<crest_> which is confusing, but could be a bug in the freebsd linux64 abi
<crest_> wait i could do it the proper unix way...
<crest_> just write a smart wrapper for ascii-xfr
<tpbsd> crest_, I can make you a new binary with pretty print for all the registers, usb and a corberstone if you like ?
<crest_> tpbsd: sure
<tpbsd> how much free flash do you think you'd need ?
<tpbsd> cause it will be fairly full by then
<crest_> 20-30k should be enough for my experiments
<tpbsd> okies, I'll start on it later today
<crest_> i've found a way to make picocom more bareable: a smarter send-cmd
<tpbsd> none of that is bearable to me
<crest_> just make $* followed by ascii-xfr -vs $* already helps
<tpbsd> this is a video of my system, it's a bit dated now and it's poorly done as it was one of my first, but youll get the idea
<crest_> what does screen offer?
<crest_> thx
<tpbsd> screen is so superior that nothing else except purpose made scripts could compete
<tpbsd> superior for Mecrisp-Stellaris that is
<crest_> i've never liked screen because i had some run ins with one of the authors which disinclined me to trust any code he touched
<tpbsd> it's poorly maintained now
<tpbsd> the remote serial connections are the main attraction for me
<tpbsd> it's taken me years to finally arrive at what I wanted
<tpbsd> while I like the traditional Forths I cant use the traditional serial terminals at all, Id rather use C and arm-none-eabi
<tpbsd> along with gdb
<crest_> wait your setup allows make to inject into screen?
<tpbsd> thats a 'remote serial connection'
<tpbsd> i never use the screen terminal unless I'm doing some experiment with the hardware, but even then I usually do that via the editor
<crest_> your video demonstrated what your setup does
<tpbsd> cool
<crest_> but not how it does it
<tpbsd> but ill redo it, that video is poor in almost every way
<tpbsd> yeah thats because it's so specific to what I do
<tpbsd> Im happy to share but no one has asked
<tpbsd> probably because the Forth users that see it already have their own system
<tpbsd> crest_, this should get you going with screen :)
<tpbsd> #!/bin/sh
<tpbsd> LIBRARY=library
<tpbsd> FILE_NAME_Y=$PROJECT.fs
<tpbsd> PROJECT=`basename $PWD`
<tpbsd> FILE_NAME_X=$PROJECT-includes.fs
<tpbsd> # FILE_NAME_Z=printreg.fs
<tpbsd> screen -p $PROJECT -X stuff "compiletoram\n"
<tpbsd> # screen -p $PROJECT -X stuff "compiletoflash\n"
<tpbsd> # screen -p $PROJECT -X stuff "forgetram\n"
<tpbsd> screen -p $PROJECT -X stuff "reset\n"
<tpbsd> screen -p $PROJECT -X stuff "72mhz\n"
<tpbsd> screen -p $PROJECT -X readreg x `./rcas.sh $FILE_NAME_X` $FILE_NAME_X.rcas.fs
<tpbsd> screen -p $PROJECT -X readreg y `./rcas.sh $FILE_NAME_Y` $FILE_NAME_Y.rcas.fs
<tpbsd> # screen -p $PROJECT -X readreg z `./rcas.sh $FILE_NAME_Z` $FILE_NAME_Z.rcas.fs
<tpbsd> screen -p $PROJECT -X paste x
<tpbsd> screen -p $PROJECT -X paste y
<tpbsd> # screen -p $PROJECT -X paste z
<tpbsd> rm $FILE_NAME_X.rcas.fs
<tpbsd> rm $FILE_NAME_Y.rcas.fs
<crest_> thx
<tpbsd> rcas.sh is that comment removal script I pasted earlier
<crest_> so tmux has send-keys maybe i can combine tmux and picocom
<tpbsd> hehe
<tpbsd> or just unix utils ?
<tpbsd> crest_, another screen problem is that it's limited to 460800 baud
<tpbsd> Id be using 1 - 2 mbs if I could
<tpbsd> crest I set out in 2014 to make a forth design env that was faster than Gcc/gdb on stm32, and I have that now so Im happy
jsoft has joined #forth
<crest_> sure a fast edit compile test cycle is important
<tpbsd> it uploads faster, it's heaps faster to design and debug because ... Forth
<crest_> and forth makes the testing a lot faster and more powerful than c
<tpbsd> it it to me
<crest_> and avoids the full cycle for each tiny change
<tpbsd> i can upload a 10k source file in about 2 seconds
<tpbsd> no openocd needed
<crest_> still a stall in your mental pipeline
<tpbsd> I do a full source upload, but I move Words that are debugged into the "includes.fs" file and they get written to flash
<crest_> but better than other eco systems e.g. compiling c++ with lots of templates
<tpbsd> i always execute development code from ram
<tpbsd> yeah, a stall is inevitable with a on chip Forth
<tabemann> oh
<tpbsd> we alos have a tethered Forth which is really interesting
<tabemann> btw, I got the ihex dumper to dump
<tpbsd> tabemann, during the upload cycle
<tpbsd> tabemann, nice
<tpbsd> tabemann, want a complex shell script to turn it into a binary ?
