Forums

Sega Master System / Mark III / Game Gear
SG-1000 / SC-3000 / SF-7000 / OMV
Home - Forums - Games - Scans - Maps - Cheats - Credits
Music - Videos - Development - Hacks - Translations - Homebrew

View topic - Meka 0.80 with new sound engine *TESTERS WANTED!*

This topic is locked: you cannot edit posts or make replies. Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next
Author Message
  • Joined: 05 Jun 2010
  • Posts: 701
  • Location: Pennsylvania, USA
Reply with quote
Post Posted: Wed Jul 24, 2013 1:13 am
You're packing with UPX? I stopped doing that for DOS binaries as it was causing issues with desyncing on multiplayer games. Not sure if it corrupts other things.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Wed Jul 24, 2013 8:59 am
I don't see why/how UPX would affect a software in a way that matter, either it was a UPX bug either the software was doing something dodgy rather low-level, or comparing on-disk files for version matching.
  View user's profile Send private message Visit poster's website
  • Joined: 05 Jun 2010
  • Posts: 701
  • Location: Pennsylvania, USA
Reply with quote
Post Posted: Wed Jul 24, 2013 9:16 am
Probably some low-level stuff. We are talking about DOS. But, I just wanted to let you know about it in case you're having some sort of weird untraceable bugs.
  View user's profile Send private message Visit poster's website
  • Joined: 14 Apr 2013
  • Posts: 552
Reply with quote
Post Posted: Wed Jul 24, 2013 9:17 am
Bock wrote
- Disassembler decode symbols given their full ROM bank mapping (if bank information is provided by symbol file).
- Debugger is always available at runtime. Obsoleted /DEBUG command-line parameter.
- Tidied up menu entries. Menus are now displaying shortcuts.
Discontinued ALT+P (=F12) and CTRL+X (=F10) shortcuts.
- Including Allegro binaries in repository, it should now be possible to compile the Windows version with Visual Studio 2008 with zero configuration step :)
- Added a batch script 'tools/dist_bin_win32.bat' to create binary Win32 package in one step, including UPX and ZIP tools in repository.

Adorable <3
But I've found a tiny bug!: When the input configuration is hidden it stills shows the check next to the menu entry in the menu.
But the disassembler symbols work great :)
Thanks a lot!
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Thu Jul 25, 2013 12:40 pm
- Fixed disassembly view from showing symbols not at PC position.
- GUI: Fixed checks position in menus.
- Text input : Pressing DEL key in a text input box delete the character ahead of cursor.
- Text input : fixed losing cursor when getting out of bound using Right arrow.

  View user's profile Send private message Visit poster's website
  • Joined: 14 Apr 2013
  • Posts: 552
Reply with quote
Post Posted: Thu Jul 25, 2013 10:02 pm
Very nice! Everything works now! :)
  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Fri Sep 13, 2013 4:18 pm
I forgot to say that I packaged the 20130725svn version of meka for Fedora. It is available in RPM Fusion:
http://rpmfusion.org/

It is still in updates-testing. So enable this repository if you want to install it. It should be moved to updates pretty soon.

I had to make a little patch to compile it thought. It is available here:
http://cvs.rpmfusion.org/viewvc/rpms/meka/devel/?root=nonfree
  View user's profile Send private message
  • Joined: 27 Mar 2012
  • Posts: 26
Reply with quote
Post Posted: Fri Oct 04, 2013 1:35 am
Bock - would it be possible to add another button to the memory editor to permanently commit changes to the rom? Like once I'm done with hex editing I click it to save the file? This would help immensely with my infinite live editing!
  View user's profile Send private message
  • Joined: 10 Oct 2013
  • Posts: 3
Reply with quote
Meka 0.80 build jul 25 2013 joypad problem
Post Posted: Thu Oct 10, 2013 8:04 pm
Hi.

