Alright! So at this point, my emulator can perform all of the basic TPDD file access functions pretty reliably!
Loading TS-DOS.CO off of the disk provided the perfect test. Any faults and it will lock up or throw some kind of error. After 10 read/write/read cycles, TS-DOS has started every single time. I'll add some documentation for the code and throw it up on GitHub as an early alpha in the near future. I'll keep the mailing list posted! On Wed, Jul 18, 2018 at 10:25 AM, c646581 <c646...@gmail.com> wrote: > So far, I haven't had any issues with delay. At the moment, my emulator > works with TS-DOS up until it starts reading data from DOS100.CO (because > I haven't implemented that part yet!). I did notice that it sends multiple > requests one after another, which will require me to revise my command > buffer routine. > > I'm doing all of my testing using a real Model 100 and a MAX3232 > converter, so I should see any issues if they arise. > > Jimmy. > > On Wed, Jul 18, 2018, 10:17 Kurt McCullum <ku...@fastmail.com> wrote: > >> The TS-DOS delay issue will be your biggest problem. For testing, you may >> want to load the TS-DOS or SARDOS ROM that I put together than has more >> delay built in. By default, TS-DOS waits for 3ms before if tries to switch >> down to 9600 baud. It's all part of the directory support that was added >> for DeskLink. When I was playing around with Bluetooth connections I had to >> tweak TS-DOS to wait a little longer for a response. Even some Android >> devices have slow USB response time. A USB device emulating a true serial >> port is going to have some lag and unfortunately there is no way to know >> how much lag. But 3ms is a tiny window to get a response back to >> TS-DOS. The ROM images are in the members file area. These have added delay >> and I bypass the code that switches the baud rate down to 9600 for a >> re-check. >> >> http://www.club100.org/memfiles/index.php?&direction= >> 0&order=&directory=Kurt McCullum/TS-DOS With Added Delay >> http://www.club100.org/memfiles/index.php?&direction= >> 0&order=&directory=Kurt McCullum/SARDOS >> >> >> Using TEENY, which has no delay issue, may be a good place to start. >> >> Keep us posted. >> >> Kurt >> >> >> >> On Tue, Jul 17, 2018, at 6:28 PM, c646581 wrote: >> >> I found the page on the wiki explaining DCD/DSR along with the other flow >> control quirks. Once I corrected that, it started behaving correctly. >> >> I'll look into that TPDD Arduino project, since that happens to be the >> same platform I'm using for my emulator. It kind of feels like I'm >> re-inventing the wheel, but there are some features I'd like to add to my >> emulator that that project lacks. >> >> Thanks for all of the extra info! I shouldn't have as much >> reverse-engineering to do now. >> >> On Tue, Jul 17, 2018, 20:54 Kurt McCullum <ku...@fastmail.com> wrote: >> >> >> Sorry guys, but I'm jumping up on the soap box for this one. I'm quite >> stunned by this email. >> >> First, there is the official protocol put out by Tandy "Software Manual >> For Portable Disk Drive" 26-3808 http://manx-docs.org/ >> mirror/harte/Radio%20Shack/TRS-80%20Model%20100%20Portable%20Disk.pdf >> There is a full writeup on bitchin100.com http:// >> bitchin100.com/wiki/index.php?title=TPDD_Base_Protocol >> And if that's not enough there is the TPDD2 sector access protocol >> writeup www.club100.org/memfiles/index.php?action=downloadfile& >> filename=TPDD2-Sector%20Access.txt&directory=Kurt% >> 20McCullum/TPDD%20Client& >> >> To suggest that somehow those of us who have created TPDD emulators are >> somehow hiding that information is silly. Ken and John have done a huge >> amount of work and documentation and that has been for the 'benefit' of the >> Model-T community. LaddieAlpha/mComm/Desklink and more are freely available >> for the 'benefit' of the Model-T community. If others want to create a low >> cost TPDD emulator, I'm all for it. >> >> We are not hiding any secrets here. Stepping down from the soap box now. >> >> Kurt >> >> >> On Tue, Jul 17, 2018, at 4:37 PM, John Gardner wrote: >> >> How about those who've benefited from the research of others, >> >> or even the researchers themselves, publishing the protocol? >> >> Nah... "8) >> >> >> >>
