KDE3-Forked Desktop Updated With Partial FreeBSD Support, Compatibility For GCC 6

[phoronix]

Phoronix: KDE3-Forked Desktop Updated With Partial FreeBSD Support, Compatibility For GCC 6

For those still preferring KDE3 to what's offered by KDE4 or KDE5, the Trinity Desktop Environment continues living on as a fork of the KDE3 code-base but with support for making use of modern components...

http://www.phoronix.com/scan.php?page=news_item&px=Trinity-Desktop-R14.0.4

[danieru]

Trinity is pretty cool, I might switch from lxde

[lumks]

is there a good reason that KDE3/Trinity needs a lot more RAM than Plasma 5 (5.8;apps 16.08)?

fresh start of trinity (without running an app)
~500MiB
Fresh Plasma
~300MiB

now start gwenview (that really often crashes on trinity -_-) with the same pic (1920x1080 Background jpg)
+300MiB
Plasma
+50MiB

Tested on Debian Sid 64bit

[tg--]

I'm not sure how KDE3 works, but older window managers (and other programs) without a compositor often cache images via the Xserver and X pixmaps, which are really memory inefficient.
Newer KDE uses compositors and will very likely not cache via X, but keep independent buffers that are handed to X.

[milkylainen]

I love how projects keep on calling compiling with GCC newer version for "compatibility" when it's actually more like "We had to fix our shitty code".

[Slartifartblast]

Nice to see some enthusiasts keeping old code alive but time does move on.

[cb88]

KDE3 has a memory vs CPU/GPU trade off. KDE3 uses a little more memory but does a lot less work on the GPU and CPU. KDE3 will run on 200Mhz or less hardware with just a little 2D acceleration for blitting. The go to Plasma and GTK3 desktops have serious issues even on a 1.7Ghz AMD netbook.

[Passso]

I did love KDE3 so much back then, but now that I see it again it makes me think about a child tablet
Anyway long live to Trinity, the revival is cool!

[birdie]

Most distros actively shun TDE because maintainers don't trust this project: instead of patching KDE 3.5.10 to make it compatible with modern distros, Timothy made some serious changes to the codebase by renaming KDE/Qt classes and their methods. This doesn't make TDE his pet project, however that makes using TDE patches for running KDE 3.5.10 impossible.

What Timothy should have done was:
1) Leave KDE moniker and core code intact
2) Up the version and start from zero, e.g. "KDE 3.6.0 Timothy Edition" or something like that
3) Create an easily mergeable patchset so that people could use their patches to run KDE 3.5.10

That would have made TDE a decent contented among other Linux desktop environments. Alas, without the inclusion in Fedora/Ubuntu/Mint/Arch by default this project will eventually wither and die. KDE 3.5.10 is still miles better than the current releases of XFCE, LXQT, Cinnamon, Mate and others. I could even say that it's better than Gnome 3/KDE 5 but this idea is hugely unpopular among KDE/Gnome zealots.

[birdie]

s/contented/contender/