05 June 2022, 22:45 | #41 |
This cat is no more
Join Date: Dec 2004
Location: FRANCE
Age: 52
Posts: 8,201
|
zoned everything I have: sources and disassembly of the exe
|
06 June 2022, 16:06 | #42 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Cool! Got it working after struggling with a really stupid error - had some wrong offsets for some exec LVOs in an include - doh!!
Attached is what I have so far. When text is being displayed it waits for VBL every fourth time the function is called (otherwise the text is too slow). In game frame interval is set to 4. Comments on those choices are welcome. Next I'll try to see if I can figure out where the graphic glitches come from. Also noticed something funny, if the game is running too fast the plane can crash during the demo and be returned to the hangar ship |
10 June 2022, 23:26 | #43 |
This cat is no more
Join Date: Dec 2004
Location: FRANCE
Age: 52
Posts: 8,201
|
I have merged your slave with mine. I had patched some cou dependent waits in the meantime.
Will try that this week-end. Work got in the way I could barely code anything that last week. |
11 June 2022, 21:49 | #44 | |
This cat is no more
Join Date: Dec 2004
Location: FRANCE
Age: 52
Posts: 8,201
|
had a quick check and it seems to run great. I'm releasing it now.
Crediting paraj for his help with the speed regulation. This will be shown in the splash screen and readme Quote:
|
|
12 June 2022, 11:04 | #45 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Haven't had much time either. Don't know if you've already fixed it, but text flashing green in external view can be worked around by:
Code:
PL_PS $27258, drawtext ;... drawtext ; Original code ADDA.L D7,A4 ;27258: d9c7 MOVE.W #$0142,D7 ;2725a: 3e3c0142 ; D1/A0 free lea $dff000,a0 BLITWAIT a0 rts Last edited by paraj; 12 June 2022 at 12:13. |
12 June 2022, 15:15 | #46 |
A3000-Fan
Join Date: Sep 2008
Location: Germany
Posts: 321
|
Would it be possible to get Their finest hour a little bit smoother? For me it's unplayable slow with a very low framerate - even on a 68030. No comparison to the DOS version...
|
12 June 2022, 15:37 | #47 | |
This cat is no more
Join Date: Dec 2004
Location: FRANCE
Age: 52
Posts: 8,201
|
Quote:
Psygore is the guy. I don't have his sources. paraj, not a problem, I can issue a lot of further releases with further fixes. CPU/blitter concurrency is a big problem because it's difficult to detect. Happens when you want to optimize your games using cpu & blitter in parallel. cpu for non-shifted screen copies, blitter for the rest |
|
12 June 2022, 16:10 | #48 |
Guru Meditating
Join Date: Jun 2014
Location: England
Posts: 2,339
|
will this release appear on whdload website, @jotd?
|
12 June 2022, 19:03 | #49 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Probably going to leave the plane popping issue as is for now. Here's what I've learned so far:
- Can't get it to happen in WinUAE (tried many combinations of JIT/no-jit and various CPUs) - Happens consistently for one (game) frame on my A1260 when external view orientation "change" (numpad input) [orientation doesn't have to actually change] - Only seems to affect the plane when when it's on the hangar ship, and only the part that's supposed to be in front - Does not appear to be a blitter issue. I've tried patching every blitter wait loop to be "proper" and also force immediate blits (waiting for completion after BLTSIZE write). No dice. Looks like polygon/object sorting isn't right for one iteration, but I have no clue how this would happen (and not be seen in WinUAE). Test case: Start game, press 2 for free flight mode, 4 to start on hangar ship, numpad enter to get external view, then numpad 3 a couple of times (each press=> plane is hidden behind landing strip on ship). Would be nice if this could be reproduced in WinUAE with some configuration. |
12 June 2022, 19:25 | #50 |
This cat is no more
Join Date: Dec 2004
Location: FRANCE
Age: 52
Posts: 8,201
|
yes if it cannot be reproduced on WinUAE this is going to be hell to fix. Like old times...
Unlike 2D games where you can figure out a lot of stuff like player x,y and work from that, 3D games are particularly difficult to understand without the source code.... e-mail sent to Wepl yesterday. He'll update the page in the next day(s). |
14 June 2022, 15:37 | #51 |
Registered User
Join Date: Nov 2014
Location: NSW/Australia
Posts: 462
|
Thanks for your work on this - is there a target accelerated config to get the right speed?
My naive test in WinUAE was to free flight from the carrier, start timing as soon as I smashed F10 repeatedly to get afterburner, then climbed to 5000ft lifting gear as soon as possible. The time taken (stopwatch) for the first pixel of my plane to hit green land in the map view was: A500 Cycle Exact: 1m44s A1200 Cycle Exact 8x (old slave): 0m54s A1200 Cycle Exact 8x (new slave): 1m03s So it's a lot nicer now and the text is readable, but the physics is still a way off. Is a fix for the physics possible? There is a youtube (not mine!) showing a more methodical approach to calculating the real world time. [ Show youtube player ] Last edited by dalek; 14 June 2022 at 15:42. |
14 June 2022, 18:32 | #52 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
TBH I don't know what the right speed is. My initial assessment was that it seemed to just update the game more as less as fast as possible, and fix that to something reasonable, testing on my 060 and in WinUAE with various combinations of settings. Also didn't do much real playtesting except flying around a bit and watching the demo.
