Ren'Py 6.11.0 Public Pre-Release

[PyTom]

With the maximize button, it's too easy to get the window to grow until the start orb covers it - at which point performance drops from ~300 fps to ~25.

How odd. I vaguely wonder if it might make sense to have maximized windowed-mode Seven switch to software mode?

Software mode doesn't support scaling the image, like GL mode does. (It supports an older, slower way of scaling that I'd like to deprecate.) Also, we will eventually be getting some GL-only features. So I really want to avoid anything that depends on software mode.

The alternative choice I had is that hitting maximize would place the game in fullscreen mode. But I don't like that, as it means the interface for entering fullscreen mode would be different than the one for leaving.

[Mihara]

Another problem. Actually it's been around forever and I'm reporting this one because a reader complained about it.

When ATL transforms are used to move sprites onto the screen like so:

Code: Select all

    transform moveinleft:
        subpixel True
        anchor (0.5,1.0) pos (-1.0,1.0)
        easein 1 xpos 0.25

and in other similar cases, they work as expected. But in skip mode, most of the time, transforms never finish moving, and sprites end up fully or partially offscreen. Upon investigation, it appears that a "show sprite <different expression>" stops the ATL transform dead. Normally it is not a problem, (though I can imagine many expected uses where it can be) but in skip mode this often happens before the sprite moved into position.

Replication follows:

Code: Select all

init python:

    e = Character('Eileen', color="#c8ffc8")

init:    
    
    image bg a = Solid("#f00")
    image eileen happy = Image("eileen_happy.png")
    image eileen concerned = Image("eileen_concerned.png")
    
    transform moveinleft:
        subpixel True
        anchor (0.5,1.0) pos (-0.25,1.0)
        easein 10 xpos 1.25

label start:
    
    scene bg a
    
    e "Starting the loop to imitate a long NVL mode text."
    
    show eileen happy at moveinleft
    
label loop:
    e "Press the skip button here and the motion will abruptly stop."
    show eileen concerned with dissolve
    show eileen happy with dissolve
    jump loop

[Mihara]

...and, to be more precise about the bug with the screens with "modal True"

Modal screens not only prevent developer shortcuts and builtin shortcuts, but also all keyboard navigation, period. It comes back if you remove the "modal True".

[Jake]

The Mac clipping bug is also fixed - both sides of the screen, partially-off and fully-off things and everything.

(Sorry it took me a while to get back, I was busy for the weekend.)

[PyTom]

Mihara >>> That's not so much a bug as a design decision. Repeating movements don't have a natural endpoint, and there are some times when you'll want to know where the character was when the user clicked, rather than their current location.

So I chose not to have movements complete just because the image is changed, but rather, the current position is preserved.

If you need to force a position, you should explictly do so.

[PyTom]

The Mac clipping bug is also fixed - both sides of the screen, partially-off and fully-off things and everything.

And with that, the 6.11.0i pre-release is out. It fixes modal and documents the new behavior, stores the result of the OpenGL shift-abort test on startup, and fixes a problem with adding padding on the top and bottom of the screen.

This is pretty close to the final version. Unless someone finds a major bug in the next day or so, I'm going to go ahead and release. (It's enough of an improvement over 6.10 that I think it's time we saw the benefits.)

[Mihara]

If you need to force a position, you should explictly do so.

Uh... I don't need to force a position, I need a sprite to slide into the position and stay there. Note that the transform given in the example is not repeating, the code loops to give you a chance to see what skip mode does to it, but the transform only runs once.

Skip mode goes by too fast for anyone to actually click and detect a specific sprite position anyway, so why shouldn't non-looping ATL run to completion before the image is replaced in skip mode?...

[Mihara]

Sorry, PyTom, here's a new one which I'm pretty sure is a screen bug.

Right-clicking on a redefined main menu screen attempts to bring up a game menu (at least I think that's what it's trying to do) and fails, but not before running config.game_main_transition. Needless to say this doesn't happen when the main menu screen is not redefined.

In fact, the _game_menu_screen is accessible from pretty much anywhere and I'm at a loss how to shut it off when it definitely shouldn't be accessible, like inside other parts of the game menu.

Replication follows.

