Builder: A New Development IDE Being Built For GNOME

[the303]

As for the IDE, Eclipse is a horrible nightmare that is slow and I have consistently had weird quirks with CDT. QtCreator is nice and my primary IDE but as the name suggests it is Qt Creator. As QtCreator is fine and accepted, with that logic there shall be a GTKCreator. Nothing surprising there.

[Guest]

As it should.

Apple develops OS X for Apple.

Apple isn't divided between KDE, Gnome and few other DEs. It seems gnome thinks it's alone, but the truth is KDE and Unity have much bigger market share.

[Akka]

Apple isn't divided between KDE, Gnome and few other DEs. It seems gnome thinks it's alone, but the truth is KDE and Unity have much bigger market share.

Market share impossible to convert to money is not worth much

[Akka]

Netbeans has GUI Builder.



Eclipse has Window Builder.



Does this new GNOME builder IDE possibly stand up to the mature IDEs which are better than the rest and can handle drag and drop GUI building? Is it planned? I'm ok with it as a plugin, it doesn't have to be native like Visual Studio.(Window Builder, for example, is a plugin) But drag and drop programming is a must. Its a dealbreaker.

(Note, these screenshots may be old.)

This is his least problem. In this case he can use glade or a glade plugin.
The big thing is support the different language. Implement a parser that understand them is a huge job. Hopefully they make something reusable for all gnome project in need of this. Maybe he use this https://wiki.gnome.org/Projects/CodeAssistance ?

[Karl Napf]

Qt does the same thing. QtWidgets are as stable as it gets and Qt is now working on the next generation product of theirs which is QML.

Sure, but QtWidgets are fully supported still, bugs get fixed, even new features get introduced occasionally. It is not where the cool stuff happens, but it definitely is not abandoned either:-)

[energyman]

I am sorry to break it to you but that is the definition of development. What do you expect when something matures, everybody just should look at the mature code all day? A little of the resources are allocated to maintain the stable versions while the new/shiny/broken/unstable becomes the new stable product (or it doesn't become so if it's not good).

no, that is not development, that is stupidity. Yes. you are expected to look at the mature code all day. Fix the bugs that are there, carefully ad features.

What you must NOT do is abandon it as soon as it is not horrible broken anymore for something else that will be incomple, incompatible and broken.

hal, *kit, original rpm... the list is long.. and it gets worse when you bring in gnomes very bad track record.

I am sure, as soon as systemd's kinks are worked out, it will be abandoned for something new, shiny and-oh-so-much-better. If it doesn't implode from feature creep first.

[Karl Napf]

This is his least problem. In this case he can use glade or a glade plugin.
The big thing is support the different language. Implement a parser that understand them is a huge job. Hopefully they make something reusable for all gnome project in need of this. Maybe he use this https://wiki.gnome.org/Projects/CodeAssistance ?

Yeap, there is the glade UI builder and that is planned to be integrated according to the slides.

The big thing is indeed support for all the different languages. But from what the slides say he does not even want to address this but use gtk sourceview instead and maybe use clang for some things and languages in other places. So you are basically guaranteed to get sucky syntax highlighting and code completion -- as bad as you get it in e.g. gedit or anjuta.

The whole architecture is unsuited to do something better: It is centered around stand alone applications sharing data via D-Bus (not exactly famous for its high bandwidth/high throughput, but of course kdbus will fix all that at some point in the future) that get integrated into one UI. That is great architecture for loosely coupled applications as you typically find in a desktop session, but not great for wiring several distinct pieces for access to one huge pool of data.

[Karl Napf]

QtCreator is nice and my primary IDE but as the name suggests it is Qt Creator.

Qt Creator is written by the same people that write Qt and of course it targets their use-cases before all others. But that does not limit the IDE to Qt development.

In fact I know several Linux kernel developers using Qt Creator. There are micro controller developers that work with 8bit machines with 16K of RAM in Qt Creator (which is pretty much as far away from Qt development as you can get;-). There even is a plugin shipped with Qt Creator enabling hardware debugging on those small machines.

[JustADirtyLurker]

Why am I saying that? A good IDE does understand the code you write.

This. I'm sure that vim or emacs or other editors can be transformed into IDEs using the suitable plugins, but please stop saying and thinking that an IDE is just a fancy editor with syntax highlight.
If their purpose is to push more, unexperienced people to code Gnome/GTK apps using this reference IDE (the experienced people will already code GTK with their preferred IDE), they need a real IDE, not a stupid syntax highlighter. Because that's what IDEs are for: to understand what the coder wants to do, to offer alternatives and corrections, to help refactoring.

This is going to be another anjuta, maybe a bit prettier. I am not going to hold my breath for this one.

I never used Anjuta, but wasn't it a GNOME project itself? Why are they scrapping all that? Was its status bad?

They should focus on having a GNOME office lightweight suite like Calligra/KOffice (for KDE), instead... sigh. The current "Gnome office" project is an awful mix n mojo of different and definitely not integrated pieces of software: Abiword and Gnumeric don't talk to each other, they miss a presentation app, a db frontend, etc etc etc...

GNOME projects are interesing, but they seem to miss the focus sometimes. I mean, with Gnome3 it's clear that they are targeting a hybrid desktop/tablet environment. So, why creating a Boxes app? Nobody with a tablet needs that. Instead, a lightweight Office GTK suite, it could be really useful.

[the303]

no, that is not development, that is stupidity. Yes. you are expected to look at the mature code all day. Fix the bugs that are there, carefully ad features.

What you must NOT do is abandon it as soon as it is not horrible broken anymore for something else that will be incomple, incompatible and broken.

hal, *kit, original rpm... the list is long.. and it gets worse when you bring in gnomes very bad track record.

I didn't say it should be abandoned. In the case of open source nothing is abandoned anyway. Everybody is free to fork it and support the old project instead of bitching about that you cannot get something you want for free. There are many forks to GNOME 2. What are you obsessed about the name GNOME and you want a rename every new major overhaul?

If we apply what you are saying to the linux kernel. LTS (which needs to be similar to final GNOME 2 in spirit) should disappear and we should feature freeze the whole kernel because it seems it is okay and stable the way it works. e.g. There was apparently no reason to add epoll to the kernel beacuse select indeed used to work fine I guess. Why kernel releases break proprietary GPU drivers or at least requires patches to the driver module source? Let's keep the old code it was working, right?

What exactly is stopping you from using GNOME 2 by the way apart from the case that whatever distro you are using makes it mandatory?

I am sure, as soon as systemd's kinks are worked out, it will be abandoned for something new, shiny and-oh-so-much-better. If it doesn't implode from feature creep first.

Nobody forced anybody to use systemd. Apparently it made things easier/better somehow. If any project comes along and it has better aspects, systemd will be obviously dumped, too. Or we should stick to the systemd no matter what happens by your logic?

Do you know what your logic brought? This X.org bullshit. Sticking to some outdated thing way longer than necessary.