Trying to provoke a discussion from wmi users here (Dusty?) - I like that they're trying to make wmi a bit prettier, but come on, if you're using it right, you don't see crap... and it seems like this guy has a thing for plan9.... look at some of the stuff for wmii... "X is from plan9 and look at feature Y (think plan9)" - I hope this doesn't kill wmi development.. but I think it did
"reduction of compile time"
==> i hate this statement! the reduction of compile time should not be cared by the coders but only of coders of the compiler!
execution efficiency is another story (i mean O() ;-) )
The impossible missions are the only ones which succeed.
I saw the new wmi (wmii aka wmi 2) a week or two ago and I think it's still interesting.
I don't know if using things from plan9 is good or not. I've never used it. I hope they know what they are doing. At least they are actively developing it :-)
In case of small wms like wmi it's always possible that someone will create a fork if something goes wrong (just look at the number of *boxes wms).
The real question is: Is the modularization good or not?
I think it might work since wmi is not for newbies and this should make it
And about ideas from plan9. Interesting things are in this wmii pdf:
http://w3studi.informatik.uni-stuttgart … m/wmii.pdf
Uses some ideas of the Acme Programming Environment
heavy and effective use of the mouse, including mouse chords
The IonWindowManager organises ordinary managed windows in a similar fashion to Acme.
And there is
Support of classical overlapping window management
All control/management operations accessible with
Tabbed window management (Ion-alike)
Tiled (non-overlapping) window management (LarsWM-,
WMII: new virtual filesystem based configuration/IPC
So I guess they are still not that far from the original idea. We'll see...
Trying to provoke a discussion from wmi users here (Dusty?)
Actually, I'm experimenting with KDE right now. I'm trying to figure out what the hell it is "normal" people like about point and click (tm) and drag and drop(tm). I figure I have to give it a fair chance and all.
I haven't used wmii at all. I understood they were going to totally rewrite wmi from scratch; I didn't realize it was going to run any differently. I have two views on the whole subject; rewriting things from scratch almost always end up in a better program because you learn from earlier mistakes. But wmi was almost perfect... I don't know why it would need rewriting.
"fewer lines of code"? Why not write it in Python? 8)
Perhaps because a python window manager would pretty much screw the "performance improvements" goal
By the way, what's the stats of cairo in arch ? I ask this because wmii depends on cairo to be built, and the two packages of cairo that I've seen are on staging and unstable, and both are
python aint that slow, of course it isnt quite there as good as C, but its damn good enough. more than good enough for most people
I know. But we are talking about a window manager here. I would like to be proved wrong (seriously), but I don't think a python wm would be really good when it comes to speed.
I saw this turning into a language war. Why didn't I keep my mouth shut?
There are dozens of python wm projects out there, mostly defunct, I believe. I don't think speed was the major problem, however.
IF there were speed critical portions of a wm that weren't up to snuff in Python, they could always be rewritten as C modules (leaving the main code in Python), or perhaps using Pyrex... I don't know if this has been done in any project.
Well, let's forget about this language stuff then.
Anyone built the latest snapshot of wmii ??? Is it any good ???
I'm compiling XFCE right now to test it out, but I don't think I'm going to use it
for more than one day. I just grew fond of wmi
And if anyone could answer my question about cairo, I would be really grateful.
How did you install cairo ??? Both the packages that are available (one in staging and one in unstable) are old, and they are missing some of the files needed (cairo-xlib.h for example).
I also love ruby, and after I found out that there are ruby bindings for e17, I'm really
thinking about giviing e17 a try (even tough it was really slow here the last time I tried).
BTW according to wmii docs cairo is optional.
yeah but I couldn't compile it out...
My only hope is that they keep the wmi functionality as it exists now...
And i wish I could cycle through all clients and not just the full screen ones...
Well, I just compiled it. Just had to change the config.mk file to tell it to use an older version of cairo.
But man, it sucks
It's not really usable right now. I'm sticking with evilwm for the time being.