TIPI Usage and Support

[+jedimatt42]

2 hours ago, fimbulvetr said:

For me it seems to be when CRUBASE is not set to>1000 and I make any bad call, whether a typo or mistake in capitalization. A soft reset isn’t good enough, I need to power cycle the PEB and sometimes the 4a if I use the reset button on my Ubergrom. I have to be careful setting the URI shortcuts as some cause a lockup if the URI is capitalized and vice versa. I have been setting CRUBASE to >1200 or >1400 when using Plato. My setup includes a ti disk controller, a new SAMS card, Corcomp rs232, and the TIPI with a pi zero w. I also have a sidecar speech synth plugged in. 

 

That said, with CRUBASE at >1000 it works and resets on its own when I make a bad call, so it really is only a problem when I mess with the jumpers.

 

Yep, I've check the design. It is WRONG for TIPI when it is not the 1st card in CRU order.
Works fine as long as no user errors are made.. but that means some of the internal error handling has bugs still too... I'll try and aggravate it. A lot of that can be cleaned up with PI side updates... 

[+jedimatt42]

2 hours ago, fimbulvetr said:

Yeah, I’m used to capitalization sensitivity with the 4a, I was just surprised it was locking up the computer and requiring a hard reset with the TIPI.

 

i think my long term solution will be a switch on the jumper. Right now I just put the jumper on as needed, and that works perfectly. Plato is a weird edge-case, and as soon as my kids can go back to school they will probably never look at it again.

 

I don't think there is a way to make Plato work with FIAD.. I see it loading the files fine.. I think it must check for a sector state, and that isn't replicated. I found some post from a while back that the directory structure on Plato disks is in an odd location so file copies won't work. 

[fimbulvetr]

It’s an ugly soldering job, but I now have a “Plato switch” on the TIPI.

[fimbulvetr]

It looks like the Plato cartridge crashes the machine before it accesses the disk drive. Normally, the cartridge starts, there is just a quick blink of the disk controller light with no access of the disk (maybe it is just polling the controller to make sure it is present?), the splash screen loads off the cartridge, then it loads the course information off the disk. With the TIPI at >1000, Plato only gets as far as the first blink then it freezes before the splash screen.

[Tursi]

When Plato starts up it sets top of VRAM forcibly to >3FFF (blowing away any previous init), then it calls CALL FILES(3) then a CALL FILES(1). Then it draws the red border and the PLATO title. 

 

After that it opens the only standard TI file on the disk: $$1, which contains some title information.

 

Before displaying it, it reads sector 359 (which contains the replacement directory lookup sector), then proceeds to find and read TIMENU from this list (manually, with disk sector reads). It reads only the first sector of TIMENU, then goes back to 359 to find the same file again, this time reading all of it. After that, it displays the disk information.

 

[J-Data]

On 2/24/2020 at 10:52 AM, jedimatt42 said:

run this: https://github.com/jedimatt42/ti994a-32kmemtest/releases

It's a rom for your final-grom that will run burn in testing on the 32k. I would expect this to show a problem on the data bus if there is one, eventually, within an hour or so.

Is their a version of the 32K RAM burn in test that can be run direly from the TIPI without a Final-Grom?  Ideally an EA5 that could be pulled by the TIPI directly from a repository.  I'm trying to run multiple burn-in tests in parallel on TIPI-32Ks but only have 1 Final-Grom. 

[+jedimatt42]

34 minutes ago, J-Data said:

Is their a version of the 32K RAM burn in test that can be run direly from the TIPI without a Final-Grom?  Ideally an EA5 that could be pulled by the TIPI directly from a repository.  I'm trying to run multiple burn-in tests in parallel on TIPI-32Ks but only have 1 Final-Grom. 

The code is designed to run out of RAM/ROM in the cartridge port space... so all the RAM in the 32k space can be tested without having to write code that moves itself around live... 

[+jedimatt42]

Fred Kaal has some memory tests: http://ti99-geek.nl/Projects/memtest/memtest.html

 

Run from a variety of places.. one is EA5 loadable... 

 

Doesn't really burn in, just a single pass.

