for X ilo
2026-07-04 15:15:17 <veltas      > But X ilo using X ran just fine
2026-07-04 15:16:10 <veltas      > The old 2D windowing / graphics APIs are probably really fast on older hardware for many things
2026-07-04 15:16:52 <veltas      > I'm just thinking about 3D stuff specifically right now, but yeah OpenGL isn't a good idea for all 2D work
2026-07-04 15:23:53 <iv4nshm4k0v > From what's been discussed on Usenet (news:comp.windows.x or news:comp.misc, I think) somewhat recently, I've got an impression that 2-D acceleration in X, if present, these days is implemented on top of 3-D acceleration features of the underlying hardware, if any.  Other than that, on x86, I'm using BIOS-initialized (i. e., VBE) framebuffer, so X there is no faster than memory access would allow.
2026-07-04 15:23:54 <iv4nshm4k0v > X on NetBSD might still support accelerated 2-D on specific video cards, but that's 30 years-or-so old hardware, and even being a retrocomputing enthusiast myself, I'm not at all sure of the practicality of using such systems.  I /think/ that, say, Olinuxino A64 is going to outperform a Pentium-based machine on pretty much any benchmark, and it's OSHW, too.
2026-07-04 15:24:42 <lofty       > https://www.freedesktop.org/wiki/Software/Glamor/
2026-07-04 16:26:26 <cleobuli_   > https://labynet.fr/videos/worms.mp4
2026-07-04 16:48:39 <veltas      > iv4nshm4k0v: I've tried running recent builds of Linux/X on older hardware and it's just dreadful.  Over years there's been different approaches to acceleration and the refactors probably ruin a lot of older hardware.  My main point is 'acceleration' doesn't imply a 3D-capable GPU, e.g. the IBM 8514/A accelerated GDI for Windows 3.x
2026-07-04 16:49:18 <veltas      > Apologies if you know this already, I don't know what you know
2026-07-04 16:51:58 <fkw         > speaking of, i've got a RISC-V machine with a not-very-powerful and proprietary iGPU but 256-bit vector/SIMD extensions
2026-07-04 16:57:03 <veltas      > Thinking of doing some rendering with the SIMD?
2026-07-04 16:58:21 <fkw         > if it's possible, using these instead of a GPU seems simpler for low-level control and it might be more powerful on this system
2026-07-04 17:10:50 <iv4nshm4k0v > veltas: So far as I can tell, 'acceleration' doesn't imply having to use proprietary software, either: I have a Intel D510MO mainboard, and so far as I can tell, its i915 GPU /does/ work with free software drivers, Linux or NetBSD.  Haven't actually tried, say, Trisquel there, but I /do/ expect it to work - deblobbed kernel and everything.
2026-07-04 17:10:50 <iv4nshm4k0v > I've had an AMD RS740-based mainboard, which also didn't seem to require any proprietary software to work, though I don't recall ever running any GL software there.  Same with the aforementioned Olinuxino A64 SBC.  Still, like I've said, nine times out of ten, it's either 'proprietary' or 'no acceleration.'
2026-07-04 17:15:33 <iv4nshm4k0v > To put it differently, I'd be interested in OpenGL once http://ryf.fsf.org/ starts listing affordable GPUs.  So long as it's not the case, there're lots to computing aside of accelerated graphics.
2026-07-04 18:44:11 <xentrac     > historically X had a lot of calls for acceleration, yeah
2026-07-04 18:44:42 <xentrac     > but as iv4nshm4k0v points out, that isn't very useful now
2026-07-04 18:46:30 <xentrac     > I guess in the end this conversation was mostly people telling other people things those other people already knew
2026-07-04 20:08:48 <crc         > I should probably update the SDL ilo at some point
2026-07-04 20:09:02 <crc         > Though I only really use it on the mac now, since I haven't done a native AppKit/Core Graphics implementation there yet
2026-07-04 20:21:00 <oak         > crc: is an "ilo" a konilo package?
2026-07-04 20:21:51 <crc         > ilo is the vm konilo runs on
2026-07-04 20:25:49 <crc         > https://konilo.org/ilo.txt is the core spec, I've written implementations in around 26 languages, and there's a few others written by other people as well.
2026-07-04 20:26:43 <oak         > Aah.
2026-07-04 20:27:55 <oak         > Oh nice, I like this.
2026-07-04 20:29:49 <oak         > Oooh, so is "the SDL ilo" an ilo VM that has I/O devices that map to SDL operations?
