Messages

In Room Editor>Areas select the hotspot and click on the "Interaction..." button.

Rt-click "Use inventory on hotspot" and choose "New action".

From the dropdown menu at the top of the dialog box that appears choose "Conditional - If inventory item was used", then use the "Change" button to set Value to be the item number. "OK".

Now there's a "Do Nothing" script under the Conditional.

Rt-click again, and choose what you want to do.
If nothing on the list is suitable, then (and only then)will you have to choose "Run script", which will show a blank page ready for scripting.

I just reread James Kay's post, and I think the textlist approach is sound.
Instead of trying to develop a system to find the best, let users browse through various hints by whatever instinct works for them.
The best part of it is that you don't have to come up with a general theory - just publish the linked info.

I've noticed that a few AGS games can have a startup program that runs AGS games kept in separate subdirectories.

How can this be acheived within AGS? Or is it in fact dependent on non-AGS programs?

Yes, the object number is used for all of those GUI functions.
Notice how the GUI functions in the manual always refer to the object, rather the the specific type of object relevant to the function.

So instead of thinking do function with button-#, you could think of it as do button-function with object-#.

I would've played RL&BAT, except that I have Win2k and no way to run pre-AGSv2.15 DOS games  
Any chance of a Windows release, CM?  

Thanks, Scorpiorus.
Now that I'm clear on the rt-lf associativity, I'll save hours of redoing calculations in search of a non-existent error.

I figured that I was probably getting the reference, but didn't expect there to be any way of type-casting it back to string.


James Kay. I think the point is that it is confined to the Win2k-type systems, including XP.

Most downloaded:
Utterly useless. At best it reveals the effect of the sales pitch on those who haven't played it yet. In any case, it only counts clicks - attempted downloads - not actual downloads.

Impulse ratings:
I don't use them at all. There are lots of reasons that they fail, not least of which is that they are no way to address my interests in a game. How could they? People squirt numbers based on they don't know what urges, thinking that 'numbers' make it more objective somehow.

Form-based evaluation:
It may be possible to work something out, but I think that it would require very clear thinking about what it's for and what it means. I've yet to see a system that was telling me anything that I could use.
I think that the key is to ensure a broad enough range of ideas for intelligence to be able to find its own answers, rather than attempting some kind of ideal closure which is doomed to the limits of some theory.

Comments:
I do use (and make) comments. The advantage is that one can get an idea of the authors' attitudes as well, which puts the comments into context, making them more meaningful, whether or not you share those attitudes.

Community and Awards:
This seems the best of all options. The games come under the weight of community review, critique, contribution etc. But of course, it's no good expecting visitors to find that stuff.
So there are awards, and various forms of easily understood recognition, and these come from the very people who are in the best position to judge.

Taste and subgenres:
A classification into kinds would be great ... but I wonder if it's possible to find one that's really meaningful to everyone.
Likewise, a way of letting user preferences and opinions contribute to a "if you liked this then you may also like that" system would go some way toward telling us something that would actually help us to decide.

[EDIT]
BUT FIRST
a) Obligatory "Download size" info.
b) AGS version (as in operating system: MSDOS, Linux, or Windows).

SetButtonPic(3, 2, 1, 391);
}
}
}//insert this
if (interface == QUITGAME) {

/*etc*/

if (button==4) InterfaceOff(7);
}
}//remove this
if (interface == INVENTORY) { // They clicked a button on the Inventory GUI


And what's this? ...

Code: ags

function interface_click(int interface, int button) {
  if (interface == ICONBAR) {
    if (button == 4) {  // show inventory
      show_inventory_window();
    }
    else if (button == 4) {   // use selected inventory
      if (character[ GetPlayerCharacter() ].activeinv >= 0)
        SetCursorMode(4);
    } 

else if (button == 4) {   // use selected inventory
will never happen because you've already caught it with if (button == 4) {  // show inventory.

Maybe you should check your button numbers.

Nice fluid and energetic motion. Main problems:

1) The arms are too long and too thick, and are following the shoulders instead of setting the pace. This, together with the large hands travelling back first instead of thumb first looks like a dance move from the late 50's.