[+InsaneMultitasker]

Is the filename "TIPI" reserved?

 

I assembled and copied a file called TIPI to my RPI tonight. Shortly thereafter, I could not load any program image files from TIPI.   Viewing DV80 files was OK, as was reading a catalog.  I renamed TIPI to TIPIP and programs started loading properly again. 

[+jedimatt42]

11 minutes ago, InsaneMultitasker said:

Is the filename "TIPI" reserved?

 

I assembled and copied a file called TIPI to my RPI tonight. Shortly thereafter, I could not load any program image files from TIPI.   Viewing DV80 files was OK, as was reading a catalog.  I renamed TIPI to TIPIP and programs started loading properly again. 

 

Oops, yes, it is... if you load a PROGRAM image file from a directory that has a file called 'TIPI' in it, then I try to read that on the backend like a DV/80 and pull DSK mapping configuration out of it...

 

Just load your program and the support mappings are enabled for you. 

 

https://github.com/jedimatt42/tipi/wiki/Auto-mapping

 

[+InsaneMultitasker]

1 hour ago, jedimatt42 said:

 

Oops, yes, it is... if you load a PROGRAM image file from a directory that has a file called 'TIPI' in it, then I try to read that on the backend like a DV/80 and pull DSK mapping configuration out of it...

 

Just load your program and the support mappings are enabled for you. 

 

https://github.com/jedimatt42/tipi/wiki/Auto-mapping

 

Ah, that is a cool feature.  I checked the Wiki but didn't see that special case or misunderstood what it meant.  Be aware that the 'TIPI' file I copied was PROGRAM not DV80, from the TI perspective.  I'm guessing TIPI read it anyway and had a field day with my oversight!

[aftyde]

Had some problems today with the TIPI locking up while accessing DSK1.  Red LED would stay on - and the TI-BASE app locks up.  Complete system reboot required.  After this happened three times in a row, I pulled everything apart and cleaned the edge connectors with alcohol pads - TI, Speech, Firehose - unseated all the PEB cards and cleaned all edge connectors.

 

Good to go!   

[aftyde]

Question about DSK access.  I can't seem to get TIPI to map any DSK resource.  It always defaults to the floppy disk controller.  So if I map DSK4=TIB (a directory on the PI containing TI-BASE files), then go to DM2K and try to catalog I get device errors.  I think there is some fundamental clue I am missing?  Ideally, I'd like to map DSK1 to TIB and DSK2 to my database directory.  Give the whole "get off floppies" for the database files another whirl!     If there any way to make this happen without pulling the disk controller?

Edited May 17, 2020 by aftyde

[aftyde]

When copying files to TIPI using the web page, it coverts SETUP/C into "SETUP¯C" on the PI which translates to "SETUP  C" on the TI as viewed through DM2K - which of course crashes the software. When I look at the file using mc on the PI - seems like the data is there; but file name conversations are all over the map.  HDX seems to have no problem mapping the file names back to something the TI likes.

 

I had heard that you could upload a disk image to the web interface and that it would create a directory with the files in it on the PI.  Is this true.  I have uploaded disk images to no obvious effect.

 

Wondering if anyone has some ideas on an approach I could take? 

[J-Data]

Has anyone tried connecting a Bluetooth mouse to the TIPI and would it work with TI artist and Stuart's browser like a USB mouse does?  I'm assuming the TIPI's mouse interface can accept any HID device regardless of which interface, but I can think of reasons this might not work.  Using the Pi's native Bluetooth capability would eliminate the USB mouse dongle, which is a very tight squeeze in my speech synthesizer TIPI and ties up the only USB port on the Pi ZeroW.  I'll test when I can, but don't have a BT mouse yet.  Any thoughts?

[wolhess]

2 hours ago, J-Data said:

Has anyone tried connecting a Bluetooth mouse to the TIPI and would it work with TI artist and Stuart's browser like a USB mouse does?

Yes, I have a logitech Bluetooth mouse connected to the PI and I use it in TI Artist and in Stuart‘s web browser.

It works fine for more than six months.