2026-07-04 20:32:15 <crc         > https://git.sr.ht/~crc_/retro-ilo/tree/master/item/ilo-vm/ilo-sdl.c and https://git.sr.ht/~crc_/retro-ilo/tree/master/item/ilo-vm/ilo-sdl-1bpp.c
2026-07-04 20:32:38 <veltas      > So yes basically
2026-07-04 20:33:03 <oak         > Thanks veltas ;)
2026-07-04 20:33:10 <veltas      > Well specifically pixel stuff and normal terminal output
2026-07-04 20:33:33 <veltas      > The source is readable though, take a look
2026-07-04 20:35:24 <crc         > I mainly use the 1bpp model, my son does most of the stuff with the bigger one.
2026-07-04 20:35:31 <oak         > Aye, I just did. I like how it's all exposed through a single I/O.. uh, port?
2026-07-04 20:35:49 <oak         > I love how concise it is. ^.^
2026-07-04 20:35:56 <veltas      > Agreed
2026-07-04 20:36:05 <oak         > Is RetroForth built on this?
2026-07-04 20:37:07 <oak         > ilo gives me very similar vibes to uxn -- I wonder if it was inspired by ilo.
2026-07-04 20:38:42 <crc         > RetroForth runs on a similar vm named nga. https://retroforth.org/nga/docs/Nga.md
2026-07-04 20:39:22 <crc         > I'm not aware of ilo or nga having any direct influence on uxn
2026-07-04 20:39:39 <oak         > Oh interesting.
2026-07-04 20:40:17 <oak         > This is really cool, crc. :)
2026-07-04 20:41:04 <crc         >  Nga has a more complex io interface, and the full RetroForth has a lot more ties to the host OS. ilo is intentionally smaller, more self contained
2026-07-04 20:42:06 <crc         > I use Retro to write services and in shell scripts on my BSD and Linux systems, and have a personal computing system written in Konilo that's my daily environment
2026-07-04 20:43:01 <oak         > 
ACTION grins

2026-07-04 20:43:05 <oak         > That sounds fun.
2026-07-04 20:44:06 <oak         > I really like how Forths can be so readily tailored to a person's personal style & preferences.
2026-07-04 22:31:11 <veltas      > crc, oak: When the author of uxn was on here last I think they said that uxn was inspired by nga or ilo
2026-07-04 22:31:35 <veltas      > I'll see if I can find the log to back this up
2026-07-04 22:44:07 <veltas      > https://forth.chat/logs/libera/forth/2024-08-24 stamp 17:44:24
2026-07-04 22:44:24 <veltas      > > <neauoire    > looking for viable candidates, I stumbled on retroforth's ilo and the J1, which became my principal inspiration for the project
2026-07-04 22:49:52 <oak         > oh wow, nice veltas
2026-07-04 22:56:28 <veltas      > 100 rabbits are very cool, I like their stuff / lifestyle
2026-07-04 22:58:34 <oak         > Same.
2026-07-04 22:58:52 <oak         > I appreciate the way they live their values so openly.
2026-07-04 22:59:48 <oak         > veltas: How did you come to use Forth? / What brought you here?
2026-07-04 23:00:35 <veltas      > I had been working on programming environments for the ZX Spectrum, and general stuff in 8-bit, and was recommended Forth by numerous people
2026-07-04 23:01:00 <veltas      > When I started my job I met someone (much older) who used to write Forth at MPE Forth and that provoked me to go actually learn it
2026-07-04 23:01:20 <veltas      > So I read Starting Forth and worked through it in gforth, wrote a few things in it
2026-07-04 23:01:56 <veltas      > Then thought I could probably write one myself so started writing one in Z80, after starting that I joined this channel
2026-07-04 23:02:09 <veltas      > To ask about the standard words or gforth or something
2026-07-04 23:02:38 <veltas      > I wrote my Forth for ZX Spectrum https://github.com/veltas/zenv
2026-07-04 23:02:51 <oak         > You finished writing the Z80 forth? :)
2026-07-04 23:03:38 <veltas      > Is anything ever truly finished?
2026-07-04 23:03:52 <oak         > This is really cool.
2026-07-04 23:04:19 <veltas      > I'm still working on it in a sense but I don't get as much time for programming currently, have couple of young kids
2026-07-04 23:05:22 <veltas      > Thanks oak, I suppose main appeal of it is that it's a relatively complete ANS Forth so it can run your ANS programs with minimal changes
2026-07-04 23:05:54 <oak         > Have you used it to build anything?
2026-07-04 23:05:59 <veltas      > But it's missing some major stuff like ability to save/load tape
2026-07-04 23:06:22 <veltas      > Not yet no
2026-07-04 23:06:42 <veltas      > Ideally I'd like to be able to build itself in itself, I think that's probably an important milestone
2026-07-04 23:08:05 <veltas      > Had been working on a block editor for it too https://www.youtube.com/watch?v=_6w-hjkCi3U