Keshl has quit [Ping timeout: 265 seconds]
<tpbsd> it's made for gnu screen
<tpbsd> i run the script and get a binary, I dont have any manual intervention
<tabemann> the thing is that screen is of little use to me because I don't have RTS/CTS because not only does the USB on STM32L476 not support it, but my USB-serial fob doesn't have RTS or CTS pins either
<crest_> tabemann: a proper usb serial adapter can be had for < 5 USD or EUR
<tpbsd> or $1 aud
rdrop-exit has joined #forth
<crest_> tpbsd: does that include shipping for a single board?
Keshl has joined #forth
<tpbsd> Unix friendly CP2102 chipset dongle, with Rx, Tx, nRTS,nCTS and other control signals, including +5 and +3.3 volts available on pins or holes for pins or wires to be soldered. Usually under $2 on eBay, and the most useful model in my opinion
<tpbsd> crest_, no
<tpbsd> crest_, my dongles came from China and I buy 10x at a time
<crest_> can you switch between 5v and 3.3v?
<tpbsd> they output +5 and +3
<tpbsd> just use what you want
<tpbsd> tabemann, I will do a patch for providing RTS on the F407 soon
<tabemann> mine has GND, RX, TX, +5 and +3.3 IIRC
<tabemann> cool beans
<crest_> tpbsd: which dev board would you recommend to get away from the blue pill?
<tpbsd> tabemann, yeah, many do only that, they usually have the other signals on the chip but it needs SMT skills to solder wires to them
<tpbsd> crest_, ouch thats a hard one
<tpbsd> crest_, does it have to be STM32 ?
<tabemann> I own two different DISCOVERY boards the STM32L476 and the STM32F407
<crest_> not really
<tpbsd> i ask because Nordic are pretty impressive
<tpbsd> and you get wireless
<crest_> it would only be for personal projects
<crest_> but it would be nice if they're cheap enough to leave them in a project
<crest_> 256k flash, 32k ram and bluetooth radio does sound great
<tabemann> one setup I've heard of people doing is doing something like a Cortex-M for their core processing and then wire it up to an ESP8266 for wireless
<tpbsd> have a look at nrf52832 ?
<crest_> serial over bluetooth would be nice :-P
<tpbsd> tabemann, ESP8xxxx ewww
<tpbsd> plus the nordic cortex's have some excellent facilities
<tpbsd> tabemann, you get 3 micrograms of CoV-19 with every ESPxxx ;-)
<tabemann> tpbsd: from the way I've heard it put with the ESP8266 is that the ESP8266 is used essentially as a peripheral to enable wireless
<tpbsd> Ive found that Chinese homegrown products really suck. American tech is light years ahead
<crest_> tpbsd: esp32 :-P
<tpbsd> tabemann, it is but it's a waste of space imho
<tpbsd> the nordic stuff has wireless and is seriously good by comparison
<crest_> tabemann: but for a lot of applications the esp chips are good enough
<crest_> they're cheap enough and include wifi support
<tpbsd> the only readon espxxxx is so popular is because arduino users could use them
<crest_> and because its easy to connect to wifi and has enough resources for tcp
<tpbsd> dont do it tabemann, the esp will drive you so nuts you'll jump out the window
<crest_> its fine unless you're battery powered
<tpbsd> ignore this undercover ESPxxxxx salesman!
<tabemann> I'll have to remember about Nordic
<tabemann> lol
<crest_> but you have to build ontop the existing crappy libs because the hardware is so buggy
<tpbsd> tabemann, zepto fort can be easily ported to the nordic I think
<crest_> i remeber the the SHA2017 badge team wanted to melt down the chips and their creators
<tpbsd> and of course the nrf52832 has a Mecrisp-Stellaris binary
<tpbsd> hahahh
<tabemann> lol
<tpbsd> I have worked with a few nordic Mecrisp-Stellaris users and was a bit jealous of that chip
<tpbsd> crest_, Im stuck on stm32f051 as I have 480 chips in stock
<tpbsd> and I love using them, way to late for me to switch to the nordic
<tabemann> why did you buy so many?
<tpbsd> AVENET had them for $0.56 ea in 2014 adn the lot was a batch of 480
<tpbsd> they price was for a single unit, but I wanted a lifetime supply
<crest_> how many do you have left?
<tpbsd> theyre 32 pin UFN
<tpbsd> um ... 478
<tabemann> they're a sunk cost!
<crest_> ignore them
<tabemann> as economists would put it
<crest_> if you need <10 chips order 10 chips next time
<tpbsd> they came in sealed aluminium bags, on chip carriers with moisture dots, full of nitrogen and a complete supply history
<tpbsd> nope
<tpbsd> I like in Australia, Ive ben doing embedded since 1974 and I know a great chip when I see one
<tpbsd> like = live
<crest_> which is great if you want to use them for anything safety critical
<tpbsd> it can be very hard to get stuff here, and considering CoV-19, how easy will it be to get chips for the next few years ?
<tpbsd> plus those same chips are about 3x as expensive right now
<tpbsd> (if you can get them)
<crest_> great but how important is that if you've used just a handful?