[+jedimatt42]

The TIPI mouse interface uses a linux mechanism that takes any modern mouse tech found and translates it to legacy PS/2 protocol in /dev/mice, this also aggregates multiple mice into a single device to read.  You can hot-plug mice all you want, and digitizer tablets, and whatever else, -- 

 

I don't know how you pair a bluetooth mouse on the PI... but any of the wireless-dongle mice should work without any configuration required.

The TIPI sd-card image is based on the 'raspbian-lite' so maybe this helps: https://www.raspberrypi.org/forums/viewtopic.php?t=214373

[+jedimatt42]

2 hours ago, aftyde said:

When copying files to TIPI using the web page, it coverts SETUP/C into "SETUP¯C" on the PI which translates to "SETUP  C" on the TI as viewed through DM2K - which of course crashes the software. When I look at the file using mc on the PI - seems like the data is there; but file name conversations are all over the map.  HDX seems to have no problem mapping the file names back to something the TI likes.

 

I had heard that you could upload a disk image to the web interface and that it would create a directory with the files in it on the PI.  Is this true.  I have uploaded disk images to no obvious effect.

 

Wondering if anyone has some ideas on an approach I could take? 

 

Sector dump disk images work better than track dump formats... my software doesn't know how to convert some track dump sizes, because it was ambiguous from the spec... and perfect didn't seem valuable..

 

once files are loose on the PI, the name in linux with the documented conversions of '.' to '/' are what matter...  So, from the windows share or whatever on the PI, rename the SETUP/C file to SETUP.C and DM2K will show it as SETUP/C... 

https://github.com/jedimatt42/tipi/wiki/File-name-rules

 

If you can send me some problem files, like this 'SETUP/C' in the original form of the file you are uploading, I can look at the FIAD headers and maybe there are some fixes to be found. I think during file upload, the source operating system filename isn't being used, and the TIFILES header is used instead... I don't think the WEB-UI should use the header in the FIAD, for consistency... so that host filename is what matters in all cases... send us some examples that break, and I'll look at improving it.

 

[+jedimatt42]

10 hours ago, InsaneMultitasker said:

Ah, that is a cool feature.  I checked the Wiki but didn't see that special case or misunderstood what it meant.  Be aware that the 'TIPI' file I copied was PROGRAM not DV80, from the TI perspective.  I'm guessing TIPI read it anyway and had a field day with my oversight!

 

I should fix this, so if it isn't a DV/80, my mapping feature ignores it... 

[+jedimatt42]

7 hours ago, aftyde said:

Question about DSK access.  I can't seem to get TIPI to map any DSK resource.  It always defaults to the floppy disk controller.  So if I map DSK4=TIB (a directory on the PI containing TI-BASE files), then go to DM2K and try to catalog I get device errors.  I think there is some fundamental clue I am missing?  Ideally, I'd like to map DSK1 to TIB and DSK2 to my database directory.  Give the whole "get off floppies" for the database files another whirl!     If there any way to make this happen without pulling the disk controller?

 

That should work... if TIB is a directory in the base of the TIPI. paths... can you browse in DM2K to TIPI.TIB. ? 

 

If it isn't in the base, then you need to add the rest of the path... TIPI.MYDEV.TIB would be DSK4=MYDEV.TIB

 

Also, TIPI's crubase will have to be lower than your floppy controllers for it to be found before the floppy drive in traditional DSRLNK based software. So crubase >1000 for the TIPI.

 

 

[aftyde]

15 hours ago, jedimatt42 said:

 

That should work... if TIB is a directory in the base of the TIPI. paths... can you browse in DM2K to TIPI.TIB. ? 

 

If it isn't in the base, then you need to add the rest of the path... TIPI.MYDEV.TIB would be DSK4=MYDEV.TIB

 

Also, TIPI's crubase will have to be lower than your floppy controllers for it to be found before the floppy drive in traditional DSRLNK based software. So crubase >1000 for the TIPI.

 

 

I can indeed...  TIPI.TIB. works in DM2K.  My tipi CRU is set to 1800, and I can see the TIB directory under tipi_disk, so it seems to be there.  I don't understand what you mean by MYDEV - I tried MYDEV.TIB and TIB, then back into DM2K but get open device error.  Can't see it. 

