Ren'Py 6.16 and RAPT 6.16 Pre-Released

[AxemRed]

How large is your APK? Pushing data to a device over adb can be really slow.

[Clayton Barnett]

The build for "The Question" is a mere 8.4M...so that should not be a problem.

I've since also tried with my daughter's Galaxy tablet, with no joy there, either. Perhaps the computer or software not recognize that something is plugged in?

In other news, I've found that unless I delete whatever is in the "assets" subdirectory first, the compiler crashes.

UPDATE: While the Ren'Py Launcher still won't find/send the file, I did manage to manually send the .apk file from the "bin" subdirectory to the tablet. I was able to install and run "The Question." W00ts all around. I shall now try to make test builds of my VNs.

[PyTom]

Did you install the android drivers for your device?

[Clayton Barnett]

Did you install the android drivers for your device?

The drivers (three of them) autoloaded when I plugged in the Galaxy Tablet. I'm giving thought to a general restart of it and my computer then hooking things back up for another try. I'll let you know in a few hours (have to put on "dad" hat now).

In better news, I've made Android builds for all five of our VNs. I'm playing through the main one, OTChi Kocchi, to make sure there are no issues...so far, no problem, but there's some things on Aoi's and Aris' paths I want to make sure work.

PS I've been at swim meets with my girls and have seen 90% of us in the stands on their phones or tablets. Your engine and our stories are going to change the world. Thank you, PyTom!

[PyTom]

The adb drivers may not be the same drivers as the ones that auto-load.

In other news, I've pre-released 6.16.0.377. There haven't been a lot of changes that affect the English side, with the major change being the new Japanese translation of the launcher, a new Japanese template game, and a small bit of infrastructure to support them.

So I'm declaring this a public pre-release. Please download and test it, and encourage your friends, families, and users to do the same.

[jack_norton]

Sorry I found a bug
Recompiled the Vera Blanc game and it always freeze (app not responding) after every ending.
The ending is just a nvl mode, and at the end of each one there's a simple jump theend

Code: Select all

label EndTrain2:
    $ persistent.EndTrain2=True
    $ renpy.music.stop()
    $renpy.music.play("snd/death.ogg",loop=False)
    scene black
    show i032 m with dissolve
    ve "Damn, it's too late--"
    scene black
    show i039 m with dissolve    
    $renpy.sound.play("snd/explode.ogg")
    "BOOOM!"
    show i039 m:
        xalign 0.5 alpha 1.0
        linear 1.0 alpha 0.5 xalign 0.5
    nvl clear
    E "The explosion completely demolishes the coach and its passengers.  Despite the train's derailment, the number of victims is surprisingly low." with dissolve
    E "After a thorough investigation, the detonation squad discovers that the bombers had only planted a plastic bomb in your coach. Oddly enough, the coach in question was practically empty; its number of passengers amounted to little more than a hand's worth."
    E "Perhaps they wanted to minimize collateral damage."
    E "Vera will never know the bomber's true identity or their motive for wanting to kill her and Brandon."
    jump theend

The contents of theend label is:

Code: Select all

label theend:
    scene black
    show i067 m at scale(.6),top with dissolve
    $renpy.choice_for_skipping()
    $showbar=0
    if GoodEnd:
        show text "{color=#FFF}{size=64}THE END{/size}\n\nCongratulations, you solved the case! You can still try again to see a different ending!{/color}" at Position(xalign=0.5,yalign=0.8)
    else:
        show text "{color=#FFF}{size=64}THE END{/size}\n\nBad luck! Don't give up though, play again and try to solve the case!{/color}" at Position(xalign=0.5,yalign=0.8)
    with dissolve
    with Pause(5.0)
    scene black with dissolve
    if GoodEnd:
        jump credits
    return

but the game freezes before reaching it. In practice it seems that stops before the last jump theend logcat didn't offer any useful info (indeed is not a traceback, the app really stops responding and you must forcequit). Of course with the previous version of RAPT was working fine (I tried that too).
Let me know if you need something else (like the two APK).

[Clayton Barnett]

More bottom-feeder questions:

1. How are the -release.apk, -release-unaligned.apk, -release-unsigned.apk different from one another? I send the -release.apk to my tablet and it worked fine.
2. Our flagship VN builds into a 427M winzip file. The Android apk is 594M. Is this growth typical?
3. Is it possible to have custom icons for Android builds? I know how to do it for W/M/L builds, but I keep getting the RenPy default icon.

I've a few more (about the "com.domain.name" and what that means) but I'll hold off a bit. Thanks!

[PyTom]

Jack - I need you to put together a replication. Basically, a game that jumps to just before the end, then fails. (This way, I don't have to play through the whole game to find the bug). Then send me the source so I can build it.

Clayton:

1) The main file is the -release.apk. The others are intermediate files produced by the android build process.

2) I don't think it should be that big, but it may be - it depends on if you have a lot of compressable files in there. Are there any files you exclude? Right now, the android build doesn't blacklist any files. If you send me the apk, I can look in it to see what's big.

3) Yes, use android-icon.png and ouya-icon.png.

[jack_norton]

OK after some careful testing I found out the culprit:
$renpy.choice_for_skipping()
if I comment that line, works! You still need the test source?

[PyTom]

If you want me to look at a bug, sure. I don't see how the choice_for_skipping line would cause a problem.

[xavimat]

Some minor things:
1- I've tested the None translation issue we spoke about in the IRC. It seems it works fine now.
2- I've seen that the Japanese template options.rpy has more configuration data that the English one (lines 491-666) and includes a sort of "error" in the build.config section: The "build.directory_name" and the "build.executable_name" have the name of the template "template-jp" and not the name I gave to the game (the same problem in "config.save_directory", that changes when the English template is used).
(By the way, I was wondering why the basic build configuration is not included in the template. I use "build.executable_name = config.name" and "build.version = config.version" and then only need to manually update this information in one place).

[nyaatrap]

(By the way, I was wondering why the basic build configuration is not included in the template. I use "build.executable_name = config.name" and "build.version = config.version" and then only need to manually update this information in one place).

This is also what I'm wondered. and it's nice idea about "build.executable_name = config.name", it helps me.
Btw, this JP template is copied form my game, so those more configuration lines are what I needed on my game personally (I think some of them rarely used on normal VN should be deleted, however, all of those extras are commented and it won't hurt anything before you uncomment it).

[nyaatrap]

After I updated 6.16, reloading causes freeze sometime.
It shows reloading game... but nothing proceed.
I have no idea what causes this yet.

[nyaatrap]

There's a serious problem. If a game made with 6.16 is placed non-ascii directory, it won't run.
For example, c:\Users\田中\desktop\newgame\newgame.exe doesn't start.

[Sslaxx]

Any chance that the location(s) where saves are stored can be user defined? I.e., being able to choose in the game directory (only), in <user directory>/.renpy (or platform equalivents) only, network share only or a combination of them (with default being all of them)?