Thanks, Fizzban. 😀 Seems that another group of "unplayable" games could change the status. I´m looking forward to test the games affected by this vesa problem after the implementation of your fix to CVS.

UPDATE: I´ve compiled DOSBox and tested it. I can confirm that with this fix SciTech Display Doctor is no more needed for Orion Burger, Riddle of Master Lu and Chronomaster. On the other hand, as you have already mentioned ( Mouse driver: Maximum virtual coordinates ), the movement problems persist, so these games are not yet playable.

Upadate2: With a new fix from Fizzban ( FPU: FPATAN incorrectly implemented ) everything works. 😁

hello there...
i'm just a bit curious... but how the heck do you do it Fizzban?

in such a short time you solved some wicked issues/bugs.... it's just that, since you don't belong to the Dosbox develop team, the at-ease with the code and the solving of the wicked bugs kind of impressed me...
a lot of assembly experience and many nights in front of a debugger ?

(note: i'm not trying to diminish the work of the Dosbox team, or make it less impressive in any way... i'm very greatfull for the work of the team, and i do find it quiet impressive... making an emulator/vm is no easy task... thx guys 😁 )

ps: if this post seemed lame... excuse me... 😒 😢 😖 but this question was killing me...

Could somebody from developers look at "int10_vesa.cpp" please?

Vesa has been fixed according to Fizzban´s instructions, but "dx" seems to be still there instead of "dl".

< else reg_ah=VESA_SetCPUWindow(reg_bl,reg_dx);
---
> else reg_ah=VESA_SetCPUWindow(reg_bl,reg_dl);

In new RJPA´s compiled DOSBox those games with vesa problem do not work correctly and this could be the cause. I´m not a programmer, so sorry if I´m wrong.

hmm it was fixed in the cvs few days ago, so i have no idea what kind of sources your looking at but they are outdated.

I´m looking at this:
http://cvs.sourceforge.net/viewcvs.py/dosbox/ … p?r1=1.6&r2=1.7

Notice the change from Bit16u address to Bit8u, so if you use reg_dx it will still only use the reg_dl value.

Strange. The day when Fizzban posted his vesa info I downloaded source from http://pcnwstage.phys.rug.nl/ manually changed the lines according to his instructions, compiled DOSBox and picture was perfect. Now I downloaded RJPA´s build from 13.1. (I presume he does not use outdated source) and the display is still truncated after the first 65536 pixels. I do not get it. So I will compile DOSBox from actual source and post here the result.

The problem is in fact in that RJPA´s CVS build dated 13.1. (strange, since vesa fixes are from 11.1.) and my speculations were incorrect. In DOSBox I´ve compiled myself those games work well. Thank you for your quick reply and explanation, Harekiet.

the source generation was broken for a couple of times I noticed in my mailbox. Maybe the source download was of a slighty older date.