Messages

Sounds good. I think if 10MB of disk space is that critical to someone, they'll be having bigger problems than building AGS games.
Edit: The engine components could be hashed if a rebuild needs to know if the engine is up-to-date.

I think it's a problem with the build process using hardlinks, a regular user can't hardlink into the AGS installation folder:

Code: dos

C:\Users\mwill>fsutil hardlink create in_my_profile "C:\Program Files (x86)\Adventure Game Studio 3.4.0\Linux\ags64"
Error:  Access is denied.

...and you can't hardlink across filesystems, so you can't have your installation on C: and your game somewhere else, like D:.

If you are working on a folder that's located in the root of a disk, you should probably disable inheritance and give yourself "Full Control" of that folder, or just use a folder that is inside your user profile instead. The built-in / free Microsoft antivirus seems to be getting a bit more heavy handed with it's on-access scanning; I'm noticing a lot of slowdown on the computers I manage at work, which disappears when I add an exception for programs that are active. Plus I've personally seen game compilation fail if the game folder is open in Explorer (must be some kind of indexing or thumbnail generation interfering with it, or could also be related to on-access virus scanning).

Anyway to sum up, after making considerations for the above, I can build a Linux game without manually copying files around.

Unless I'm making a classic noob error, The Secret Of Monkey Island?

I've played through the updated version, thank you for the quick fix. I'm quite intrigued by what could be in the other chapters.

I was playing on Linux and the game got stuck, I think trying to pitch up audio that was no longer playing.

Spoiler

I'd killed Hal, looked in the mirror, went to the boarded up door (where the toys were now on the floor). The game locked up when I turned around and tried to go back down the stairs. I don't think any music was playing since I killed Hal. The cursor was stuck in the busy/wait state on the stairs, so I couldn't do anything but force the engine to close

[close]

I liked the bit I played though.

Good to see the Baker's Kilobyte on the sign.

I don't think using Mate instead of anything else would affect the screen being spread over two monitors, and I think ScummVM has the the same issue. Going fullscreen should go to the primary display. For the sound, I've only seen the audio break-up or stutter because of the display scaling increasing the CPU usage of Xorg; you might want to suggest running the game with no scaling and see if that fixes the audio.

[Daily Chthonicle: Editor's Edition]

I'd like to give Daily Chthonicle: Editor's Edition a go, if it's still available.
Thanks for being so generous.

I'm pretty sure it uses the character's speed to predetermine the distance it needs to move per frame, for the x and y directions. I think the offset you are seeing is probably because the speed is enforced so animation frames always look correct (rather than it snap the character to the destination at the end of the walk, at a speed that doesn't match the animation frames).

If you change the character speed to 1 pixel per frame, does it still miss the destination?

However, about System.AudioChannelCount: isn't it supposed to be the total number of channels, at all times (constant) ?

Yes, as far as I know it's always constant but since the manual isn't entirely specific about it, and the way the crossfade works isn't documented at all, I thought I'd check that it remained at 8 during whilst the two music tracks were playing. There is always the possibility that it is using an idle channel, but it's hidden from any of the audio functions, I guess someone would have to check the code to clarify (it's probably quite hard to playback 8 sounds at once and also hear that a crossfade is still working).

I can't verify that it's using only one channel, but it works even if you have set Music to use only one channel, so I would assume so.

I've just done a quick test, crossfading two (music) audio clips with only a single music channel available.

Code: ags

//return the first channel index for a given audioclip
int get_audioclip_channel(AudioClip* audioClip)
{
  int index = 0;
  while (index != System.AudioChannelCount)
  {
    AudioChannel* audioChannel = System.AudioChannels[index];
    if (audioChannel.IsPlaying && audioChannel.PlayingClip == audioClip)
      return index;
    index ++;
  }
  
  return -1;
}

This returns -1 for the audio clip that is fading out, and the actual channel number for the one that is fading in. So if it using another channel it's not accessible. System.AudioChannelCount still returns 8 during the crossfade, so I would guess that it isn't wasting a channel to do the crossfade.

please someone confirm there is no stuttering in the ambient sounds channel.

I had the stuttering when I used an mp3 file. With ogg everything always plays seamless in my experience.

Yes, I think MP3 files are unlikely to loop seamlessly because of the way the data is padded. I've also had no issues when looping OGG audio, so I would stick with that.

This is a big issue for me. If I play 2 ambient sounds in the room, multiplied by two to cross-fade them, plus one channel (or two!) for the footsteps for each character, that's already 6+ channels used. They're running out quickly...

For the ambient sounds you should be able to reduce the number of channels you need, as long as you don't need to dynamically implement the crossfade.

• load the sample into an audio editor
• split it into two pieces where you would want to crossfade (just not too near the edge)
• swap the position of the two pieces (left part moves to the right, right part moves to the left)
• crossfade the two pieces (where they meet in the middle) and render that out as a new sample

You could do the same with the other ambient sample and optionally render them both together, so you would only be using 1 or 2 channels for ambient sounds. I'm pretty sure Audacity can do all of the above. MP3 format isn't good for looping sounds, but OGG should loop seamlessly between the start and end (where you made the initial split).

I'm pretty sure this was implemented before and it was removed, as it seemed to break the input queue rather than improve it.
https://github.com/adventuregamestudio/ags/commit/661637b070612bd8675d80b23d8f8b2bc6f4a864

[at]

I think it would be "at screen resolution".

It's nothing to do with antivirus, I rewrote this page on the wiki because someone else also thought this is an antivirus issue (I didn't know if I could / should rename the page).
http://www.adventuregamestudio.co.uk/wiki/Prevent_AGS_from_triggering_antiviruses
(i.e. ignore the page name)

I've not looked very closely at the BASS-style script, but I'm pretty sure that (by default) inventory clicks are processed for you by the engine, and so it makes assumptions about what a left and right click does. If you want to process these clicks yourself (where you can decide what a left and right click does) you need to set an engine option. These options are in the manual, but are quite hard to find.

OPT_HANDLEINVCLICKS is described as "Handle inventory clicks in script (0 or 1)", so you would set it like this:

Code: ags

SetGameOption(OPT_HANDLEINVCLICKS, 1);

Oh, right. Fixed that.
Also another thing was that a variable was assigned outside of the function, which is not supported by AGS as far as I remember.

I think setting a default value is allowed (and in some cases is important to know what it going on - declaring variables in AGS will also magically initialise them to a value rather than null), but you can't call a function and assign the result.

Code: ags

bool IsBoomBoxOn = true;

or

Code: ags

bool IsBoomBoxOn = false;

...should be okay.

I think that the true and false values may need to be switched around, as at the moment IsBoomBoxOn will never be set to false.

I'll create a ticket on the new tracker (unless there is somewhere else?) with an initial plan. Even if it's changed entirely (I'm not overly familiar with the engine internals and data structures) hopefully it can be starting point.