Hi Fabian, On Di 03 Jun 2014 14:36:00 CEST, Fabian Greffrath wrote:
The MATE desktop environment is a fork of what was formerly known as the GNOME v2 desktop environment. The MATE upstream developers have performed a really good job in integrating the old GNOME code with latest technologies like DConf an GSettings. The next upcoming releasePlease don't get me wrong, but I am still not sure if this tremendous effort would not have been better invested in "GNOME Classic" or "GNOME flashback" instead of creating (and packaging) another fork. Speaking of packages, how does e.g. mate-panel 1.8 differentiate from gnome-panel 3.8?
Please test that yourself. The MATE panel is quite similar to the GNOMEv2 panel. When I switched from GNOMEv2 to GNOMEv3 classic, I sort of missed several configuration options around panels. With MATE these options have come back now.
of MATE (which will be the 1.10 series) will also have GTK3 support (if things go well!). During the last 6 months several people have worked on the provisioning of MATE packages in Debian.Same question for other packages: If e.g. engrampa 1.10 (MATE's pendant to file-roller, the archive utility) gets ported to GTK3 and features DConf and GSettings support, how does it differentiate from GNOME's file-roller? Why not rather use the original actually?
The MATE desktop consists of several components that closely interact with each other. Only those GNOMEv2 components absolutely necessary got forked by the MATE project. Updating to newer backend technologies like dconf, gsettings or gtk3 is simply about moving with time and development of shared libraries.
It's like propping up a VW T3. The appearance stays, but the engine under the hood got quite refurbished.
Again, I don't want to degrade the MATE project or your packaging effort, not at all. But my questions and concerns are serious. I still do not get the point of the whole MATE desktop now that it tries to run behind and catch up with GNOME again.
Let me name you several of my personal reasons for giving so much time into packaging MATE (several tens of hours during the last six months):
o MATE works perfectly in X2Go sessions (remote desktop solution for Linux servers),
[Note that I am member of the X2Go upstream development team...]
o MATE scales well without graphical hardware acceleration
o The resources footprint is much smaller than that of GNOMEv3
o Many of my customers run GNOMEv2 on Debian squeeze. Many users
don't want to
change. With MATE, I can provide a desktop shell that allows
smooth migration
of those customer setups without enforcing too much change on users.
o GNOME classic/fallback/flashback has become obsolete by upstream AFAIK
There are goods and bads about forked code. GNOMEv3 is a completely
different product compared to GNOMEv2. So, a fork at that time, like
it was done by the MATE people, absolutely made sense (to me).
If a group of devs comes together and they do a good job in maintaining the code, then a fork certainly is a way to go.
And to my impression, the MATE upstream team does a really GREAT job!!! I have been working with them closely together for the last six months and the interaction has been really good and also very professional.
Greets, Mike -- DAS-NETZWERKTEAM mike gabriel, herweg 7, 24357 fleckeby fon: +49 (1520) 1976 148 GnuPG Key ID 0x25771B31 mail: mike.gabriel@das-netzwerkteam.de, http://das-netzwerkteam.de freeBusy: https://mail.das-netzwerkteam.de/freebusy/m.gabriel%40das-netzwerkteam.de.xfb
Attachment:
pgpTF1H4SwGSP.pgp
Description: Digitale PGP-Signatur