<tpbsd> in 2014 I had no isea how many Id use
<tpbsd> this is australia ...
<crest_> but isn't your time more important?
<tpbsd> there are none available here, I cant get fedex to send me some overnight unles I want to pay $100 +
<crest_> is it really cheaper use external support chips for bluetooth etc.?
<tpbsd> sure, my time is the most important thing to me, especially at age 66
<crest_> okay i'm spoiled by 1-2 day shipping and affordable overnight shipping for for a lot of things in germany
<tpbsd> it's never cheaper to use external chips except perhaps for a hobbyist
<crest_> and are you a hobbist these days?
<tpbsd> yeah, retired mostly
<tpbsd> i still make gear, cant help it after a lifetime of doing so
<tpbsd> I have a ton of projects
<crest_> the nRF52840 dk board is a bit expensive to just put into hobby projects like a door sensor
<tpbsd> and I'm still in the infatuation stage with Forth having only started with it in 2014
<tpbsd> crest_, hmm maybe thats the wrong one
<crest_> the annoying fact is that at the price point of 40-50€ you can grab a raspberry pi (or some other pi)
<tpbsd> the units people were using cost $10
<tpbsd> maybe this one: Nordicsemi nRFgo Starter Kit nRF51822
<crest_> and that runs forth as well almost every other language you want to throw at it
<tpbsd> crest_, have you met matthias koch ?
<tabemann> back
<tabemann> I'd buy a more expensive board just because I can't see myself soldering
<tabemann> so I'd opt for a board with pins on it over a board for holes for soldering
<tpbsd> tabemann, sure
<crest_> tpbsd: yes i met him at 32c3
<crest_> he gave talk on mecrisp for msp430
<tpbsd> tabemann, and you have some reasonable boards, even if the discovery boards are pretty cheap
<tpbsd> crest_, yes, it's a shame that he is virtually deaf
<tpbsd> tabemann, there are more expensive boards that are really nice, gold pins, plenty of room
<tpbsd> Olimex make some of them
<tpbsd> but the discos and nusleos have all the basics
<tpbsd> and are dirt cheap
<tpbsd> nucleos
<tabemann> the only things these boards really lack which I'd like is wireless
<tpbsd> crest_, Ive seen it
<tpbsd> tabemann, then a nordic board may be in your future ?
<crest_> tpbsd: the last few years we managed to get a good team of volunteer translators for all german and english presentations
<tpbsd> crest_, nice!
<tpbsd> matthias gets by, his written English is very good
<tabemann> tpbsd: possibly
<tpbsd> tabemann, the ESP stuff is pretty horrible imho
<crest_> but the real time translations widen the audience and relief the pressure on speakers to present in english
<tpbsd> tabemann, for instance the flash is loaded into ram before it runs but the flash is a lot bigger than the fam ...
<tpbsd> ram
<tpbsd> crest_, yes, thats very organised, just what I'd expect
<crest_> i see you haven't been there
<crest_> its quite chaotic and a lot of things go wrong, but there are enough people to catch most of it in time
<tpbsd> crest_, hahah, no Ive only ever ben to Hongkong on a technical course in 1985
<tpbsd> crest_, I drive a Audi 80 quatro
<MrMobius> ive often wished for a chip with a lot of ram and small flash but virtually all microcontrollers are the opposite
<tpbsd> it's nothing special in germany but theyre rare in Australia
<tpbsd> MrMobius, true
<tabemann> these nRFgo boards are expensive
<crest_> and not really useful at that price point for hobbists
<tabemann> whereas these disco boards are cheap enough I can easily afford two of them
<crest_> because there are cheaper boards with more features (and a higher power consumption as well)
<tpbsd> Generic evaluation and development platform for Nordic ultra low power radios
<tpbsd> Built-in support for nRFProbe hardware debug solution for Nordic radios with embedded microcontrollers
<tpbsd> Two motherboards with standard socket for nRFgo radio modules
<tpbsd> USB, RS232, ISP and hardware debug interfaces
<tpbsd> Multiple power supply options: battery, external supply and USB
<tpbsd> Multiple I/O ports and buttons for prototyping
<tpbsd> On-board flash I/O controller provides USB interface to PC
<tpbsd> Upgradable board firmware
<tpbsd> Four 10-wire patch cables and eight 2-wire patch cables included in the kit
<tpbsd> Seven segment status LED for board ID
<tpbsd> Battery pack for three AAA batteries
<tpbsd> Extension module socket for advanced prototyping
<tpbsd> One nRFgo Display extension module with 16x2 alphanumeric display and joystick
<tpbsd> nRFgo Studio for RF evaluation and testing
<tpbsd> Auto detection and hot plugging of nRFgo radio modules
<tpbsd> Rich set of configurable RF and link tests
<tpbsd> Easy access to available radio module configurations
<tabemann> whereas the cost of the nRFgo boards is such that unless I have a really good reason to own one....
<tpbsd> Comprehensive set of documentation
<tpbsd> you kids ....
