Messages

It's possible that every line that Conventionary says is also said by another character.

Ah... now I understand. No, everything he says is unique. BUT his name is used once previously as a separate String. Suddenly it all makes sense. All the missing characters are major characters - they will each have been mentioned previously in the game. The remaining names are minor characters who never appear elsewhere. (It's odd how the character list comes after the modules, but whatever.)

This is VERY good news. It means there's no danger of the same String being translated in different ways, since unique Strings only appear once.

Hats off to pumaman, AGS continues to surprise and impress.

It appears that the translation source (in 2.72) misses out some character names. Any idea why? I attach a screenshot: character list at right, resulting translation source at left.


My translator pointed out, for example, that "Conventionary" - a major character in the game - doesn't appear in the translation source at all, in any place. Any ideas what might be happening?

This deletion from memory is only supposed to be done with standard sprites that are part of the game files and can be re-loaded from disk when needed -- but there was a bug where in certain situations it could delete DynamicSprites that you created in the script too.

This thread has come at just the right time. Thanks!

I just started selling my game, and for the first time ever found this bug. It's Murphy's law - some bugs won't show up until people pay money. I'm using 2.72, and obvious it's been fixed since then, but it's nice to know what it was. I think I know where it happens in my game, and can work around it. For the record, SSH's Simple Snow module makes heavy use of dynamic sprites, and the crash I get is usually soon after a snow scene, with the same error message listed earlier in this thread.

I finally got the game finished. Well, the first story at least. Hopefully the second story won't take as long! I won't be promoting it until the second story is ready (in July) because it's been a learning experience and I'll do a lot of things differently in the next installment: Dante's Inferno.

But anyway, it's done! Woohoo! Hopefully there aren't too many rough edges, but  I think people got sick of testing it, so it's time to kick it out to the world and move on.  For screenshots, demo, the full game, and other stuff, see


EnterTheStory.com

Many thanks to all those who left feedback about early test builds. Anyone who's helped out gets a free game. If you helped and don't get an email from me in the next couple of days, just let me know.

Oh, and happy 2009!

I'm waiting impatiently for the Divine Comedy...

So am I !! All I want for Christmas is to get Les Miserables out the door and never see it again (i.e. no more bugs to fix).

It has to do with multitasking mode.
http://www.adventuregamestudio.co.uk/manual/SetMultitaskingMode.htm

Thanks. Are there known issues with this? I have it on the default setting (0)- do you recommend changing it (to 1)? Or providing that option to users?

What's about "look at" codes for the objects? You've mentioned them in the blog.

That just refers to right clicking a person, then right clicking a different person or object. Most objects and some minor characters only have one response. But major characters are supposed to have a lot of different things to say. If you just get the same response each time, please let me know.

(The one exception if if the response is obviously a clue. E.g. first look at "I was thinking of X", second look at "maybe I should look at X", third look at "I must look at X right now!!"

A couple of users have complained of ALT-TAB problems. For example:

When I alt-tabbed out, I couldn't go back in, it said mouse = 0 click = 0 or something, and got me stuck.

It works fine for me. Any idea what's happening and how to avoid it?

useful files:
The game he tested
the latest version
the latest version without music (smaller download)

Thanks for the feedback! I really appreciate you taking the time.

It would be cooler to play as Jean Valjean but that would actually make us miss part of the story.

Yes, that was one of the big challenges. In the original book, not much happens for the first few chapters. But you do get to control Valjean by the end of "puzzle" number 3. (There are 120 "puzzles" in the game)

2) Parts of the game are ripped

Do you mean
(a) broken?
(b) attacked by critics?
(c) sudden?
(d) copied?
(e) intoxicated?
(f) muscular with almost no fat?
(g) other?

Effort: 10
Result: 8.4

I'm very happy with that. The first game is where I make all the mistakes. That's why I'm not promoting it heavily until the second story is added in June - hopefully that one will be effort 8, result 10

Well, make a completed games about it!!

It's not completed until it's tested.

The final, final test version is up.

If anyone has time to look at it I'd be grateful. Needless to say, I've done a lot of work on it since October. If there are no major bugs then I'll declare this first story finished at the end of this week, then begin work on to the second story.

Why bother with your manual backups (which are excellent to have, to don't get rid of them) first? Have you tried the .bak files?

~Trent

Yes, that's how I finally recovered most of it. For some reason I used two of the three.baks and forgot the third one.

You can in 2.72. I haven't tried that yet - be interesting to see if it changes just one instance of a number or both.

However, good news (I think) - after a lot of trying different backups, I think I've got a fixed version that won't need me to replace many sprites (hurrah!)

My game shut down while saving yesterday, resulting in corrupted sprite files. So I'm busy adding sprites from earlier backups. For speed, I save the entire sprite folder from the old backup, then 'import multiple sprites'. But strangely, this results in TWO identical sprites (with identical numbers) for each imported sprite. Is there any way to remove the second sprite? The game won't let me delete it because the number is of course used.

I don't have this problem when pasting over corrupt sprites one by one (using the clipboard), but this takes literally hours, since every single change means every other sprite in the folder is loaded again, with multiple error messages for each one (many of my sprite folders have over a 100 sprites in each).

Is there an easier way to recover sprites from a backup, without changing any other parts of the game?

A thousand thank yous. Nay, a million. I have Puppy Linux, not Ubuntu, so wasn't able to test this. I was starting to panic there, as my game relies very heavily on RunAGSGame. (Some future installments will be unplayable without it). Guess I'll sleep easy tonight.

According to the old thread on the Linux runtime,

You cannot tell ags which file to use as the game file. It looks for ac2game.dat in the current directory.

Does this mean that runAGSGame cannot work in Linux? Or is there a workaround -  can Linux detect one game's request for a game in another location?

I'm just finalising details for getting my game to run in Linux, and feeling pretty stupid that I never thought through the implications last year (I did ask about RunAGSGame in Linux back then, but nobody answered so I assumed it wasn't an issue.)

(I'm assuming you're using AGS 3.0 and higher)

That explains it. I saw the 'button ==' code in other threads, but it doesn't compile in 2.72 (which I use for Linux compatibility).

I have a compromise where RepExAlways constantly updates 'lastMouseButtonClicked' and the GUI button code looks at that. It seems to work, but I wondered if there was a solution that could be completely contained within the GUI button.

I want to make a GUI button behave differently, depending on if it's the left or right mouse that called it. I tried mouse.IsButtonDown(eMouseLeft) but that doesn't work (I'm guessing because the button is triggered by the mouse being released).

Do I need to use repExAlways and constantly check if the mouse is pressed over a button? Or is there a simpler way?

You could:

make a ChrisTChangeRoom function which sets a global variable "nocleanup" to 1. Use this in all scripts. Then in your on_event, only cleanup  if nocleanup is 0. Then always reset it to 0 in the on_event handler, too.

Ah, that would work! I don't fancy going through all 700 "normal" hotspots, but it would be much less work to go through the 50 or so "special" hotspots. Thanks.

Which leads to the next question... is there a way to access the code in each hotspot from a script? Name, walkToX, etc., are accessible, but I don't now how to see inside the interaction without opening every single hotspot one at a time and tunneling down the layers of boxes for each one (80 roomx x 10 hotspots on average... and I can't remember which ones have scripts

Can't you do the cleaning up in the on_event for eEventLeaveRoom?

No, because that can't tell how the person left the room. If a character left via "player go to a different room" then I need to reset all variables. But if she left via "run script" then I need to preserve the variables for the next room.