With most version after Meka 0.73, including the latest one from July 25 2013, as soon as I enable the joypad, the emulator slow down and become "jerky" and I suddenly see a lot of "PushSampleRequestBufs overflow" . No problem with keyboard or mouse. Meka 0.73 work just fine with my joypad. Tried to switch to opengl and 22050 hz, same thing.

I have practically every emulator ever released on my PC an non of them have this issue. If I unplug every joystick, still same issue. All drivers are up to date.

The strange thing is if I run Meka 0.80 in VMware, no problem at all. I can enable joypad, map my joystick and play a game with my joystick, no slow down.

My spec:

Windows 7 ultimate 64 fully updated
Asrock Z68 Extrem4 GEN3 (USB 2 = Intel, USB 3 = Etron EJ168)
Intel 2500K
16GB RAM
Nvidia GTX 660
3 X LCD monitors
PSX USB adapter joypad
Razer Onza tournament edition Xbox 360 joypad
Gravis Extrminator digital gamepad
Anker 5000dpi laser gaming mouse

Thanks!
  View user's profile Send private message
  • Joined: 23 Mar 2013
  • Posts: 573
  • Location: Copenhagen, Denmark
Reply with quote
Window focus?
Post Posted: Sat Oct 19, 2013 8:17 pm
I'm testing the newest version of Meka. Sometimes this happens: When I hit 'enter' after I type stuff into the debugger or the memory editor, sometimes Meka does something to the window (a little like alt+enter for a kind of full-screen, at least it seems somehow similar), instead of interpreting my 'enter' as a keystroke to enter the stuff into the editor in question. I'm sorry for not being able to explain this clearly, but maybe this occasional behavior is already a known issue? It does not get in the way of the actual debugging/gaming experience of Meka, because the next 'enter' is always interpreted in the regular way, so I'm never stuck or anything. Anyway, is Meka a 'windowed' (?) application? My Meka is always in a kind of full screen state, with no regular buttons for window control. My 'old' version of Meka has the regular windows buttons, and is in a regular window?

My system: Win 7
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Sun Oct 20, 2013 9:31 pm
It should switch between windowed and full-screen if you press ALT-ENTER.
I wonder if the ALT key are stuck and for what reason. Does it have something to do with ALT-tabbing back and forth from Meka?
  View user's profile Send private message Visit poster's website
  • Joined: 23 Mar 2013
  • Posts: 573
  • Location: Copenhagen, Denmark
Reply with quote
Post Posted: Mon Oct 21, 2013 4:12 am
Yes - it might be related to Alt-Tabbing, as this does not always work for me. Meka's icon becomes faded, and is not accessible by Alt-tab? But I think I have never managed to get it in windowed mode. I have tried alter-enter.
?
  View user's profile Send private message Visit poster's website
  • Joined: 10 Jan 2013
  • Posts: 12
  • Location: Wyoming, USA
Reply with quote
Post Posted: Sat Dec 07, 2013 12:45 am
I downloaded the latest binary and started up Alex Kidd in Miracle World, only to discover that something is wrong with the controls. Upon starting the game, Alex automatically started running to the right, and there's no way to make him go left. I started up another game, same problem. I have discovered that it seems like the emulator constantly thinks that right is pressed on the control pad. Pressing left only makes whatever you're controlling stop, until you let go and it starts moving to the right again. There is no way to move to the left.
  View user's profile Send private message
  • Joined: 14 Apr 2013
  • Posts: 552
Reply with quote
Post Posted: Sat Dec 07, 2013 9:28 am
Arjak wrote
I downloaded the latest binary and started up Alex Kidd in Miracle World, only to discover that something is wrong with the controls. Upon starting the game, Alex automatically started running to the right, and there's no way to make him go left. I started up another game, same problem. I have discovered that it seems like the emulator constantly thinks that right is pressed on the control pad. Pressing left only makes whatever you're controlling stop, until you let go and it starts moving to the right again. There is no way to move to the left.