<crest_> which is worth it
<tpbsd> in 1974 we were paying $287 aud for 1 16 bit MCU
<crest_> but its the classic old style dev board
<tpbsd> the 6800 was $178 USD
<tpbsd> no clock, no peripherals, 8 bits, multiple voltages, 1Mhz
<crest_> but hardware got cheaper and if less then optimally designed boards can be had for between $2 and $12
<crest_> and all it takes to make them work is open source software that was already written
<tpbsd> well the blue pill is an anomaly, but I grew up with the 6800 pr`icing
<tpbsd> but thats not the board the nordic mecrisp users are using
<tpbsd> theirs cost $10 and has a usb connector on one end
<crest_> a nanopi neo2 costs just $20
<tpbsd> it's the size of a usb stick
<tpbsd> yeah, thats for broadcomm lovers
<crest_> that includes a quad core cortex a53
<crest_> allwinner h5 in this case
<tpbsd> and and a untouchable sustem supervisor
<tpbsd> oops, Im out of date already!
<tpbsd> not for me, Im small embedded only, lots of IO, machine control, no wifi
<tabemann> the thing is this - unless I need wireless, I have little reason to justify buying a nordic board when an STM disco board does everything else just as well and is much cheaper
<crest_> https://www.friendlyarm.com/index.php?route=product/product&path=69&product_id=132&sort=p.price&order=ASC
<crest_> 256mb ram, four cores, ethernet and a micro sd socket
<crest_> for most hobby projects power and long term availablity aren't that important ... until they are
<tabemann> that's essentially a lower price, lower capability version of the raspberry pi
<crest_> exactly
<tpbsd> this is the one I'm thinking of: nRF52840 MDK USB Dongle
<crest_> and most allwinner chips are supported by freebsd
<tpbsd> have a look at the pricing for that
<tabemann> I own a raspberry pi but haven't opened the packaging because I'd rather work on something that can truly run on the metal rather than something that needs an OS to work, because then I might as well use a PC
<crest_> the larger ones are even selfhosting in a meaningful way
<tpbsd> tabemann, I have a few rpi's but use them as higher end running Linux
<crest_> tabemann: the original raspberry pi is a deeply flawed design
<crest_> the gpu bootstraps the cpu
<tpbsd> that's the guy who we helped port Mecrisp-Stellaris to that chip
<crest_> tpbsd: that looks a lot more reasonable
<crest_> i don't need a two line lcd display, dpad input, piezo buzzer etc.
<tpbsd> thats the one i was thinking of, and that MCU is crazy
<tpbsd> fast, heaps of flash, wireless, and the cortex-m peripherals are really nice
<tpbsd> the ones that nordic use
<tpbsd> I mean any stm32 is great, but maybe nordic are just a bit better
<tpbsd> NXP is also excellent, but pricier
<crest_> wait is the 802.15.4 radio also zigbee compatible?
<MrMobius> heh, are the peripherals any less confusing than other cortex m stuff from other manufacturers?
<tpbsd> crest_, stay away from ATMEL SAM, their flash is a problem for Forth
<crest_> tpbsd: what's so bad about their flash?
<crest_> because it would be intersting to mess with my ikea tradfri lights from a forth system
<tpbsd> crest_, min flash size is too big
<crest_> and to little sram or just a waste of money?
<tpbsd> it's just the flash
<tpbsd> matthias refuses to port to it for just that reason
<tabemann> for Forth you want a minimum flash write size of something like 16 bytes or less
<crest_> oh flash write size
<crest_> not total flash capacity
<tabemann> and even for 16 bytes you want some kind of write deferral system like what I've coded for STM32L476
<crest_> i assumed it was just a core with 8k sram and 1mb flash or something like that
<tpbsd> matthias spent weeks with SAM trying to work around this
<tpbsd> crest_, it's the SAM flash controller thats the problem
<tabemann> STM32F407 has the problem of really give flash erasure blocks, but that's not quite as bad
<tabemann> *big
boru` has joined #forth
boru has quit [Disconnected by services]
<crest_> an additional sub ghz radio for longer range would be nice
boru` is now known as boru
<MrMobius> you usually have a lot more than 8k sram on those arm chips that have 1mb flash
<crest_> MrMobius: exactly
<crest_> it would be a waste of flash
dave0 has quit [Quit: dave's not here]
<tpbsd> yeah, 196kN on the 32F407
<tpbsd> yeah, 196kB on the 32F407
<tpbsd> of ram
<tpbsd> The nRF52840 Micro Dev Kit USB Dongle features a programmable user button, RGB LED, up to 12 GPIOs and 2.4G Chip antenna on board.
<tpbsd> The USB Dongle can be used as a low-cost Bluetooth5/Tread/802.15.4/ANT/2.4GHz multi-protocol node or development board. Alternatively, the USB Dongle can be used as a Network Co-Processor(NCP) with a simple connection to a PC or other USB enabled device.
<crest_> at first glance it can be a lot of other things too
<tpbsd> sure
<crest_> you just have to solder a few headers to it
<tpbsd> thats a adafruit advert I think
<crest_> and you could solder a cable to the usb pads to make life easier until the radio works for everything
<tpbsd> it's not tempting you is it ???
