This Week in FujiNet – Week 26

This comes out on July 4th and I hope everyone is having a good 4th in the US. If you are from EU, just bear with us.
Screen Shot 2022 07 04 at 11 15 35 AM

This is Week 26 – the last week in June 2022.

General

TCH announced that his TNFS servers would be collapsed into one system, with top level directories for each platform. Sort of like mine, but imitation is the highest form of flattery.
I started a conversation in General asking about memory for the 64k machines and how they functioned with ROMs (that seem to overlay the address space) and we found out from dely that the 800XL is really a 62k RAM system, not 64K.

​​dely:With C64 6502 can see whole 64 KB. With Atari “only” 62 KB – $d000-$d7ff is always ROM.6510 in C64 has extra IO pins, therefore can completely expose the full 64 KB

The answer to my query is: io ports are available at the start of the systems address space, and the manipulation of those bits maps the ROMs (BASIC and or Kernel) in and out, exposing or hiding the RAM in the same address space.

TL;DR

  • Moz finished and ordered the Rev0 FujiNet boards for AppleII SmartPort
  • JeffP has a FujiNet talk at KansasFest (in 15 days)
  • Schadret has completed his port of Atari CONFIG back into the multi-platform code base
  • Thweasel and This Old Nerd are just going medieval on the SPI interface to the Z80
  • idolpx will have a FN table at the upcoming VCF SE on July 15- C64 for FN?

You can find more info at these locations:

FujiNet Flasher is here – keep your FN up2date.

Platform Notes

ADAM

Moz fixed an issue with the ADAM newsreader – he needed to switch to a longInt as the number of news articles never stops growing. He did some optimization to his back end and everything is back working.
Newsreader adam

Apple II

Jeff Piepmeier has a FujiNet talk scheduled for this year’s KansasFest! He is gathering video and information for his talk. I can’t find a schedule yet for his presentation but KansasFest is in 15 days (despite the kerfuffle that swirled around yes, Twitter, for a few days).

Smolderer is also doing a KansasFest presentation! Don’t miss KF this year- it’s cheap to attend virtually.

I’m going to present a little game i wrote for the apple iigs using the (16:9) text modes from an apple ii video card (vidHD)

https://www.kansasfest.org/

Moz has finished his design for the FUJIApple Rev0 prototype board! Real actual hardware for the Apple II systems will be here soon. The boards have already been run and assembled and are headed back across the ocean to those who reserved one. These are Rev0 and will be used by the team to test and shake out the hardware. Closely following them will come the Rev 1 boards with cases and we’ll have our official in production, ready to buy, Fourth FujiNet platform device (Atari, ADAM, Lynx and Apple II). Amazing teamwork to get this to this state.

These Apple FujiNets are SmartPort devices, meaning they do not work as simple Apple II disk devices (yet). They will work with SmartPort equipped systems- so any IIGS, IIC or IIC+ will work just by plugging them in. You also can obtain SmartPort ROMS and use an existing Apple II card like a Grappler to bridge to SmartPort.

Work is already being discussed on how to extend the current design and create a AppleII disk emulation to create a FuijNet that will work without SmartPort.

FujiApple 0 3D

Robj continued his work with the FujiNet and the AppleIII. He has built a working version of CONFIG that runs on the AppleIII

I built the latest config program and tested with the /// (and my IIc+) and it worked great.

Ron Klein is doing some nice testing with the FujiNet in a loaded IIGS. He has a Transwarp, ethernet and a MicroDrive Turbo card. A build of CONFIG wasn’t loading on his system till he removed the MicroDrive. Discussion ensued about SmartPort enumeration of the buss and if the current build was not happy with another SmartPort device on the IIGS. Getting the Apple FN device to work 100% with other SmartPort devices is still in progress, and things like this are helping remove the edge cases for the project.

20220703 183330

Ron eventually got it to work by removing the MicroDrive card from the IIGS -but found another issue with directory transversal- it would do any. Moving any images to the root directory however works until this bug was fixed.

Ron has been using GEOS 6.x on his MicroDrive (and didn’t want to give it up) but he realized that the FN could theoretically mount the GEOS disk image. This should work but there are some issues and Ron is continuing to work on getting GEOS to mount from the FujiNet. Smolderer and Rob discussed getting GEOS booting from the FN SD card and the speed of devices like the MicroDrive which use DMA to achieve near or better than DISKII speeds.

TCH fixed the issue with the AppleII CONFIG and updated the AppleII bootdisk then moved the updated disk image into the main FujiNet code repo. Ron was able to re-flash with the fixed image and CONFIG was back working again with the MicroDrive Turbo installed in his system.

Atari 8bit

Moz found the very first boot of the FujiNet on Atari- the famous Jumpman boot. He uploaded it to YouTube for anyone to witness. Intriguing bookmark for just how far this project has come in the past few years.

