openmpt - empty qt ui fork?

doveraudio · January 31, 2011, 03:59:57

Just a thought i had the other day. Suppose someone simply began the process of mimicking the gui of modplug, and collecting lists of what the gui controls need to be targeting etc.

. i'm just now finally reading c++ how to program, and not quite up to the task alone, but i thought this would be an idea anyway. If people begin the layout, maybe some of the programmers can help with lists of what the parameter names are and such.

good idea? bad idea? too confusing?

doveraudio · January 31, 2011, 05:08:12

Another thought, perhaps a peek at some of the issues surrounding the use of qt in the project lmms "linux multimedia studio" might be useful for the future of a platform free modplug. From what ive seen of gtk in its various forms, i'd think it would provide modplug with a chance to maintain the original look, as well as provide some cool features like docking and floating the tabs, full screen tracker window, and native integration.

Saga Musix · January 31, 2011, 08:13:51

I've more or less been collecting (imaginary) lists of which GUI controls have to be coded in which way in Qt, and where possible problems could arise, and at the moment there are quite a few things which would introduce problems if not planned carefully.

doveraudio · January 31, 2011, 16:18:32

i'll be be more involved physically when i finish reading "c++ how to program" by h. deitel, but for now im at least aware of some of the concepts that provide the strength of c++. In this case, if the the actual literal interface was more separate from the implementation, this would be easier to accomplish.
My idea is that if there was an ongoing project for creating the gui in qtcreator or glade, then at least SOME of the work on this project would be getting done. In the actual OpenMPT project, perhaps a folder in docs or somewhere could be slowly developed in which the member functions are described.
But ;like i said in kinder words, i'm still a wannabe when it comes to real programming. i made a few projects in synthedit, but have a ways to go in reality.

probably the first thing to do if anyone wants to second the motion is create a project outline.
this page might help with the planning procedure.
http://en.wikipedia.org/wiki/ISO_12207

if htis is a thing that should happen perhaps an idea is to get the code documented and sorted well enough that openMPT can have a gtk and qt flavor.
"IT IS POSSIBLE. IT IS JUST CODE AFTER ALL."

Saga Musix · January 31, 2011, 17:28:16

Those are all great ideas in theory, but they don't work out with a small unpaid development team. Of course it would be nice if OpenMPT's code was documented properly, but that is exactly something I have been doing since I'm working on it, and there's not much progress simply because there are much more important things to do. For reference: according to ohloh.net, the source code : comment ratio in the OpenMPT code raised from 12% to 13% in less than one year. That is quite slow, considering how much I thought I was adding comments. And just adding comments for the sake of having code documentation is not a good idea either.

Rakib · January 31, 2011, 20:11:53

Quote from: doveraudio on January 31, 2011, 16:18:32
probably the first thing to do if anyone wants to second the motion is create a project outline.
this page might help with the planning procedure.
http://en.wikipedia.org/wiki/ISO_12207

if htis is a thing that should happen perhaps an idea is to get the code documented and sorted well enough that openMPT can have a gtk and qt flavor.
"IT IS POSSIBLE. IT IS JUST CODE AFTER ALL."

I think looking at test-driven development would gain this project even more. It's more fun to develop, less errors in code. Write test first and then code!

doveraudio · February 01, 2011, 08:41:59

i'm trying to get WxWidgets to compile right now. on my second run. i'll wipe the system if i have to... argh. Apparently it aids in the task of cross platform issues by being awesome in the true sense of the word.

I also need to openly seek mentoring in c/c++ if i want to accomplish my own goals.

what i understand about c++ so far is that everything that does work should be able to be manipulated by another, separate object. so i hope i'm not offending anyone here, i'm not being a critic. this is a complex project in general, never mind the gui issues. the main devs i'm sure have been long insane just from file formats.

I'm so tired right now, lol!!