2026-10-24 09:31:32 From Liang Ng to Everyone: My slides today omnixtar.github.io/svfig SVFIG 2026-10 Omni3D.pdf 2026-10-24 09:32:57 From mike to Everyone: Email to send abstracts to? 2026-10-24 09:34:14 From Program Chairman Kevin to Everyone: forther@comcast.net 2026-10-24 09:35:51 From Dave Jaffe to Everyone: "abstracts" should be brief & concise - just a couple sentences 2026-10-24 09:37:35 From mike to Everyone: mike@softwarewrighter.com - 2026-10-24 09:41:36 From Dave Jaffe to Everyone: I had an APL daisy wheel for my Diablo printer! 2026-10-24 09:45:37 From mike to Everyone: Replying to "mike@softwarewrighter.com -": FORTH topics - on an IBM 1130 emulator I am writing. asm demo only so far - on COR24-TB development board - 4 .fth/.s implementations - live demos < favorite - in WASM in browser - live demo - on a new OS - live demo - to implement a new language, Tuplet - to run apps on GPUs - to generate Verilog for FPGAs to define ML circuits - on Arudino Uno Q for testing GPU eventually 2026-10-24 09:46:26 From Nelo Mitranim to Everyone: Brevity is well and good, and building a lot of useful math operations into the language is extremely useful, but I never understood the fascination with operators. Regular ASCII function calls like log(123 234) are perfectly clear, cost barely more space, and are scalable in the sense that you can make more names from regular letters, without having to mess with hieroglyphics or keyboard layouts. 2026-10-24 09:46:33 From Dave Jaffe to Everyone: I attended in-person presentations given by Ken Iverson & Grace Hopper in Chicago 2026-10-24 09:48:21 From mike to Everyone: Replying to "Brevity is well and good, and building a lot of us...": I have implemented APLs, with/without special keyboard, with/without greek/symbols. I have live demos. I do have a custom keyboard, too and a 2741 emulator. 2026-10-24 09:52:37 From Nelo Mitranim to Everyone: Replying to "Brevity is well and good, and building a lot of us...": How would you feel about the idea that it would be nice for everyone if math in general got closer to programming in ASCII (with less weird notation)? I realize this goes against centuries of tradition, but using one function call syntax, with names, feels to me like it would make math more understandable in general, compared to the traditional trend of over-complicating the notation in favor of brevity. 2026-10-24 09:52:59 From Dave Jaffe to Everyone: Try APL! - https://tryapl.org 2026-10-24 09:53:55 From Liang Ng to Everyone: Replying to "Brevity is well and good, and building a lot of us...": https://urldefense.com/v3/__https://github.com/udexon/SymForth__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6K-FrmR4Q$ 2026-10-24 09:54:19 From Liang Ng to Everyone: Replying to "Brevity is well and good, and building a lot of us...": I think this may have what you want .... 2026-10-24 09:55:11 From mike to Everyone: Replying to "Brevity is well and good, and building a lot of us...": I’m working on this problem. I have a Rosetta Stone approach to switch between math, notation, and visuals. The source code can be: equations, tensor pipelines, my notation (typed as ASCII, rendered with ligatures) 2026-10-24 09:55:26 From Stephen Adels 1 to Everyone: Nelo : Or perhaps a dual syntax. Yes, words would be useful for beginners. But the tersity has its own usefulness. 2026-10-24 09:56:26 From mike to Everyone: Replying to "Brevity is well and good, and building a lot of us...": You can read the source as-typed, so few symbols (on a standard keyboard) 2026-10-24 09:57:36 From mike to Everyone: Isn’t gnu-all close to APL2? 2026-10-24 10:04:13 From Curtis Jones to Everyone: curtis_jones@ieee.org 2026-10-24 10:04:29 From mike to Everyone: I implemented APL\360 ca. 1970 version, and IBM 5100 APLSV ca. 1975 - https://urldefense.com/v3/__https://sw-apl.softwarewrighter.com/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6Le2M6NoQ$ usable on a phone or in the browser, also works at CLI 2026-10-24 10:06:32 From Andrew Bias to Everyone: may not belong here; but what made people choose this instead of say FORTRAN? was it based off readability or something else? 2026-10-24 10:06:51 From Liang Ng to Everyone: Replying to "Brevity is well and good, and building a lot of us...": SymForth is FORTH on C++. Now part of LLASMA Large Language Stack Machine Architecture https://urldefense.com/v3/__https://github.com/llasma/llama.cpp/tree/20260421/CHANGES__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6LQVf4bcw$ 2026-10-24 10:07:14 From Dave Jaffe to Everyone: IBM 5100 https://urldefense.com/v3/__https://thumb.wikimedia.org/wikipedia/commons/thumb/2/2d/IBM_5100_-_MfK_Bern.jpg/250px-IBM_5100_-_MfK_Bern.jpg?utm_source=en.wikipedia.org&utm_campaign=parser&utm_content=thumbnail__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6K8PpOBZQ$ 2026-10-24 10:08:19 From Dave Jaffe to Everyone: IBM 5100 https://urldefense.com/v3/__https://en.wikipedia.org/wiki/IBM_5100__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6KA6aHY6A$ 2026-10-24 10:08:48 From Martin Andronikos to Everyone: nvidia perhaps 2026-10-24 10:09:25 From Brian Stuart to Everyone: Dave, was that Chicago meeting NCC '81 with the founder's day celebrating the UNIVAC? 