WinAPE 2.0 Beta 1

[Executioner]

Works fine! Got the name "Winape IDE"

Good, at least you didn't get "iWanepI ED", stupid big-endian specs.

[pelrun]

provided it follows the laws of BOMDAS!

Sounds like a Doctor Who villain...

[Audronic]

Winape and Parados V1.2
When using Parados Version 1.2 in Slot 6
It thinks Drive B: is A:
I have tried this on 2 emulators and 1 MegaFlash and the effect is the same.
Thanks   Ray

[CraigsBar]

Winape and Parados V1.2
When using Parados Version 1.2 in Slot 6
It thinks Drive B: is A:
I have tried this on 2 emulators and 1 MegaFlash and the effect is the same.
Thanks   Ray

1.2+ works fine for me on the 6128+ and 4128+


Both drives A and B are correctly accessed.


I don't currently have a CPC 6128 set up so cannot test that just yet, will see what can be done at the weekend.


Craig

[Audronic]

@CraigsBar
Ok Thanks I don't have a Plus to test it with.
Did you put Parados 1.2 in ROM Slot 6.
I can  check what happens in a couple of emulators.


Thanks   Ray

[Docent]

Hi,
I played a bit with timers and have a question: help file says that the granularity of a timer is max #ffff (16 bits?), while mine show values over 170000 or so during testing, are they 32bit or 16?
I have also an enhancement request - would be great to have the possibility to use also strings as timer ids, for example timer_start(main_loop) etc, It would be much easier to identify a timer while profiling.

[AMSDOS]

Hi,
I played a bit with timers and have a question: help file says that the granularity of a timer is max #ffff (16 bits?), while mine show values over 170000 or so during testing, are they 32bit or 16?
I have also an enhancement request - would be great to have the possibility to use also strings as timer ids, for example timer_start(main_loop) etc, It would be much easier to identify a timer while profiling.

32bit I would of thought. The CPCs internal timer system, as used by KL TIME PLEASE (&BD0D), uses DEHL to return a four byte time count of the time elapsed, with D holding the most significant byte to L holding the least.

[Executioner]

I played a bit with timers and have a question: help file says that the granularity of a timer is max #ffff (16 bits?), while mine show values over 170000 or so during testing, are they 32bit or 16?

Internally, parts of the timer are 64 bit, but the expression evaluation in the breakpoint conditions is only 16 bit, so the maximum returned value is 65535. I could perhaps add timer_high and timer_low functions to get 16 bits at a time where it's necessary to time over 65535 cycles.

I have also an enhancement request - would be great to have the possibility to use also strings as timer ids, for example timer_start(main_loop) etc, It would be much easier to identify a timer while profiling.

It would be possible, but all breakpoints are compiled into byte-code for faster evaluation, so the IDs would need to be given a unique 16 bit ID, then all breakpoints which use them recompiled.

[Executioner]

32bit I would of thought. The CPCs internal timer system, as used by KL TIME PLEASE (&BD0D), uses DEHL to return a four byte time count of the time elapsed, with D holding the most significant byte to L holding the least.

They don't have anything to do with the CPC really, the interface definition looks like:

    property ID: Word read FID;
    property Start: Int64 read FStart;
    property Last: Int64 read FLast;
    property Min: Int64 read FMin;
    property Max: Int64 read FMax;
    property Total: Int64 read FTotal;
    property Count: Integer read FCount;
    property Average: Double read GetAverage;

[AMSDOS]

They don't have anything to do with the CPC really, the interface definition looks like:

    property ID: Word read FID;
    property Start: Int64 read FStart;
    property Last: Int64 read FLast;
    property Min: Int64 read FMin;
    property Max: Int64 read FMax;
    property Total: Int64 read FTotal;
    property Count: Integer read FCount;
    property Average: Double read GetAverage;

We're clearly talking about two different things. 


Been a long time, so I'm guessing that's Delphi.

[Executioner]

Been a long time, so I'm guessing that's Delphi.

Yup..

[Docent]

Internally, parts of the timer are 64 bit, but the expression evaluation in the breakpoint conditions is only 16 bit, so the maximum returned value is 65535. I could perhaps add timer_high and timer_low functions to get 16 bits at a time where it's necessary to time over 65535 cycles.

Is the 16-bit limit forced by the emulation? perhaps you could change the expression evaluation to support 32 bits - that would be enough for the timer.

It would be possible, but all breakpoints are compiled into byte-code for faster evaluation, so the IDs would need to be given a unique 16 bit ID, then all breakpoints which use them recompiled.

how about a 16 bit hash from the text ? the actual text of id would be needed only for displaying on ui...

[Executioner]

Is the 16-bit limit forced by the emulation? perhaps you could change the expression evaluation to support 32 bits - that would be enough for the timer.

It is probably quite easy to support 32 bit expression evaluation, I'll look into it.

how about a 16 bit hash from the text ? the actual text of id would be needed only for displaying on ui...

