Linked by Thom Holwerda on Wed 6th Apr 2011 17:50 UTC, submitted by Cytor
Gnome The day is finally here, the day that the GNOME team releases GNOME 3.0, the first major revision of the GNOME project since 2002. Little of GNOME 2.x is left in GNOME 3.0, and as such, you could call it GNOME's KDE4. We're living in fortunate times, what, with two wildly divergent open source desktops.
Thread beginning with comment 469424
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE[6]: Tested it for a few hours
by kaiwai on Thu 7th Apr 2011 05:41 UTC in reply to "RE[5]: Tested it for a few hours"
kaiwai
Member since:
2005-07-06

gtk3 is completely incompatible with gtk2 source or otherwise.

Edit:

Well you may have a point. But developers who maintain an application larger than "Hello World" would definitely have their work cut out for them.

I can't remember the details of the changes needed to port from QT3 to QT4, but I do know for most gtk developers (especially the ones that use the language bindings, there are lots of them these days) would definitely need to do a rewrite of their apps.

Migration Guide:
http://developer.gnome.org/gtk3/stable/gtk-migrating-2-to-3.html


I kind of wonder where OpenOffice.org/LibreOffice fits into the equation given that there is a degree of GTK+/GNOME integration but how much of a PITA is it going to be with the migration from gtk2 to gtk3. I've always wondered to what extent is the open source world better off stripping off the abstraction layer from OpenOffice.org/lIbreOffice in favour of replacing it with native front ends as to have the best native experience rather than the half baked pseudo integration that exists today.

IIRC a lot of the applications have already done the preparation work mentioned in the documentation with the last part of the migration being pretty straight forward. Although I am seeing GNOME and many open source projects becoming more and more 'Linux centric' with the lack of contributions by non-Linux developers (why isn't there a native FreeBSD backend to the GNOME features that hook back into the system?) but in the long run if it means consolidation where Linux becomes a strong number 3 competitor in the desktop world then hopefully it'll translate into more vendors will to consider providing software on said platform.

With that being said, others have pointed out that GNOME 3.x has become more of a a complete platform for developers to aim against with the desktop abstracting things such as Bluetooth and printing which should mean application writers aim for the GNOME desktop and the GNOME desktop API's take care of the rest. If GNOME keep working down this road then I see in the long term third parties seeing it as a viable environment to target.

Edited 2011-04-07 05:47 UTC

Reply Parent Score: 2

Mystilleef Member since:
2005-06-29

I share your sentiment on LibreOffice. I'd much prefer it be split into a backend and frontend and have each community design the frontend in their native widgets (Qt, Gtk, Windows, Quartz, etc). The frontend should of course should be designed to an accepted specification. How well this will work in practice is open for debate. I know old school unix apps back in the day used to be designed with multiple toolkits/interfaces (especially the command line interface) in mind. It wasn't hard find an application with a cli, motif, tcl/tk, gtk, gtk2 and qt interface. Maybe it's time to resurrect that culture. LibreOffice's emulation layer will eventually be ported to GTK3 but GTK2 isn't going anywhere anytime soon.

I'm sad the BSDs are second class citizens in the free desktop ecology. I don't think their marginalization is intentional. The reality is that it takes resources to port and maintain projects to multiple platforms. At the moment the resources and momentum are skewed towards Linux. The developers have held their end of the bargain. They've worked hard to make free software available. It's now left to the community and corporations to provide funding and/or resources to make sure the software runs on ALL free software platforms. At the very least, the BSD community if they are interested in GNOME should find a way to influence the decisions in the GNOME project.

What constitutes a GNOME 3 application is not solidified. Heck GNOME 3 isn't even solidified. What we're really celebrating today is GNOME Shell and the vision that GNOME 3 might become. A lot of so called "GNOME libraries" have been pushed into GTK3. So today, if you wrote a GTK3 application you can easily get away by calling it a GNOME 3 application. I think eventually what will make an application a "real" GNOME 3 application is how well it integrates into and interacts with GNOME Shell.

From a technical perspective you can have a look at what the GNOME Platform consist of. http://developer.gnome.org/

Reply Parent Score: 2

kaiwai Member since:
2005-07-06

I share your sentiment on LibreOffice. I'd much prefer it be split into a backend and frontend and have each community design the frontend in their native widgets (Qt, Gtk, Windows, Quartz, etc). The frontend should of course should be designed to an accepted specification. How well this will work in practice is open for debate. I know old school unix apps back in the day used to be designed with multiple toolkits/interfaces (especially the command line interface) in mind. It wasn't hard find an application with a cli, motif, tcl/tk, gtk, gtk2 and qt interface. Maybe it's time to resurrect that culture. LibreOffice's emulation layer will eventually be ported to GTK3 but GTK2 isn't going anywhere anytime soon.


Personally I'd sooner have each platform to be designed in a way that conforms to the operating systems UI specifications rather than a situation where each front end might use a native widget kit but they all try to 'feel' the same through everything laid out the same way. For me I'd sooner there to be a shared back end but when it comes to the front end that the GTK+ one behaves and laid out like a GNOME application or the Mac front end is laid out like a Mac OS X conforming application.

I've never really investigated much into OpenOffice.org/LibreOffice as so far as the code base itself but from what I have read in terms of what programmers are saying about it on the mailing lists, it isn't a pretty thing to work with. Lots of dependencies for example, code that some people don't know what it does, very few programmers from a UI centric background whose focus is on fit in finish rather than solely on a utilitarian solution (does it get me from a to b), no real long term solution to move SAL further down and replace the UI parts of it with native widgets etc.

I'm sad the BSDs are second class citizens in the free desktop ecology. I don't think their marginalization is intentional. The reality is that it takes resources to port and maintain projects to multiple platforms. At the moment the resources and momentum are skewed towards Linux. The developers have held their end of the bargain. They've worked hard to make free software available. It's now left to the community and corporations to provide funding and/or resources to make sure the software runs on ALL free software platforms. At the very least, the BSD community if they are interested in GNOME should find a way to influence the decisions in the GNOME project.


Well the big thing is the majority of the contributors are Linux developers so they're going to focus on the platform that they run, the responsibility for FreeBSD support in KDE and GNOME falls on the shoulders of developers who use FreeBSD as their platform of choice rather than it being the responsibility for Linux developers. I know for me if I was an x86 user I'd be more than happy to put up a bounty for a FreeBSD developer(s) to write FreeBSD backends to GNOME and KDE but since I'm a Mac user it wouldn't make much sense. If someone said to me, "hey, lets put together a bounty to pay for it" I'd be happy to throw some cash their way simply to see it happen but I'm really in no position these days given Mac OS X is my primary platform.

What constitutes a GNOME 3 application is not solidified. Heck GNOME 3 isn't even solidified. What we're really celebrating today is GNOME Shell and the vision that GNOME 3 might become. A lot of so called "GNOME libraries" have been pushed into GTK3. So today, if you wrote a GTK3 application you can easily get away by calling it a GNOME 3 application. I think eventually what will make an application a "real" GNOME 3 application is how well it integrates into and interacts with GNOME Shell.

From a technical perspective you can have a look at what the GNOME Platform consist of. http://developer.gnome.org/


With the move to the GNOME shell, are they going to update or replace the GNOME HIG? If they're going to change the paradigm that much then wouldn't it necessitate a revision of the HIG?

Reply Parent Score: 2