1 |
Nikos Chantziaras <realnc@×××××.de> posted gm2gv4$45g$1@×××××××××.org, |
2 |
excerpted below, on Sat, 31 Jan 2009 23:48:34 +0200: |
3 |
|
4 |
> Duncan wrote: |
5 |
>> But... while I have kde-4.2 merged now, I've not yet played with it |
6 |
|
7 |
Well, I did a marathon session of playing and configuration today... |
8 |
|
9 |
> I also was waiting for a "better" version when KDE 4.0 came out and was |
10 |
> sticking to KDE 3. Then KDE 4.1 came out but I still preferred 3. Two |
11 |
> days ago I tried KDE 4.2.0 and it seems that this is the version that |
12 |
> can replace KDE 3 for me. So far, I didn't login to KDE 3 again after I |
13 |
> had 4.2.0 running. |
14 |
|
15 |
KDE 4.2 is definitely better than 4.1, and I think after I get it |
16 |
customized it'll be actually usable, definitely the first time I've been |
17 |
able to say that on the 4.x cycle, but let's put it this way... |
18 |
|
19 |
I can't say that Linus was wrong switching away for awhile, as it's |
20 |
certainly not up to 3.5.10 yet and has a quite a way to go, and given |
21 |
that, on distributions like Fedora that aren't making it easy to continue |
22 |
to run 3.5, Gnome may unfortunately be one of the least-worst options. |
23 |
|
24 |
> The only feature I'm missing from KDE 3 is sticky |
25 |
> keyboard layout switching and HAL mount options. |
26 |
|
27 |
I'm still missing several. Speed for one. |
28 |
|
29 |
My video card is an older Radeon 9200SE as that was pretting much the |
30 |
best available with freedomare drivers for several years, and even still, |
31 |
full support on the newer cards isn't yet stable -- tho it's improving |
32 |
very fast and looks to be headed there by the end of the year. The |
33 |
problem is that the R2xx series chips (of which the Radeon 92xx is one of |
34 |
the last) only support upto 2048x2048 OpenGL acceleration, and at |
35 |
1960x1200 stacked for 1960x2400, the bottom 352 pixels of my viewport |
36 |
aren't OpenGL accelerated. |
37 |
|
38 |
That means none of the fancy OpenGL stuff works reliably. On older |
39 |
versions it would crash. On 4.2, it simply doesn't enable them and I use |
40 |
XRender. |
41 |
|
42 |
The problem is that while it disables the effects themselves, there's |
43 |
little if any hint in the GUI what depends on OpenGL. I can checkmark |
44 |
them on and only by actually trying to invoke them and "nothing happens" |
45 |
do I see (rather, am left to /guess/) that said features require OpenGL. |
46 |
|
47 |
While "not crashing" in such cases is a VAST improvement over 4.1, "not |
48 |
crashing" really isn't good enough. They need at least a description |
49 |
saying they only work with OpenGL, and would in most traditional |
50 |
instances simply be disabled, since it won't let me enable OpenGL. |
51 |
|
52 |
And the features that work... without good OpenGL, are still relatively |
53 |
slow. Sure, they can be turned off, but there goes much of the reason |
54 |
one might otherwise find kde 4.x better, to date at least. |
55 |
|
56 |
Without OpenGL and with some of the composite effects turned off that |
57 |
actually work reasonably well in 3.5 because they're too slow in 4.2... |
58 |
yes, it's now at least functional, but there's really not all /that/ much |
59 |
left to justify the extreme hassle and tens of hours of work (I'm |
60 |
guessing it'll be roughly 40 hours here, a full-time work-week |
61 |
equivalent) recustomizing! |
62 |
|
63 |
> I also didn't have a single crash/segfault (yet) in KDE 4.2.0. Plasma |
64 |
> and KWin used to crashed like hell in 4.1.4. It's very stable on my |
65 |
> machine; running straight for about 20 hours now. |
66 |
|
67 |
I DID get a couple things to segfault, but I was pushing it pretty hard |
68 |
(BECAUSE I HAD TO GUESS BECAUSE THEY CAN'T DISABLE WHAT WON'T WORK |
69 |
ANYWAY). But I expect once I get thru customizing everything and have it |
70 |
configured (BY GUESSING AND TRIAL AND ERROR, SINCE THEY CAN'T BE BOTHERED |
71 |
TO DISABLE WHAT WON'T WORK!) more reasonably within the limits of my |