<crest_> it sure is
<tpbsd> have you had a look at the chip specs yet ?
iyzsong has joined #forth
<tpbsd> :)
<crest_> i never meant to say that those cheap buggy boards are a good solution just that they are available and easy to get started with
<tpbsd> The ARM® Cortex-M4 processor with floating-point unit (FPU) has a 32-bit instruction set (Thumb®-2
<tpbsd> technology) that implements a superset of 16- and 32-bit instructions to maximize code density and
<tpbsd> performance.
<crest_> its not hard to write an image on a micro sd card, ssh into the system and modify python examples
<tpbsd> crest_, actually I hated the 'bluepill' from the start and dont have any, but after writing the diag binaries on my site, Ive come to appreciate parts of the STM32F103 and it's still *very* capable
<crest_> tpbsd: i lack the experience to compare it other modern chips because i'm a sysadmin and high level programmer by trade
<tpbsd> crest_, I still have twenty STM32F103's from AVENET in sealed aluminium bag full of nitrogen that I'll probably never use, paid $3.50 a chip at the time
<tpbsd> crest_, aha, I thought youd be in that category
<tpbsd> crest_, but you seem to be pretty good at hardware also ?
<tabemann> back
<crest_> not really but i do enjoy optimizing code and playing with obscure systems
<tpbsd> crest_, Im at the complete other end of that scale
<crest_> e.g. i have a rca 1802 membership card, but haven't gotten around to building a programmer for it
<crest_> and toggling programs in in binary gets frustrating quickly
<tpbsd> tabemann, the nordic stuff is M4 as well, so you could port zeptoforth easy enuf
<crest_> but the rca 1802 instruction set is so insane
<tpbsd> I had one a long time ago, threw the chip out iirc
<crest_> and relatively well suited to forth
<crest_> or i would be better to say you can't run anything but forth on such a strange design
<tpbsd> it's amazing how many programmers are interested in old hardware, I *hate* old hardware
<tpbsd> probably as I *had* to use it at the time
<tabemann> tpbsd: the only thing that makes me question the wisdom of that is just how much those boards cost
<tpbsd> so naturally I *love* new hardware
<crest_> i tought myself assembler with a 1802 emulator because i couldn't get ahold of introductory texts on powerpc assembler
<tpbsd> tabemann, did you see the nRF52840 Micro Dev Kit USB Dongle ?
<crest_> tabemann: the dongle costs just 10 bucks which is fine
<crest_> and someone already go a radio repl working: https://github.com/rabbithat/nRF52_wireless_Forth_REPL
<tabemann> tpbsd: no I didn't
<tpbsd> crest_, there is a more detailed account of his Forth journey here: https://sourceforge.net/p/mecrisp/discussion/general/thread/e29549a8d5/
<crest_> tpbsd: after the rollercoster of the last few years i just want to run when i see sourceforge
<tpbsd> tabemann, it's a bargain for $10 considering the MCU and wireless + usb
<crest_> adding malware^Wadware to opensource software installers
<tpbsd> crest_, I run when I see github
<rdrop-exit> oops
<tpbsd> rdrop-exit, ZEn Guru of Forth!
<rdrop-exit> good morning Forthwrights
<rdrop-exit> hello Forth Master Technician (tm)!
<tpbsd> hehe
<crest_> gn8
<tpbsd> rdrop-exit, crest_ is using one of my older Forth bootable binaries!
<rdrop-exit> cool
<tpbsd> i think crest_ makes #2 !
<rdrop-exit> hehe
<tpbsd> #1 bombed badly, arduino had melted his brain cells
<tpbsd> poor guy, I think it drove him to drink
<rdrop-exit> catching up on the logs, you guys had a high traffic night
<tpbsd> we did
<tpbsd> we had a 'hardware night' on #forth, something that hasnt happened since 1969 when Chuck was here ;-)
<tpbsd> tabemann, acyually that high quality nordic board with lcd etc is probably EXACTLY what you need as you cant solder
<MrMobius> hmm, 1802 assembly is probably not bad with macros
<MrMobius> always thought loading in 16 bit constants 8 bits at a time looks daunting otherwise
<rdrop-exit> 1802 is funky but fun
<rdrop-exit> (as an intellectual exercise, wouldn't want an 1802 for anything else)
<tpbsd> too funky for me, I bought a chip in the 70's and when I read the datasheet I ran around screaming and rending my clothes
<tabemann> back
<tpbsd> tabemann, actually that high quality nordic board with lcd etc is probably EXACTLY what you need as you cant solder
<tabemann> I just don't want to shell out $100+ for a single board
<tabemann> especially since I see myself buying a bunch of boards in the near futrue
<tpbsd> The COVID-19 pandemic has become one of the most complex and unpredictable events to affect our world in our lifetime. IBM has expanded the 2020 Call for Code Challenge to seek innovative solutions to the pandemic's impacts.
<tpbsd> Developers and problem solvers: join us in building open source solutions that make an immediate and lasting impact.
<tpbsd> The deadline to submit your solutions is July 31st.