Could you elaborate on what you mean by physics way off (not doubting you as I've seen the demo crash with certain settings)? I've dug a bit more, and the game actually does seem to use a real time clock for some updates, I'll have to watch the video, and see if sheds some light on what the expected speed is and perhaps come up with a solution. EDIT: Watched the video, and I think it mostly confirms what I wrote above. If the game is updating as fast as possible and that's only every 5/6/7/8th frame on a stock A500 then the updated slave (which is limited to 1/4th update) will run almost twice as fast if that's possible. Though no faster than that. If that's the case and accurate speed is desired then the solution is to even further limit the game update frequency, but then you'd not get a faster frame rate (without a time consuming rewrite of the game). Also at 1:20 the popping plane issue is seen on the right-most monitor Last edited by paraj; 14 June 2022 at 19:15. |
15 June 2022, 00:32 | #53 |
Registered User
Join Date: Mar 2019
Location: Melbourne, Australia
Posts: 422
|
Amazing work guys.
|
16 June 2022, 13:42 | #54 |
Registered User
Join Date: Nov 2014
Location: NSW/Australia
Posts: 462
|
What I mean about the physics is that it just feels too fast - like you climb to high altitudes really quickly and can't fly slow properly. I haven't tried the rescue pod mission but I expect it would be quite difficult at 2x speed as it requires you to slow right down and drop a pod at low altitude.
The game seems to be written for an NTSC A1000 / A500 type system - actually my tests above were PAL so I've retested as a stock A500 in NTSC and it's actually pretty much the same. Growing up I played this game for many hours (it was the first and only game I had for months) so I'm just used to how it felt on my PAL A500. Playing on an accelerated Amiga just didn't work well for me so my latest replay was on my MiSTer at stock A500 speeds. I think if you can cap /get the frame rate to twice as fast as an A500 but keep the time it takes to fly between 2 points the same as a stock A500 (or within say 10%) that would really make the game much nicer. It sounds like that may be the only possible solution? I had though of other ways to test flight time - like setting up keyboard macros but I think a margin of error is fine. But say 1m44s vs <1m really doesn't make for a playable game if you are used to it. I wonder what other people's thoughts are? I'm sure there are a lot of people who only ever played it on an A1200 never realising it was running faster than it should... |
17 June 2022, 17:39 | #55 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Unfortunately it looks like the only (easy) solution is to cap the update frequency (which also limits the frame rate). Maybe, just maybe, it's possible to do but it would probably require extensive play testing to make sure something else isn't broken.
A "compromise" of sorts might be to have an option to limit the update frequency to something that better matches A500 ("real") speed. That would at least enable playing the game closer to the original experience on an accelerated Amiga. I've attached a few different versions with different frame caps (can be seen in the file name). If either gives a better experience it would be easy to include an option to play at that speed. Even if none of them are "good" knowing which one is closest to the wanted speed would make it easier to try to aim for A500 speed later on. |
18 June 2022, 17:33 | #56 |
Registered User
Join Date: Aug 2006
Location: Scunthorpe/United Kingdom
Posts: 2,000
|
This keeps coming up and nobody seems to believe me but there is a version of F/A 18 Interceptor out there that was fixed for the A1200 and gets a much better framerate without accelerated physics.
I was a massive fan of the game and was really disappointed when my copy failed to run at all on my new A1200. I got in touch with a friend who supplied me a fixed cracked version and it was just glorious. I guess whoever bought my A1200 off me way back in about 96 might still have the disk but it's a long shot. It's pretty easy to recognise as it had a green-screen style ASCII crack intro bolted on at startup. |
18 June 2022, 18:34 | #57 |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Would be nice if that version could be found. It's not beyond belief that somebody made great A1200 version. There could be a simple trick we're missing or they just put a lot of effort into it.
|
18 June 2022, 18:49 | #58 | |
Going nowhere
Join Date: Oct 2001
Location: United Kingdom
Age: 50
Posts: 8,997
|
Quote:
Just for reference, FLT had absolutely FUCK ALL to do with it, but somehow it go spread on the BBS's back in the day, people thought it was legit, I think it even made it onto the FLT CD as well. Its on the file server, just do a search for "Interceptor FLT" and see if thats this mysterious version. |
|
18 June 2022, 19:31 | #59 | |
Registered User
Join Date: Feb 2017
Location: Denmark
Posts: 1,107
|
Quote:
Could you put the ADF in the zone or PM me slightly more precise instructions (I think I know how to access the file server)? Thanks. |
|
18 June 2022, 19:45 | #60 | |
Going nowhere
Join Date: Oct 2001
Location: United Kingdom
Age: 50
Posts: 8,997
|
Quote:
if it fails to load addbuffers simply type LOADWB hit return, and then ENDCLI and hit return. I didn't have anything to do with this, execute the file X1, someone bolloxed it up, but doing this makes it work. |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
F/A-18 Interceptor | qzxcvbn | support.Games | 4 | 16 April 2022 23:31 |
F/A-18 Interceptor | jotd | Nostalgia & memories | 51 | 21 January 2020 09:36 |
F/A-18 Interceptor | whiteb | project.SPS (was CAPS) | 12 | 12 May 2013 19:44 |
F/A-18 Interceptor | _-IceMan-_ | support.Games | 3 | 13 July 2008 19:18 |
F/A-18 Interceptor | VermillioN | support.Games | 4 | 15 October 2002 06:58 |
|
|