Posted By
Luca
on 2003-05-13
 |  [YAPE] My god!
Oh mygodmygodmygod!
Please someone must keep me away from this 0.46 version, can't stop to play with it.
Attila, what a great great job you did! 
|
|
Posted By
Csabo
on 2003-05-13
| Re: YAPE
It kicks major ass, doesn't it? My jaws were on the floor when I saw Silence loading up. I messed with that demo so much, I wanted to make a hack so that it would run as separate parts, but couldn't... And now it runs fine! Exciting times, friends.
|
|
Posted By
Gaia
on 2003-05-13
| Thanks guys :-)
It really warms the heart when I read your words, guys. Thank you very much, it keeps me motivated More to come soon, I hope... The code is in place to support the rest of the turbos potentially, but it's probably bugged and it does not work yet. 
Luca: there's a surprise for ya in the sound config dialog  I haven't added that to the news because it was a last minute change.
|
|
Posted By
Luca
on 2003-05-13
| Bugs...
Ok, let cooperate for destroy the bugs:
- with GDI option on, if you alt+enter for fullscreen, and alt+enter again to come back, you'll loose GDI
- 256K emulation doesn't work, try it with Dream world, e.g.
- +4 Mania Forever stilll doesn't run
have you got other of those insects?
|
|
Posted By
MIK
on 2003-05-13
 | ..
Ati well done mate!! I've yet to take a peek but reading the news its unreal and can't wait to see what she can do
Many thanks!!!
|
|
Posted By
SVS
on 2003-05-14
 | Attila: only a word
T H A N X
|
|
Posted By
Luca
on 2003-05-14
| Us it as a test
Attila, probably you shouid use Fatal Mirror demo for testing the next version: its second fli plasma totally fails, and the same for its morphing fx...
BTW, I saw that Silence to has some weird rasterbars and, most of all, that strange fx during the "LandScape"'s piclogo vertical scrolling...
Indeed, we're talking about our beloved YAPE, but the complexity level has completely changed after the release of this stunning 0.46, really damn good job!
|
|
Posted By
SVS
on 2003-05-15
| (De)bugs...
I experienced the following bugs:
a) Drive LED missing;
b) Repeat print on screen (the same char), when Joystick setting is on.
c) Confuse Capital-Lock key (sometimes when hit, you have lower chars)
d) LOAD executed from a running prg. do not work properly (variables are reset)
e) ALT-D do not "remember" last set folder.
f) Returning from full-screen to windowed mode, you lose GDI set (as Luca said)
Staying at disposal for possible further info, I wish a good work!
|
|
Posted By
Csabo
on 2003-05-15
| Re: Bugs
SVS, the drive led appears on the Quick Debug info. Press SHIFT + F9 to turn it on. It's a one character symbol right before the T/S (track/sector) values of the drive.
What the joystick emulation does is emulates the joystick movements when we press the cursor keys. On a real C16 or +4, when you move the joystick, the same thing happens (3WA4, or 5RD6 keys are 'typed'). So with joystick emulation on, the cursor keys won't move the cursor, but type these letters, this is by design. Is this what you were referring to?
|
|
Posted By
SVS
on 2003-05-15
| Re: Bugs
>>>So with joystick emulation on, the cursor keys won't move the cursor, but type these letters, this is by design. Is this what you were referring to?
No: what I want to say is that *sometimes* I have a continuous key' repeat when I hit CapsLock where Joystick is set (for example wwwwwwwwwwwwwwwwwwwwwwwwwwwwwww...). To stop it, I have to unlock CapsLock on keyboard, or unselect Joystick from Yape. When the bug does not appear, the joy emulation works properly (sending the right chars as you said).
Secondary, it seems that Yape goes "into confusing" keyboard processing, because after the bug appears, the Caps works on a contrary way (Caps on produces low chars; Caps off produces high chars).
