Activity for Peter Jakacki
-
Peter Jakacki created a blog post
TachyonESP files released for ESP32-C3
-
Hi John, Welcome to Tachyon Forth . Firstly here's a tip: Always best to start a new thread so the request is not lost and easy to find again. So rather than replying to this message, start a new thread (subject) and let me know the source you used. In the meantime use the files from my working set and try again: https://www.dropbox.com/scl/fo/n1tbumt8v8td1y780cylj/AM3BQqgabH8sSDjWUxgWL7M?rlkey=fyhlk6yai7yqydyt8ju7wimcv&st=esmqrrld&dl=0
-
Hi Andy, maybe I haven't been using the reply button. Read my latest post.
-
Hi Andy, I guess prerequisites are a fact of life and we have to start somewhere. But that's also where we ask questions and look for feedback too. I think I've also mentioned so many times the different terminal emulators that one can use, but I use minicom because it works properly and is a fully functional ANSI terminal emulator that can also keep up with high baud rates too. But when it comes to Windows I used to use and still use TeraTerm. But Mecrisp on the RP2040 only uses the hardware UART,...
-
Hi Andy, The Tachyon extensions to Mecrisp on the RP2040 were done as a bridge at the time I was looking at writing a Tachyon kernel for it from scratch. I played with the RP2040 for a while and I should come back to it I guess. I will dig out the board and check the files as well. Do you have some suggestions ?
-
Format any SDHC compatible card to FAT32 with variable cluster size
-
Format any SDHC compatible card to FAT32 with variable cluster size
-
Just to let you know then that TAQOZ is going through a revamp to V3 which also reduces the compiled kernel to the minimum to support loading EXTEND which will have make it a lot easier to make improvements. I have already redone all the number processing and dump routines, and now I am implementing the ABORT/QUIT console routines which will allow for interactive line edit and history as well.
-
Just had a quick look and seems that I chose to write it to the RTC on DATE! which means DAY! only writes to the buffer and depends upon a following DATE! Other than that you can make it to a 3 RTC! directly if that suits. I will have to check it out later.
-
Hi Maciek, hope things are going well for you. I must confess that I have a drawer full of the bare HC05 modules I bought one time to mount on my own little adapter to plug straight into my 4x2 Prop Header which of course includes power. However I have been considering the C3 since it is so low cost and has BT5/BLE and WiFi as well. I must get some and see how suitable they are for BT serial. My module has a cheap n tiny micro on board to intercept a reset sequence so I could still reset the Prop...
-
Hi Maciek, hope things are going well for you. I must confess that I have a drawer full of the bare HC05 modules I bought one time to mount on my own little adapter to plug straight into my 4x2 Prop Header which of course includes power. However I have been considering the C3 since it is so low cost and has BT5/BLE and WiFi as well. I must get some and see how suitable they are for BT serial. My module has a cheap n tiny micro on board to intercept a reset sequence so I could still reset the Prop...
-
Ok, that makes sense although there should have been an error reported in TAQOZ if it didn't find FMAKE. BTW, since I preallocate FAT32 sectors/clusters for files I needed something to specify that and I just haven't picked a good name for that yet. But FSIZE! or FMAKE also set the permissions for creating a new file though. Funny thing I find is that FAT32 filesystem checking software flags preallocated sectors as some kind of error, but preallocation is my way of ensuring defragmentation and efficient...
-
Ok, that makes sense although there should have been an error reported in TAQOZ if it didn't find FMAKE. BTW, since I preallocate FAT32 sectors/clusters for files I needed something to specify that and I just haven't picked a good name for that yet. But FSIZE! or FMAKE also set the permissions for creating a new file though. Funny thing I find is that FAT32 filesystem checking software flags preallocated sectors as some kind of error, but preallocation is my way of ensuring defragmentation and efficient...
-
Ok, that makes sense although there should have been an error reported in TAQOZ if it didn't find FMAKE. BTW, since I preallocate FAT32 sectors/clusters for files I needed something to specify that and I just haven't picked a good name for that yet. But FSIZE! or FMAKE also set the permissions for creating a new file though. Funny thing I find is that FAT32 filesystem checking software flags preallocated sectors as some kind of error, but preallocation is my way of ensuring defragmentation and efficient...
-
First off, as soon as you load code it encounters the TAQOZ word which loads to memory so you can go as fast as your little serial will go, no delays necessary. Can you do a DIR to see if it has written FRED.CFG and TEMP to card? I'm also doing some extra little checks to make sure it starts up one way or another. However once it has been setup with files to open it seems to work well. If you can start off with some text file first then try that. Include the name after ED (ED HELP.TXT) The Forth...