That could be done but two strings could have the same hash code, so maybe a list of unique identifiers and IDs would be better. I don't think there's currently support for Strings in the expression evaluator. Perhaps an alternative would be to add a name on the Timers screen so you can just assign it a name.

[Docent]

It is probably quite easy to support 32 bit expression evaluation, I'll look into it.

Great - thanks in advance!

That could be done but two strings could have the same hash code, so maybe a list of unique identifiers and IDs would be better. I don't think there's currently support for Strings in the expression evaluator. Perhaps an alternative would be to add a name on the Timers screen so you can just assign it a name.

crc16 or Pearson hash would probably do and provide enough uniqueness, but if you prefer to avoid using these functions, you could just build an internal table of all timer numerical ids and texts and use the table index as a unique id in breakpoints. In this case, modifying the timer name do not require the recompilation of breakpoints, only the text in the table will be changed.

[Beaps]

Anyone know what this error is about, it happens when I try and run it full screen 

It happens with all WinApe releases.

[arnoldemu]

I'm guessing you chose a screen resolution and specific bitdepth and your graphics card doesn't support it?
Try choosing another.

[Executioner]

Try choosing another.

Unfortunately WinAPE currently only supports 800x600 in 8 or 16 bit, and a lot of more recent graphics cards have dropped support for this resolution. I'll look at adding at least INI support for other modes in the next major release.

[arnoldemu]

Unfortunately WinAPE currently only supports 800x600 in 8 or 16 bit, and a lot of more recent graphics cards have dropped support for this resolution. I'll look at adding at least INI support for other modes in the next major release.

I didn't realise you couldn't choose the mode

I think 800x600 is still available in 24-bit and 32-bit colour modes?

[Maskedman]

Thanks a lot for a great new release of this great emulator!

Just a quick question, I've got a Comptetition Pro USB joystick and since I'm left handed, I press the number 2 fire button with my right hand.

Is there any way to change this to fire button 1 in the emulator without having to use programs such as joy2key ?

Was feeling a bit nostalgic and played some Bomb Jack for 45 minutes and now my right wrist is killing me 

[SuSke]

Great achievement! Lot's of work. Also ParaDOS 1.2+ looks nice, has Vortex format. That's great.
One thing I realized: X-DDOS 2.10 still does crash the system. Would be nice if this bug can be cured.

Yep, same here; I could not get the X-DDOS 1.11 running as well.. (did not try all possible combinations yet  )
ParaDOS 1.2+ does not yet recognize the X-DDOS data format entirely (which was slightly different from the Vortex one as I remember)
For some 'protected' games I still have to disable ParaDOS 1.2+ to get them working again.
But hey, this version is already great to work with

[Executioner]

Yep, same here; I could not get the X-DDOS 1.11 running as well.. (did not try all possible combinations yet  )
ParaDOS 1.2+ does not yet recognize the X-DDOS data format entirely (which was slightly different from the Vortex one as I remember)
For some 'protected' games I still have to disable ParaDOS 1.2+ to get them working again.
But hey, this version is already great to work with

Why the tiny text? A few protected games overwrite the ParaDOS drive definition bytes (BAFE/F) and it can cause problems. It is quite rare though.

[SOS]

The CPC-Display in the Beta-1-Version looks a little bit squishy/blurred compared to the a18-Version (The settings are equal in both versions).
Why?

(It's a Win8.1 x64-PC with a superb Intel HD Graphics 4600)

[Executioner]

The CPC-Display in the Beta-1-Version looks a little bit squishy/blurred compared to the a18-Version (The settings are equal in both versions).
Why?

There was actually a bug in the previous version where it wrote incorrect frame sizes for DirectX Stretch mode. If you upgraded and continued to use the same INI file it may look squished, try resizing the window until you find the region where it is correct (it sort of snaps there).

[TFM]

Yep, same here; I could not get the X-DDOS 1.11 running as well.. (did not try all possible combinations yet  )
ParaDOS 1.2+ does not yet recognize the X-DDOS data format entirely (which was slightly different from the Vortex one as I remember)
For some 'protected' games I still have to disable ParaDOS 1.2+ to get them working again.
But hey, this version is already great to work with

Would be too great to see X-DDOS running (works with all other emus btw). As I remember VDOS only has Vortex format, whereas X-DDOS support Vortex format and in addition the 0.7 MB Data and System format. (Dunno if ParaDOS supports the last too, Vortex works now, but needs to be selected). 

[dragon]

This game (This crack exactly). Not work in winape and arnold. But it work in other emulators, the crack jump to memory localitation, but its found nothing the memory empty. And cpc resets.

http://cpcrulez.fr/GamesDL_arc/index.php?download=aLPhzd3TgJjk3si_gLHNu7eIm7vl&p=aLPi0diuhsXpm8y7jbPt5cTAkMTY0ZPClb8=