This bug is not a new one, I signalled it to Attila time ago. (I don't think it's mine keyboard anyway).
|
|
Posted By
SVS
on 2003-05-15
| (De)bugs...2
Other detected insects:
g) 256K management: bit 2 of $FD16 is always set to "1". And it becomes "1" even if you just set it to "0";
h) Banking read operations work bad (you always obtain the value of the same bank, when you want to read a cell of a different one);
Sorry to bore you Attila, my friend. Anyway I discovered the SEQ read operations work as well (finally!), and also the MegaAssembler Turbo-load runs
|
|
Posted By
Gaia
on 2003-05-15
| OK, let's see
Hi guys. Thanks for all the exhaustive bug reports and also the compliments. I'm trying to answer them below.
1. Fatal Mirror I use for testing and I am aware of the bugs in the morph. It's a very trick demo, I'll have to investigate the most inner workings of the TED to fix it (where exactly some registers change state etc.)
2. +4 Mania Forever: I'll have a look
3. CAPS LOCK: Note that I am not detecting the state of the CAPS LOCK but the state change. So if you set the CAPS LOCK outside of Yape then it will be inverted. It's a low priority issue at the moment, but if it really annoys you, I'll see what I can do.
4. Keyboard repeat is occuring when the WM_KEYDOWN and WM_KEYUP Windows messages get mangled or intersected by another Windows message. In that case the plus/4 buffer might be forgotten to be emptied and the intersecting message gets implemented as a keypress in the emulated machine. Since no respective WM_KEYUP message is detected, the key gets interpreted as being pressed continuously.
5. AL+D: I don't understand the problem... I assume you might misinterpret the goal of that option. The Set drive folder is meant to set the folder for the 1551 emulation if you're willing to use a directory as your disk (instead of a D64). Note that this is also the only manner for disk writing operations at the moment.
6. LOAD bug: what do you mean? It's not clear what variables got deleted? Also what LOAD you mean? The BASIC command?
7. Fullscreen vs. GDI: you are right on that, but it has so far been explicitly set to stay in DirectX mode instead of GDI once you've used fullscreen mode. Again, if you think it's annyoying I might reconsider this.
8. Raster glitches: again some of those raster tricks are very timing sensitive. ALthough Yape is rather accurate already, it's still not 100%. Even a few cycles difference can cause some raster effects to flicker. One known bug is that sometimes the raster timings get wrong on bad lines (you can see it as every 8th rasterline effect is "out" of synch), but I still could not find the reason for it...
I am trying to best to sort things out, but there's still a lot to do... For example, I want a proper GUI for the drive settings (I'm now working on true 1551 emulation by the way...), for the joystick settings and so on. I also want to further improve disk emulation, the external monitor and the remaining TED glitches....
Moral of the story: I should update the README.TXT and maybe include a Known Bugs section too. Is there any volunteers for making a nice User Manual for Yape?
|
|
Posted By
Ulysses777
on 2003-05-15
| Another (minor) bug?
If you try to access a D64 within 1 second of loading or resetting YAPE (either with "Autostart Image" or with BASIC commands, the emulation freezes (although the emulator itself is still working). I've noticed that the drive LED comes on for 1 second when loading or resetting YAPE, perhaps that might have something to do with it (although this bug happens whether the LED is displayed or not). I don't have a disk drive, so I don't know if this happens on a real machine or not...
I've also noticed that when you use "Autostart Image" on a D64 file, it now uses the C64 command (LOAD"*",8,1) instead of the 264 command (DLOAD"*"). It still works anyway, but any reason why it's changed?
|
|
Posted By
Gaia
on 2003-05-15
| Drive emulation
The freezing is there on the real machine either. When the drive's CPU is still in its init/reset sequence, your computer may freeze if you try accessing the drive too quickly. Just wait with any disk activity until the LED burns out at startup.
The DLOAD command has been replaced by the LOAD"*",8,1 because the latter loads to the address where the program was originally saved from and not to the start of the BASIC area. Some machine code programs will fail to load to the desired memory location and probably crash.
|
|
Posted By
SVS
on 2003-05-15
| To clarify some bugs descriptions:
>>>>5. ALT+D: I don't understand the problem... I assume you might misinterpret the goal of that option. The Set drive folder is meant to set the folder for the 1551 emulation if you're willing to use a directory as your disk (instead of a D64). Note that this is also the only manner for disk writing operations at the moment.
a) When you set a new default-folder, the prompted folder is where YAPE resides. You navigate till you set a different folder. Well: if you attempt to modify it, you are prompted the YAPE folder again, and not the previous set folder.
b) I'm not able to run a DIRECTORY command of the default-folder, if I have not load something from it yet (by [F8]).