-
Oh, I changed "compile" function to F8 as I seem to have a problem on this PC with F10 mapped as F12 for some strange reason on the terminal, but I will dig deeper.
-
It seems to be working fine now and I added some what-to-do's when it first starts up or when you close all the tabs etc. You can also ED <name> to directly open a file from the command line too. Tomorrow I will enable the highlighting again although I want to handle that differently.</name> Code takes 6.5kB. here's the screenshot back at the command line to show that it is running on TAQOZ
-
It seems to be working fine now and I added some what-to-do's when it first starts up or when you close all the tabs etc. You can also ED <file> to directly open a file from the command line too. Tomorrow I will enable the highlighting again although I want to handle that differently.</file> Code takes 6.5kB. here's the screenshot back at the command line to show that it is running on TAQOZ
-
Hi Stefan, this is my bad, I totally missed that you were talking about the P2 somehow although I had been thinking the other day to actually get it up and running on the P2 (which it wasn't). So maybe today I will grab this latest RP2 version and get it running on the P2. I will also post back here when I'm done.
-
Hi Stefan, have you tried that version yet? I will update Dropbox now but otherwise everything should be ok. The only difference I found was that I removed an annoying smart tab feature and also added a smart delete shift+DEL to delete leading spaces between the cursor and the next character so that it left justifies a line.
-
Hi Kevin, if it is any consolation I will try to make it less painful for the next person 😄 You can open a standard text file which will look a mess as FRED ignores its line endings but you can SAVE AS with &S (ALT+S) and it will create a new formatted page file or overwrite an existing one. (Previously you had to manually create a new file and import it from the command line) One of the first files that you should import this way is the help file. On the PC copy the HELP.TXT file to your card and...
-
Hi Kevin, if it is any consolation I will try to make it less painless for the next person 😄 You can open a standard text file which will look a mess as FRED ignores its line endings but you can SAVE AS with &S (ALT+S) and it will create a new formatted page file or overwrite an existing one. (Previously you had to manually create a new file and import it from the command line) One of the first files that you should import this way is the help file. On the PC copy the HELP.TXT file to your card and...
-
General information: Use Sandisk Ultra cards if you can't get your card to work. Use an ANSI terminal emulator such as minicom or picocom (rather than the better looking but less capable GUI terminals) or TeraTerm on Windows.
-
Btw, my earlier Tachyon produced very narrow clock pulses which were a problem for some cards so please try my latest from Dropbox. Normally I use Sandisk Ultra cards. Don't use old old cards especially those less than 4GB as they are the old SD standard rather than the newer sector based SDHC.
-
Hi Kevin, if you do a screenshot then you can attach that to your post so that we can see what you see. The very first time you use this on an SD card it won't have a FRED.CFG file which it uses for session info but it will create one. However I must add the extra code for it to work without a file open. Quick way around this is to do a DIR before you start FRED to make sure it works and load one of those files with ED <file>. However if it is blank (but formatted) or they are not text files then...
-
Hi Kevin, I just tried FRED.FTH at 5ms and it wasn't able to sustain loading at that speed for the whole file. So I tried it at 10ms and it works fine. Perhaps I should fix up the fastloader but still it only took 8 seconds to load it that way. ./s FRED.FTH USB0 10 My "s" script is just: ascii-xfr -sn -l $3 Forth/$1 > /dev/tty$2 Normally I edit and load FRED from the SD card and that takes 1.8 seconds to load. Then I also have an earlier stable version called FREDA or WILMA in case I stuff up FRED...
-
What's the : doing on the line after the PRINT PRINT: I see you only have part of the Tachyon extensions, especially the TACHYON QUIT which regulates handling not only of echo/response but also little things like maintaining the stack even after an error etc. Some people think you have to have an "ok" but I find that doesn't convey anything and it really doesn't stand out which is one of the reasons I use colors and a clear non-alpha --- separator between input and response. Don't forget the useful...
-
What's the : doing on the line after the PRINT PRINT: I see you only have part of the Tachyon extensions, especially the TACHYON QUIT which regulates handling not only of echo/response but also little things like maintaining the stack even after an error etc. Some people think you have to have an "ok" but I find that doesn't convey anything and it really doesn't stand out which is one of the reasons I use colors and a clear non-alpha --- separator between input and response. Don't forget the useful...
-
You are using 2rot but that is for a double number which means at least 6 items on the stack. There is ROT and -ROT (reverse). Btw, if you try to access a parameter that isn't on the stack then it will still access whatever garbage is at the indexed location. Your test: .S Stack: [5 ] 42 17 20 30 19 TOS: 14 *> drop drop swap rot 2rot ok. meant that you had 5 items, you dropped 2 and then you had 3 for swap and rot but not enough for 2rot - so you ended up with "rot" Testing 2ROT R00# 1 2 3 4 5 6...
