Forum TI-Planet.org

Hello

nothing finished tonight but a good piece work in progress to report for this week Vlog.

Multiple windows are now correctly managed, as well as the methods to pick one window from the background and move it to the foreground (if not locked by a master window).
Also started to add Menus, but there are still at their very beginning. MenuBarWidget is coded as well as MenuItemWidget, one more brick is needed (MenuPaneWidget corresponding to the part of the menu that popup when a menu item is selected), so all of this should happen during this week.

As usual some pictures of the current status of the work (no need to say these are screenshots from the current revision still under degugging, and some visuals (by instance the IDs) will not appear in the next revision, this is just for coding purposes) :



For Bisam : please look at the second picture hereafter, you will see your algorithm at work for comparing the WidgetID of the window (in red in the left top corner of the windows) with one calculated from the DepthBuffer (in Green close to the mouse pointer). As the window move from back to top, it means it works fine





There is still a lot to do, especially tracking memory leaks cause I experienced some warning messages for low memory while using the Toolkit heavily (so some housekeeping will be mandatory).

Ciao

Sly

Very nice ! Glad to have been a (tiny) part of this.

Bisam a écrit:Very nice ! Glad to have been a (tiny) part of this.

Everything is made of small atoms ... nothing is small enough not to be dully considered.

Without you help, i would still be turning around and looking at how to make the window manager working correctly

Ciao and thanks

Sly

In which language are you programming this ?

Hi Bobb,

All is coded in C++, this is fully object oriented concepts with a lot of pointers for dynamic allocation.

BR

Sly

I think there is almost the same tool in C by epsilon5 for CEs, but I think it is not public:

https://www.cemetech.net/downloads/files/2095/x2227

Well that's not the same, though, the CE one is much more hardcoded for its own purpose, as far as I know, and has less features. But then again, it didn't have the same goal

Hello,

I was not aware of the project on CE, but from what I understand, it is rather a shell with lot of parameters that can be changed than a toolkit for developing applications in a easy way. So my understanding is the same than one of Adriweb.
I do not know exactly if it is based on an internal toolkit (that's a possibility, but I am really not sure), but for sure it does not serve the same purpose.

My goal with the GUI Toolkit is to offer a simple framework to develop graphical applications/utilities for which the developer can spend his/her time on developing the app rather than loosing his/her time on making the visuals (cause it can be very time consuming). In any case, it is not designed as a shell replacement, this is absolutely not my goal. (but if somebody would like to use the GUI Toolkit to make a shell, I am very happy with that).

Cheers

Sly

Yes, I understood the purpose of your project, and I think it is a nice thing to do a project to help others. What I was saying is that the author of the VYSION shell designed his shell, as well as another CONTAGION game with a tool (I think it is a library using GRAPHX) that he programmed for himself, without publishing it.
I wasn't talking about the shell, I was talking about the library that Epsilon5 used to design his shell and his other game. I just think it is a pity that he did not publish his tool.

Yes, agree that looking at his other productions, it seems to be all based on the same kind of widgets/interface elements, so we can reasonably think that he is using a library/toolkit not to re-code everything from scratch.
I also agree it is a pity that he is not sharing his work, it would help a lot of developers.
But we cannot force him to share. Anyhow, if you are interested to use the GUI Toolkit, please do. I just ask users to report bugs when they find some
Ciao

Sly