2023-08-02 03:40:19 <veltas      > Regarding lack of the 'ok' prompt?
2023-08-02 03:40:43 <veltas      > I'm still upset when you start a Forth and it doesn't greet you with 'hi' on the first prompt
2023-08-02 03:41:13 <veltas      > Or repeat back words it doesn't understand like 'CUISINART?'
2023-08-02 03:45:12 <iyzsong     > i'd like forth speak to me, without high degree glasses, i'm blind...
2023-08-02 03:52:31 <veltas      > iyzsong: What Forth do you use, and do you have any preferred speaking programs?
2023-08-02 03:52:34 <veltas      > And what OS?
2023-08-02 03:59:16 <iyzsong     > honestly i haven't starting using forth yet.. For OS, I use GNU Guix.
2023-08-02 04:06:17 <iyzsong     > also haven't tried speaking program, just thinking programming could be more comfortable/health without staring at the screen.
2023-08-02 04:07:22 <veltas      > gforth is available on Guix
2023-08-02 04:08:48 <veltas      > It's an interesting question with Forth, because it's conventional for all Forth words to have a "spoken equivalent"
2023-08-02 04:09:07 <veltas      > Like DUP is pronounced 'dupe', U< 'U less', etc.
2023-08-02 04:09:48 <veltas      > But it's not usually programmed into the system. But with an accessible Forth (or accessible layer on another Forth) I suppose you would program those pronunciations in
2023-08-02 04:10:33 <iyzsong     > yes, that sounds powerful
2023-08-02 04:10:42 <veltas      > Sounds like a worthy project, hopefully someone does something with this
2023-08-02 04:42:36 <gordonjcp   > yeah it sounds like it wouldn't be hard
2023-08-02 04:43:01 <gordonjcp   > which of course means it would be
2023-08-02 04:43:10 <gordonjcp   > but, you'd break the back of it by just doing string matching
2023-08-02 04:43:35 <gordonjcp   > "If you're not within quotes, pronounce ! as 'store'" kind of ruels
2023-08-02 04:43:37 <gordonjcp   > *rules
2023-08-02 04:50:33 <iyzsong     > yeah, a semantics/struct aware (like colorForth, call it speakForth?) speak form would be convenient, cann't imagine the forms, have to actually try o.O
2023-08-02 06:09:47 <veltas      > I've thought a little about it and I wouldn't do any string matching, instead I would allow a bypass mechanism and get the interpreter loop to be 'aware' of pronunciation
2023-08-02 06:10:07 <veltas      > And provide the same words so the user can use custom or Forth word pronunciation as needed
2023-08-02 06:10:58 <veltas      > But I would redirect TYPE to just speak everything else that doesn't bypass, so you can hear normal unintelligent output
2023-08-02 06:44:32 <next4th     > make sense.  also if defintion are not too long, it can speak back the full once reach ; and ask for a confirm..
2023-08-02 06:53:48 <next4th     > and use a hierarchical db instead of text source, like https://holonforth.com/
2023-08-02 08:06:15 <KipIngram   > That is an interesting idea.
2023-08-02 08:54:04 <veltas      > Well I thought I might look on lunch break but I can't get audio to work in my Linux VM :)
2023-08-02 08:54:43 <veltas      > I don't think this is especially hard though, I think it's probably relatively easy for someone who understands Forth systems well
2023-08-02 08:55:18 <drakonis    > which linux distribution?
2023-08-02 08:55:47 <veltas      > Ubuntu
2023-08-02 08:55:55 <veltas      > I use Ubuntu on my work VM
2023-08-02 08:58:19 <veltas      > Looks like audio device crashed
2023-08-02 08:58:27 <veltas      > I'll have to reboot so that's a no today
2023-08-02 09:03:27 <drakonis    > ah, that's not too complicated
2023-08-02 09:03:45 <drakonis    > this a windows or mac host?