2026-10-24 10:10:25 From Dave Jaffe to Everyone: No, I was in CA by '81 -- It had to be prior to '79. 2026-10-24 10:10:42 From Martin Andronikos to Everyone: have you all heard about Uiua per channce? 2026-10-24 10:11:56 From Brian Stuart to Everyone: I was wondering because I was at Chicago in '81 and I've never found a list of who spoke that day, much less any transcripts, 2026-10-24 10:13:23 From Brad Nelson to Everyone: What is the I-Beam syntax mentioned? (a system services operator of some sort?) 2026-10-24 10:13:26 From Stephen Adels 1 to Everyone: My question was, what do you think about CoSy, which has a distinct Forth flavor, as it is based on Reva Forth. CoSy is here: https://urldefense.com/v3/__https://cosy.com/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6LS6AZvRQ$ 2026-10-24 10:13:49 From paul durao to Everyone: Inn abandoned smalltalk for Java Eclipse was written in smalltalk Don’t know smalltalk machines ran apl 2026-10-24 10:14:21 From Dave Jaffe to Everyone: One clarification - the presentations I attended by Iverson & Hopper were separate events 2026-10-24 10:17:10 From paul durao to Everyone: Who was the person that implemented apl ? For historical purposes 2026-10-24 10:17:50 From paul durao to Everyone: Smalltalk did real numbers of arbitrary precision 2026-10-24 10:18:02 From Curtis Jones to Everyone: Link to slides on APL Meeting Notices https://urldefense.com/v3/__https://docs.google.com/presentation/d/1Q1MPxLUYlloye2mxrYrxEbLGlNJsCC9kZXrzTzyTU3E/edit?usp=sharing__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6LJP47jhA$ curtis_jones@ieee.org 2026-10-24 10:18:57 From paul durao to Everyone: If ibm abandoned smalltalk for Java you’d think they’d abandon apl as well But there must be some reason for not abandoning apl as much as smalltalk 2026-10-24 10:19:45 From Curtis Jones to Everyone: Replying to "Smalltalk did real numbers of arbitrary precision": APL\360 was implemented by Larry Breed, Roger Moore and Phil Abrams (I think) with support from Ken Iverson and Adin Falkoff - and doubtless others at IBM. 2026-10-24 10:20:14 From mike to Everyone: Replying to "What is the I-Beam syntax mentioned? (a system ser...": https://urldefense.com/v3/__https://docs.dyalog.com/20.0/language-reference-guide/primitive-operators/i-beam/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6Lg8su5ag$ 2026-10-24 10:21:18 From mike to Everyone: Replying to "Who was the person that implemented apl ? For hist...": https://urldefense.com/v3/__https://en.wikipedia.org/wiki/Kenneth_E._Iverson__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6LA1Or_-A$ 2026-10-24 10:23:15 From Curtis Jones to Everyone: Replying to "What is the I-Beam syntax mentioned? (a system ser...": IBM had paying customers using APL. 2026-10-24 10:23:58 From paul durao to Everyone: Fpga could be interesting for apl as well 2026-10-24 10:24:59 From paul durao to Everyone: APL never made it to a gui always stayed on tty ? 2026-10-24 10:25:56 From Curtis Jones to Everyone: Replying to "Who was the person that implemented apl ? For hist...": APL\360 was implemented by Larry Breed, Roger Moore and Phil Abrams (I think) with support from Ken Iverson and Adin Falkoff - and doubtless others at IBM. (I put your answer in the wrong place.) 2026-10-24 10:26:20 From mike to Everyone: Replying to "APL never made it to a gui always stayed on tty ?": APL supported comminicting via shared variables which could be a file system, a graphics display, etc. 2026-10-24 10:28:45 From Curtis Jones to Everyone: Replying to "APL never made it to a gui always stayed on tty ?": There was a GUI auxilliary processor that came with APL/SV about 1975. There has been GUI support ever since - in most APLs. 2026-10-24 11:10:17 From paul durao to Everyone: Who was the person in the group implementing retro apl etc 2026-10-24 11:12:14 From mike to Everyone: Forth and APL on 24-bit RISC FPGA dev board 2026-10-24 11:15:45 From paul durao to Everyone: What filtering 2026-10-24 11:16:18 From Nelo Mitranim to Everyone: Bots are great for info search, good at looking for bugs (won't find all), useful for mechanical refactoring, and terrible at design. 