Is that correct? clib4.library version 2.2 is older than 2.1? But I never had that version, I copied the one from the github Odyssey archive to LIBS: .

I wonder if this is across OS 4.1 Final E., or only on some hardware or configurations, I'm using A1XE.

So 'mount RAD:' followed by 'remrad RAD: force' seems fine. However, if I mount and reboot and then try to remove it, it will come up with a DSI. Then, if I reboot the same way again, without pressing the reset button or picking that full boot option in the error requester, the system never boots up.

@Gregor

Yes, only pasting is apparently fine, I did the copying before starting the browser. Copying text from a webpage rendered in Odyssey also might work.

Crash log for task "ConClip"
Generated by GrimReaper 53.19
Crash occured in module Odyssey at address 0x7BAF0844
Type of crash: DSI (Data Storage Interrupt) exception
Alert number: 0x80000003

Register dump:
GPR (General Purpose Registers):
0: 7BAF082C 6B930D70 63E51178 00000000 00000000 00000010 6B930DA8 00000004
8: 661D8494 00000000 7DE198BC 7B8D87A0 000008E4 C1603FD4 6C12B840 00000FA0
16: 6C1F0000 6C1220E4 6C122050 6C1220D4 68E3F894 6FFA05D0 01CB2D54 00000000
24: 00000004 62D6E910 68E3FD14 00000000 00000001 6FFF9300 6B930D88 00000000


FPR (Floating Point Registers, NaN = Not a Number):
0: -3.47938e-245 -4.87717e+56 2.47505e+185 2.87327e-233
4: -2.24298e+56 1.01026e+132 8.4585e+292 -6.94785e+31
8: -8.51649e+274 7.90407e+275 -1.71056e-68 4.27746e-247
12: -8.90289e-266 -7.64092e-78 5.02031e-98 -3.47519e-109
16: -8.39395e-276 -1.25847e+131 6.85225e-183 -1.13928e+139
20: 0.00143348 -0.0541526 -2.32971e+48 4.17892e-179
24: -6.92348e+301 3.38166e+285 -2.9405e+184 4.92368e+303
28: -1.51981e+202 9.73077e-72 -1.00967e+257 -4.0539e-194

FPSCR (Floating Point Status and Control Register): 0xB93E1F9D


SPRs (Special Purpose Registers):
Machine State (msr) : 0x0000D030
Condition (cr) : 0x63046DF0
Instruction Pointer (ip) : 0x7BAF0844
Xtended Exception (xer) : 0x0143DE3C
Count (ctr) : 0x00000000
Link (lr) : 0x00000000
DSI Status (dsisr) : 0x0145E804
Data Address (dar) : 0x00000000



680x0 emulated registers:
DATA: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
ADDR: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
FPU0: 0 0 0 0
FPU4: 0 0 0 0



Symbol info:
Instruction pointer 0x7BAF0844 belongs to module "Odyssey" (HUNK/Kickstart)

Stack trace:
module Odyssey at 0x7BAF0844 (section 0 @ 0x1B3820)
module Odyssey at 0x7BAF082C (section 0 @ 0x1B3808)
module DEVS:clipboard.device at 0x7FADDC9C (section 0 @ 0x1C78)
module DEVS:clipboard.device at 0x7FADE3F4 (section 0 @ 0x23D0)
native kernel module kernel+0x00039d90
module LIBS:iffparse.library at 0x7FFD212C (section 0 @ 0x2108)
module LIBS:iffparse.library at 0x7FFD05B8 (section 0 @ 0x594)
module LIBS:iffparse.library at 0x7FFD1CD0 (section 0 @ 0x1CAC)
module ConClip at 0x7FF89904 (section 0 @ 0x18E0)
module ConClip at 0x7FF89BB4 (section 0 @ 0x1B90)
module ConClip at 0x7FF88BF4 (section 0 @ 0xBD0)
module ConClip at 0x7FF88068 (section 0 @ 0x44)
native kernel module kernel+0x00059b2c
native kernel module kernel+0x00059ba4

PPC disassembly:
7baf083c: 39088494 subi r8,r8,31596
7baf0840: 394a98bc subi r10,r10,26436
*7baf0844: 91030000 stw r8,0(r3)
7baf0848: 7d4903a6 mtctr r10
7baf084c: 38600000 li r3,0

System information:

CPU
Model: IBM PowerPC 750 FX V2.2
CPU speed: 800 MHz
FSB speed: 133 MHz
Extensions: performancemonitor

Machine
Machine name: AmigaOne
Memory: 524288 KB
Extensions: bus.pci bus.agp

3.3.3 is years old and the latest, but the keyboard handling does not recognize either ctrl key. This is for information displays which are attainable slower without those shortcuts, and rebinding is another option. Also quitting the game hangs where it will lead to freezing of the system at least when interacting with Intuition.