-
You are using 2rot but that is for a double number which means at least 6 items on the stack. There is ROT and -ROT (reverse). Btw, if you try to access a parameter that isn't on the stack then it will still access whatever garbage is at the indexed location. R00# 1 2 3 4 5 6 7 8 --- R08# .s --- 0 : $00000008 8 1 : $00000007 7 2 : $00000006 6 3 : $00000005 5 4 : $00000004 4 5 : $00000003 3 6 : $00000002 2 7 : $00000001 1 R08# 2rot .s --- 0 : $00000004 4 1 : $00000003 3 2 : $00000008 8 3 : $00000007...
-
You are using 2rot but that is for a double number which means at least 6 items on the stack. There is ROT and -ROT (reverse). R00# 1 2 3 4 5 6 7 8 --- R08# .s --- 0 : $00000008 8 1 : $00000007 7 2 : $00000006 6 3 : $00000005 5 4 : $00000004 4 5 : $00000003 3 6 : $00000002 2 7 : $00000001 1 R08# 2rot .s --- 0 : $00000004 4 1 : $00000003 3 2 : $00000008 8 3 : $00000007 7 4 : $00000006 6 5 : $00000005 5 6 : $00000002 2 7 : $00000001 1
-
Not quite sure what you are expecting as it seems to be correct. I also have GPCR etc defined like this: (Btw, I prefer operator symbols for operators, names for functions i.e. PRINT instead of . ) --- GPIO STATUS : GPSR ( pin -- ) 3 << $40014000 + ; --- GPIO CONTROL : GPCR ( pin -- ) 3 << $40014004 + ; Tachyon will print the stack as a stack but with the value in ASCII, HEX, and DECIMAL. You can even hit ^Q to print the stack or ^S to clear it at anytime. Take note though, normally Forth will automatically...
-
Not quite sure what you are expecting as it seems to be correct. I also have GPCR etc defined like this: (Btw, I prefer operator symbols for operators, names for functions i.e. PRINT instead of . ) --- GPIO STATUS : GPSR ( pin -- ) 3 << $40014000 + ; --- GPIO CONTROL : GPCR ( pin -- ) 3 << $40014004 + ; Tachyon will print the stack as a stack but with the value in ASCII, HEX, and DECIMAL. Your can even hit ^Q to print the stack or ^S to clear it at anytime. Take note though, normally Forth will automatically...
-
No, it's really a bridge I built to host my development tools treating the RP2040 much like a PC so I can build my own custom Tachyon kernel. I keep getting waylaid but I have many low level utilities and the virtual memory FAT32 SD and a multi-tab editor and multichannel logic analyzer that operate over an ANSI terminal connection. When I finally stop playing I will finish my resident macroassembler and then rolling out kernels will be a such a breeze and without having to rely on awkward PC tools....
-
No, it's really a bridge I built to host my development tools treating the RP2040 much like a PC so I can build my own custom Tachyon kernel. I keep getting waylaid but I have many low level utilities and the virtual memory FAT32 SD and a multi-tab editor and multichannel logic analyzer that operate over an ANSI terminal connection. When I finally stop playing I will finish my resident macroassembler and then rolling out kernels will be a such a breeze and without having to rely on awkward PC tools....
-
With my Tachyon extensions you can select any pin and specify frequency and duty cycle. Here are some methods: 21 PIN 1 KHZ ( square wave ) 25 PIN 33 PWM% ( 1/3 duty at 1 Khz default ) 10 KHZ 333 1000 DUTY Remember that there is a PWM module per odd/even pin pairs so while you can change the duty independently, they share the same divider so changing the frequency of one will affect the other. UART channels are selected in a similar way and just as easy to use too btw. Here's a screenshot from SPLAT...
-
With my Tachyon extensions you can select any pin and specify frequency and duty cycle. Here are some methods: 21 PIN 1 KHZ ( square wave ) 25 PIN 33 PWM% ( 1/3 duty at 1 Khz default ) 10 KHZ 333 1000 DUTY Remember that there is a PWM module per odd/even pin pairs so while you can change the duty independently, they share the same divider so changing the frequency of one will affect the other. Here's a screenshot from SPLAT showing the output (I reduced the channels down to 20 to 25)
-
I was a little surprised by the lack of questions but sometimes when we take a quantum leap rather than incremental steps it takes a while for most to catch-up. So it is of no surprise though that one person "complained" that the Tachyon layer didn't have a consistent response for ZeptoCom to handshake with. After having already demonstrated that it is possible to download 3,000 lines of dense code serially in under 2 seconds or loading a file directly off SD card or an editor tab, it probably was...