DSK1 is CRU 1100 so all should be good.  Basically, tipi is highest CRU in system with SCSI and all else at lower CRU numbers.  Here is what I see in the web interface...  (attached image)

 

Will give the file rename a shot and re read the link on file naming.

 

 

 

[aftyde]

One other item I noticed.  If I set my router to hide the ESSID, the PI won't connect to it.  I had made that change last night, and when I came in the PI was offline.  I telnet in from the TI to see if I had gotten a different IP address (though that is fixed by dnsmasq) and found wlan0 was not associated.  Everything else looked good... so I went back to the router, re-enabled the ssid broadcast, rebooted the PI and all was well.

 

Just FYI...

[aftyde]

BTW - thank you for making this.  I have not had this much fun TI hacking in ages!!!    And I do actually run a production workload on the thing...  My TI has hosted a parts inventory for 30+ years.

[aftyde]

Hi Jedi, attempting to paint a complete picture of the file naming issues.  I have studied the link you sent.  But I'm still unable to work out a best practices.  I must assume that other people don't have problems copying or reading files and that I am an idiot, missing some crucial bit of into, or have a broken setup - possibly all three...  ?

 

Here we see HDX1.TIB (left) copied to TIPI.TIB (right), all looks good.  In ti99dir the file names and designations match.  HDX1.TIB is TI-BASE in a fully running configuration.  I can read all files using View both at source and destination.

 

 

Here is how the PC see the files...  both on the TIPI and on the SAN containing the HDX1 source directory.  So far, all match up.

 

 

Looks good so far, then - we go into DM2K to have a look at the directory on TIPI and try to edit SETUP/C (to point TI-BASE at TIPI instead of HDX1).

 

 

WTH???    All screwed up.  So I went ahead and renamed all the files with overscores per suggestion.  TIDIR however now sees all the TI Names as wrong, the /(C or H) is missing - for kicks, I left one TEST(overscore)H which still shows as TEST/H.

 

 

However, DM2K is loving it...  It sees all the files as properly named now... except for TEST of course, which is as Ti99dir likes...

 

 

However, none of the text files are readable from DM2K - oddly, I can execute TIBASEW - which defaults to DSK1 in the absence of a readable SETUP/C file.

 

 

So this is where I am at the moment.  I'll look around at other programs to see if I can view or edit this file - very close to having it work.  If I can make this happen, and then move the DB from diskette.  Life will be better.  I am still completely unable to map DSK4 to any dir on the TIPI.  Again, I must be an idiot or something is broken... because it seems to work for everyone else.  For example...

 

Never results in a working DIR...

 

In any event - not sure if this helps in any way.  I guess the question I would have is does this actually work for anyone else?  And if so, what tools have you used to make it work?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[aftyde]

In the Happy Happy Joy Joy department...

 

Though a 8 hour routine of hacking file names and editing binaries with WinHex - to point TI-BASE from "HDX1.TIB." to "TIPI.TIB." and data files to "TIPI.AFTDB" I finally managed to get the software to run from the PI perfectly, and have it access the DB files (also on the PI) perfectly.

 

Interestingly, this has never before worked anywhere else.  Only diskettes.  All past attempts to place the DB on a hard disk of any type, or HDX resulted in data corruption.  This is sort of a life changing moment for my little TI universe.  It's fast - at least 3x.  My storage limitations are probably a thing of the past.  I used to juggle indexes on and off RAM cards and such - now it all works at high speed (for a TI) in one place.

 

I could not be happier!  I'm still disappointed that DM2K seems to not like it and there must be bugs in Ti99dir as well (they seem to map filenames in different ways) - and that file name conversions are painful and manual.  I'll test with this for a while, and if good...  I can lose a lot of hardware and maybe put some interesting old stuff up on eBay!

 

Thank you Matt, Greg, and everyone who has helped me along the way with this!  I'll probably continue to ask dumb questions - but pizza and beer is on me should our paths ever cross!

 

Cheers, Arthur...