Code: Select all

screen main_menu:
    
    vbox:
        textbutton _("Start Game") action Start()
        

init python:
    config.game_main_transition = Fade(0.5,0,0.5)

    e = Character('Eileen', color="#c8ffc8")

init:    
    image bg a = Solid("#f00")

label start:
    
    scene bg a
    
    e "Meh."

Also, the screen examples in the documentation now seem to either confuse "modal" and "tag", or the syntax of those changed without actually reflecting that in the manual itself where it describes "modal" and "tag". My project using the original syntax still works.

[PyTom]

Right-clicking on a redefined main menu screen attempts to bring up a game menu (at least I think that's what it's trying to do) and fails, but not before running config.game_main_transition. Needless to say this doesn't happen when the main menu screen is not redefined.

Fixed it. Here's a new copy of 00library.rpy that you can drop into the common directory to see the fix, in lieu of a new testing release.

In fact, the _game_menu_screen is accessible from pretty much anywhere and I'm at a loss how to shut it off when it definitely shouldn't be accessible, like inside other parts of the game menu.

I'm not sure what you mean by this. Can you check it with the new code, and see if the behavior seems appropriate, and then describe what you're doing in more detail? Also, are you showing the screens yourself, and if so, are you using Show or ShowMenu?

Also, the screen examples in the documentation now seem to either confuse "modal" and "tag", or the syntax of those changed without actually reflecting that in the manual itself where it describes "modal" and "tag". My project using the original syntax still works.

Fixed. Thanks!

[Mihara]

I'm not sure what you mean by this. Can you check it with the new code, and see if the behavior seems appropriate, and then describe what you're doing in more detail? Also, are you showing the screens yourself, and if so, are you using Show or ShowMenu?

The new code fixed the main menu.

With the rest of it, however, I might simply be doing something wrong, it's hard to extricate this into a minimal example.

I have a chapter select screen which comes up if the reader has previously read to the end:

Code: Select all

screen chapter_select:
    tag menu
    modal True
...
        hbox:
            vbox:
                textbutton "1. Silence in the Library" action Return("entry01")

which is called like this

Code: Select all

label start:

    if not persistent.main_complete:
        call s00preface
    else:
        call screen chapter_select
        $ renpy.block_rollback()
        jump expression _return

label entry01:
    call s01welcome
    call openingmovie
    
label entry02:
    call s02morning
    call eyecatch_clockless
    ....

The logic calls individual scene blocks in this manner throughout, they return, it continues by calling the next scene, chapter select just jumps in at appropriate points.

Now, when I rightclick on the chapter select screen, game menu comes up. The idea is that it should be disabled, i.e. it's completely useless to save, turn on auto mode or skipping at this moment. Well, the game menu comes up regardless of whether the chapter select screen has the "tag menu" or not...

[Mihara]

And, another thing, not so much a bug as it is an omission.

To make every part of a screen-based menu use a redefined navigation menu screen, you need to redefine navigation, preferences, load and save with screens. I needed to redraw the preferences screen anyway since the default one doesn't fit properly in a widescreen project, even though there's very little I did to the example in the manual beyond shuffling the elements around to fit.

Now, the manual does not give me an example of how to redefine the joystick preferences screen, and the various actions in the Preference() group don't allow me to set the behaviour of individual joystick buttons. I'd leave it as is, but my beef with it is that it doesn't use my navigation menu...

[rinrin]

OpenGL mode doesn't work on my laptop (I tried resizing the window - can't). I find this rather strange, since this laptop otherwise supports OpenGL (I just updated the graphic drivers to the latest version and can even see an OpenGL settings screen).

Any ideas what might be causing this / what to do about it?

Here's the comp's basic data just in case any of it is relevant:

Windows XP Home Edition
Version 2002
Service Pack 3

Intel(R) Atom(TM)
CPU N270 @ 1.60 GHz
798 MHz, 0.98GB of RAM

Display: Mobile Intel(R) 945 Express Chipset Family
max. resolution: 1024 by 600

Ren'Py 6.11.0i

[Mihara]

...and another minor bug: Trying to Shift+I a style of an element inside a viewport produces no results as the style inspector stops on the viewport and doesn't inspect further down.

[PyTom]

Any ideas what might be causing this / what to do about it?

Can you post the opengl.txt file created by Ren'Py? (It's in the same directory as renpy.exe.)

[rinrin]

All it says is this:

GL Disabled.