<tpbsd> and win $100,000 USD
<tpbsd> tabemann, but this will be a quality board the like of which youve never seen
<rdrop-exit> veltas, p@ i.e. port fetch, pb@ for port byte fetch
<tpbsd> tabemann, in fact it's TWO boards
<tabemann> tpbsd: I'm going to just get that USB dongle for now, maybe get the uber high quality nordic board later
<tpbsd> tabemann :)
<tpbsd> tabemann, uber quality and uber packed with resources
<tpbsd> tabemann, it comes with all the leads you need, everything
<tpbsd> it's like night and day to other cheap boars
<tabemann> back
reepca has quit [Read error: Connection reset by peer]
<rdrop-exit> wb
* tabemann just purchased the dongle
reepca has joined #forth
<MrMobius> covid 19 has been very helpful for my hobby productivity :P
<tabemann> it hasn't been for me, because my work expects me to work as if I were employed in the normal fashion, but just at home instead
<tpbsd> tabemann, at least you get paid for it ?
<tabemann> yes
<rdrop-exit> Not mine, I've been doing stuff around the condo
<tpbsd> tabemann, nice, sadly nordic may seduce you away from stm!
<tabemann> lol
<tpbsd> rdrop-exit, if you go outside Duarte may be there with his M6 and shoot at you!
<tpbsd> M16
<rdrop-exit> I duck when I'm on the balcony
<tpbsd> hahah
<tabemann> I hope they at least allow you to go to the grocery store
<rdrop-exit> The neighborhood is so quiet since the lockdown
<tpbsd> tabemann, only id he's faster than bullets
<rdrop-exit> We've been ordering groceries online when needed
<tpbsd> same here
<rdrop-exit> Even though we have 8 or 9 shopping centers within walking distance
<tpbsd> initially I could only order a 'basic box' for $80 and the supermarket puts whatever they like in it
<tabemann> here the grocery stores are still open but they're advising people to stay six feet apart and in the lines by the checkouts they put markers down telling people where to stand while they wait
<tpbsd> the post office guy delivered today, no mask, no nothing
<rdrop-exit> we get fresh fruits and vegetables delivered from the countryside
<tpbsd> I get fresh air delivered from the countryside
<tabemann> nice
<rdrop-exit> better than volcano ash
<tpbsd> it wafts from across the farmers paddocks across my street
<tpbsd> thats a fact!
<tabemann> why is it that the docs for the STM32F407VG say that it has 192K of RAM, but when I configure it to use 192K of RAM it crashes
<tpbsd> $30000 ?
<tpbsd> tabemann, the nRF52840 contains 1 MB of flash and 256 kB of RAM
<tabemann> I just wrote a RAM test
<tabemann> it crashes when it attempts to read from $20020000
<tabemann> indicating that that's the actual top of the RAM
<tpbsd> .equ Kernschutzadresse, 0x00008000 @ Darunter wird niemals etwas geschrieben ! Mecrisp core never writes flash below this address.
<tpbsd> .equ FlashDictionaryEnde, 0x00100000 @ 992 kb Platz für das Flash-Dictionary 992 kb Flash available. Porting: Change this !
<tpbsd> .equ Backlinkgrenze, RamAnfang @ Ab dem Ram-Start.
<tpbsd> .equ FlashDictionaryAnfang, 0x00008000 @ 32 kb für den Kern reserviert... 32 kb Flash reserved for core.
<tpbsd> lol, mathias says 992 kB and has allocated 1048576 ram for it
<tpbsd> oops thats flash!
<tpbsd> and thats right, matthias never makes a mistake!
<tpbsd> tabemann, ok, matthias also says 128 kb ram
<tpbsd> it must be a error in the doc
<tabemann> and mind you this was from more of a technical overview rather than a true datasheet
<tabemann> okay, well, I'm gonna hit the sack, so g'night
<rdrop-exit> bonne nuit tabemann, stay healthy
<tpbsd> night tabemann
reepca has quit [Ping timeout: 265 seconds]
webchat9 has joined #forth
reepca has joined #forth
actuallybatman has quit [Ping timeout: 256 seconds]
actuallybatman has joined #forth
gravicappa has joined #forth
Keshl has quit [Quit: Konversation terminated!]
dddddd has quit [Ping timeout: 256 seconds]
Keshl has joined #forth
mtsd has joined #forth
webchat9 has quit [Ping timeout: 260 seconds]
xek__ has joined #forth
dys has joined #forth
webchat9 has joined #forth
_whitelogger has joined #forth
john_cephalopoda has joined #forth
jsoft has quit [Ping timeout: 256 seconds]
dys has quit [Remote host closed the connection]
mtsd_ has joined #forth
mtsd has quit [Ping timeout: 265 seconds]
dys has joined #forth
rdrop-exit has quit [Ping timeout: 240 seconds]
dave0 has joined #forth
mtsd_ has quit [Quit: Leaving]
rdrop-exit has joined #forth
mtsd has joined #forth
dddddd has joined #forth
rdrop-exit has quit [Quit: Lost terminal]
iyzsong has quit [Ping timeout: 265 seconds]
dave0 has quit [Quit: dave's not here]
cheater has joined #forth
webchat9 has quit [Ping timeout: 265 seconds]
mtsd has quit [Quit: Leaving]
john_cephalopoda has quit [Ping timeout: 260 seconds]
john_cephalopoda has joined #forth
webchat9 has joined #forth
jsoft has joined #forth
Lord_Nightmare has quit [Quit: ZNC - http://znc.in]
john_cephalopoda has quit [Read error: Connection reset by peer]
john_cephalopoda has joined #forth
Lord_Nightmare has joined #forth
Lord_Nightmare has quit [Quit: ZNC - http://znc.in]
Lord_Nightmare has joined #forth
proteus-guy has quit [Quit: Leaving]
proteus-guy has joined #forth
jsoft has quit [Ping timeout: 265 seconds]
jsoft has joined #forth
Zarutian_HTC has quit [Ping timeout: 240 seconds]
<crest_> tpbsd: do you want translations of the german comments?