2023-08-02 09:16:31 <veltas      > Windows host
2023-08-02 09:16:43 <veltas      > Reboot means I open all my documents and terminals up again
2023-08-02 09:16:59 <drakonis    > ah yes
2023-08-02 09:17:01 <veltas      > Don't really have time for that
2023-08-02 09:17:05 <drakonis    > well, you can try wsl2
2023-08-02 09:17:16 <veltas      > Not on my enterprise computer I can't :)
2023-08-02 09:17:24 <drakonis    > fair enough
2023-08-02 09:17:49 <veltas      > Things are getting very corporate and locked down here, and I literally need to get someone from IT to walk down here and add the groups manually for that
2023-08-02 09:17:52 <veltas      > Can't be bothered
2023-08-02 09:18:16 <veltas      > All this 'security' nonsense, miss the good old days
2023-08-02 09:18:33 <veltas      > When it was the wild west and your company would just get wiped out by ransomware
2023-08-02 09:42:46 <KipIngram   > My wife's company had a cyber attack a few months ago.  Bad one, and I'm still not sure they've told the employees everything about it.
2023-08-02 09:44:08 <KipIngram   > At work we're in the process of adding ransomware detection and mitigation to our system level storage products.
2023-08-02 09:44:16 <KipIngram   > AI-based somehow.
2023-08-02 09:44:23 <KipIngram   > Watches for suspicious traffic patterns.
2023-08-02 09:44:38 <KipIngram   > I've no exposure to the details at all.
2023-08-02 09:44:58 <KipIngram   > I work with the individual drives, not the system level.
2023-08-02 11:23:38 <DKordic     > PIC32MM and STM32C0 are meant to replace 8-bit MCUs.
2023-08-02 12:32:01 <KipIngram   > Still no luck getting 32655 output to show up in a console connection run from the bash prompt.
2023-08-02 12:32:13 <KipIngram   > Only in the "debug terminal" inside Segger Embedded Studio.
2023-08-02 12:32:57 <KipIngram   > Maybe I'm misunderstanding something.  But when the studio tells me that it's successfully download my program to the board, I assume that program is FLASHED IN.  I'd then expect to be able to close the studio, re-plug the board, and have that program run on startup.
2023-08-02 12:33:17 <KipIngram   > I guess the program my be written to require some "go" command that the studio sends.
2023-08-02 12:33:38 <KipIngram   > And if that is a correct assumption, I'd expect it to run again every time i push the reset button.
2023-08-02 12:33:59 <KipIngram   > So what I'd like to see is for my flock of debug prints to repeat in that console connection every time I reset.
2023-08-02 12:36:30 <thrig       > maybe there's some "debug mode" vs "run mode" distinction being made?
2023-08-02 13:17:18 <KipIngram   > Yeah.  I made sure I was building and downloading using "release" configuration instead of debug, but that didn't change the behavior at all.
2023-08-02 13:17:31 <KipIngram   > But there could be something like that.
2023-08-02 13:18:54 <KipIngram   > Meanwhile, on the Maxim SDK front (which I also have installed), there's a path error going on.  It looks for an Include directory with a board.h file in it.
2023-08-02 13:19:07 <KipIngram   > in the place the SDK expects to find the Include directory, I don't have one.
2023-08-02 13:19:29 <KipIngram   > Instead, I have a pair of board directories, each one corresponding to one of two different eval boards you can get for this guy.
2023-08-02 13:19:34 <KipIngram   > One of those matches mine.
2023-08-02 13:19:49 <KipIngram   > And it's inside THAT directory that I actually have an Include directory.
2023-08-02 13:19:59 <KipIngram   > So, the SDK is off by one level.
2023-08-02 13:20:16 <KipIngram   > Of course, I could create a soft link or move things around, but I suspect there's a setting for this somewhere and I haven't found it yet.
2023-08-02 13:35:29 <KipIngram   > Aha!
2023-08-02 13:35:38 <KipIngram   > I got it to work using the Maxim SDK.
2023-08-02 13:35:59 <KipIngram   > Built a hello world project there, and it does EXACTLY what I expected.
2023-08-02 13:37:02 <KipIngram   > It's blinking an LED and sending one "Hello world" followed by an incrementing "count = %d" over the serial link.