If I remember correctly this is a known issue related to using the meka.inp file of an old version of meka with the new version with meka. So you could just use the default Meka.inp file and setup your individual keys again.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Happy new year build!
Post Posted: Fri Jan 03, 2014 8:08 pm
- Updated to latest stable Allegro 5.0.10
- Updated checksum and compatibility lists.
- Fixed out of the box building for VS Express.
- Debugger: Added shortcuts button in interface for "Stop/Continue", "Step Into", "Step Over" and "Step Out".
- Debugger: Renamed "CONTRET" to "STEPOUT" and fixed it to properly take account of stack-pointer (actually stepping "out").
- TextInput: support for text selection, hold SHIFT+Left/Right keys to select text.
- TextInput: support for CTRL+Left/Right to jump to previous/next word.
- TextInput: support for clipboard (CTRL+X,CTRL+C,CTRL+V) shortcuts, tied to OS clipboard.
- FileBrowser: alphabetical shortcuts ignore disk drivers under Windows.

In the Memory Editor although you cannot select individual bits of text you can still use CTRL+C to copy a value and CTRL+V to paste one (hold to repeat copying, I adjusted the repeat rate to what I thought was nice).


meka-20140103.png (7.26 KB)
meka-20140103.png

  View user's profile Send private message Visit poster's website
  • Joined: 04 Nov 2011
  • Posts: 95
  • Location: Scotland
Reply with quote
Post Posted: Sun Jan 19, 2014 12:05 am
I don't mean to sound like a broken record but I still get the periodic freeze thing happening. I load the emu and as the menu is scrolling across the top it freezes for a second and then it continues. This 'freeze' happens maybe every second. I reported it before and have mentioned it a couple of times.

The really annoying thing is that I'm now on a different machine with different OS (i7 CPU, nVidia 540M/Intel 3000 and Windows 7 Pro 64 bit) and it still happens so it can't be anything to do with my previous computer... Just thought I should give you another heads up. :)
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Sun Jan 19, 2014 5:54 pm
Just to confirm, are you running NTSC emulation at 60 Hz ? (press F3/F4 to adjust the speed)
  View user's profile Send private message Visit poster's website
  • Joined: 04 Nov 2011
  • Posts: 95
  • Location: Scotland
Reply with quote
Post Posted: Sun Jan 19, 2014 10:34 pm
Yeah. Just running it 'straight out the box'. No changes have been made. After your post I switched region and TV type but it doesn't make any difference. :)
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Thu Jan 23, 2014 12:38 am
Last edited by Bock on Sun Jun 22, 2014 8:43 pm; edited 3 times in total
Various fixes on the debugger. Namely it display symbols at runtime for IX/IY indexed instructions.

- Disassembler: Display symbol+offset on JP/JR/CALL/RET/DJNZ instructions.
- Disassembler: IX/IY offset are always as signed and in decimal (which seems to be some sort of convention).
- Disassembler: Display symbol on IX/IY indexed instructions given the current value of IX/IY.
- Disassembler: Fixed disassembly trackback for prefixed instructions.
- Disassembler: Opcodes shown in lower-case, increased spacing for readability.
- Debugger: Fixed bug when disable CPU-space execution breakpoint.
- Debugger: Fixed handling of symbols >= 0x4000 which were broken when bank wasn't explicitly specified in symbol file (e.g.: TASM symbols, or manually input symbols). If bank is missing the symbols are treated in CPU-space instead of ROM-space.

*EDIT* Use build in the next post!


Capture.PNG (10.55 KB)
Capture.PNG

  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Thu Jan 23, 2014 11:17 am
Note that I made a handful changes in the disassembler and I'm not 100% confident it is bug free yet. Any report on using the latest build is welcome!
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Thu Jan 23, 2014 9:25 pm
I have uploaded a fixed .exe of the build above. There was a layout bug where Z80 registers wouldn't appears. If you downloaded yesterday please download again!
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Fri Jan 24, 2014 12:56 am
Fixed a few more disassembler bugs that creeped in yesterday's build.