<crest_> tpbsd: maybe some of the peripherals have buffers they added to get to 192kB
<crest_> or an additional m0 core with its own puny sram?
Zarutian_HTC has joined #forth
xek__ is now known as xek
TheCephalopod has joined #forth
john_cephalopoda has quit [Read error: Connection reset by peer]
<tpbsd> crest_, thanks but I have the english pdf's
<tpbsd> crest_, I never use this chip myself anyway, I fired it up to help test tabemann's Forth
<webchat9> : w1 POSTPONE [ ; IMMEDIATE
<webchat9> : test1 w1 S" worked1" TYPE w2 ; \ nothing is printed
<webchat9>
<webchat9> : test2 [ S" worked2" TYPE ] ; \ worked2 is printed
<webchat9> : w2 POSTPONE ] ; IMMEDIATE
<webchat9> how to make w1 and w2 work as [ and ] ?
gravicappa has quit [Ping timeout: 265 seconds]
<veltas> webchat9: Remove POSTPONE
<veltas> I think
<veltas> No that's dumb
<veltas> webchat9: Okay The issue is that ] should not be postponed because you want its interpretation syntax
<veltas> Remember that after w1 you are in interpretation mode, so w2 is then going to compile ] (which never happens usually, if you use ] after [ )
<webchat9> I think w2 isn't going to compile ] , it will just execute it
<veltas> Hmm you might be right
<veltas> Maybe it's because POSTPONE doesn't have interpretation semantics
<veltas> Oh you're not interpreting it
<veltas> I'll just shut up before I get anymore of this wrong, as you can see I am not experienced with the compiler magic
<webchat9> no problem, thanks for giving it a thought
<webchat9> I've been struggling with it for some hours
<veltas> Thanks for hitting me with it because the more *I* think about it the better I'll understand it ;)
<webchat9> I learned that in spf-forth one can do : my [ ALSO ASSEMBLER ] [ NOP NOP NOP A; ] [ PREVIOUS ] ;
<webchat9> \ to add 3 asm instructions
<webchat9> now I want to wrap [ ALSO ASSEMBLER ] into a new word called ASM[
<webchat9> so I need to solve this mystery
<veltas> But... I think I *can* say that 'IMMEDIATE' on w2 is erroneous, but I don't see how it will hurt it either
<veltas> webchat9: What if you use an already defined word that prints something?
<veltas> I mean instead of S" .." TYPE
<webchat9> lemme try
<webchat9> : hello S" hello_worked" TYPE ;
<webchat9> : test3 w1 hello w2 ; \ nothing printed
<webchat9> hm strange
<webchat9> in gforth they get printed
<webchat9> but not in spf-forth
<veltas> Yes that is very strange
<webchat9> ok it turns out I was in a different VOCABULARY
<veltas> Ah
<DKordic> IIRC ' .( is an IMMEDIATE variant of ' ." .
<webchat9> which replaced stuff like usual words
<webchat9> so things were behaving strange
<veltas> Good that you solved it!
<webchat9> DKordic: right
<webchat9> thank you so much guys
<veltas> I will be careful around vocabs lol
<veltas> nasty
rdrop-exit has joined #forth
<rdrop-exit> webchat9:
<rdrop-exit> : w1 POSTPONE [ ; IMMEDIATE
<rdrop-exit> : w2 ] ;
<webchat9> right . veltas' comment was correct: "Remember that after w1 you are in interpretation mode, so w2 is then going to compile ] (which never happens usually, if you use ] after [ )"
<rdrop-exit> correct
<tpbsd> rdrop-exit, Zen Guru of Forth!
<tpbsd> rdrop-exit, are you awake early or late ?
<rdrop-exit> I'm late for bed, just winding down
<tpbsd> heh, winding down with some Forth :)
<rdrop-exit> just finished binging on news and videos with the wife
<rdrop-exit> you're up late
<rdrop-exit> (or very early)
<tpbsd> i just woke up, it's 0330 here
<rdrop-exit> 0130 here
<rdrop-exit> going to hit the sack, catch you tommorow when I have my morning coffee
<tpbsd> yeah, we stopped 'daylight saving' here on the 5th
<tpbsd> cya!
