Windows bug for small screens / fullscreen

Forum for discussion about the documentation project.
Tourist Trap
Posts: 2958
Joined: Jun 02, 2015 16:24

Windows bug for small screens / fullscreen

Post by Tourist Trap »

Hello,

the example at the Pmap (graphics functionality) would be clearer if it used screenres 640, 480 rather than screen 12. Otherwise things are a little more obscured than necessary I think.
By the way, I have a bug when I set a screenres 100,100 on win10. A vertical thick white band appears near right side, and cls doesn't remove it. I don't known if it's known?

Thank you

edited: title changed according to the most important among affairs treated around here.
Last edited by Tourist Trap on Feb 06, 2019 18:12, edited 1 time in total.
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

Tourist Trap wrote: By the way, I have a bug when I set a screenres 100,100 on win10. A vertical thick white band appears near right side, and cls doesn't remove it. I don't known if it's known?

Thank you
This weird behavior only appears with fbc 1.06 and 64-bit.

Version 1.06.0
[fixed]
- Github #113: When a WINDOW was active, the top/right edges (bottom/right for WINDOW SCREEN) were transformed out of the viewport.
???

[edit]
Similar behavior for a similar code with fbc 1.05 and 64-bit.
dodicat
Posts: 7976
Joined: Jan 10, 2006 20:30
Location: Scotland

Re: Documentation of PMAP function

Post by dodicat »

It looks OK with 32 bits, but it behaves badly also in a different way.
(1.05)

Code: Select all


screenres 100,100,32
color , rgb(200,0,0)
dim as integer mx,my
do
    getmouse mx,my
    screenlock
    cls
    line(0,0)-(99,99)
    
    locate 1,1
    print mx;my
    screenunlock
    sleep 1
    loop until len(inkey)

sleep
 
Tourist Trap
Posts: 2958
Joined: Jun 02, 2015 16:24

Re: Documentation of PMAP function

Post by Tourist Trap »

fxm wrote: This weird behavior only appears with fbc 1.06 and 64-bit.

Version 1.06.0
[fixed]
- Github #113: When a WINDOW was active, the top/right edges (bottom/right for WINDOW SCREEN) were transformed out of the viewport.
???
It's my version. I took it from the last WINFBE package. I'm not sure if it's related to WINDOW. Can't investigate too deeply those days.
dodicat wrote:It looks OK with 32 bits, but it behaves badly also in a different way.
Anyway, here is what I get with your test under 1.06:
Image
dodicat
Posts: 7976
Joined: Jan 10, 2006 20:30
Location: Scotland

Re: Documentation of PMAP function

Post by dodicat »

With 32 bits -gen gcc or gas the screen is drawn as is 129 by 100 if set for 100 by 100.
A white line doesn't show.
I tested on fb .24 as well as the latest official (1.05) build.
A good find TT.
I think all graphics bugs should be resolved in fb, gfx graphics are fb's main strength.

I have no idea how to view Github #113 - sorry.
counting_pine
Site Admin
Posts: 6323
Joined: Jul 05, 2005 17:32
Location: Manchester, Lancs

Re: Documentation of PMAP function

Post by counting_pine »

dodicat wrote:I have no idea how to view Github #113 - sorry.
It's at https://github.com/freebasic/fbc/issues/113.
Hmm, we have bbcode for bugs on SourceForge, but not for Github. Maybe that's worth adding..

Anyway, I don't think the above bug is related. The fixes should only affect the resulting image in the screen buffer, not how it's rendered in the window.
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

No problem (with 100x100) when there is no frame (flag: FB.GFX_NO_FRAME), or with a shorter title bar (flag: FB.GFX_NO_SWITCH).
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

Code: Select all

screenres 60, 60
do
    dim as integer mx ,my
    getmouse mx, my
    screenlock
    cls
    line(0, 0)-(59, 59)
    locate 1, 1
    print mx; my
    screenunlock
    sleep 1
    loop until len(inkey)
sleep
Observed behavior with fbc 32-bit:
- Regardless of the graphics window width requested, a minimum width is always imposed (when the GFX_NO_FRAME flag is not used) to allow the display of a minimum window bar.
- In such a case, on the x-axis, the view mapping for any drawing keyword does not match the physical mapping (only the GETMOUSE statement always follows the physical pixel): as if the image was internally stretched.

It seems that for fbc 64-bit, this behavior does not work properly (if we consider that the behavior with fbc 32-bit is the wanted).
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

Tourist Trap and the others:
I can, if you allow, fill in a bug report on this behavior, but from my point of view as described in my post above.
Otherwise, do it yourself.
counting_pine
Site Admin
Posts: 6323
Joined: Jul 05, 2005 17:32
Location: Manchester, Lancs

Re: Documentation of PMAP function

Post by counting_pine »

So far, it seems like this is a problem with Windows refusing to make the window as small as the graphics area.
Are you saying that on 32-bit it stretches the graphics to fill the window, while on 64-bit it is not stretched, but the remaining area is basically filled with nothing useful?
I've tested quickly on FB 1.05 64-bit on Windows 10. I see a similar effect to the screenshot, although the white bar fills the height of the window.
Tourist Trap, the separate PMAP issue seems to have been forgotten here. I suggest starting a new thread if you want to talk about it, and rename this one.
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

counting_pine wrote:Are you saying that on 32-bit it stretches the graphics to fill the window, while on 64-bit it is not stretched, but the remaining area is basically filled with nothing useful?
Yes (including the printed characters which are also stretched on fbc 32-bit).
I observed that behavior on Windows 10 (64-bit), testing both with fbc 32-bit and fbc 64-bit.
counting_pine wrote:Tourist Trap, the separate PMAP issue seems to have been forgotten here. I suggest starting a new thread if you want to talk about it, and rename this one.
The day before yesterday, I added a comment after 'SCREEN 12' to highlight the dimensions (640x480) of the corresponding window:
  • KeyPgPmap → fxm [Added a comment on the first line of the example]
counting_pine
Site Admin
Posts: 6323
Joined: Jul 05, 2005 17:32
Location: Manchester, Lancs

Re: Documentation of PMAP function

Post by counting_pine »

I see now, just a documentation shortcoming.
The simplest solution is probably just to use ScreenRes, as suggested.
fxm
Moderator
Posts: 12081
Joined: Apr 22, 2009 12:46
Location: Paris suburbs, FRANCE

Re: Documentation of PMAP function

Post by fxm »

KeyPgPmap → fxm ['ScreenRes 640, 480' rather than 'Screen 12' in example to explicit the window resolution]
Tourist Trap
Posts: 2958
Joined: Jun 02, 2015 16:24

Re: Documentation of PMAP function

Post by Tourist Trap »

fxm wrote:KeyPgPmap → fxm ['ScreenRes 640, 480' rather than 'Screen 12' in example to explicit the window resolution]
Thank you.

By the way, as we evocated issues about screen sizes, FBSHOOT , and my own test, tends to show that GET_DESKTOP_SIZE fails with 1920×1080 res on win10. It seems that Lachie has also encountered this behaviour. I have no time to go in deep for the instant, so take it as just a warning.
Tourist Trap
Posts: 2958
Joined: Jun 02, 2015 16:24

screen sizes...

Post by Tourist Trap »

counting_pine wrote: Tourist Trap, the separate PMAP issue seems to have been forgotten here. I suggest starting a new thread if you want to talk about it, and rename this one.
I tnink tbe pmap affair is solved, just left the screen sizes difficulties.
Post Reply