  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Sun Jun 22, 2014 8:27 pm
This version vastly improve the disassembly track-back, you can now semi-freely scroll in the disassembly! Also added a TRACE command as requested. It automatically records up to 65536 past instructions along with the registers and you can display them at anytime with TRACE. When the demanded output is large it goes directly to the log file without being displayed in the debugger console.

Debugger: Added TRACE command to trace past execution.
Disassembler: uses recorded Z80 execution history to allow semi-freely scrolling up and down in disassembly.
Disassembler: Fixed symbol resolving in RST instruction.
Debugger: fixed bug/crashes with enabling/disabling breakpoints.
Updated checksum and compatibility lists.


trace.png (16.79 KB)
trace.png

  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 19 Oct 1999
  • Posts: 13819
  • Location: London
Reply with quote
Post Posted: Sun Jun 22, 2014 8:54 pm
Awesome! This seems much more useful than a generic spam-to-disk trace feature, and would have been useful for me last week :)
  View user's profile Send private message Visit poster's website
  • Joined: 23 Mar 2013
  • Posts: 573
  • Location: Copenhagen, Denmark
Reply with quote
Scroll column cover color?
Post Posted: Wed Aug 20, 2014 9:06 am
Is it possible that Meka does not emulate the "scroll column cover" color? I'm talking about the cover column (column 0?) that you can switch on with a VDP register (or maybe more correctly: when you blank out the leftmost column), so the background and sprites can safely scroll out to the left (the column that covers the leftmost 8 pixels on the screen). In Meka this column is always black, but it should be the same color as the overscan/border color, right?
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Wed Aug 20, 2014 9:11 am
This is correct, it isn't emulated by MEKA.

It's an old old missing feature. At the time I thought that things looked better in black because on an emulator the outside of the screen would be black anyway. Filling the entire side with border color seems odds.

But I *should* add it and then perhaps add an option to disable it, it would make more sense.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 19 Oct 1999
  • Posts: 13819
  • Location: London
Reply with quote
Post Posted: Wed Aug 20, 2014 9:16 am
The shift when the column is hidden in the first stage water part in AKMW has always bugged me a bit. It would be useful to be able to show the overscan colour when using it as a profiling proxy, more so if we can see "all the way" around the overscan area - but that would look weird normally.

Feature request: assuming people will use the overscan colour index for profiling, compute times for each colour as a percentage of the inactive display area and as a percentage of the total frame time, maybe with min/max/mean stats for some history of frames. If the screen is off, don't compute stats.
  View user's profile Send private message Visit poster's website
  • Joined: 05 Sep 2013
  • Posts: 2976
Reply with quote
Post Posted: Wed Aug 20, 2014 9:25 am
Bock wrote
At the time I thought that things looked better in black because on an emulator the outside of the screen would be black anyway. Filling the entire side with border color seems odds.


Few minutes ago in another topic I was requesting to support the overscan area so that I can see the color I set there... so, if you can, add these two -closely related- features, eventually optionally activated :)
  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Tue Sep 02, 2014 1:07 pm
Bock wrote
This version vastly improve the disassembly track-back, you can now semi-freely scroll in the disassembly! Also added a TRACE command as requested. It automatically records up to 65536 past instructions along with the registers and you can display them at anytime with TRACE. When the demanded output is large it goes directly to the log file without being displayed in the debugger console.

Debugger: Added TRACE command to trace past execution.
Disassembler: uses recorded Z80 execution history to allow semi-freely scrolling up and down in disassembly.
Disassembler: Fixed symbol resolving in RST instruction.
Debugger: fixed bug/crashes with enabling/disabling breakpoints.
Updated checksum and compatibility lists.



I tried to update the meka RPM using this version (svn r472) but I get some errors:
https://dl.dropboxusercontent.com/u/12575912/meka_build.txt

Bock, can you have a look? I'm really not comfortable with C++.

