Re: [Gzz] SCUMM, PUI and FenPDF

[Matti Katila]

On Wed, 23 Apr 2003, Tuomas Lukka wrote:
>> Today we were talking with Asko about these issues and he talked a lot 
>> about mouse cursor shape changing while actions and after our discussion 
>> I really think we do need it.
> I agree, and have agreed for a long time.  The question is: when can 
> we schedule this for whom.

Yes, I definitely agree. We have already discussed about this. Idea is 5 
min. coding etc. 5 months * Math.PI.
 
> The problem is that there is a small number of people doing lots
> of jobs.

Yes and I think there's a problem because nothing is ever finished :)
I think you are very good at startting things, like pp was at Autumn.
So let we students to keep the work going and give some responsibilities 
of things. If we are too lazy write some deadlines to our journals ;)
"Nothing is done before deadline" - "Deadlines are the thing to keep us 
going", so you can use this 'advantage' freely for our own good =)
 
> It would be really nice if we could start discussing priorities so that 
> it wouldn't be only my job to keep in mind all these "wouldbenice"ties

Yes, for example give more responsibility and the things which are under 
his shoulders are always low priorized and anything with higher 
priorities goes before that in journals. There would always be  
some work for the areas of responsibilities if no other priorities.

> and try to think what's so important that it should be done
> now by someone.

You are the leader :)
 
>> Some of these are currently implemented in prue java pp so if you don't 
>> know what I'm talking about you should try 'ant clean compile pp' in 
>> fenfire module. 
> Only the pure java version?? Why not also GL version?

Both of course! (and now GL uses setOrtho thing and is quite fast ;)

There still a lot of people who cannot actually run gl programs and are 
not interested to try it if they don't have to. 
Benja saw pp first time when I started it to him a week ago. And I don't 
say it's benja's fault or something but currently he can live with 
storm/alph/navidoc/loom and our project without gl side so why should he 
try to keep gl in running state with all difficult nvidia drivers etc?
Although I think we should be interested in other people work =)
 
>> Next waypoint:
>>    - How we do area/thing selecting?
>> Issues:
>>    -We cannot paste, since we don't know are we using left or right 
>>     linking and this is *important*.
> Well, that can be done by Ctrl-V creating a special flashing object
> and mouse motions take it to right or left, or something?

This is one option. The special flashing object sounds quite funny :)
 
>> I propouse 'adventure games' like mouse handling because I don't want for
>> example push modifier keys at the same time (in macosX and one button 
> It cuts both ways: for any extra button or feature, 50%-80% of people
> will *never* touch it.

Yes, we really need to keep things simple and usable.
Well, we can start the application with example rdf where is simple 
guide written :)
 
>> So there's some free actions:
>>    -button3(right) click is free. This is good for changing 'mode', i.e.:
>>   ...3 modes...
>> Mouse pointer shape must be changed between these modes.
> I have to say that I dislike this idea. Modes are usually bad, especially
> at such a central thing.

Ok, but having in mind your comment about modifiers this is as bad as it.
Too difficult in any sense.
 
> A modifier key has the advantage that you always *know* whether you're
> pressing it or not.
>
> A changed mouse cursor shape just makes people feel weird, unless
> it's directly associated with something they do.

There's a people who construct and surf the web and people who only surf 
the web. It would be easy to keep these different because of this but I 
think our users can not be thought as advantage users/newbie users and 
so be separated to people who construct and surf on structure or people  
who only sur. So we need to think as constructing is something which even 
newbie can do. Pressing modifier isn't that easy.

>> so the algorithm would be:
>>     -select area
>>     -select another area (probably from another pdf/paper (I like to call 
>>      these as canvases))
>>     -after this, first area is seen with two buoys, left and right side of 
>>      the selected area and link can be made by clicking one of these 
>>      floating buoys.
>>     -or abandon by clicking in other place than the link making buoys.
>
> A different idea: the selection floats at a "clipboard".
> When you select a new area you then drag a line out of it, left or right,

So you need to know that you need to drag? Not a good thing even quite 
different idea. We can put left/right arrow for mouse pointer so that 
would make sence. Hmm.. afterall sounds interesting.

> and then leave the line and drag the thing from the clipboard to the line?

I don't understand this :)


   -Matti