But since the game uses more resources than the 1.x versions, the maths stuff breaking somewhat under JIT is a bigger issue. So for new characters, a spell or other attack might have a nonsensical value and this will crash the game when it's applied.
You can try a necromancer as they typically have Frost Bolt, or to avoid the crash, look at .K.ick statistics. A ratling monk with strength 8, here, shows up as "Kick: +4 bonus to hit, 1d10-16385 damage". So somehow the integer goes wrong, perhaps with the wrong type (un/signed), I don't know how that works, exactly.

It requires ixemul. I tried both the PPC and 68k v49.17, which seem to produce similar results.

@redfox

Though whether there is a memory leak without exception... I would reboot in any case.

@Kirirur

Hello right back. That's a nice A500 you had, I still have one. It's in working order but basically without upgrades. AmigaOS is and has been actually quite nice for most of its run, despite the lack of development.

I think with "AT" type layouts, Scroll Lock is Help by default.

@Capehill This is useful information.

@Maijestro No, I have not tried 3.0.0git.

edit: I downloaded the Buildbot daily build "stable" from
https://buildbot.scummvm.org/#/dailybuilds
and while it does eat a lot of RAM, I got it running at least a SCUMM game.

So looks like no public version really works, your mileage may vary: the newlib ScummVM on OS4Depot has a dependency problem except I suppose if you did not update SDL 2. The amigasoft dot net build has some problem, hard to tell because there is no error message or anything else.

@manga303

Quake 2 is indeed different than Quake:
" The Quake engine was the only game engine to use QuakeC. Following engines used DLL game modules for customization written in C, and C++ from id Tech 4 on."
(Wikipedia)

In other words, for Quake the work was already done when the first port was released. There was no code in mods that would have needed work.

@Gregor

This is because the 1..20 is not supposed to do that, instead it changes the balance of the emulation between the custom chip emulation and the 68k emulation.

As for AmigaInput, I would expect it to treat the two sticks the same and have two boxes, in addition to the hat switch graphic, in the calibration dialog. Instead it has that box and then separate axis for the right stick. So when calibrating the horizontal axis on the right, AmigaInput says "up/down" instead, from what I remember, having done this today.

But I think it's best to not start with the mode button. For instance, in the FPSE emulator (an AmigaInput plugin, to be clear) everything just works if the LED is not lit. A lot of software will not be like this, but will instead support only axes and buttons. Pressing the mode button makes the hat switch pretend to be two axes, but then you lose the fine control.

I think calibrating the unit works better when the LED is off, and doing with the LED on is just for observing.

@AmigaSociety

Is your executable file version 1.1.1 AOS4 Final, from December 2010? That's what is on OS4Depot.

I don't think it has anything to do with calibration.

Even with the keyboard it doesn't behave very well, as can be seen especially in the high score entry view where you can move many steps left and right.

I was able to reach the Leviathan on a controller, an old F310 that seems to function the same but looks slightly different everywhere. So it is not completely unplayable.

edit: I noticed on OS4Depot there is a version a few months newer than what I have installed. 1.1.1 not 1.0. But the highscore functionality is broken in both of them.

While 1.0 joystick handling was bad, 1.1.1 at 640×480 resolution is too slow for my system, which is probably why the input lag is, uh, extreme.

On quitting the 3.0 alpha it produces this. Workbench was still showing more than 40 MB RAM available.

I updated with Amiupdate without any issue, I suppose. Got A1 here through all of the public OS4 versions. I haven't used it that much yet but the USB mass storage seems to do okay.

@daveyw

I booted from a SATA drive in 2011 and probably before that (v. 1.1.1), maybe you mean a mechanism that doesn't need the env variable?

@Elwood

I might have an inkling of the arc that you are running. Still it's more likely that I have it wrong.

Well, you can try https://html.duckduckgo.com/ .
Since you mentioned searching.

Basilisk II uses UAE 68k emulation, so adding JIT is plausible.