-
Hi Kevin, yes, I have the Si5351 built into my P2 module as a programmable clock mainly because there were some problems with the PLL in the Parallax P2 prototypes but the silicon was updated plus a few extra instructions and now I don't really need it. However, Bob Edwards G4BBY is using the P2 and the Si5351A with radio and of course TAQOZ - my Tachyon for the P2. I tried to get him to do a presentation this time but maybe another time. Anyway, he has also written some synth software in Forth so...
-
Hey Maciek, looks like someone wasn't quite ready for their presentation this month. Would you like to do 10 minutes talking about something or your Forth experience?
-
I will do a presentation of FRED as an Rapid Embedded Development editor and IDE on Forth2020 zoom this Saturday. 11pm my time or14 UTC July 9th These are the Tachyon extensions to Mecrisp on the RP2040 and will show how with the addition of an SD card (4 I/O) and an ANSI terminal we can host development on the chip itself with a multitab editor and compile and execute from with that etc.
-
I will do a presentation of FRED as an Rapid Embedded Development editor and IDE on Forth2020 zoom this Saturday. These are the Tachyon extensions to Mecrisp on the RP2040 and will show how with the addition of an SD card (4 I/O) and an ANSI terminal we can host development on the chip itself with a multitab editor and compile and execute from with that etc.
-
There are better ways of downloading serially than pandering to the raw and cluttered console and echoing of every character that bog up the screen and the serial channel. All I want to see are the names of new words (optional) and any messages so it is easy to spot the error. I'm using Mecrisp on the RP2040 without any handshakes or delays and I have the baud rate set to 921600. Even loading 3,000 lines of dense source and 86kb later, it only takes a couple of seconds. Now that I have FRED fully...
-
There are better ways of downloading serially than pandering to the raw and cluttered console and echoing of every character that bog up the screen and the serial channel. All I want to see are the names of new words (optional) and any messages so it is easy to spot the error. I'm using Mecrisp on the RP2040 without any handshakes or delays and I have the baud rate set to 921600. Even loading 3,000 lines of dense source and 86kb later, it only takes a couple of seconds. Now that I have FRED fully...
-
There are better ways of downloading serially than pandering to the raw and cluttered console and echoing of every character that bog up the screen and the serial channel. All I want to see are the names of new words (optional) and any messages so it is easy to spot the error. I'm using Mecrisp on the RP2040 without any handshakes or delays and I have the baud rate set to 921600. Even loading 3,000 lines of dense source and 86kb later, it only takes a couple of seconds. Now that I have FRED fully...
-
There are better ways of downloading serially than pandering to the raw and cluttered console and echoing of every character that bog up the screen and the serial channel. All I want to see are the names of new words (optional) and any messages so it is easy to spot the error. I'm using Mecrisp on the RP2040 without any handshakes or delays and I have the baud rate set to 921600. Even loading 3,000 lines of dense source and 86kb later, it only takes a couple of seconds. Now that I have FRED fully...
-
There are better ways of downloading serially than pandering to the raw and cluttered console and echoing of every character that bog up the screen and the serial channel. All I want to see are the names of new words (optional) and any messages so it is easy to spot the error. I'm using Mecrisp on the RP2040 without any handshakes or delays and I have the baud rate set to 921600. Even loading 3,000 lines of dense source and 86kb later, it only takes a couple of seconds. Now that I have FRED fully...
-
Even after I fixed those fast timings I had a card that would work in my external adaptor but not in the Maker Pi Pico socket, and that was the second board too. So I remembered that I had been using an older SanDisk card, just a plain black one before the Ultra A1 - mainly to test for things like this. Anyway, I grabbed an Ultra and copied the files onto that and popped it into the internal socket - and it worked! So I'm guessing that the other I/O that's connected to the unused pins (in SPI) are...
-
It looks like the problem is with the way Mecrisp automatically optimizes code and it means that the SD clock is only high for a single cycle = 8ns. Combine that with the LED MOSFET gate capacitance and it's a wonder anything works. So Mecrisp is great for speed but not for timing where you need to control the width of pulses etc. I'm playing with the code now to fix it up and still have speed (and timing).
-
It looks like the problem is with the way Mecrisp automatically optimizes code and it means that the SD clock is only high for a single cycle = 8ns. Combine that with the LED MOSFET gate capacitance and it's a wonder anything works. So Mecrisp is great for speed but not for timing where you need to control the width of pulses etc. I'm playing with the code now to fix it up and still have speed (and timing).