72 |
hardware, segfaults, etc, will be pretty rare. |
73 |
|
74 |
For a .2 release, that's... well, honestly, I'd expect MS to be pulling |
75 |
stuff like that, and from what I've read, they actually did, with Vista. |
76 |
Heh, at least I didn't pay good money for the privilege of getting |
77 |
crapped on, like all those "Vista Ready" folks did! |
78 |
|
79 |
But it's still nothing the FLOSS community can be proud of, especially |
80 |
for an X.2 release (where X > 0)! And it's nowhere near the smooth |
81 |
functioning of 3.5.10, and by X.2 of the next version, it should really |
82 |
be better. |
83 |
|
84 |
Really, the experience would have been /much/ better of they had just |
85 |
disabled in the GUI all the effects that weren't actually available since |
86 |
I couldn't enable OpenGL. That's something they obviously know, whether |
87 |
an effect requires OpenGL or not, and the infrastructure is already |
88 |
there, so why are they still forcing trial and error to find out what |
89 |
works and what doesn't, especially by X.2 (I could see it for x.98 betas, |
90 |
feature complete but rough aroudn the edges)? That makes no sense at all! |
91 |
|
92 |
> So far, it looks good. Make sure you have upgraded to |
93 |
> x11-libs/qt-gui-4.4.2-r2 though and then rebuild kdelibs and |
94 |
> plasma-workspace (to make sure)! It fixes crashes that |
95 |
> occur with 4.4.2-r1. |
96 |
|
97 |
I did! |
98 |
|
99 |
|
100 |
Meanwhile, khotkeys sort of works now... but the "extra" key I had most |
101 |
of my app-launchers stacked on (two-key sequence invocation), XF86WWW, |
102 |
works fine in 3.5.x and has since before 3.5, and it's detected on 4.2, |
103 |
but the message when I try to set it is, "Qt 4.x doesn't support the |
104 |
selected key." or something to that effect. WTF!? It can SEE the key |
105 |
and knows I pressed it in ordered to give me that message. It worked |
106 |
perfectly fine in 3.5 (and before)! What's this about not supported?! |
107 |
|
108 |
Same story, different page, and it's getting tiresome. It's that sort of |
109 |
stuff that should NOT still be happening in an X.2 version. |
110 |
|
111 |
... And 4.2 was what they were billing as finally worth switching for! |
112 |
<shakes head> Well, It might finally be barely tolerable if that's the |
113 |
best option available, but it's frankly extremely disappointing. This |
114 |
would arguably be 4.0 material, but more accurately, I'd call it x.0-rc |
115 |
material. It's really a cryin' shame! |
116 |
|
117 |
And... I don't see any way yet to enable anything parallel to the |
118 |
ksysguard kicker applet, which was how I did most of my system |
119 |
monitoring. The various system monitor plasma applets sort of get there, |
120 |
but aren't very flexible and so far, seem way dark (or way bright, |
121 |
presumably, in a different scheme). |
122 |
|
123 |
... |
124 |
|
125 |
As I expected, 4.2 does finally seem functional enough to actually run |
126 |
and bug report on, functional enough I can start configuring and |
127 |
hopefully keep it instead of blow it away as I did with the 4.1 config, |
128 |
to prevent bugs, etc. But it's still vastly disappointing. Hopefully |
129 |
when I upgrade video cards and get full OpenGL support, probably about |
130 |
the 4.3 timeframe, the combination of the two will finally be 3.5.x |
131 |
parity functional replacement, even improvement. But I wouldn't call |
132 |
this there yet. But, as I said, at least it's functional enough I can |
133 |
run it and not be filing bugs on essentially the entire thing, now! |
134 |
That's vast improvement from 4.0 and 4.1! So they're getting somewhere. |
135 |
It's just not really "there" yet, for me and my obviously way too extreme |
136 |
for them power user demands. But 3.5.10 could do it, and I remain |
137 |
hopeful for 4.3, and 4.4, and.... |
138 |
|
139 |
-- |
140 |
Duncan - List replies preferred. No HTML msgs. |
141 |
"Every nonfree program has a lord, a master -- |
142 |
and if you use the program, he is your master." Richard Stallman |