Altirra v2.60 test 29 [27.12.2014]
- Disk: Added workaround for XFD images improperly named .ATR.
- Debugger: Added option to randomize application memory after EXE load (direct EXE load only).
- HDevice: H: can now be hot-started or stopped without a reboot.
- HDevice: H: can now also be installed as D:.
- Printer: Added SIO-level 820 emulation.
- Serial: Added software T: handler and support for 1030 handler bootstrap.
- Display: Improved quality of NTSC high artifacting mode.
- HLE: Hook page for H:, P:, R:, T:, and E: hooks is now better at dodging conflicting hardware windows.
- Serial: Many fixes to 1030 T: device command handling.
This includes a pretty big rework of a bunch of the device emulation code. Changes:
- The Add Device dialog now has detailed descriptions of each device.
- The printer has been converted to a device, and now supports SIO-level 820 emulation so it works with SpartaDOS X's whacked PRN: device.
- All of the SIO/CIO/OS acceleration options have been consolidated into an Acceleration submenu.
- 850 and 1030 have been converted to devices. For the 850, you will need to add one or more modems as sub-devices.
- Lots of fixes to 1030 and T: emulation -- command handling fixed, better heuristics for detecting dial-outs, and proper support for translation/parity in the HLE T: handler. AModem 7.5 now works.
- Added a new T: software driver so the 1030 device now supports T: bootstrap. The Additions disk also now has a 1030 bootstrap tool (boot1030.com) and a relocatable version of the T: handler (thnd1030.com). (Yes, DTMF tone dialing is actually implemented.)
- All emulated handlers (H:, R:, and T:) can now be hot-loaded or hot-unloaded without a reset, as long as they're not being actively used. The emulator now fixes up HATABS and open IOCBs on the fly.
- The host device (H:) can now also register itself as D:, for direct-loaded programs that need disk access.
- Non-disk devices now have options for SIO patch acceleration and SIO burst transfers.
- Disk Explorer now has the ability to view text files on disk image (right-click > View), with EOL conversion and inverse video support.
- The debugger now has an option to randomize application zero-page and memory (MEMLO to MEMTOP) when directly loading executables. This can catch uninitialized memory usage that would otherwise be hidden by the OS memory clear.
I'd be interested in hearing if anyone has a real 1030 and can do testing with it -- in particular, what kind of bootstrap it supports and the actual binaries for its embedded ModemLink and T: handler software. This doesn't seem to be floating around the Interwebs like the 850 stuff is. I have the XM301 disk handler, but am not sure if it is the same as the one built into the 1030.
Also, there is a problem with XE-Term not working properly when the 1030 is set to full emulation mode. The problem is that XE-Term downloads the T: handler from the 1030, but never inits it and displays the XM301-not-working message. For some reason, it appears to issue a valid 1030 bootstrap and then complains if it succeeds. I'm not sure why yet.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 41
Altirra v2.60 test 41 [24.02.2015]
- MMU: Axlon banking register aliasing is now optional.
- Issue: The "NTSC Artifacting (Standard)" option appears to have broken between versions 2.60-test32 and 2.0.60-test33.
- Fixed -- it was 15 times too bright after the luma ramp change.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 40
Altirra v2.60 test 40 [10.02.2015]
- Issue: Pressing the "menu" key on a 104(+) key keyboard. Using Altirra 2.60t39 on Windows 8.1 Pro 64-bit. Altirra 2.50 final also crashes with this as do both the 32 and 64 bit versions of both. Obviously an easy work around is just not press that key, but I have a feeling Avery would want to know that and mop that up.
- Menu key crash was just using the wrong window. Error 5 problem was due to a subtle issue when doing a direct putchar -- SDX expects ICAX1Z to be set, and because it wasn't, it was advancing the directory read pointer when writing to the file (!).
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 39
Altirra v2.60 test 39 [04.02.2015]
- Issue: The new XEP80 driver works better. There's an issue with it not handling keypresses though. It doesn't process BRK or Ctl+1 (pause) until an operation is complete. For example, copy a file to E: and if you press the Break key, after the listing is complete you get the Break Abort line.
- Yay, someone's using it! Fixes
- adds placeholders for BlackBox and MIO firmware.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 38
Altirra v2.60 test 38 [31.01.2014]
- Devices: Added MidiMate emulation.
- Friend of mine handed me a bag of obscure hardware he pulled out of his ass garage. Adds MidiMate emulation. No actual MIDI in/out and just pipes basic MIDI events to the system MIDI player, but it's enough to get MidiTrack III to play.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 37
Altirra v2.60 test 37 [30.01.2014]
- Devices: Added SX212 modem emulation.
- Fixed Autoselect not sticking for firmware mode.
- Firmware is now sorted internally so auto is more consistent.
- ATOS: Fixed keyboard repeat timer and character escaping compatibility issues.
- Added SX212 emulation support. Additions disk now contains an SX212 handler.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 36
Altirra v2.60 test 36 [23.01.2014]
Argh... introduced a zero page location overlap. Fixed in ATBasic 1.36
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 34/35
Altirra v2.60 test 35 [19.01.2014]
- XEP80: Set horizontal/vertical position commands now update the cursor address.