-
I would have thought that perhaps the gate capacitance (both to drain and to source) had more to do with it. Simply using a series resistor to reduce the effect of the capacitance would have helped with this as well as the ESD diode. I will put the scope on this and check it out as I need to get to the bottom of it.
-
My most bleeding edge version is not in the cloud, it's on the card. Every day or so I print it to the terminal, copy that, and paste it into FRED.FTH where the editor trims any trailing spaces etc. So go use that but now if you want a version that you can also edit with FRED, then you can grab the one from the SD CARD folder and copy it onto the card. Alternatively, copy FRED.FTH to the card and with a blank FRED (no extension) file or the one from the folder do an import across. btw, IMPORT does...
-
I am starting to wonder if the extra I/O for 4-bit mode and the transistor led drive circuit is causing some problems. I have tried pullup settings on unused pins etc but I will investigate further to find out exactly where it fails. btw, the Maker Pi Pico I got already had a Pico mounted so I can't see why it wouldn't work either way. One thing I though was cut the 3.3V track on the GP0,1 grove connector and hook that across to Vsys so that I could supply an external 5V. I also mark this connector...
-
It was simply FRED.NEW or something but I thought why not call it FREDA just to differentiate it. When I'm working on FRED and get a compile or runtime error, I need a stable backup to edit the FRED again. So it really is just for FRED. Maybe I should call it WILMA? :)
-
Don't forget to look in the SDCARD FILES folder and make a point of copying these files in particular. FRED (you can fload this of the FRED.FTH file but this is a page file that can be edited) IMPORT (to include in FRED but only used from the command line at present) HELP (keyboard shortcuts etc - use F1 in FRED (but ^Z to close) or just open it in a tab) *optional files FREDA (used to edit FRED if it isn't loading) SPLAT.F FRED.LOG (changelog etc) FRED will create a FRED.CFG as needed which at present...
-
There's a HELP.TXT that you can import into a HELP page file but I may add some functionality to display variable line length text files. I will update the other folders shortly otherwise my Tachyon Dropbox is the bleeding edge. Actually I do have a folder for github that has stable releases so you can check that too as sometimes my bleeding edge files have temporary bugs as I make changes.
-
Yeah, it would have been so easy to allow for an SD socket on the Pico, even in SPI mode. I started backporting FRED to TAQOZ a couple of weeks ago but I have to start all over again since I have added lots of features, mainly the file tabs. So with FRED on the P2 I could use VGA and PS/2 for a standalone system. It would also be interesting to use a dedicated Pico as a video terminal/editor as well so that we wouldn't need a PC, just a Pico. Of course that Pico PC could still be programmed locally...
-
I see the problem with serial download of FRED in that I had removed the IMPORT section and made an INCLUDE IMPORT that works only with FLOAD. Sorry about that. This section is not needed normally so I just commented that line out and loaded it serially without any problems at all. I thought you would have noticed the INCLUDE straight after EDEDIT in the source! :) I will make INCLUDE detect that it is being loaded serially and simply issue a warning EDIT: just made it a patch in FRED for the moment...
-
I see the problem with serial download of FRED in that I had removed the IMPORT section and made an INCLUDE IMPORT that works only with FLOAD. Sorry about that. This section is not needed normally so I just commented that line out and loaded it serially without any problems at all. I thought you would have noticed the INCLUDE straight after EDEDIT in the source! :) I will make INCLUDE detect that it is being loaded serially and simply issue a warning
-
I see the problem with serial download of FRED in that I had removed the IMPORT section and made an INCLUDE IMPORT that works only with FLOAD. Sorry about that. This section is not needed normally so I just commented that line out and loaded it serially without any problems at all. I thought you would have noticed the INCLUDE straight after EDEDIT in the source! :)
-
A local Forth2020 fiend gave this one to me which was running a nice video demo, but the board does nothing else otherwise. I had some quirks with the SD card on the Maker Pi Pico so I added a couple of pins to the VGA board since it had SD. I will be using the board for VGA work later although I don't like the way they have used all those pins and 16-bit color is not practical with so little memory. Someone else had problems on the Maker board and cut the power link for the LEDS to make it work....
-
A local Forth2020 fiend gave this one to me which was running a nice video demo, but the board does nothing else otherwise. I had some quirks with the SD card on the Maker Pi Pico so I added a couple of pins to the VGA board since it had SD. I will be using the board for VGA work later although I don't like the way they have used all those pins and 16-bit color is not practical with so little memory. Someone else had problems on the Maker board and cut the power link for the LEDS to make it work....
