[hatari-devel] Fwd: Re: Hatari 2.0 Patch + Videl problem

[Nicolas Pomarède <npomarede@xxxxxxxxxxxx>]

Hi

Here's a mail sent to me by Mark Fechtner, including a patch for 
tmpfile() that requires privilege rights under recent Windows + an 
interesting videl effect when number of pixels is less that screen width.

CCing Mark too as I don't know if he's subscribed to our ML or not 
(which would be a shame :) )

Nicolas

-------- Message transféré --------
Sujet : Re: Hatari 2.0 Patch + Videl problem
Date : Fri, 11 Nov 2016 22:57:47 +0100
De : Nicolas Pomarède <npomarede@xxxxxxxxxxxx>
Pour : Mark Fechtner <mf@xxxxxxxxx>

Le 11/11/2016 à 20:42, Mark Fechtner a écrit :
> Hello Nicolas,
>
>  thanks to you and your team for the great 2.0 release. I can honestly
> say that I probably wouldn't have coded anything on ST nowadays if it
> weren't for Hatari and how it can be integrated in one's workflow
> (Natfeatures, Debugger, Logs etc.).
>
>  Speaking of which: one of the features I use _very_ often is starting
> PRG/TOS files directly off the hosts file system (autostarting). This
> works nice on OSX and I noticed the regression of this under windows (re
> needing admin rights for the temp file) a while back, hacked up a fix
> and forgot about it. Until now. Please see attached patch - not pretty,
> but works™.

Hi

thanks a lot for the patch, doesn't look not so pretty to me and always 
nice to get contribution from windows users :)

>  While tinkering around with Hataris F030 emulation, I fired up
> Terrorize Your Soul and noticed that the hardware trick I used in the
> Z-Buffer thingy before the rotator doesn't work (see attached
> PNG). Hatari stretches that column all over the screen, instead of
> repeating it. IIRC I set the videl to a very narrow display width (64
> pixels or so, memory's very hazy on that) - which resulted in the videl
> repeating that amount of pixels until the right border. That's why it
> looks like this on real hardware: https://youtu.be/SonTpLHEHKw?t=305
>  I'm not sure if anyone else used this back then, AFAIK one of the very
> rare hardware tricks on Falcon. Unfortunatly I don't think I have
> sources left of that particular part. Hopefully that behaviour isn't not
> that hard to emulate?
>
> Thanks again,
>  Mark (aka tIn/newline/absence./the independent yaddayadda)

I don't know that videl trick, I would need to search if it's documented 
elsewhere, but still good to have this demo as a reference.

If you don't mind, I will forward your mail to hatati mailing list ? 
Just to keep a trace for other hatari devs, some of them know falcon HW 
better.


Nicolas


-------- 2nd mail -----------------

Hi Nicolas,

> thanks a lot for the patch, doesn't look not so pretty to me and
> always nice to get contribution from windows users :)
Well, just a tiny one at that :) Let me know if I can test anything 
windows specific. Being an OSX and Linux user, too, I could even cross-test.

> I don't know that videl trick, I would need to search if it's
> documented elsewhere, but still good to have this demo as a reference.
Speaking of references: here's another one that Hataris videl doesn't 
emulate correctly - it uses raster splits with resolution changes: 
https://demozoo.org/productions/60283/
AFAIK the videl emulation isn't built to support splits?

And last but not least: https://demozoo.org/productions/60680/ uses a 
vblank/vsync fader IIRC for the truecolor reality.sys corrupted picture 
- but that's probably out of scope for an emulator do emulate that.

> If you don't mind, I will forward your mail to hatati mailing list ?
> Just to keep a trace for other hatari devs, some of them know falcon
> HW better.
Sure.

Thanks again!
 Mark.

Attachment: hatari_f030hwrepeatbug.png
Description: PNG image