Lazarus
Programming => Widgetset => GTK => Topic started by: LazaruX on May 01, 2011, 10:54:23 pm
-
Hi folks,
as I can see from the repository of my Linux distribution (PClinuxOS) and from Wikipedia, GTK 3.0 as been released. I have no idea what is different in GTK 3.0 frmo GTK 2.x but still the question is: Will Lazarus support GTK 3 soon?
-
Hi folks,
as I can see from the repository of my Linux distribution (PClinuxOS) and from Wikipedia, GTK 3.0 as been released. I have no idea what is different in GTK 3.0 frmo GTK 2.x but still the question is: Will Lazarus support GTK 3 soon?
nope. AFAIK the idea is to write a clean widgetset for 3.0, not inheriting the load of gtk1 and 2
-
Good call there. Yes, Gtk3 should probably be re-written from scratch although alot of the gtk2 code can be reused (and cleaned) in the process.
-
Yes, that's exact idea. But first, someone must generate gtk3 units, and then implement gtk3lcl.
-
Yes, that's exact idea. But first, someone must generate gtk3 units, and then implement gtk3lcl.
And reimplement just about the whole WinAPI.... LCL is just too much Windows centric. But I must add, the "starting a new widgetset implementation from scratch" is a very good idea!
-
I apologize for my English. Don't speak.
I usually talk at the forum of the national communities FreePascal and Lazarus. Read the news in the "road map" (http://wiki.lazarus.freepascal.org/Roadmap) about the implementation of GTK3 version of 1.4. I decided to write here. This is a cry of the soul.
This is one of the two most anticipated innovations. I am desperate to wait at least the beginning of works on introduction of GTK3 in the LCL. I didn't have experience, that would be a member of this work. But I am ready to be an active tester GTK3 in the LCL.
I am very afraid that this innovation will not come to a realization. Or will be implemented too late. Necrophilia on GTK2 - is a dead end. Especially against the background of active introduction of Wayland. There are not talking about support for GTK < version 3.
GTK3 is indeed a very important priority. GTK3 used not one year. Without GTK3 Lazarus loses its appeal for most Linux developers.
I really ask You to bring this innovation to the finals. I would like to be well informed on the implementation of GTK3. Ready to help at least testing.
-
http://www.mail-archive.com/lazarus@lists.lazarus.freepascal.org/msg38545.html (http://www.mail-archive.com/lazarus@lists.lazarus.freepascal.org/msg38545.html)
-
I apologize for my English. Don't speak.
I usually talk at the forum of the national communities FreePascal and Lazarus. Read the news in the "road map" (http://wiki.lazarus.freepascal.org/Roadmap) about the implementation of GTK3 version of 1.4. I decided to write here. This is a cry of the soul.
This is one of the two most anticipated innovations. I am desperate to wait at least the beginning of works on introduction of GTK3 in the LCL. I didn't have experience, that would be a member of this work. But I am ready to be an active tester GTK3 in the LCL.
I am very afraid that this innovation will not come to a realization. Or will be implemented too late. Necrophilia on GTK2 - is a dead end. Especially against the background of active introduction of Wayland. There are not talking about support for GTK < version 3.
GTK3 is indeed a very important priority. GTK3 used not one year. Without GTK3 Lazarus loses its appeal for most Linux developers.
I really ask You to bring this innovation to the finals. I would like to be well informed on the implementation of GTK3. Ready to help at least testing.
I've commited gtk3 in july into lazarus trunk (most handles (widgets) are implemented basically), so you can play with it. Most heavy task is correct implementation of cairo in winapi (look at the sources).
-
I tried to compile under GTK3, but lacks some packages OS. Distribution: Linux Mint15 Cinnamon. Please tell me what packages you need to install.
But the essence of my message even not in this. I understand that now GTK3 in Lazarus - it is a "headache". The last real edit the GTK3 in SVN was about 2 months ago. The emergence of GTK3 purpose for versions Lazarus 1.4 gives hope. But plans are not always brought to finals. Therefore, I ask to pay attention GTK3. This is really very much in demand!
-
This is really very much in demand!
Please remember that you are writing to voluntary open source developers now.
The best way to speed up the development of GTK3 bindings is by offering your help.
Study the code, create code, make a patch...
Every developer has started by doing that same thing, sending patches.
-
Please remember that you are writing to voluntary open source developers now.
The best way to speed up the development of GTK3 bindings is by offering your help.
Study the code, create code, make a patch...
Every developer has started by doing that same thing, sending patches.
Yes I can understand it. If my knowledge of Pascal and LCL would have been enough, I would have started with writing patches. But I do not have adequate knowledge for solution of such tasks.
Unfortunately, that is all I have left.
-
I tried to compile under GTK3, but lacks some packages OS. Distribution: Linux Mint15 Cinnamon. Please tell me what packages you need to install.
You must install *-dev packages of gtk3, pango and cairo.
-
You must install *-dev packages of gtk3, pango and cairo.
Unable to find the appropriate packages. Please tell me the approximate package names, distribution of Linux Mint 15.
-
Unable to find the appropriate packages. Please tell me the approximate package names, distribution of Linux Mint 15.
I have not tried it yet myself but looking at Mint packages I would try libgtk-3-dev.
-
Finding these types of packages is easy using something like Synaptic, which should be available in the Mint repository.
Simply type "gtk", "3", and "dev" in the Find dialog as separate words, with Look in set to Name. That will bring up all the packages related to the group GTK, 3, and dev. You do it that way because the naming conventions aren't consistent. GTK3 and GTK-3 are both possibilities.
