2026-01-23 02:05:28 lispmacs[work], the RPI pico is the cheapest fully assembled board in history, to get cheaper you have to go to individial cpus on a board of your own making 2026-01-23 02:07:20 lispmacs[work], but be aware that buying the cheapest pico is just asking for problems as they can use really cheap SPI flash that wear out quickly. Ive had users temm me their SPI flash has suddenly 'died' after a few flashes 2026-01-23 09:22:38 lispmacs[work] Look at Mecrisp Forth.  These micros are very cheap and board level development kits are available from many manufacturers including ti.com 2026-01-23 12:18:47 what do you prefer of these 2 syntax? https://bpa.st/GBQQ 2026-01-23 12:59:58 The one using : and ;, by far. I see no need whatsoever to "verbose things up" with written out words. 2026-01-23 13:01:38 But in fairness I have a strong preference for APL as opposed to "wordy" equivalents like J. 2026-01-23 13:01:55 To me the terse symbols are an important part of what APL *IS*. 2026-01-23 13:02:24 It's supposed to be "extended arithmetic," and no one would advocate replacing + with PLUS and - with MINUS. 2026-01-23 13:02:41 But in spirit that's exactly what J does. 2026-01-23 13:03:45 If an elementary school student complained that it was so much easier to know what to do with PLUS and MINUS, the most common response would be "JUST LEARN THE SYMBOLS." That's my APL vs. J response. 2026-01-23 14:51:28 https://www.electronicdesign.com/blogs/nonlinearities/blog/55352019/electronic-design-states-ban-guns-mags-and3d-printers-and-cnc-machines?o_eid=9835B9248256I1T&rdx.ident[pull]=omeda|9835B9248256I1T 2026-01-23 14:51:31 Ugh. 2026-01-23 18:21:51 KipIngram: Checkmate black market gun makers; what are you going to do, not tell us about your CNCs? 2026-01-23 18:22:21 Reminds me of the UK approach to student loan repayment for foreign students where they have to tell them that they're earning money abroad to repay 2026-01-23 18:23:05 I'm sure they will contact SLC immediately upon getting work abroad, don't worry HMRC 2026-01-23 18:30:45 rendar_: Here's some design I made not long ago https://pastebin.com/raw/09WqA7Fr 2026-01-23 18:31:24 I think either you can make the recursive syntax look a lot like Lua or C and it will work out 2026-01-23 18:31:45 Especially if you use tokens instead of whitespace delimiting 2026-01-23 18:32:30 veltas: wow great, why you use ( ) ? 2026-01-23 18:34:40 They're optional, just either statically or dynamically (on debug build) check that the stack increases by exactly 1 by the second paren 2026-01-23 18:35:08 Based on this throwaway concept https://veltas.co.uk/blog/forth-expr.htm 2026-01-23 18:46:49 I think I decided using { and } for structure instead of lots of words was too similar to Rust; and likewise I just thought the semicolons were too weird after multiple nestings 2026-01-23 18:52:50 But really, if you go too complicated with any RPN then eventually I wonder what the point of it being an RPN is 2026-01-23 18:53:10 Given it's already a huge concession to be an RPN in the first place, in the name of reducing the size and complexity of the implementation 2026-01-23 18:53:28 Some might like the style but most (including myself) don't prefer it, even if I enjoy it sometimes 2026-01-23 18:53:56 So really I think the most 'practical' RPN will always be a minimal application-driven Forth, not an overly 'comfortable' or 'general' programming environment 2026-01-23 19:18:04 isene: Done anything else with Simplicity OS? 2026-01-23 19:23:49 It looks from the logs like a lot of time is spent debugging? 2026-01-23 21:02:39 veltas: I have a few dozen concurrent projects, so it's gotten less attention lately. And it was in interesting run with changing the Forth to something I like better - so more tweaking and adjusting than just debugging. When it grabs hold of me again, I'll move it further 2026-01-23 21:08:16 veltas: We made a lot of hay over the years here in the U.S. about "innocent until proven guilty." That is, the PRESUMPTION should be that people are going to behave legally, and we deal individually with the ones that don't and refrain from stepping on everyone's freedom in an attempt to make it "impossible" to commit crime. 2026-01-23 21:08:24 Some of us still think that way. 2026-01-23 21:08:55 The barrel will have some bad apples - you deal with them individually. 2026-01-23 21:10:30 Besides, statistics I've seen (from 2015 - it's just the latest I know data from, because I generally don't just chase this kind of thing) show that only one privately owned gun in the U.S. out of every TEN THOUSAND is ever used to commit a violent crime. So for every gun that is, there are 9,999 that are not. 2026-01-23 21:11:48 I'm sorry - I misspoke. Not "violent crime." Actual deaths by illegal gun use. 2026-01-23 21:12:12 In 2015 there were about 36,000 such deaths, and in that same year the "private gun quantity" was estiamted at 350 million. 2026-01-23 21:12:49 So that factor's really even bigger than 10,000 if you take into account that in many cases multiple deaths attribute to a single gun. 2026-01-23 21:14:48 Also, making my Forth look like C (or similar) is exactly what I do NOT want to do. :-) 2026-01-23 21:15:35 I push hard for (fairly short) single-line definitions, organized into small groups that relate to a given functionality. 2026-01-23 21:16:08 And in a hierarchical way - within such a group (which might be 15-ish lines) there are sub-groups. Factor, factor, factor. 2026-01-23 21:16:45 Usually only one or two words in a 15-ish group produces a name that I retain long-term in the dictionary - the rest are "helpers" that I prune out. 2026-01-23 21:21:34 ACTION is currently designing a thermometer that uses a STM32G030 instead of his older STM32F051. Its a learning experience as a all the register absolute addresses have changed and a lot of the CMSIS syntax between the two models 2026-01-23 21:22:39 as FURS changes syntax to absolute addresses, the new syntax stands out when running the old source thu it 2026-01-23 21:35:18 Unless I'm misremembering what CMSIS is, I've always used Libopencm3 instead for STM32F MCUs. Not that I have much experience with Cortex M, or ARM in general, though. 2026-01-23 21:37:43 iv4nshm4k0v, I use Mecrisp-Stellaris Forth for my embedded projects with FURS for the Libopencm3 equavilent functionality per model 2026-01-23 21:38:57 iv4nshm4k0v, https://mecrisp-stellaris-folkdoc.sourceforge.io/furs/blog-furs.html 2026-01-23 21:45:08 AHA! My phone service is Google Fi, and it turns out that I can receive calls in my browser from the Fi website. Amazon did have a "verify by call" option - I chose that and got myself logged in. 2026-01-23 21:46:42 And it looks like if I WANT to I can arrange to have my texts go to Google Chat. I'd deliberately refrained from doing that initially, but maybe after this debacle I'll reconsider. I don't know if I can set it up now, though because it looks like establishing the link requires the phone. 2026-01-23 21:57:56 ... That MCUs may not have enough flash for having all the various constants in the dictionary is something I didn't quite consider... 2026-01-23 22:03:26 iv4nshm4k0v, I'm sure some do, especially with the larger 1 -2 MB flash models, but how does the programmer get these into the dictionary ? type them all by hand ? 2026-01-23 22:04:52 iv4nshm4k0v, mard selected constants from a huge list ? perhaps select constants by peripheral, and load all the register constants in a peripheral ? 2026-01-23 22:05:01 mard = mark 2026-01-23 22:05:41 iv4nshm4k0v, Ive done all these methods years ago, theyre inefficent and wasteful