2) She needs hips.

Everything else looks spot on.

Not worth individual posts, but I haven't found a worthwhile reason to mention them elsewhere.

(General, Win2k)
1) Many AGS games do not unload properly on exit - at least, not on my system. i.e. they remain as active tasks that don't do anything or respond to system messages.
It may be the legacy of earlier versions, or a common (but unnoticed) scripting error/conflict in the games, or just Windows up to its old tricks.
Someone needs to figure out why so that it can be addressed at whatever level the fault lies.

(AGS v2.56d)
2) Arithmetic has unconventional parsing:
a-b-c-d = a-(b-(c-d)))
a/b/c/d = a/(b/(c/d)))
It took me ages to debug an occurrence of the first case (if it had been the second, a DIV0! might have clued me in).

3) I'm still experimenting with types. It seems that some illegal (or misleading) code is accepted by the compiler even though it doesn't work.
So here is what I believe. Feel free to clear up any misconceptions:

I haven't done much with char or short, so I'll assume they have the same uses as int.

struct may only contain int and int array
array may only contain int

string can be passed (by reference), but not returned by functions.

neither struct nor array can be passed as parameters, returned by functions, or assigned directly. (Passing them by reference would be nice, though).

4) Can't block-comment #defines, and /*/ appears to be interpreted as /**/.

That's all for now.

I've put a new topic on drawing systems in the critics lounge.

When you haven't got the feel for something, it can be good to just follow exact instructions until it becomes second nature.

It's not necessary to follow some strict rule about what is 'correct', but it is important to be able to figure out why something isn't working and improve it.

[if]

It's a brilliant idea. The sort of thing that could spread like wildfire ... if it's not too difficult :
The background uses planometric drawing, but this idea could work with other systems too.
(P.S. "Isometric" means 'same scale' for each dimension, so it only applies to the everything-at-30-degrees technique).

I found this SWF presentation on technical drawing systems.
It covers the basics well enough to start using them (and also provides the correct names for each system).

The inclusion of chapters with new objectives is great, but synchonizing with the music may be a bit too extreme for me :o

[g(y-v)+abs(x-u)]

I made the diagonal loops suggestion in another thread.

The baselines I think present a bigger obstacle in terms of the routine demands on the engine.

The easiest approach that I can think of just happens to be compatible with my suggestion in the other thread.

For each room define a value, g=horizontal/vertical. (g = 2 or 3 being the most likely)
For each walkbehind set a point=(u,v) corresponding to the bottom of the "\/" shape where the angled 'baselines' meet.
Then, for the bottom-centre of a character/object, (x,y),
g(y-v)+abs(x-u) is positive when in front, negative when behind.

Contrast this with y-v, for flat baselines.

You could probably use this formula to script a variable baseline, but I imagine that it could get messy.

I'm pretty sure that setting the bubbles baseline just below the water-cooler baseline should be enough. It would certainly be the most direct solution.

I've been sticking objects on walls with no trouble - Walk behind or in front works as expected.
I wonder if it's animation related; e.g. the baseline gets reset at some point. I'll have to try it.

Just putting a ^ in the middle below the shoulder blades should be enough. Darken it on the side where the arm moves back for added depth.

You may even want to try variable shadows/highlights on the trunk to reflect arm movement and body twist, just to make it more fluid. Depends on how far you want to go with it.

I like the whole style as it is. Even the completely flat area is very distinctive.
The only things that I think need attention are
a) consistency. The rightmost leaf and edge of the island are just too smudgy (overly reworked errors would be my guess), and
b) Stronger hue difference in the shadows.

big brother - read the (html)manual and watch the (swf)intro.

Yeah, the letters were brilliant; I could really feel how they stretched over time.

There was one puzzle that didn't seem right to me though --

Spoiler

I didn't see the need to go through the doctor's files. I thought I had enough info to begin researching, and missed any hint that only the confidential records were of any use.
But more than that, since when is a stopwatch an alarm?

[close]