Do similar with pango and cairo.
From that point it should be easy to pick out which packages to try. I would start with Juha's suggestion.
-
Pango and cairo and other libs are dependencies for libgtk-3-dev. They will be installed automatically. (Just checked).
-
I'm still waiting for GTK3. But updates LCL GTK3 are extremely rare (long gone). Usually collects Lazarus GTK3 program successfully. But they are not efficient. Scrolling the mouse leads to a fall. TreeView looks and does not work properly. TOpenGLControl do not support GTK3. The situation with GTK3 still abysmal.
By the way our bugtracker GTK3 not among widgets, although stated as a goal for Lazarus 1.4.
-
Implementing a new widgetset is hard work, all done by a few (1 or 2 ATM IIRC) volunteers.
You are more then welcome to suggest patches/implementations for GTK3 widgetset.
As a bonus you get your name on the contributers list O:-)
Bart
-
Bart, i understand the difficulties associated with lack of resources. But I have no knowledge that would participate in the development GTK3. However, I am very concerned about this issue.
-
Bart, i understand the difficulties associated with lack of resources. But I have no knowledge that would participate in the development GTK3. However, I am very concerned about this issue.
1.Gtk3 is alpha state.
2.I don't have enough spare time to commit more stuff to gtk3
3.Gtk3 WS skeleton is setted up and it's pretty clear, so you can for example implement brush and pen in TGtk3DeviceContext (cairo stuff), or MouseWheel event (gtk3widgets) or whatever and send patch if you're in hurry with this.
-
Unfortunately I do not have enough experience and knowledge to join the development GTK3 in LCL. I am closely following the update GTK3 in LCL. Trying to compile the demos in GTK3. In the bugtracker no tag GTK3 widget. I will post here some observations on the work of GTK3. I hope this will be useful.
Again compiled the demo. Overall pleased with how GTK3 automatically scaled for HiDPI. But (subjectively) think that the buttons become too (excessively) large. This leads to the fact that the space of the screen is not very rational.
Unfortunately GTK3 still far from at least test use. The program falls on standard manipulation interface. Many controls are not supported.
P.S.: Waiting for GTK3 like nothing else. Pray to God about one... save us from GTK 4... at least 50 years... otherwise LCL will not have time...
-
Waiting for GTK3 like nothing else. Pray to God about one...
Praying God does not help now, providing code to implement it would help.
If you really wanted this, you would dive into the code, learn and improve it. Clearly you don't want it badly enough.
GTK3 is indeed a very important priority. GTK3 used not one year. Without GTK3 Lazarus loses its appeal for most Linux developers.
That was from your earlier message.
Sure in future GTK3 will be more important than GTK2, just like GTK1 was replaced by GTK2.
Still, it is not a "very important priority", it is only one widgetset among many others. The desktop Linux already has LCL-QT which is well maintained. Other Linux based gadgets like Android require a CustomDrawn widgetset. I think CustomDrawn should be the highest priority now.
When GTK3 finally replaces GTK2, I believe somebody will have enough motivation to complete LCL-GTK3. It has not happened yet.
-
I think CustomDrawn should be the highest priority now.
Just curious. Is CustomDrawn still developed?
-
Clearly you don't want it badly enough.
Some people are able to do a good paint, other people using these paint - paint pictures. That can be done by one person, but the reality is otherwise.
just like GTK1 was replaced by GTK2
This happened in about 1-1,5 years before the appearance of GTK3. So I pray to God
Pray to God about one... save us from GTK 4... at least 50 years... otherwise LCL will not have time...
The desktop Linux already has LCL-QT which is well maintained.
This actuality qt5? And it does not depend on third-party libraries? And TOpenGLControl works without "crutches"? Still have not heard about this. I would appreciate a link to the information.
Other Linux based gadgets like Android require a CustomDrawn widgetset.
Android and Linux is not the same thing. People who are interested in Android are not interested in Linux for desktop systems. Those who are interested in Linux those not interested in Android.
-
Just curious. Is CustomDrawn still developed?
Slowly yes. I believe contributors would be welcome there, too.
-
The desktop Linux already has LCL-QT which is well maintained.
This actuality qt5? And it does not depend on third-party libraries? And TOpenGLControl works without "crutches"? Still have not heard about this. I would appreciate a link to the information.
There's no such information since we don't have qt5 C bindings. LCLQt will always depend on C bindings. Take it or leave it.
Same applies for opengl. Currently qt implementation uses QWidget, and not QGLWidget , so there might be some problems.
Other Linux based gadgets like Android require a CustomDrawn widgetset.
Android and Linux is not the same thing. People who are interested in Android are not interested in Linux for desktop systems. Those who are interested in Linux those not interested in Android.
That's not everyone opinion. Only yours.
-
There's no such information since we don't have qt5 C bindings.
So we no GTK3. That is, do not have any actual widget Linux... maybe it's really an not urgent task. As a developer, you know better.
But I have no doubt: the lack of current widgets:
1. reduces the level and attractiveness of Lazarus as a development tool;
2. reduces the attractiveness and competitiveness of the developed programs in Lazarus.
That's not everyone opinion. Only yours.
Judging by the response I could not (by means translate.google.com) to convey the idea.
Surely you misunderstood me. I talked about the following: developing for Linux and development for Android - they are two different platform, two different target groups of developers. They do not intersect. It's a fact ... he did not is connected with my opinion.