-
After noticing that you posted about 3 hours ago I thought I should have a look at that but as I started to I decided to refresh the browser - and there you go - all working. It would be interesting to find our why that PC plays up though. So at some point I will be able to open a non-page formatted text file like FRED.FTH and have it imported automatically as a paged text file (127 characters + CR per line). For the moment you could quit to the console and type IMPORT FRED.FTH MACIEK then you could...
-
Oh yes, I forgot about the time zone, just like I do when I am still up on zoom at 4:30 in the morning!? If I go to bed at midnight, I am having an early night! There was a problem with FOPEN$ that didn't like null strings and it would throw an address exception. Install the latest Tachyon but remember that once Mecrisp resets it will lose RAM and so you have to reload FRED. R01# Mecrisp-Stellaris RA 2.6.1 with M0 core for Raspberry Pico by Matthias Koch TACHYON Mecrisp extensions 2206081130 - Peter...
-
Maybe you are loading FRED serially and there is nothing on the card. Try something like: 128 KB FSIZE! EDNEW MACIEK btw - you can chat with me via messenger or better still skype etc
-
Coincidentally I was just testing FRED for a totally fresh start with no CFG file etc. There were some hiccups. I'm just making a few changes at the moment and then I will update the main Forth Dropbox folder. However, I'm fairly certain that once you load FRED you can type EDME or ED <file> to start with,. Mecrisp though wipes everything when there is an address fault so you have to reload. I try to fix everything so that it doesn't get there but thanks for the feedback!</file>
-
good news then. Those LED circuits were designed to be non-intrusive by the look of it but I suspect that gate to drain capacitance and the 3.3M pulldown etc contribute to some side-effects. I will check my board and it may just be a software tweak.
-
Regardless of whether the card is formatted or not, it should still respond and be readable. With the SDPINS just remember that DI means the SD card's DATA IN, not the Pico, and DO means the output from the card. So the pins should be in this order CS DO DI SCLK So $02050304 means 4 is the clock to the card, 3 is the data to the card, 5 is data received from the card and 2 is the card select and detect. MOUNT will go a step further and read the MBR and mount the filesystem. However, here is a quick...
-
FRED was missing an INCLUDE directive so I created a file stack and now INCLUDE file can be nested up to 4 deep. I will also add some conditional compile directives such as #IFDEF etc. Interestingly although I normally use VScode in conjunction with minicom and ascii-xfr on Linux, I forced myself to use FRED to edit and compile FRED (and other files) so that I would aware of bugs and missing features etc. Now I am very happy to edit and compile solely from FRED running on a tiny RP2040 chip. The...
-
FRED was missing an INCLUDE directive so I created a file stack and now INCLUDE file can be nested up to 4 deep. I will also add some conditional compile directives such as #IFDEF etc. Interestingly although I normally use VScode in conjunction with minicom and ascii-xfr on Linux, I forced myself to use FRED to edit and compile FRED (and other files) so that I would aware of bugs and missing features etc. Now I am very happy to edit and compile solely from FRED running on a tiny RP2040 chip. The...
-
Don't forget to let the system know which board you using by defining an INIT. By default the INIT calls PICO which does not have an SDCARD. If your board is a MAKER PI PICO then do this after you have loaded TACHYON. C2F : INIT MAKERPICO ; C2R SAVE This will compile to Flash the new INIT etc. Now hit the enter again and you can use a ^C as the shortcut to reset. After that it will init the sd card and report etc.
-
assuming that this is not done old pre SDHC card but a SDHC card from 4GB upwards formatted to FAT32 then it should work. Of course Forth debugs hardware so the low level "0 0 cmd" readers there card to spi mode and returns with a 1 if successful although they may need a few retries normally. !SD does that and the rest of the card initialization and returns with a flag value etc. MOUNT does that and also mounts the FAT32 volume
-
assuming that this is not done old pre SDHC card but a SDHC card from 4GB upwards formatted to FAT32 then it should work. Of course Forth debugs hardware so the low level "0 0 cmd" readers there card to spi mode and returns with a 1 if successful although they may need a few retries normally. !SD does that and the rest of the card initialization and returns with a flag value etc. MOUNT does that and also mounts the FAT32 volume
-
FRED is not just an editor anymore - it is my Forth Rapid Embedded Development environment. Edit easily on the RP2040/Pico with multiple file tabs open etc. Now I've added session save into FRED and with the new look you can see which files tabs are open and the active one is in red. The latest files are in my Dropbox in the Forth folder while the github folder there should just be the latest stable version that might not be the most current. https://www.dropbox.com/scl/fo/5tub3mkkfb2x6cve404re/h......