2026-10-24 11:16:23 From Bob Armstrong to Everyone: APL had an interactive ` workspace from the beginning , in the days of punch cards 2026-10-24 11:16:47 From Andrew Bias to Everyone: Newcomers prospective: Thus far my prespective is that the community is nice; documentation is great. While I understand stacks, I have found university professors to steer away/tend to express concerns about security. 2026-10-24 11:20:34 From paul durao to Everyone: These ai can’t drive Plus they have censorship 2026-10-24 11:21:34 From Nelo Mitranim to Everyone: I have issues with using the term "AI" for the current crop of bots. All for actual artificial intelligence but we don't have that yet. We have non-intelligent bots whose limitations must be recognized to use them effectively. 2026-10-24 11:21:53 From paul durao to Everyone: Often ai leaves the most interesting stuff out 2026-10-24 11:22:31 From paul durao to Everyone: lol 2026-10-24 11:24:19 From Bob Armstrong to Everyone: " Yesterday , Today & tomorrow " w Sophia Loren 2026-10-24 11:28:56 From paul durao to Everyone: Our group has had very good speakers over the years 2026-10-24 11:34:53 From paul durao to Everyone: Holon forth was interesting 2026-10-24 11:35:03 From paul durao to Everyone: What computer was this 2026-10-24 11:35:12 From paul durao to Everyone: Kay lack was good 2026-10-24 11:36:55 From paul durao to Everyone: America has a lot more computing hardware than Europe 2026-10-24 11:37:47 From paul durao to Everyone: When is forth day link ? 2026-10-24 11:37:52 From Brian Stuart to Everyone: If I weren't on the opposite coast, I'd love to take you 11. 2026-10-24 11:40:26 From paul durao to Everyone: Holon forth was interesting with the smalltalk style tcl tk 2026-10-24 11:44:01 From Liang Ng to Everyone: Replying to "I have issues with using the term "AI" for the cur...": We may measure intelligence with cardinality of FORTH words (total number of FORTH words in the system) -- classify robots or AI according to tests with increasing difficulties. 2026-10-24 11:44:05 From paul durao to Everyone: Temple os could mix multiple objects 2026-10-24 11:45:13 From paul durao to Everyone: For real ai u need a feedback loop control system type thing … 2026-10-24 11:47:26 From Brad Nelson to Everyone: AppEngine 2026-10-24 11:47:59 From Brad Nelson to Everyone: Though over the years it's become just a layer on Google Cloud 2026-10-24 11:48:58 From Nelo Mitranim to Everyone: @Mike Wright I like that your APL demo is less than a megabyte of WASM 2026-10-24 11:49:38 From Nelo Mitranim to Everyone: Seen so, so much worse 🙂 2026-10-24 11:51:23 From Mike Wright to Everyone: https://urldefense.com/v3/__https://sw-vibe-coding.github.io/sw-fth-wasm/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6KepO9kOA$ 2026-10-24 11:51:36 From Mike Wright to Everyone: https://urldefense.com/v3/__https://sw-embed.github.io/web-sw-cor24-forth/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6KGqInt8g$ 2026-10-24 11:51:45 From Mike Wright to Everyone: https://urldefense.com/v3/__https://swtos.softwarewrighter.com/forth/__;!!G92We9drHetJ8EofZw!dufgZkip4zlCFeOVo84Q0JfFYejjmMpKF_LfN8lU9Ml56qElAFRbHOgPsoxsAd2wfpWNuGQrnDpYC6JzMty1Hw$ 2026-10-24 11:56:37 From Nelo Mitranim to Everyone: Replying to "I have issues with using the term "AI" for the cur...": Just like disk capacity doesn't correspond to any kind of intelligence, dictionary size doesn't correspond to that either. Nor to sentience. We need different metrics for that. It's an annoying tradeoff between software and wetware (and hardware design) that for anything artificial, you want the simplest possible design, but for living creatures to function, high complexity seems to be required... Still, a simple metric like vocabulary, while tempting to be used as a proxy metric, I don't believe it can reflect intelligence or sentience. 2026-10-24 11:57:14 From paul durao to Everyone: No. Y is rust being pushed 2026-10-24 11:58:36 From paul durao to Everyone: There is a new ai called Jarvis That takes actions With Claude underneath Scary when they can take actions 2026-10-24 11:58:57 From paul durao to Everyone: What was the question point about rust ? 2026-10-24 12:00:28 From paul durao to Everyone: Ai often does not cover the most interesting novel points It’s always trying to be non controversial 2026-10-24 12:00:33 From Nelo Mitranim to Everyone: Rust comes with a very useful mental model of ownership and lifetimes, worth adopting when programming in any language. Comes with major tradeoffs too, like overcomplicated syntax, compiler internals, compilation times. The model is worth adopting, but other language authors (see Swift, Nim) disagree with inflicting such a cost on the users. 2026-10-24 12:01:18 From Liang Ng to Everyone: Replying to "I have issues with using the term "AI" for the cur...": Onus probandi -- we must start with a positive hypothesis.