Re: [MiNT] XaAES

[Jo Even Skarstein <joska@online.no>]

Helmut Karlowski skreiv:

> > I can mail you a test-program if you like.
>
> Go ahead.

Will do it tomorrow, my Milan is busy compiling ATM...

> > try it somehow. Do you have a 060-binary I can test on my Milan?
>
> Yes. BTW: If you can mail you have internet? Cant you use 
> Nullmodem+ZMODEM to transfer?

Of course I have internet, but only on my PC's. I haven't set up a 
network yet, just a WAN for my PC and my wife's laptop.

> There is a lot of code for the resolution-change, but its deactivated. I
> removed it from app_man.c(?) and made a seperate file of it. Isnt this
> present already in teradesk? Maybe someone could make a km for this, so
> it will not be present in XaAES and teradesk simultaniously.

N.AES simply started the program pointed to by RSMASTER. XaAES and 
Teradesk could do the same, both saving some bytes. Then it would be 
easy to support all sorts of hardware, "just" write a new 
resolution-change-app.

> > - Make the G_SLIST-object available to applications :-)
>
> YES! I was planning this as the next major enhancement, somehow forgot
> to mention it, maybe it is so high on the todo-list that I could not see
> it :)
>
> I guess magic has something similar, maybe one could make the XaAES
> compatible to this (at least).

MagiC only have the WDialog listbox, which is "only" a listbox. It 
doesn't handle trees like G_SLIST. And XaAES already supports this. 
WDialog is a pain to use anyway, I'd rather be able to just use the 
G_SLIST-object in any RSC, supported by some simple AES-calls.

> > - A better Control-Alt-Tab handler. One that puts the previously top 
> > application in front of the list, and skips accessories. Or provide a
>
> I dont understand what you mean. If there are 4 apps (* ist top): *1 2 3 4.
> Now C-A-TAB switches to 2, the next to 1 again? One could probably control
> the direction by using shift, so C-A-S-TAB goes back ...

Yes, the idea is that the previously topped application always is the 
first in the list when switching using alt-ctrl-tab. This is how it 
works in most (all?) other operating systems. It's very convenient when 
you're constantly switching between two applications.

Using shift-alt-control-tab is not very userfriendly though. I do have 
more than three fingers on my left hand, but pressing four keys 
simultaneously is... well... not easy.

I'd rather use just Alt-Tab instead. Would that be a problem? Are there 
any apps at all that use Alt-Tab for anything?

> > clean hook into Control-Alt-Tab so I can enable the handler in Taskbar 
> > again ;-)
>
> We must watch out not to mix the features of XaAES-taskmanager and 
> Taskbar. I admit I
> do not know taskbar very well yet.

I did have a Control-Alt-Tab handler in Taskbar years ago. It used the 
hotkey-feature of Clocky for this, and it worked really well until I got 
the Milan. Could never get Clocky to work properly there, so when I was 
no longer able to test properly I fazed out all code that depended on 
Clocky.

My handler was quite neat. It displayed a dialog with a list of the 
current applications, and gave visual feedback of what application the 
next Control-Alt-Tab would switch too. An additional shift reversed the 
order - this is how I know that four keys at the same time sucks ;-)

However, I agree that such a thing belongs to the AES and not an 
application.

> This taskmanager-wishlist is more or less science-fiction, I was just
> dreaming ("Would be nice...").
>
> Its on the very bottom of my todo.
>
> > (like Alt+Help will execute althelp.sys today). You can find all the 
> > necessary information in /kern anyway (well, not network traffic 
> > perhaps).
>
> The information is there but its some work to make a tree-list-window of 
> it.

Well, I probably shouldn't say this (in case of failure!), but this is 
one of the things I'm working on now. Some sort of GEMified top with a 
treeview and a couple of graphs showing CPU and RAM usage.

> My idea to proceed is:
>
> - Implement your fixes
>
> - I would then send you the 060-binary, so you can test
>
> - Fix the things you find

Yes please :-) I might also be able to help you out with some coding. 
Unless gcc gives me a stroke that is.

> - Find a victim to send the patches to

I would suggest submitting patches after each successful iteration. 
Would make it easier for the victim ;-)

Jo Even