The horizontal/vertical position commands were updating the cursor position but not the cursor address. This version should fix it.
Altirra v2.60 test 34 [18.01.2014]
- Debugger: Added .diskreadsec and .diskwritesec commands.
- Debugger: Added dbi (dump bytes with INTERNAL text) command.
- Simulation: Added support for more realistic power-on memory patterns.
- ATBasic: PTABW is now supported.
- AltirraOS: Cassette boot now reads EOF record.
- AltirraOS: Fixed diagnostic cartridge handling.
- Cassette: Fixed inconsistent motor control state when using C: patch with short IRG mode.
- Debugger: Debug display now reflects CHACTL bits.
- GTIA: Fixed DMA reads from hardware registers.
- XEP80: Move Cursor to Start of Logical Line ($DB) no longer moves cursor to left margin.
- Added support for 512K 5200 cartridge. (Bryan, do you have a name for this?)
- Debug display now uses CHACTL
- Fixed XEP80 "Move to logical line start command" -- DOS CP now works
- Fixed diagnostic cartridge support in ATOS
- Fixed ANTIC reads from GTIA registers
- Fixes for cassette boot bugs exposed by XEX2CAS (ATOS wasn't reading EOF record, and C: acceleration was shutting motor off without updating PIA).
- ATBasic updated to 1.35 -- added PTABW support and fixed crash when DOS command fails. Did you know that JMP (DOSVEC) can return on a MEM.SAV error?
- Added support for more realistic DRAM power-on patterns. This affects software that fails to properly initialize high RAM under the OS or extended RAM. This supercedes the old randomize memory debugging option and is now under System > Memory Config > Power-on Pattern. The three patterns implemented are from my 800XL, 130XE, and 800XL + U1MB.
- Lots of changes to the DOS on the Additions disk, but I don't recommend using this just yet.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 33
Altirra v2.60 test 33 [08.01.2014]
- Problem with new burst I/O code. Also adds PAL-60/NTSC-50 support, a tweaked luma ramp based on hardware measurements, and reverts the gamma correction setting default back to 1.0. I tried to find a way to auto-compensate brightness for scan line mode, but couldn't find a satisfactory algorithm. Oh well, it's tweakable. I changed the registry path for the gamma correction value so it'll universally reset to 1.0 once even if it's been set manually, since previous builds would have auto-set it; you can change it back again if desired.
- Simulation: Added support for mixed NTSC/PAL ANTIC+GTIA combinations (PAL-60 and NTSC-50)
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 32
Altirra v2.60 test 32 [05.01.2014]
- Fixes and updates to filesystem routines. DOS 2.5 ED and MyDOS formats are now supported for read/write access and directory creation is now supported on MyDOS and SDFS disks. The file viewer now attempts to use Unicode to display some control symbols (works better on Vista+ due to better font support).
- Disk Explorer: Added read/write support for MyDOS and DOS 2.5 disks.
- Disk Explorer: Added create directory (new folder) option for MyDOS and SDFS disks.
→ NOWSZY [ATARI] Altirra x86 i x64 2.60 test 30/31
Altirra v2.60 test 31 [31.12.2014]
- Problem: The 64 bit version of Altirra wont run on Win XP 64 Bit Pro, it gives a message "not a valid 32 bit application". Most other 64Bit Programs will run on the same Xp 64 Bit, But som other programs have had the same error. On Win 8.1 the 64 bit version of Altirra Runs, The Pc with XP 64 bit i use for older games, Emulators etc...
- Answer: XP64 for old games? The mutt halfway between XP and Windows Server 2003 that has terrible driver support and still can't run 16-bit applications? Seriously? Why aren't you running XP32? The "not a valid 32 bit application" happens when the application is built for Vista or above.
Altirra v2.60 test 30 [31.12.2014]
- AltirraOS: Fix timeout for SIO operations.
- AltirraOS: Fix ESC-Clear handling in E:.
- ANTIC: Improved emulation of phantom P/M DMA effects
- 1030 ModemLink bootstrap is now supported. Required firmware image is $2880 (10368) bytes, consisting of the 128 byte boot sector with the $2800 ModemLink and T: handler software afterward.
- Added workarounds so SDX PRN: can run in intercepted mode (no line limit).
- Emulated CIO devices hook properly with OS-B again.
- Phantom player only DMA effects are now emulated (DMACTL=$26, GRACTL=$07).
- Fixed 850 firmware hotloading.
- Fixed incorrect initial translation mode for R:.
- Fixed some bugs in built-in OS (short SIO timeout, Esc+Clear not printing).
- Command-line images now set the default for the boot/open dialog.
- Fixed a bug in Altirra BASIC's BGET and BPUT commands (1.34 attached).
The 1030 boot sector initializes the T: handler before returning to the OS, so yeah, the type 3 poll is too late. Need a redump from just the SIO read, without the boot. No rush, though -- thanks for the dump work so far. As far as I know, the 1030 doesn't use the motor line but still uses the command line for control, so I'm not sure why using a cassette drive with it is inadvisable.