2023-08-02 13:37:18 <KipIngram   > I can see that in the bash console screen.   And if I hit reset it starts over.
2023-08-02 13:41:21 <KipIngram   > Ok, nice.  I'm ready to start playing in earnest, I think.
2023-08-02 14:13:58 <KipIngram   > It still presents that directory issue above, but it's a warning and not an error.
2023-08-02 14:14:19 <KipIngram   > I'll spend a little more time looking for a setting, and if I don't find it I'll just soft link my way around it.  Just for "cleanliness."
2023-08-02 14:18:00 <KipIngram   > I think what I want to do now is try poking my own settings into the serial port and see if it still works.  Change its baud rate, etc.
2023-08-02 14:25:47 <KipIngram   > Ok, a soft link made that warning go away.
2023-08-02 15:13:55 <DKordic     > Wouldn't #include "missing_dir/board.h" be proper solution?
2023-08-02 15:14:54 <DKordic     > Try printf-ing HTML5 :P
2023-08-02 15:40:21 <KipIngram   > Ok, I think I understand why the Segger studio isn't sending output to the com port.
2023-08-02 15:40:26 <KipIngram   > It's a FEATURE.
2023-08-02 15:40:43 <thrig       > funny way to spell BUG
2023-08-02 15:41:04 <KipIngram   > They have this feature they call RTT - Real Time Transfer.  All printf calls are re-routed to a proprietary channel that finds its way to the debug terminal in the studio.
2023-08-02 15:41:10 <KipIngram   > And NOT sent to the com port at all.
2023-08-02 15:41:16 <KipIngram   > I'm searching for how to disable that.
2023-08-02 15:41:57 <KipIngram   > So it probably IS running every time I punch the reset button - but just sending the output where I can't get at it outside of the studio.
2023-08-02 15:42:26 <KipIngram   > Although I thought an LED was supposed to blink too, and I'm not seeng that.  But it may fill up some buffer and block, waiting for the studio to unload it.
2023-08-02 15:48:01 <KipIngram   > The purported advantage (which I can't entirely disagree with) is to allow debug prints without having the UART pins of the chip tied up.
2023-08-02 16:35:54 <KipIngram   > I'm not sure Segger supports a UART interface in any of their delivered examples - it looks like if you want to use the UART you write the code.  They provide minimal examples and it looks like a project that does it is availble for download.
2023-08-02 17:29:36 <KipIngram   > Ok, the Maxim SDK only makes .elf files, but I figured out how to use an arm variant of objcopy to turn it into a .bin file, and I can drop that on the board's storage device to program it.  That works, confirmed.
2023-08-02 18:08:55 <MrMobius    > DKordic: unfortunately you cant program any of the pic32mm parts with the free compiler
2023-08-02 18:10:24 <KipIngram   > Now that sounds sucky.
2023-08-02 18:10:53 <KipIngram   > I've never liked hardware companies trying to make money on the necessary tools for their hardware.
2023-08-02 23:01:00 <KipIngram   > You know, a time or two we've (I've, I guess) mentioned Norman Wildberger and his dissatisfaction with how modern math handles infinities.  He thinks we should use only rational numbers in our mathematics.  But it just occurred to me this evening that that means he'd have no place in his "mathematical house" for e and pi.  And it's pretty hard for me to see doing without those.
2023-08-02 23:01:14 <KipIngram   > I wonder if he's talked anywhere specifically about that.
2023-08-02 23:01:56 <KipIngram   > e and pi are even worse than irrational - they're actually transcendental, which means they are solutions of any polynomial equation with integer coefficients.
2023-08-02 23:02:07 <KipIngram   > I'm sorry - "aren't."
2023-08-02 23:14:53 <DKordic     > Zeno's paradoxes?  Grandi's series?
2023-08-02 23:22:37 <DKordic     > KipIngram: Two's Complement ?  W/P-adic_number#Quote_notation ?
2023-08-02 23:39:49 <xelxebar    > At least e and pi are computable, though. It's mildly disconcerting that we only have countable computable reals.