<rdrop-exit> goodnight :)
rdrop-exit has quit [Quit: Lost terminal]
<tpbsd> g'night
<webchat9> is there a way to declare a word INLINE ?
<tpbsd> inline, ?
jsoft has quit [Ping timeout: 258 seconds]
<webchat9>
<webchat9> SEE BP
<webchat9> 589653 CC INT 3
<webchat9> 589654 C3 RET NEAR
<webchat9> : my ['] BP [ INLINE, ] ; \ I'd like to INLINE it
<webchat9> so that SEE MY also gives the same definition as SEE BP
<webchat9> (the above doesn't work)
<tpbsd> does INLINE need to be in square brackets ?
<webchat9> without square brackets: SEE my
<webchat9> 589667 8945FC MOV FC [EBP] , EAX
<webchat9> 58966A B853965800 MOV EAX , # 589653
<webchat9> 589677 C3 RET NEAR
<webchat9> 589672 E815A2FDFF CALL 56388C ( INLINE, )
<webchat9> 58966F 8D6DFC LEA EBP , FC [EBP]
<webchat9> END-CODE
<tpbsd> 589672 E815A2FDFF CALL 56388C ( INLINE, ) is that BP ?
<tpbsd> looks like it's a call and not inlined
TheCephalopod has left #forth ["Leaving"]
<tpbsd> perhaps you have the rigt syntax but BP cant be inlined for some reason ?
<tpbsd> not everything can be inlined
<webchat9> ok
<tpbsd> I'm guessing only, I'm a noob myself
<tpbsd> i know with my own Forth that I cant inline certain things
Zarutian_HTC has quit [Read error: Connection reset by peer]
<webchat9> BP is just a single asm instruction , a very simple ordinary word, I think it should be inlined
<tpbsd> but mine has auto inlining so if it can be inlined, it already is
Zarutian_HTC has joined #forth
<tpbsd> i do know when Ive tried to inline word(s) that werent automatically inlined, it wouldnt do it
<tpbsd> webchat9, have you inlined any words yet in other code ?
<webchat9> no,I never tried to inline words
<webchat9> I'm new to forth
crest has joined #forth
lonjil2 has joined #forth
<tpbsd> webchat9, generally you specify INLINE within BP
crest_ has quit [Read error: Connection reset by peer]
lonjil has quit [Quit: No Ping reply in 180 seconds.]
Croran has quit [Quit: No Ping reply in 180 seconds.]
<tpbsd> not within the later word that uses BP
Croran has joined #forth
<tpbsd> then when BP is called it gets inlined
<crest> tpbsd: did you get around to building a new usb enabled mecrisp st image for bluepills?
<tpbsd> crest, no, havent started yet
<tpbsd> Ive made a lot of changes to my system since that old binary and I have to bring thats stuff up to date yet
<tpbsd> I tend to do all my improvements on cortex-m0 then backport it to M3 as I don't actually use the M3 in any projects
<tpbsd> so it will take a while to do that
<tpbsd> I've significantly changed my SVD2FORTH which means a fair bit of M3 code needs to be updated
<crest> so it's a lot more work than just running ./assemble ;-)
<crest> btw what does the -ra suffix for some targets signify?
<tpbsd> well it's easy enuf, just a fair bit to do
<tpbsd> The RA (Register Allocator) Kernel contains an analytical compiler which keeps track of the top five stack elements and maps them to registers whenever possible.
<veltas> Darn I missed rdrop, was hoping to ask him about his port fetch/store words
<veltas> Another time
<veltas> So I was right the first time about [ and ] apparantly, that is the most shocking part to me :P
<tpbsd> hahha, some people just have the knack!
<veltas> Well webchat9 easily convinced me I was wrong, and I feel like I deserve little credit if my answers lack that conviction to say "no u"
<tpbsd> such is life, I find Im easily swayed by a logical argument, just proves we have open minds!
<webchat9> actually I was convinced by the answer but then forth's tricked me
<webchat9> so I started believing it's not that way
<tpbsd> I always look to my notes before I assert that my opinion is correct with code examples
<webchat9> it was interesting to observe, how logical the answer seemed at first, then how it felt wrong, and finally it turned out it was actually right
<tpbsd> fact is Forth is very complex, it's "easy to learn but difficult to master"
<tpbsd> plus every Forth is different
<webchat9> catch you later friends
webchat9 is now known as webchat9_bbl
<tpbsd> cya
webchat9_bbl has quit [Ping timeout: 250 seconds]
reepca has quit [Read error: Connection reset by peer]
reepca has joined #forth
dddddd has quit [Ping timeout: 256 seconds]
Zarutian_HTC has quit [Read error: Connection reset by peer]
Zarutian_HTC has joined #forth
kori has quit [Ping timeout: 260 seconds]
mark4 has joined #forth
jsoft has joined #forth
xek has quit [Ping timeout: 258 seconds]
dddddd has joined #forth
mark4 has quit [Quit: Leaving]
jsoft has quit [Ping timeout: 265 seconds]
WickedShell has joined #forth
actuallybatman has quit [Ping timeout: 260 seconds]
actuallybatman has joined #forth