Moreover, Fedora and almost all major Linux distributions do not have allegro5.1.x available because it is still marked unstable. They only have 5.0.x. Do you think meka latest svn versions will still work with this?
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Tue Sep 02, 2014 5:03 pm
Try to update now, I think I have fixed the errors.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Tue Sep 02, 2014 5:25 pm
It should work with Allegro 5.0. I only upgraded to 5.1 because I was testing something but I think I'll revert.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Wed Sep 03, 2014 7:46 am
I have reverted to Allegro 5.0 for now.
  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Wed Sep 03, 2014 9:33 am
Bock wrote
Try to update now, I think I have fixed the errors.


I still can't build the latest svn version. I get this error now:
https://dl.dropboxusercontent.com/u/12575912/meka_build.txt
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Wed Sep 03, 2014 7:34 pm
Fixed...
Could you change the warning level from -Wall to -W3 ?
  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Thu Sep 04, 2014 7:59 am
Bock wrote
Fixed...


Now it builds fine. Thanks!

Bock wrote
Could you change the warning level from -Wall to -W3 ?


BTW, there is no -W3 option available. If you meant the equivalent of -Wall -Wextra, meka still builds fine.

For your convenience, I put the log here:
https://dl.dropboxusercontent.com/u/12575912/meka_build.txt
  View user's profile Send private message
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Thu Sep 04, 2014 8:40 am
Oh.... I still have problems with Fedora 21 and Fedora 22 (they use different flags than Fedora 20).

Build log here:
https://dl.dropboxusercontent.com/u/12575912/build.log
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Fri Sep 05, 2014 7:04 am
Remove the -Werror=format-security compilation parameter. Meka pulls format strings from the .msg file which is totally unsecure in the sense that you can make it crash with bad data, but it's not a software which you run with special privilege so it doesn't matter.

Meka really needs a Linux maintainer tho.
  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Fri Sep 05, 2014 7:07 am
Bock wrote
Remove the -Werror=format-security compilation parameter. Meka pulls format strings from the .msg file which is totally unsecure in the sense that you can make it crash with bad data, but it's not a software which you run with special privilege so it doesn't matter.


Unluckily I cannot do that. I had to follow Fedora packaging guidelines and among them there is a guideline that states packages must honour rpmoptflags, and starting from Fedora 21, there's included -Werror=format-security.
  View user's profile Send private message
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Fri Sep 05, 2014 12:06 pm
I made a patch to fix this issue:
https://dl.dropboxusercontent.com/u/12575912/meka-format-string.patch

It is based on the instructions written here:
https://fedoraproject.org/wiki/Format-Security-FAQ

Now meka compiles but I still have to test it at run time though.

Anyway, I think it would be easier to handle translations using gettext:
https://fedoraproject.org/wiki/How_to_do_I18N_through_gettext

It is also available for Windows:
https://www.gnu.org/software/gettext/FAQ.html#windows_woe32

In this way you will always have the original English text in the source code without having to remember mnemonic keywords.

There are many translation tools for gettext available, such as:
http://poedit.net/

They ease the life of the translator a lot.
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Sat Oct 04, 2014 7:27 pm
Last edited by Bock on Sat Oct 04, 2014 7:40 pm; edited 3 times in total
musuruan wrote
Unluckily I cannot do that. I had to follow Fedora packaging guidelines and among them there is a guideline that states packages must honour rpmoptflags, and starting from Fedora 21, there's included -Werror=format-security.

I have applied your patch to fullfil the Fedora zealots but it's clearly useless and simply giving a false sense of security. It's not safer with the path, Meka can probably be made to crash one way or another. For instance a lot of format strings are still loaded from file. Being a client-side app carrying all its files its probably not really a problem in the first place. The directive are fine is you are heading toward making software that's unexploitable but it's not the case of MEKA and probably not the case of many software running on Liniux distributions.