-
FRED is not just an editor anymore - it is my Forth Rapid Embedded Development environment. Edit easily on the RP2040/Pico with multiple file tabs open etc. Now I've added session save into FRED and with the new look you can see which files tabs are open and the active one is in red. The latest files are in my Dropbox in the Forth folder while the github folder there should just be the latest stable version that might not be the most current. https://www.dropbox.com/scl/fo/5tub3mkkfb2x6cve404re/h......
-
I mentioned SPLAT before which is the logic analyzer. Here's a video using FRED to setup a configuration and stimulus file and then switching to SPLAT from within FRED etc.
-
Here's a dropbox link and look inside the Forth folder. In the UF2 folder I have 2.6.1 defaulting to 921600bd. In a separate linux terminal I startup minicom in ANSI mode 921600bd and ^AU to get it to add a CR (grumble grumble). After booting I do a 1 SAVE# to make a backup of the clean kernel, load TACHYON.FTH with my s script that uses ascii-xfr (no need to close minicom as Linux is happy to pipe while open) RP2040$ ./s TACHYON USB0 7 ( I use visual code studio and type that into it's built in...
-
I forgot to show the copy&paste selection process too.
-
Here's a dropbox link and look inside the Forth folder. In the UF2 folder I have 2.6.1 defaulting to 921600bd. In a separate linux terminal I startup minicom in ANSI mode 921600bd and ^AU to get it to add a CR (grumble grumble). After booting I do a 1 SAVE# to make a backup of the clean kernel, load TACHYON.FTH with my s script that uses ascii-xfr (no need to close minicom as Linux is happy to pipe while open) RP2040$ ./s TACHYON USB0 7 ( I use visual code studio and type that into it's built in...
-
Here's a dropbox link and look inside the Forth folder. In the UF2 folder I have 2.6.1 defaulting to 921600bd. In a separate linux terminal I startup minicom in ANSI mode 921600bd and ^AU to get it to add a CR (grumble grumble). After booting I do a 1 SAVE# to make a backup of the clean kernel, load TACHYON.FTH with my s script that uses ascii-xfr (no need to close minicom as Linux is happy to pipe while open) RP2040$ ./s TACHYON USB0 7 ( I use visual code studio and type that into it's built in...
-
Totally functional, running, leaping and jumping. As for "more work" that is always true of any software and I am adding more capability and features to them all, all the time. Finished off last night by starting to add a variable width window while still maintaining the 128 byte lines for small terminal or video screens. So autowrap or horizontal scrolling might be in order for that. Editing and then hitting F10 for a fast load along with the compact summary (rather than the raw echo) is so quick...
-
I am using a MAKER PICO with an SD socket as my dev board and I have my Tachyon extensions along with FAT32 that was originally meant to host dev tools including a multipass macro assembler for the RP2040. Most of this stuff is in place but I needed a resident editor just to complete it. This is all in prep for writing my own Tachyon Forth kernel but the existing tools are hopelessly mangled or crippled for assembly level coding, hence my own assembler. So yes, I can create/edit source code files...
-
I am using a MAKER PICO with an SD socket as my dev board and I have my Tachyon extensions along with FAT32 that was originally meant to host dev tools including a multipass macro assembler for the RP2040. Most of this stuff is in place but I needed a resident editor just to complete it. This is all in prep for writing my own Tachyon Forth kernel but the existing tools are hopelessly mangled or crippled for assembly level coding, hence my own assembler. So yes, I can create/edit source code files...
-
I am using a MAKER PICO with an SD socket as my dev board and I have my Tachyon extensions along with FAT32 that was originally meant to host dev tools including a multipass macro assembler for the RP2040. Most of this stuff is in place but I needed a resident editor just to complete it. This is all in prep for writing my own Tachyon Forth kernel but the existing tools are hopelessly mangled or crippled for assembly level coding, hence my own assembler. So yes, I can create/edit source code files...
-
So far today I have implemented auto highlighting for comments, for bold headings as well as other attributes. Just now I added some code highlighting. It needs tweaking as well as working in the colors with the theme but that's the fun of playing with it here and there.
-
As part of my Tachyon extensions for the RP2040 running Mecrisp I have a load of tools including FAT32, and SPLAT, the LA analyzer that runs in an ANSI terminal, and now FRED, the Forth Ready EDitor that runs in an ANSI terminal too. It is surprisingly snappy too, especially at the 921600bd default I use. You can edit a file and hit F10 to load it. There's a HELP file built-in and I am adding Mecrisp help files including a glossary as well. Just type HELP and full documentation appears that you can...
-
I forgot to show the copy&paste selection process too.