Schadret continued his journey with bringing the Atari CONFIG back into the common repo now used for ADAM and AppleII. He was working on the perfect build flow with FujiNet pc and apc (the creator of FujiNetPC) helped with some comments on how to run multiple instances of the magic which is called FujiNetPC:

apc: yes, ports can be altered, multiple [Altirra]-[bridge/hub]-[FujiNetPC] on single desktop is supported 🙂
apc: It’s a puzzle, I should make some howto.
Four involved components needs to be configured:
1) Prepare multiple device files (modify line with port in netsio.atdevice)
2) Create Profiles in Altirra, use different custome device file in each profile
run Altirra with /profile:PROFILE_NAME option
3) Prepare multiple fnconfig.ini files (modify NetSIO port, optionally devicename)
run fujinet with [-u URL] [-c config_file] [-s SD_directory]
4) Run netsiohub with –netsio-port NETSIO_PORT –port ATDEVICE_PORT

Schadret squashed a big bug in his code and was rewarded with some nice advances in the Atari port of CONFIG.

Config debug screen 1

And then, he was ready for testing…

Schadret (github: frachel): I think I’m pretty much done with the Atari implementation of config in the fujinet-config-adam framework. Since I could only do my testing on fujinet-pc I’m anticipating some bugs on the real hardware (especially in the wifi screens since that’s the piece that differs most between -pc and -platformio). Therefore, I could use a brave soul or two to test on hardware and report issues..

Moz took his fork and tested the CONFIG. Found some bugs. Schadret fixed them and introduced a new io function to test for wifi_enabled. TCH will stub that out for the other platforms. Atari is now back in sync with CONFIG app! Impressive job done by schadret over the past few months. The teamwork on this project keeps making it a win.

Feoh had some issues with SpartaDosX and FujiNet CONFIG- which still don’t play together perfectly. There are some configuration options in the FN Web interface which help with SDX interop, carefully choosing these options and using the N: commands available in SDX from TCH. Check out my guide here for more info- Pg.7 of the Andy’s Atari Reference has a nice list of the N commands you can use to manage your FN via SDX.

https://www.atariorbit.org/guides/

Lynx

Moz has started a new layout for his super improved version of a Lynx PowerSupply. He is spec’ing USB-C, LIPO charging and an auto load switch.

Commodore

The C64 channel has been quiet, but we know idolpx is working behind the scenes as he merged over the master branch of FujiNet into his iec (based on his meatloaf project) branch on Wednesday of last week- which means he is cleaning up code and getting iec ready to party with FujiNet for his table at VCF South East.
VCF South East is coming up July 15th and more information is available at the VCFed Website.

Vintage Computer Festival Southeast 11.0 — 2024

FujiNetPC

Schadret, in his work to port back the Atari CONFIG into the common repo has been using FujinetPC and Altirra for all his work. During this he has contributed back some nice test cases for FNPC in terms of mocking multiple wifi networks and adding code to use the wifi_enabled call, same as the other platforms. Apc happily merged in these changes for FujiNetPC.

PC-88

Seatsafetyswitch ordered a SP1 board for PC-88 and hopes to try and hack it to communicate with the ESP32 to start the journey towards a FujiNet device for this platform.

ZX-Spectrum

Thweasel and That Old Nerd (TON) continue, again, to go full bore for another week pushing the design and testing of getting the ESP talking to a Spectrum bus. I really don’t have the facilities to follow them and do them justice. I will say they continue to operate at an amazing level of output and are pushing the project further than I’ve seen any other platform implementation go. I’ll try and sprinkle some highlights here.

thweasel: @That Old Nerd I have added some details to the signaling for the Z80 bus. Let me know if I have put something on a Latch (Flip-flop) that should have been on a GPIO pin. All the operation/sources are set using a formula, so its a correct by construction approach, which means we can bounce pin allocations about no problem

See? I know something happened there, but not sure what….

TCH dipped in and helped with the discussion around triggering a GPIO pin on the ESP32, and the best way to handle that.
Skonkfactory mentioned the new Raspberry PI PicoW, which is nice but still unobtainium and there are licensing restrictions for commercial use of the Pi.

thweasel: Draft on the sanity check sheet. For each of the Z80 bus operations (row), the idea would be to copy the strip, then document how the proposed circuit is driven to achieve the operation.

E.g, ESP Write memory:
_BUSRQ, ESP GPIO3 low.
_MEMRQ, ESP GPIO4 low.
Addr, ESP SPI1 16 bits out to Sregs 1 & 2.
Data, ESP SPI1 8 bits out to Sreg3.
_WR, ESP GPIO5 low.

The list will expand when we play with NMIs or ROM swapping. Plan being once we work out a signaling order for an operation, we can document it as a row; then as we iterate on designs we can check the design can do as needed.

20220630 002811

Thweasel and TON are currently working on trying to debug SPI packets and getting good scopes time to debug over 20Mhz. TON’s logical analyzer should go 100Mhz once he sorts out how to properly use it.

Wrapup

And so that was another week of FujiNet. Another full week of good progress on a number of fronts- the summer time (the heat?) seems to be energizing many people in the project to put in the time necessary to push FujiNet forward towards its ultimate goal as the “only peripheral needed for any 8bit platform.”
I made that goal up, but it’s not far from what people expect from this amazing team.

Screen Shot 2022 05 06 at 7 06 53 PM

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.