Translation works fine. The fact that the gettext library gz file is 16 megabytes and contains five million gives gives me no incentive to look at the library, it is guaranteed to be less portable than what Meka uses, aka a few C functions.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Sat Oct 04, 2014 7:35 pm
Emulation: Added support for Sega Graphic Tablet v2.0 prototype hardware.
Fixed mouse control fpr paddle, sportspad, Etc. broken in 0.80 branch.
Fixed Light Phaser / Graphic Board mouse position when game window is scaled.
Fixed moving game window when using mouse controls in GUI mode.
Various portability fixes.
Updated checksum and compatibility lists.

  View user's profile Send private message Visit poster's website
  • Joined: 14 May 2006
  • Posts: 34
Reply with quote
Post Posted: Sun Oct 05, 2014 9:42 pm
I've just submitted svn r496 to Fedora/RPM Fusion. It should be available soon in updates-testing for F19+.

BTW, to compile I had to apply the following patch because not all format strings were fixed in svn:
http://cvs.rpmfusion.org/viewvc/*checkout*/rpms/meka/devel/meka-0.80-format-stri...
  View user's profile Send private message
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Sun Oct 05, 2014 9:58 pm
Applied extra patch. Must have failed applying last, TortoiseSVN patching helper is a bit confusing when reporting errors. Thanks!
  View user's profile Send private message Visit poster's website
  • Joined: 10 Oct 2013
  • Posts: 3
Reply with quote
Post Posted: Sun Oct 05, 2014 10:16 pm
Bock wrote
Emulation: Added support for Sega Graphic Tablet v2.0 prototype hardware.
Fixed mouse control fpr paddle, sportspad, Etc. broken in 0.80 branch.
Fixed Light Phaser / Graphic Board mouse position when game window is scaled.
Fixed moving game window when using mouse controls in GUI mode.
Various portability fixes.
Updated checksum and compatibility lists.


Fantastic stuff! Unfortunately, the bug introduced after 0.73 built 2010-05-30 14;40:12, is still there for me.

The "static screen" on the interface, to simulate a TV with no signal, is "fluid" on 0.73. Meaning the "static" animation is clean and uninterrupted.

With every version of 0.80, including the last one you just posted, that static screen is freezing every 1.2 seconds or so, for about 1/4-1/8 of a second. This is the first sign that something is wrong as when you load a ROM, that freezing/skipping happens too, making the game unplayable.

I was able to trace back the issue to my USB joysticks. This is the list of my joysticks:

Xbox 360
Ni****do Game Cube
Ni****do 64
Sony Dual Shock 2
Sega Saturn
Gravis Xterminator digital

Other USB Devices on my PC:

Ducky shine DK 9008S Keyboard
Anker 5000Dpi 98AN5000-BA laser Mouse
Cyberpower CP850PFCLCD UPS
Lian-li CR-36 media reader / USB2 hub

I unplugged my joysticks 1 by 1 to isolate which one was causing the issue, but as soon as ANY of them is plugged, the issue comes back! Tried USB2 and USB3 port too, direct to PC and on a hub. The only way to make it go away is to unplug ALL joysticks!

No other emulator (and I have them all, believe me...), application or game have this issue. Only Meka, only after 0.73.

My system:

ASRock Z68 Extreme 4, Bios 2.30
Intel 2500K
16GB RAM
Intel chipset INF driver: 10.0.22 WHQL
USB2: Intel, latest driver (comes with the INF)
USB3: Etron EJ168 on board controller, driver 0.119
Windows 7 ultimate SP1 64 bit, fully updated
Aero is off

I hope this help!
  View user's profile Send private message
  • Joined: 22 Dec 2004
  • Posts: 151
  • Location: Rouen, France
Reply with quote
Post Posted: Sat Nov 01, 2014 10:55 am
Things are evolving quite positively, thanks to Bock and people who are willing to help. While I don't use a lot the debug features, I have some remarks and wishes about the current version (build 2014-10-05, used with Fedora 20/Gnome), some of which may be OS-dependant or already known, but I prefer to be comprehensive.

-Alt+enter closes Meka (instead of previously breaking the session), so it's impossible to set this display mode without editing the config file.