>>>>>6. LOAD bug: what do you mean? It's not clear what variables got deleted? Also what LOAD you mean? The BASIC command?
To obtain automatic multi-loads, somebody uses the trick to test a variable each time the "booter" BASIC program is run. The first time the control-variable is = 0: the program set it to 1, then LOAD the second segment. As you know when run-mode executed , "LOAD" command performs a GO 0, that is re-executes the BASIC program existing in BASIC area (that is the Booter). When executed for the second time, the booter does not go into an endless loop, because of the control-variable that is different (1 instead of 0), ... and so on.
I checked this with Yape 0.46 but the control-variable still remains 0. Maybe when the booter runs the second time, it clears the variable-area (?).
Hoping you understand *my* English. If not, do not hesitate to ask me for further info, OK?
Ciao
|
|
Posted By
Luca
on 2003-05-15
| YapeBench
Ok, this is mi idea:
we should select a resticted bunch of +4 stuff that use all the TEd features you picture in mind, something like 3DMark2001 or similar benchmarks.
Oh, apropo: morphing won't work in all the demos that support it, and other diskdemos don't load as well...
|
|
Posted By
Crown
on 2003-05-15
| Crash bug
I turned on 1541 emulation, turned off 1551 emulation and as soon as started to load something, it crashes. Crash address is 004010AB.
|
|
Posted By
Gaia
on 2003-05-16
| Re: crash bug
Hi Crown! Could you send me the D64 you've tried, or just give a URL? The crash address alone does not help too much. Thanks.
|
|
Posted By
Crown
on 2003-05-16
| Re: crash bug
I've tried Fatal Mirror, and Silence downloaded from here, but it happened also when there was no disk attached. If you use MSVC and compiled with a pdb files enabled then you can load up the exe into MSVC, and check the address back to the source.
I couldn't really start either of them, with other settings even. Silence would show the first screen, then nothing happened.
|
|
Posted By
Gaia
on 2003-05-16
| Re: Re: crash bug
I'd check it... but it's strange 'cause both are running without problems here, so I can't really reproduce a problem Please try deleting the INI file, maybe that helps...
|
|
Posted By
Crown
on 2003-05-16
| Re
I just checked it on XP and it worked, the crash was happening on a WinME box, I'll try to redownload the binary tonight and check it on a few OS.
|
|
Posted By
Luca
on 2003-05-16
| A significant bug
I will try YAPE 0.46 with many demos.
I tried it with 5 Magics too, and the 42 columns screen in the Crown's part (the 3rd) fails!
It seems to be something fundamental...
|
|
Posted By
Csabo
on 2003-05-16
| Re: A significant bug
Sure it's fundamental. Running that demo properly would require that the WRITING of $FF1E is emulated. We are not quite there yet (to put it mildly). That particular thing is probably in the top 3 of 'most difficult thing to emulate ever'. Very few demos use it (in fact I can't think of another one besides Crown's).
|
|
Posted By
J. Robertson
on 2003-05-16
| Compliment and suggestion
Great emulator Attila - thanks for the work!
I don't have a bug report more of a request . Would it be possible some time in the future to include NTSC support as well? (ie. TED timing, resolution, processor speed).
Jason
|
|
Posted By
Luca
on 2003-05-16
| Re: A significant bug
I've seen it at the end of Last from GFW demo, when thee is the tv-off fx: whereas vertical screen depth increases, YAPE shows a nice beautiful black border on left and right sides.
|
|
Posted By
Gaia
on 2003-05-17
| Re: NTSC
Thanks Jay NTSC is certainly possible only I need to find the time to implement it. Also it'd require some architectural changes in the code. So it's not going to happen in a few weeks
|
|
Posted By
J. Robertson
on 2003-05-18
| Re: NTSC
Understood, it's not an urgent need anyway. Well, back to running some programs...
|
|
Posted By
MC
on 2003-05-28
 | Re: [YAPE] My god!
I'm so haaaappppy to read that my Fatal Mirror demo is being used as a main test sample for TED
|
|