-
As part of my Tachyon extensions for the RP2040 running Mecrisp I have a load tools including FAT32, and SPLAT, the LA analyzer that runs in an ANSI terminal, and now FRED, the Forth Ready EDitor that runs in an ANSI terminal too. It is surprisingly snappy too, especially at the 921600bd. You can edit a file and hit F10 to load that file. There's a HELP file built-in and I am adding Mecrisp help files including a glossary as well. Just type HELP and full documentation appears that you can navigate...
-
Ah, sorry, I am late answering this question. I started V6 but hadn't gotten very far although there are things I want to change in Tachyon that rather annoy me now. I'd add a standard line edit buffer for interactive use and perhaps a very small history buffer, at least the previous line. I may revisit this sooner than later as I will be using edit buffers for interactive use with TAQOZ and some other Forths too. The other thing is that I will have a proper QUIT routine that can be user revectored...
-
Maciek - that's it! The true spirit of adventure - look no further than the fridge! 😄 Next time you are in a "dollar" store, or any stationary section anyway, you will find them. My colored clips are very strong and were from Daiso. btw, Daiso have the best and cheapest USB cables. I have some really light and flexible 1m microUSB and R/A USB-C cables that have a very low voltage drop comparable to the short stiff cable that Parallax ships with the P2-EVAL. Like most things in Daiso, they are only...
-
Maciek - that's it! The true spirit of adventure - look no further than the fridge! 😄 Next time you are in a "dollar" store, or any stationary section anyway, you will find them. My colored clips are very strong and were from Daiso.
-
These mag clips are used to clip papers to the fridge I think but there are all shapes and sizes as you can see the bigger one has no problems with a board stack. Seems these are called "Whiteboard magnets" from office supply places (and discount stores etc) .
-
My vices are my vice :) Magnetic paper clips and a steel bookend. I can use one clip or two and move it around as I need to quick and smart.
-
Maciek, I had a friend who had trouble with his board until he realized he was trying to run it at 5V. Maybe? btw, I see you like using proper matrix board and soldered kynar wire like me :) though I would have placed the pin headers right next to the chip and just solder blobbed across or use stripped kynar wire as there is not need for insulation on straight runs. tip: I use the backside of the flush cutters to strip wire since it more of a right angle edge rather than the cutting side which is...
-
I had a friend who had trouble with his board until he realized he was trying to run it at 5V. Maybe? btw, I see you like using proper matrix board and soldered kynar wire like me :) though I would have placed the pin headers right next to the chip and just solder blobbed across or use stripped kynar wire as there is not need for insulation on straight runs. tip: I use the backside of the flush cutters to strip wire since it more of a right angle edge rather than the cutting side which is V sloped...
-
Maciek, just to help you diagnose what your problem is with terminal garbage, bear in mind that your terminal can be flipped into an alternative character set with 14 emit or back to normal with 15. Check that. 14 emit �Ŀ. �Ŀ. ._ S�ñ�: [ 0 ] TOS: 42 *> �Ŀ. 15 �� ok. ok. .s Stack: [ 0 ] TOS: 42 *> ok. btw, that is if you have VT102 mode selected but I set mine to ANSI (^ATD).
-
Maciek, just to help you diagnose what your problem is with terminal garbage, bear in mind that your terminal can be flipped into an alternative character set with 14 emit or back to normal with 15. Check that. 14 emit �Ŀ. �Ŀ. ._ S�ñ�: [ 0 ] TOS: 42 *> �Ŀ. 15 �� ok. ok. .s Stack: [ 0 ] TOS: 42 *> ok.
-
Maciek, that's a mystery why it won't work at 921600 in Linux. Did you try another machine or even ramping the baud with CONBAUD to see if you machine breaks? I mostly use Linux Mint, some MX and now giving pop_os a good go. All good so far. Do you have a shot of the 921600 "not working"?
-
Okay, maybe you went a bit EXTREME with UHS cards :) It is possible that they need a lower operating voltage which they negotiate with the host during initialization and they really need to operate in 4-bit mode. I've never tried one so I will have to pick one up and check it out. Also, you may find that cards that are 32GB and over are preformatted to exFAT rather than FAT32 so that is why they may complain if it tries to mount the filesystem itself (after the SD is initialized) I do have a 128GB...
-
When you said 64MB I though that was a typo until I looked at the photos, and indeed it is a 64MB card. That must be 20 years old! Don't use any cards that are 2GB or less because they are not SDHC compatible. The old SD cards were byte addressable and have a slightly different initialization sequence. SDHC and beyond are block addressable where a block is 512 bytes. But I would be interested to see the CID of the Sandisk 32GB Ultra although a real one should be red and grey. Do any of these cards...
1 >
