Home Page Forums MAME MAME Horizon Ramdisk on TI-99/4a

While playing around with the Horizon ramdisk, I attempted to create a 16mb drive, however no matter what ROS I try to load onto it, it just freezes up the emulation. I am also still getting corruption on my HD images since switching to 0.147. I can read the drives without problems, and the data that was written to them seems to be fine since converting them to v5. However if I try to write to them, the data gets corrupted. For example, I tried to put a backup copy of the Horizon MENU program on the HD so that I could experiment with the various Horizon Ramdisk sizes. However, when I copy the backups from the HD image to the Horizon using Myarc's disk manager, the data is always corrupted. This corruption does not take place however using just floppy .dsk images. I'll keep experimenting, I'm sure it's just some stupid mistake I'm making, but it's frustrating none-the-less.

Hi dave,

try to reproduce the problems you are experiencing with the HD access and post them here or send them to me personally.

What I need is a step-by-step description how you get from a proper image to a corrupted one. And you should isolate that from possibly different issues coming from Horizon or other cards.

Have you tried with smaller sizes for Horizon? I know about a bug in the initializer program for sizes larger than 2 MiB. Also, consider the dip switches with the different modes.

Michael

Michael,
I don't seem to have any problems at all using a 2mb Horizon, it works perfectly! As to the HD troubles, they seem to be intermittent. I ran the emulator for about 4 hours straight last night, organizing the main HD for the TI. Most of the time it seems ok. It only seems to give me trouble when I transfer certain files from a floppy image over to the HD. I have been thinking it might have something to do with using Myarc's disk manager. I started using Birdwell's utilities last night and didn't seem to have any problems at all.
On another note, I noticed that the keyboard and joystick routines don't seem to be quite as smooth as they used to be. For example joystick response on cartridge games seems to be a little jerkey (not drastic, but noticeable), could that have something to do with the changes you said you were going to make to the interrupt routines?
And finally, I'd like to ask for a Christmas present , could it be possible to request an emulation of the P-Gram card? It is the one last piece of hardware I used to have for my TI that I really miss.
Thanks for all your prompt responses and I will keep farting around until I can give you a definitive answer as to how to reproduce the HD troubles I seem to be having. Like I said before, I think it's just something stupid that I am doing, (or did in the past when I originally set up the drive image).

A very Merry Christmas to you and all the developers of MAME/MESS.

Dave

Hi Dave,

the Horizon RDs should all work properly from the hardware point of view. However, there were only few of them available with more than 2 MiB; the 2 MiB version was expensive enough. Hence, the ROS was never thoroughly tested with such big RDs, and neither was the initializer tool which is used to upload the ROS to the card.

When I wrote the HRD emulation and tested it, it was right then when we discovered the bug in the initializer. IIRC it failed to set the memory banks correctly. I remember the author said he fixed it in the meantime; I did not try again since then. For the time being, you are safer with the 2 MiB setting.

Second, the Myark Disk Manager does have some bugs, and maybe you found one of them. Try to work with a different disk manager in the meantime. If it is MDM's fault there is not much I can do about it. I'm mainly working with the Geneve emulation and its MDOS and luckily I never had issues with the file system.

My TIImageTool java program has a feature to check the file system. You should try that. It can save its report to a file. The check includes checks of allocations, detecting cross-allocations and lost blocks. Once you have a defect in the file system the effect may aggravate over time.

I am sorry to tell you that the P-GRAM will not be ready for this 2012 Christmas season. But maybe next year.

Michael