-Alt+Tab displays the icons of the currently open windows, but Meka is named "Inconnu" (unknown)

-Alt+Tab used while in fullscreen doesn't allow to display another window

-FM sound is still jerky, timing is irregular, thus sound effects often soud broken. No none seems to complain about this, is it just me, or an already known issue?

-No graphic filter seems to be available, either in the GUI or in the config file. On a big screen, it looks very blocky. :)

-There doesn't seem to be an option that disables the screensaver while playing.

-Zooming the tile viewer could be useful, especially on wide screens (8x8 tiles are TINY on these). Maybe merely changing the display resolution may do the trick, as the font too looks tiny.

-Just an idea: allow to set a simultaneous button 1 + button 2 key, which may be useful on some games (may be hard to do, though).
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Post Posted: Sun Nov 02, 2014 6:02 am
Hroþgar SCRB wrote
-Alt+enter closes Meka (instead of previously breaking the session), so it's impossible to set this display mode without editing the config file.


Could you elaborate on that? What do you mean by "previously breaking the session" ?
If you toggle fullscreen mode in the .cfg file it works but Alt+Enter doesn't allow to switch back and forth? Sounds like MEKA may be aborting for some reason. Perhaps if you can compile MEKA and have access to a debugger to look try to look at what's wrong.
  View user's profile Send private message Visit poster's website
  • Joined: 22 Dec 2004
  • Posts: 151
  • Location: Rouen, France
Reply with quote
Post Posted: Mon Nov 03, 2014 6:32 pm
Well, with "previously breaking the session", I mean it is a now fixed issue, featured in a previous post, but now the same operation results in Meka closing. Maybe the issue is related, maybe not, that's why I mentioned it here, just in case.

Yes, I can toggle fullscreen mode through the cfg file, but the keyboard shortcut just causes Meka to close with no warning whatsoever.

Hmm that's the theory, because I cannot reproduce the bug anymore: everything works smoothly today. -_-

Anyway, I'll be back if I have the problem again. Such a nuisance, lol
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
New build
Post Posted: Sun Nov 30, 2014 12:57 am
Inputs: Fixed dead keys when switching focus in/out of the application (namely with ALT+TAB).
Inputs: Cleaned up inputs inconsistencies, ALT+F9 --> CTRL+F9 (Sega Keyboard toggle), ALT+F11 --> CTRL+F11 (Background display toggle)
Debugger: fixed disassembly/backtracking bug with IX/IY instructions using offset + constant.
Debugger: STACK command more generous with evaluating symbols.
Debugger: STACK command shortcut added in debugger interface.
Debugger: fixed visual glitch on most-right side of disassembly.
Debugger: Added "Step Frame" (CTRL+F12) item in Debug Menu.
Debugger: Added "Load State & Continue" (CTRL+F7) item in Debug Menu.
  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 08 Jul 2001
  • Posts: 8407
  • Location: Paris, France
Reply with quote
Fonts
Post Posted: Mon Dec 01, 2014 12:37 pm
I was getting frustrated with the debugger (and generally MEKA) using such a small font and started reworking MEKA to allow for different fonts usage. It is still a bit early and so much of the code is hardcoded for certain size, but I am getting there.
fonts.PNG (67.27 KB)
fonts.PNG

  View user's profile Send private message Visit poster's website
  • Site Admin
  • Joined: 19 Oct 1999
  • Posts: 13819
  • Location: London
Reply with quote
Post Posted: Mon Dec 01, 2014 4:30 pm
TTF? Antialiasing? Unicode? :) I deliberately bought a laptop with a crappy 1366x768 screen in order to make it possible to use Meka. My next laptop will hopefully have a ridiculous pixel density and Meka will run with 100px fonts.

Also, using non-alphabetical characters in labels is like black magic. Doesn't WLA DX's parser get confused and think you're multiplying?
  View user's profile Send private message Visit poster's website
This topic is locked: you cannot edit posts or make replies. Goto page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next



Back to the top of this page

Back to SMS Power!